Reed Beverly M

THE EFFECTS OF STUDYING THE HISTORY
OF THE CONCEPT OF FUNCTION

ON STUDENT UNDERSTANDING OF THE CONCEPT
A dissertation submitted to the

Kent State University Graduate School
of Education, Health and Human Services
in partial fulfillment of the requirements
for the degree of Doctor of Philosophy
by
Beverly M. Reed
December 2007
© Copyright by Beverly M. Reed 2007
All Rights Reserved
ii
A dissertation written by
Beverly M. Reed
B.A., John Carroll University, 1975
M.A., John Carroll University, 1979
Ph.D., Kent State University, 2007
Approved by
___________________________________ Co-Director, Doctoral

Ed Dubinsky Dissertation Committee
___________________________________ Co-Director, Doctoral

Michael Mikusa Dissertation Committee
___________________________________ Members, Doctoral Dissertation Committee

Austin Melton
___________________________________
Anne Reynolds
Accepted by
__________________________________, Interim Chairperson, Department of

David Keller Teaching, Leadership, and Curriculum
Studies
__________________________________, Interim Dean, College and Graduate School

Donald L. Bubenzer of Education, Health, and Human Services
REED, BEVERLY M., Ph.D., December 2007 TEACHING, LEADERSHIP AND
CURRICULUM STUDIES
THE EFFECTS OF STUDYING THE HISTORY OF THE CONCEPT OF FUNCTION

ON STUDENT UNDERSTANDING OF THE CONCEPT (377 pp.)
Co-Directors of Dissertation: Ed Dubinsky, Ph.D.

Michael Mikusa, Ph.D.
The purpose of this study was to examine the mathematical learning that occurred
when students studied the history of a mathematical concept. In particular, the focus was
on the concept of function. Students experienced an in-depth study of the history of
functions during a 5-week unit in the junior-senior level History of Mathematics course.
They completed a series of worksheets, readings, and problems.
The research methodology was a teaching experiment and the framework for
analysis of data was APOS (Action, Process, Object, Schema) Theory. All 17 students
enrolled in the course completed an extensive initial questionnaire and 6 were selected to
participate in an in-depth interview to reveal their understanding of the function concept.
During the unit, each student wrote a series of reflections about his or her understanding.
After the unit, students completed a second questionnaire and participated in another in-
depth interview to discern the changes in their thinking about the concept.
The findings support the notion that studying the history of a mathematical
concept enables a deep reflection of ideas. Four of the six participants notably
strengthened their function conceptions. Two moved an entire APOS level. Five of the
six exhibited an increased ability to recognize a function in a given scenario. Growth was
most profound in the area of graphical representations.

ACKNOWLEDGMENTS
First and foremost, I would like to thank my co-director Ed Dubinsky. To have
someone of his caliber work with me on this project was indeed an honor. As the
developer of APOS theory, his judgment in the field is unparalleled. He was prompt with
comments and suggestions, encouraging whenever I had questions, and offered valuable
insights. He challenged me. Skeptical at first that studying the history of math would
cause significant growth in understanding, he acted as the true scientist, letting the data
speak. Thank you, Ed.
Secondly I couldn’t have completed a project of this magnitude without the
encouragement and support of my husband, Jim. There were many times that “life got in
the way,” but his gentle prodding kept me going. The endless hours he spent keeping the
home front intact were instrumental in allowing me time to do the research. Always ready
with a listening ear, his emotional support was invaluable.
Special thanks go to my colleagues Austin Melton and Bathi Kasturiarachi who
worked with me on the pilot study. Their interest and encouragement kept my motivation
alive and working with them made research most enjoyable. Austin also provided
invaluable assistance with his editing suggestions and with his continued interest,
friendship, and encouragement throughout the entire process.
Thanks also to Michael Mikusa, my co-director for keeping me on track, to Anne
Reynolds for her timely and encouraging remarks, and to Judie Melillo for her helpful
iv
editing suggestions. Sharon Smith, overall editor, did an outstanding job proofreading
the final document.
Thanks to my son, Joe, who often reminded me “if it was easy everyone would
have one” (a Ph.D.) and to Jen, my daughter, for repeatedly telling me how proud she
was of me and how cool she thought my research was.
Such a long journey would not have been as enjoyable without the support of dear
friends. Carol Steiner, for your endearing sense of humor; Deb DeBenedictis for those
wonderful walks and talks; and Mary Beth Rollick, for just being there… a heartfelt
thanks.
Lastly, I would like to thank my dear mother, who, though not a college graduate
herself, knows the value of education and kept encouraging me to continue.
Posthumously, I owe a tip of the hat to my dad, who instilled in me as a youngster a love
of learning, a creative can-do spirit, an inquisitive mind, and strong work ethic.
v
DEDICATION
To my children, Joe and Jen,
and my students,
may you continue your life-long journey of learning
vi
TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS ................................................................................................. iv
DEDICATION................................................................................................................... vi
LIST OF FIGURES ......................................................................................................... xiv
LIST OF TABLES........................................................................................................... xvi
CHAPTER
I. INTRODUCTION TO THE STUDY............................................................................. 1
Claims About the Value of the History of Mathematics ........................................... 1
Statement of the Problem .......................................................................................... 3
Purpose of the Study ................................................................................................. 5
Theoretical Basis for the Study: APOS Theory ........................................................ 6
The Action Construct....................................................................................... 7
The Process Construct ..................................................................................... 8
The Object Construct..................................................................................... 10
The Schema Construct................................................................................... 10
Instructional Implications........................................................................................ 11
Conclusion............................................................................................................... 13
Significance of the Study ........................................................................................ 14
Definition of Terms ................................................................................................. 15
vii
Summary of Chapter 1 ............................................................................................ 17
II. REVIEW OF RELATED LITERATURE ................................................................... 18
Four Theoretical Frameworks: History and Learning............................................. 18
Piaget and Garcia........................................................................................... 18
Luis Radford .................................................................................................. 20
Anna Sierpinska............................................................................................. 22
Anna Sfard ..................................................................................................... 25
Summary........................................................................................................ 29
Studies Using History of Mathematics as a Tool.................................................... 30
Studies About Learning the Concept of Function ................................................... 45
History of the Concept of Function......................................................................... 56
Origins ........................................................................................................... 56
The Fourteenth Through Sixteenth Centuries................................................ 58
Seventeenth and Eighteenth Centuries .......................................................... 60
Rene Descartes ..................................................................................... 60
Newton and Leibniz ............................................................................. 61
Leonhard Euler..................................................................................... 62
The Vibrating String Problem.............................................................. 62
The Nineteenth Century................................................................................. 64
Fourier Series ....................................................................................... 64
Cauchy and Dirichlet ........................................................................... 66
Riemann and Weierstrass..................................................................... 68

viii
Function as Correspondence ................................................................ 69
Conclusion ........................................................................................... 70
Summary ................................................................................................................. 70
III. METHODOLOGY ..................................................................................................... 72
Introduction ............................................................................................................. 72
Design...................................................................................................................... 73
Sample ..................................................................................................................... 75
Procedure................................................................................................................. 76
Triangulation ........................................................................................................... 77
Inter-Rater Reliability ............................................................................................. 78
Data Analysis .......................................................................................................... 81
Finite Sequences (#8, #18) ............................................................................ 81
Character Strings (#5, #6).............................................................................. 82
Graphs (#2, #13, #19) .................................................................................... 82
Sets of Ordered Pairs (#1, #16, #17).............................................................. 83
Tables (#12) ................................................................................................... 83
Equations (#3, #4, #11, #14, #15).................................................................. 84
Statements (#7, #9, #10) ................................................................................ 84
Pilot Study ............................................................................................................... 84
Summary of Chapter 3 ............................................................................................ 88
IV. DATA ........................................................................................................................ 89
Introduction ............................................................................................................. 89
ix
Initial Conceptions of Functions ............................................................................. 90
Participant 1: DB ........................................................................................... 90
General Function Conception .............................................................. 90
Understanding of Graphical Representations....................................... 99
Participant 2: CW ........................................................................................ 104
General Function Conception ............................................................ 104
Understanding of Graphical Representations..................................... 109
Participant 3: MJ.......................................................................................... 112
Participant 4: BG ......................................................................................... 127
Participant 5: MS ......................................................................................... 138
Participant 6: CS .......................................................................................... 149
Summary of Participants’ Initial Conceptions ...................................................... 154
APOS Theory Applied to Understanding Graphs ................................................. 155
Brief Summary of the Literature.................................................................. 155

x
APOS Characterization................................................................................ 156
The Pre-Function Construct ............................................................... 157
The Action Construct ......................................................................... 159
The Process Construct........................................................................ 160
The Object Construct ......................................................................... 162
Conceptions of Function After the Unit on History of Functions......................... 166
Participant 1: DB ......................................................................................... 166
Participant 2: CW ........................................................................................ 178
Participant 3: MJ.......................................................................................... 188
Participant 4: BG ......................................................................................... 204
Conclusion ......................................................................................... 212
Participant 5: MS ......................................................................................... 215
xi
Participant 6: CS .......................................................................................... 229
Summary of Chapter 4 .......................................................................................... 248
V. DISCUSSION ........................................................................................................... 250
Introduction ........................................................................................................... 250
Findings................................................................................................................. 250
Definitions of Function................................................................................ 251
Functions in Situations ................................................................................ 252
APOS Level Changes .................................................................................. 256
Changes in Understanding Graphical Representations................................ 257
What Caused the Improvement ............................................................................. 258
Comparisons to Other Studies............................................................................... 261
Contributions to the Field...................................................................................... 264
Future Studies........................................................................................................ 267
Cautions With Interpretations ............................................................................... 269
Conclusion and Summary ..................................................................................... 270
APPENDICES ................................................................................................................ 271
APPENDIX A. QUESTIONS ABOUT FUNCTIONS: INITIAL
QUESTIONNAIRE ......................................................................................... 272
APPENDIX B. QUESTIONS ABOUT FUNCTIONS II: SECOND
QUESTIONNAIRE (Q2) ................................................................................ 279

xii
APPENDIX C. WORKSHEET #1 GRAPHS OF FUNCTIONS.......................... 283
APPENDIX D. WORKSHEET #2 THE ROLE OF FERMAT’S ANALYTIC
GEOMETRY ................................................................................................... 293
APPENDIX E. WORKSHEET #3 DESCARTES AND THE LOCI WITH
RESPECT TO THREE OR FOUR LINES ..................................................... 304
APPENDIX F. WORKSHEET #4 LEIBNIZ AND THE HISTORY OF
FUNCTIONS................................................................................................... 310
APPENDIX G. WORKSHEET #5 RABBITS, RABBITS EVERYWHERE!
WORKSHEET................................................................................................. 314
APPENDIX H. WORKSHEET #6 EVEN MORE FIBONACCI FINDING A
FUNCTION TO REPRESENT THE FIBONACCI SEQUENCE .................. 318
APPENDIX I. WORKSHEET #7 EULER ON FUNCTIONS ............................. 323
APPENDIX J. WORKSHEET #8 FOURIER SERIES REPRESENTATIONS
OF FUNCTIONS............................................................................................. 329
APPENDIX K. WORKSHEET #9: DEFINITIONS OF FUNCTIONS ............... 335
APPENDIX L. READINGS FOR WORKSHEET #9 .......................................... 342
APPENDIX M. LETTERS OF PERMISSION TO REPRINT............................. 350
REFERENCES ............................................................................................................... 365
xiii
LIST OF FIGURES
Figure Page
1. Oresme’s Uniform Velocity and Uniform Acceleration......................................59
2. Graph for Task 13 ................................................................................................93
3. Graph for Task 19 ................................................................................................94
4. Table for Task 12 giving club members’ names and the dues they owe .............96
5. Graphing Task 3 showing speed vs. time for two cars ......................................100
6. Graphing Task 5 showing speed vs. distance along a race track .......................101
7. Graphing Task 1 showing position vs. time of two objects...............................102
8. Graphing Task 2 showing speed vs. time of a cyclist .......................................103
9. Graph for Task 19 ..............................................................................................106
12. Graph for Task 2 showing an arbitrary graph with no labels ............................118
13. Graph for Task 19 ..............................................................................................123
16. Graphing Task 1 showing position vs. time for two objects..............................126
17. Graph for Task 2 showing an arbitrary graph with no labels ............................128
18. Graph for Task 13 showing an arbitrary graph with no labels ..........................132
xiv
22. MS’s sketches showing function as mapping ....................................................140
24. Graph for Task 19 ..............................................................................................147
26. Graphing Task 5 showing speed vs. distance along a race track .......................153
27. CS’s sketch for graphing Task 4........................................................................154
33. DB’s sketches for Q2 Task 12 ...........................................................................169
34. DB’s sketch for graphing Task 4 after unit on history of functions ..................171
36. Graphs for Q2 Task 6 showing a caterpillar’s path ...........................................179
37. CW’s sketch for graphing Task 4 after unit on history of functions .................186
38. Graphs for Q2 Task 6 showing a caterpillar’s path ...........................................220
39. Graph for Task 19 showing discrete points .......................................................223
xv
LIST OF TABLES
Table Page
1. Description of APOS Levels................................................................................. 51
2. Summary of Research Design............................................................................... 79
3. APOS Levels of Participants Before Studying the History of Functions ........... 155
4. APOS Characterizations for Understanding Graphical Representations of
Functions............................................................................................................. 157
5. Conceptions of Graphical Representations Before the Unit on History of
Functions............................................................................................................. 166
6. Summary of DB’s Growth in Understanding Functions..................................... 178
7. Summary of CW’s Growth in Understanding Functions.................................... 187
8. Summary of MJ’s Growth in Understanding Functions ..................................... 203
9. Summary of BG’s Growth in Understanding Functions..................................... 215
10. Summary of MS’s Growth in Understanding Functions .................................... 228
11. Summary of CS’s Growth in Understanding Functions ..................................... 248
12. Participants Holding the Notion of “Function as Formula” Before and After
History of Functions Unit ................................................................................... 253
13. Appropriate Answers to Specific Tasks Before History of Functions Unit........ 254
14. Correct Answers to Specific Tasks after the Unit on History of Functions........ 255
15. APOS Conception Before History of Functions Unit......................................... 256
xvi
16. APOS Conception After History of Functions Unit ........................................... 257
17. Interpretation of Graphs Before History of Functions Unit................................ 258
18. Interpretation of Graphs After History of Functions Unit .................................. 259
xvii
CHAPTER I
INTRODUCTION TO THE STUDY
This chapter presents the background and the statement of the problem and the
purpose of the study. Also included is a discussion of the theoretical basis for the study.
Claims About the Value of the History of Mathematics
With the recent addition of a History of Mathematics course into the pre-service
secondary mathematics teachers’ requirements, one may wonder if, indeed, such a course
is of significant benefit to the students. Concern arises if such a course is a good
replacement for an upper division “regular” mathematics course. A number of writers in
mathematics education, however, reap praises upon studying the history of mathematics:
“Without deep, penetrating motivations, all education is lost, and it is in providing these
that I’m sure that the history of mathematics can be of great service in mathematical
education” (Grattan-Guinness, 1977, p. 122).
I have more than an impression—it amounts to a certainty—that algebra is made
repellent by the unwillingness or inability of teachers to explain why . . . there’s
no sense of history behind the teaching, so the feeling is given that the whole
system dropped down ready-made from the skies, to be used only by born
jugglers. (Barzun, cited in Jones, 1989, p. 1)
1
2
“A sufficiently concrete and detailed tracing of the history of the development of an idea
is one of the best ways to teach an appreciation of the nature and role of generalization
and abstraction” (Jones, p. 13).
The most common reasons given in the literature are the following.
• History can help increase student motivation and develop a positive attitude
toward learning mathematics (Barbin, 2000b; Liu, 2003; Rickey, 1995);
• Past obstacles in the development of mathematics can help explain why a
certain topic is difficult for students (Avital, 1994; Barbin, 2000a; Liu, 2003),
increasing a teacher’s understanding of barriers to student understanding
(Liu), and helps them understand the stages of learning (Barbin, 2000a). It
may thus lead teachers to change the way they think about their students’
errors and develop a more constructive attitude toward them. As a result of
this understanding, teachers may develop different teaching strategies (Avital,
1994; Barbin, 2000a; Liu, 2003) and focus on producing a variety of
responses to a given problem (Barbin, 2000a), incorporating problem solving
and problem posing while teaching (Avital, 1994). Learning the historical
development may affect teachers’ perceptions of the time students spend
while struggling with coming to understand a concept (Avital, 1994; Barbin,
2000a; Swetz, 1982): “If it took several centuries for mathematicians to be
able to make explicit our current concept of limit, for example, it is going to
take a considerable time for our students as well” (Barbin, 2000a, p. 65).
3
• It can provide an opportunity for developing an idea of what mathematics is;
that is, it helps students understand the reasons behind the development of
algorithms and concepts and which problems mathematics helps us to solve
(Barbin, 2000b). It leads to the idea that mathematics is not a sequence of
“discrete chapters” but is “an activity of moving between different ways of
thinking about mathematical concepts and tools” (Barbin, 2000a, p. 65). It
demonstrates that the development of mathematics is a “necessary human
activity” (Swetz, 1982, p. 696).
• Historical problems can develop students’ mathematical thinking (Liu, 2003).
• Shows that mathematical ideas evolve over a period of time, are struggled
with and subject to change (Avital, 1994; Swetz, 1982). It shows that
grappling with mathematics is a common experience and humanizes the
subject (Avital, 1994).
• Since history may change the teacher’s perceptions and understanding of
math, it would influence the way she or he teaches it and thus affects the way
students perceive it and understand it (Barbin, 2000b).
Statement of the Problem
Backing all these claims, however, are few research studies documenting the
effect that studying the history of mathematics has on a student’s learning of
mathematics. Many are anecdotal and others deal with changes in attitudes, not learning.
Po-Hung Liu (2003) claimed that NO empirical studies indicating that learning history
helps students perform better on standardized exams exist. “Although studying the
4
history of mathematics may improve students’ attitudes toward mathematics, the linkage
between attitude and achievement is neither linear nor straightforward” (p. 420). Barbin
(2000b) noted that “there exist no successful studies where the impact of an historical
dimension can be measured by using a battery of tests for determining the competencies
of students” (p. 66).
The National Council of Teachers of Mathematics (NCTM, 1998) recommended
study of the history of mathematics for teachers. The Mathematical Association of
America (MAA, n.d.) has an online magazine providing resources to help teach
mathematics using history. Resources abound for using history in the classroom
(Brummelen, 1995; Calinger, 1996; Fauvel & Van Maanen, 2000; Katz, 2000; Katz &
Michalowicz, 2005; Swetz et al., 1995), but research-based conclusions on its
effectiveness are scarce. Furinghetti (1997) claimed that not only are the training and
education of teachers in the history of mathematics non-homogeneous, but the
experiences of such teachers are quite scattered, with no “organized network of classes
and teachers carrying out analogous experiments” (p. 55). Comparing results and
establishing trends in the research of the effectiveness of teaching history is therefore
difficult. Though teachers who use the history of mathematics are enthusiastic about its
effect, these opinions are subjective, not a result of regular and systematic studies
(Furinghetti, p. 55). In 2000, the ICMI (International Commission on Mathematics
Instruction) published a study volume to survey and assess the state of the field, to
provide a resource for those interested in the relations between the history and pedagogy
of mathematics, and to suggest lines of future research activity. In this volume, Barbin
5
(2000c) identified a need to “collect questionnaires and interviews of teachers and pupils
about mathematics” (p. 90).
Purpose of the Study
The purpose of this study is to discern if students learn mathematics by studying
its history. In particular, it investigates the changes in pre-service secondary school
teachers’ thinking about functions resulting from their studying the history of the
concept. That four of six participants of this study are mathematics students who are pre-
service secondary school teachers is important in light of the directives from the MAA
and NCTM that these students study the history of mathematics. A secondary purpose is
to evaluate their current understanding of functions. This study addresses the following
questions.
• What is the function conception of junior and senior level pre-service high
school teachers?
• Does studying the history of the concept of function deepen a student’s
understanding of the concept in any way and if so, in what way? In particular,
does studying the history facilitate his or her move from an action level
understanding to a process level understanding as described by APOS theory
(Action, Process, Object, and Schema)?
• In what ways can studying the history of a mathematical concept be used to
deepen a student’s understanding of the concept?

6
• Does a student’s studying the history of the concept of function facilitate his
or her move from a process level understanding to an object level
understanding?
Theoretical Basis for the Study: APOS Theory
The purpose of this section is to explain APOS Theory, a constructivist approach
to the learning and understanding of mathematics at the post-secondary level. The section
begins with an exposition of its principle components, including comparisons to von
Glasersfeld’s radical constructivism and Piaget’s work. APOS theory and constructivism
are theories that are simultaneously about knowing and coming to know, that is,
knowledge and learning. A look at instructional implications of the theory concludes the
section.
The developers of APOS theory wanted to use the idea of theoretical cognitive
structures from Piaget and relate them to observable behaviors in college-level students
(Asiala, Brown, et al., 1997). They created a model for conducting research in
mathematics education. APOS is the cognitive aspect of the model. It guides the
theoretical analysis of a student’s understanding of a mathematical concept. This analysis
suggests the design of instruction, which, in turn, provides an opportunity for gathering
data and reconsidering the initial theoretical analysis. APOS is an attempt to model the
epistemology of the concept.
The acronym APOS stands for Action, Process, Object, and Schema—mental
constructions made by students in their attempts to understand mathematics. Actions lead
to processes, which must come before seeing a concept as an object. The origins of this
7
interpretation of mathematical understanding lie in Piaget’s work and parallel many
concepts in von Glasersfeld’s radical constructivism. The developers of APOS see their
work as “the result of reconstruction of our understanding of Piaget’s theory leading to
extension in its applicability to post-secondary mathematics” (Asiala, Brown, et al., 1997,
p. 41). To them, the following best describes what it means to learn and know something
in mathematics:
An individual’s mathematical knowledge is her or his tendency to respond to
perceived mathematical problem situations by reflecting on problems and their
solutions in a social context and by constructing or reconstructing mathematical
actions, processes and objects and organizing these in schemas to use in dealing
with the situation. (p. 40)
They assumed that what a person knows and is capable of doing is not necessarily
available to him or her in any given moment in any situation. As did Piaget, therefore,
they see reflective abstraction as crucial to a student’s construction of mathematical
process and objects. Reflection involves paying conscious attention to operations
performed. Such reflection is significantly enhanced in a social context (Asiala, Brown,
et al., 1997). APOS theorists refer to the literature supporting the importance of a
student’s social interaction while learning and also to a research mathematician’s need for
interaction with colleagues before, during, and after doing creative work in mathematics.
The Action Construct
The Action construction is similar to Piaget’s action schemes. A student who has
an action understanding of functions sees an algebraic expression as a command to

8
calculate. Such a student can carry out a transformation only by reacting to external cues
(textbook directions, teacher suggestion, etc.) that give exact details on what to do. This
conception is like a recipe and they must apply it to some number before it will produce
anything. They do not necessarily see the recipe as an object in itself, that is, a result of
its own application (Thompson, 1994). Though it is considered the lowest level of
abstraction, it is a necessary beginning to the understanding of functions. The reason so
many students have trouble understanding piece-wise functions, composition, and
inverses of functions and sets of functions, for example, is that the learner is not able to
go beyond an action understanding of functions (Asiala, Brown, et al., 1997).
Interestingly, Breidenbach, Dubinsky, Hawks, and Nicols (1992) found students
who did not display any aspect of a function concept, not even an action conception. The
meaning these students gave to the term function was not useful to them as they
attempted to deal with activities related to mathematical functions. The researchers
categorized such students as having a pre-function conception.
The Process Construct
A process is an internal construction that performs the same transformation as the
action, but it is internal and hence under the control of the individual. She no longer
needs the external stimuli, no longer needs to actually evaluate an expression to think of
its result. She can reflect on, describe, or reverse the steps of a transformation without
actually performing those steps. A good example is an understanding of the function cos
x. Since no explicit recipe exists for evaluating this function at a given value, one needs
to imagine the process of associating a real number with its cosine. With this
9
understanding, a student can link one or more processes to construct a composition, or
reverse the process to obtain inverses of functions.
When a student moves from an action understanding to a process one, APOS
theorists say the student has interiorized an action to form a process. This process relates
to von Glasersfeld’s internalization of a topic, which, according to Battista (1999b, p. 3),
is “the process that results in the ability to re-present a sensory item without relevant
sensory signals being available.” It is at this level, according to von Glasersfeld, that a
concept has been formed. To be considered a concept, these constructs must be stable
enough to be re-presented without requiring perceptual input (von Glasersfeld, 1995).
Achieving a process conception of function is non-trivial for students (Thompson,
1994) and research has shown that many students do not achieve this level without
specific instruction geared specifically to that end (Thompson). An example of such
instruction is having students write named processes in a programming language called
ISETL. These processes serve as functions. Students then can use the name of the
function to direct its application to individual numbers or a set of numbers. Note here that
the students are using the NAME of the function rather than its defining process while
working with it.
Once students have practice working with processes, groundwork is laid for them
to begin thinking about sets of inputs in relation to sets of corresponding outputs. APOS
theorists say students are then ready to begin to reason more formally about functions—
they encapsulate the process to form an object. A process understanding of function

10
corresponds to Piaget’s “operations” whereas the next level—object construct—
corresponds to Piaget’s objects.
The Object Construct
An object understanding of a concept sees it as “something to which actions and
processes may be applied” (Selden & Selden, 1992, p. 19). One indication that a student
is functioning at the object level in understanding functions is her ability to reason about
operations on sets of functions (Thompson, 1994). It is often necessary, however, to de-
encapsulate objects back into processes. For example, one can think about adding or
multiplying functions or forming sets of functions, but to actually find these sums or
products of sets requires the student to de-encapsulate them back to the processes from
which they come.
Encapsulating processes into objects corresponds to Steffe’s and Cobb’s first level
of interiorization. Interiorization is “the most general form of abstraction; it leads to the
isolation of structure (form), pattern (coordination) and operations (actions) from
experiential things and activities” (Battista, 1999a, p. 3). Asiala, Brown, et al. (1997)
claimed that reaching this level of abstraction is incredibly difficult for students and has
found few pedagogical strategies to be effective in achieving this end. They claimed the
reason for this difficulty is the fact that there is very little in our experience that
corresponds to performing actions on processes.
The Schema Construct
The schema construct is the highest level of abstraction and closely mirrors
Piaget’s schemata construct. This idea is the least studied of all constructs in APOS
11
theory. It can be thought of as a coherent collection of processes and objects that are
organized in a structured manner. Schemas themselves can be organized into higher-level
schemas. When this happens, APOS theorists say the schema, itself, has been thematized
to an object. Asiala et al. (1997) suggested that the difference between schema and other
constructs can be understood by comparing it to the distinction between an organ and a
cell in biology. The organ (schema) provides the organization required for the cells
(objects) to function for the benefit of the organ (schema). It is the totality of knowledge
about a particular mathematical concept for an individual. A student will have a function
schema, an integral schema, a field schema, and so forth.
Instructional Implications
In attempting to find pedagogical strategies that encourage students to move from
one level of abstraction to the next, APOS theorists have developed a particular theory of
instruction. They claim that students do NOT learn course material in a logical,
organized, and sequential order—the way most textbooks are arranged. Though students
need to experience the levels of abstraction in the order previously described, they do not
necessarily understand the content of the material in such a neat, organized fashion.
Students gain partial knowledge, repeatedly return to the same knowledge, periodically
summarize and tie related ideas together (Asiala, Brown, et al., 1997). Students will often
move from one level of abstraction back to a lower level, depending upon the demands of
a given problem. Also, a student cannot always recall every piece of knowledge in her
repertoire. For a student to forget something on an exam, only to remember it shortly
after she hands in her paper or after only a slight suggestion from her instructor, is a
12
common occurrence. The instructional approach developed from APOS theory takes into
account this learning pattern by using what Asiala, Brown, et al. called a holistic spray.
In this approach, teachers give students activities (not a lecture) that deliberately
cause disequilibrium in their thinking and expose them to as much about the topic as
possible. The manner is holistic, not sequential. Such perturbations cause students to
question their current constructs and force intellectual growth as they try to figure out
“what’s going on.” Different students may pick up different pieces and parts of the
picture; hence the benefits of collaborative work become apparent. Each student shares
her partial understandings with the group, thus adding to other group members’
understandings. As a result, more pieces and parts are available to each student than that
provided by individual study. This approach is certainly consistent with constructivist
viewpoint that knowledge is obtained by constructing and reconstructing understandings
in a social setting.
Students are organized into permanent (i.e., for the semester) cooperative groups
of three or four students and do all course work in these groups. Research has shown that
students are more likely to reflect on their procedures when working in a cooperative
group (Asiala, Brown, et al., 1997) and numerous research studies have shown the
benefits of collaborative learning (Springer, Stanne, & Donovan, 1998). Instructors
utilize the ACE Teaching Cycle which has the following components:
• Informal activities (NOT a lecture) introduce a concept, frequently through
the investigation of examples. The purpose of these activities is to provide
students with an experience base, not lead them to the correct answer.
13
Students gain experience with a concept that will be developed later in class.
This is a good example of an inquiry, as opposed to discovery approach to
learning.
• Classroom tasks and discussions then pull together the main ideas developed
in the activities. The instructor leads discussions that encourage students to
reflect on the activities mentioned previously. In some instances, the instructor
may provide definitions, give explanations, and summarize.
• Students then work on exercises, which include traditional homework
problems that require application or extensions of the concept and thus
reinforce the concept being studied.
Asiala, Brown, et al. (1997) employed a particular strategy in employing the ACE model.
They required students to write and interpret code in the mathematical computer
programming language, ISETL (Interactive Set Language). In the current study, students
read and analyzed primary sources in the historical development of the concept of
function.
Conclusion
APOS theory is a theory of knowing mathematics that has its roots in
constructivism. It was developed as an extension of Piaget’s work so researchers could
understand the nature of learning in college mathematics students. It firmly holds to the
basic tenets of constructivism: (a) knowledge is not passively received, but built up by
the cognizing subject; and (b) the function of cognition is adaptive and serves to organize
the experiential world, NOT the discovery of an ontological reality (Von Glasersfeld,
14
1995). Knowledge is obtained by reflecting on problems and by constructing and
reconstructing actions, processes and objects in a social setting. It has specific
implications for instructional strategies that are based on inquiry learning.
Significance of the Study
The professional mathematical societies are deeply concerned about the
mathematical education of our teachers and are continuing to search for effective means
to deepen students’ understanding of fundamental mathematical concepts (Conference
Board of Mathematical Sciences, 2001). National reports call for better preparation of our
mathematics teachers (RAND Mathematics Study Panel, 2003; U.S. Department of
Education, 2000). For the past several years, The Mathematical Association of America
(MAA) has participated in a National Science Foundation-funded initiative entitled
Preparing Mathematicians to Educate Teachers (PMET), the purpose of which is to
strengthen the education of American teachers by sponsoring workshops, collaborations,
regional networks, resources, and mini-grants for mathematics faculty.
The concept of function also takes center stage when it comes to mathematics
education. Guershon Harel and Ed Dubinsky argued (1992) that
The concept of function is the single most important concept from kindergarten to
graduate school and is critical throughout the full range of education. Arithmetic
in early grades, algebra in middle and high school, and transformational geometry
in high school are all coming to be based on the idea of function. (p. vii)
The National Council of Teachers of Mathematics (2000) emphasized functions in their
algebra recommendations for grades 9-12. College textbook authors frequently take a
15
functions approach to the study of entry-level algebra. The concept is ubiquitous in
mathematics and mathematics education.
Though some researchers have obtained positive results for student construction
of a process conception of function (Breidenbach et al., 1992), others have not (Sfard,
1992) and still others note the continued difficulty students have with the concept
(Breidenbach et al., 1992; Carlson, 1998; Even, 1993; Norman, 1992; Sierpinska, 1992;
Wilson, 1994). The question of the value of learning history of mathematics remains.
School curriculum designers, as well as university mathematicians designing programs
for pre-service mathematics teachers, will be interested in the results of this study.
Historians of mathematics will also take note as their interest in history as a pedagogical
tool is either validated or not. Similarly, professional societies currently reaping praise on
the benefits of studying history shall benefit from the results.
Definition of Terms
Abstraction. Abstraction is the process by which the mind selects, coordinates,
combines, and registers in memory a collection of mental items or acts that appear in the
attentional field (Battista, 1999b, p. 2).
Action. An action is a transformation of objects which is perceived by the
individual as being at least somewhat external (Asiala, Brown, et al., 1997, p. 42).
Continuous Function. A function, f, is continuous at a number a if
lim f ( x) = f ( a ) . Geometrically, one can think of a continuous function as a function

x→a
whose graph has no breaks in it.

16
Formula. A formula is an algebraically expressed relationship between varying
quantities.
Function. A function, f, is a process which takes any element, a, of a set A and
transforms it into a unique element f(a) of a set B. The set A is called the domain and the
set B is called the co-domain of f.
Graph. A graph is a drawing that exhibits a relationship between two sets of
varying quantities.
Internalization. Internalization is the process that results in the ability to re-
present a sensory item without relevant sensory signals being available (Battista, 1999b,
p. 3).
Interiorization. Interiorization is the most general form of abstraction; it leads to
the isolation of structure (form), pattern (coordination). and operations (actions) from
experiential things and activities (Battista, 1999a, p. 3).
Object. An object understanding of a concept sees it as something to which
actions and processes may be applied (Selden & Selden, 1992, p. 19).
One-to-one function. A function, f, is one-to-one if it never pairs two different
elements of the co-domain with the same element in the domain, that is,
f(a) = f(b) implies a = b for all a,b ∈ A.
Prefunction. A pre-function conception is a lack of any mathematical
understanding of the concept; a meaning which was not at all useful while performing
tasks needed when dealing with mathematical activities related to functions.

17
Process. A process is an internal construction that performs the same
transformation as an action, but it is internal and hence under the control of the
individual.
Reflection. The conscious process of mentally replaying experiences, actions, or
mental processes and considering their results or how they are composed (Battista,
1999b, p. 4).
Summary of Chapter 1
This chapter presented the background, statement of the problem, purpose of the
study, research questions to be addressed, and a discussion of the theoretical basis for the
study. Also included was the study’s significance and definitions of terms. The next
chapter takes a closer look at the research related to this study.

CHAPTER II
REVIEW OF RELATED LITERATURE
This chapter consists of four parts. The first describes four theoretical frameworks
for analyzing the link between the history of mathematics and the learning of
mathematics. The second reviews studies concerning the effectiveness of using the
history of mathematics as a tool to learn mathematics; the third reviews the literature
concerning students’ understanding of the concept of function. The chapter ends with a
look at the history of the concept of function.
Four Theoretical Frameworks: History and Learning
This section discusses theories concerning the relationship between student
learning and the historical development of a mathematical topic. The principal theorists
in this area are Jean Piaget and Rolando Garcia (1989), Luis Radford (1997), Anna
Sierpinska (1992), and Anna Sfard (1991, 1992),
Piaget and Garcia
Jean Piaget was a well-known cognitive psychologist and Rolando Garcia a
physicist. The objective of their combined work (1989) was to develop a synthesis that
can explain the evolution of knowledge, both at the level of the individual and that of
social evolution. The authors claimed that the mechanisms mediating the transition from
one historical period to the next are analogous to those mediating the transition from one
psychogenetic (cognitive) stage to the next.
18
19
The mechanisms they describe are of two types and the first deals with their
constructivist view of learning. Reflective abstractions, those resulting from the subject’s
actions and operations; empirical abstractions, those drawing information from objects
themselves; and interaction between the subject and object are the principal mechanisms
of this type. A second type of mechanism is a “dialectical triad” which they described as
a transition from an intra phase of understanding to an inter phase to a trans phase. They
summarized this mechanism as follows:
The intra phase leads to the discovery of a set of properties in objects and events
finding only local and particular explanations. The “reasons” to be established can
thus be found only in the relations between objects, which means that they can be
found only in “transformations.” These, by their nature, are characteristic of the
inter level. Once discovered, these transformations require the establishment of
relations between each other, which leads to the construction of “structures,”
characteristics of the trans level. (pp. 273-274)
The authors noted that though inter-object analysis and trans-object analysis
operate together, conscious reflection in psychogenesis is originally oriented to the
outside rather than internally (Piaget & Garcia, 1989). The result is that correspondences
remain independent of transformations for a long time. Piaget and Garcia saw this pattern
historically as well and claimed that, at all levels, “the thematization of an operation
always occurs later than nonreflective use of the same operation” (p. 12). This analysis is
consistent with the APOS framework used in the present study, in which an action level
of understanding precedes the process level, which precedes the object level.
20
Piaget and Garcia (1989) dealt with concept development in four areas:
elementary concepts of physics—in particular that of motion, the understanding of space
as it relates to development in geometry, basic concepts of algebra, and those of
mechanics. Garcia presented the historical development of each topic in a separate
chapter, followed by Piaget’s analysis of the cognitive development in an individual.
Their general conclusion was that the mechanisms of intra-object analysis, inter-object
analysis, and trans-object analysis were found in all domains of cognitive development
and at all levels.
They found these mechanisms in the historical development as well. Algebra, for
example, for centuries was restricted to finding the solutions of individual equations. In
the 16th century, Francoise Viète facilitated a transition from analyses of these individual
equations dealing only with numbers, essentially an intra stage, to the inter stage by
noting general properties of numbers and developing a purely symbolic approach. In
doing so, he created a new discipline, a synthesis between the geometric analysis of
Pappus in the 4th century and the arithmetic methods of Diophantus. Even though the
concepts of transformations and invariants were not thematized at this time, they still
played a fundamental role, facilitating the further development of symbolic algebra.
According to Piaget and Garcia (1989), Galois and the development of group theory
represent the first thematized structure in the development of algebra.
Luis Radford
Radford (1997) believed that the history of mathematics offered a type of
“epistemological laboratory” (p. 66) in which to explore the development of

21
mathematical knowledge, though he claimed that several problems existed in doing such
research. He cautioned researchers against viewing history as a sequence of “clumsy
efforts” leading to our modern conceptualizations. For example, some historical
researchers criticize Diophantus’ Arithmetica (2nd century B.C.E.) because he is “just
looking for answers” to specific problems rather than general methods (Radford, p. 11).
According to Radford, Diophantus’ work becomes quite sophisticated when one
considers the socio-cultural era during which he lived. Another difficulty is that of being
objective while interpreting historical data. Radford claimed that
(a) the “real” meaning of a past concept is unattainable; it will always be
“filtered” by our framework and by our modern socio-cultural conceptions of the
history, and (b) given that any historical investigation puts in contact two different
horizons, and that the present horizon is always in movement, the history of any
concept or of any theory will always be rewritten. (p. 12)
Radford (1997) supported a socio-cultural view of the history of mathematics,
criticizing the work of Piaget and Garcia (1989) and the notion of “recapitulation.”
Ultimately, he saw today’s culture as radically different from that when such concepts
were developed, so assuming that “ontogenesis (the development of a subject’s idea)
recapitulates phylogenesis (the historical development of it)” is naïve (p. 13). Perhaps he
misunderstood Piaget’s and Garcia’s work, though, since they found evidence of the triad
mechanism across cultures. Radford concluded that “the configuration and the content of
mathematical knowledge is properly and intimately defined by the culture in which it
develops and in which it is subsumed” (p. 17) and saw the need for understanding both
22
the negotiations and the conceptions of the culture that underlie mathematical meanings.
He admitted that the process by which an ancient method was developed may indeed help
us understand old meanings and that one may then redesign these processes to be
compatible with modern curricula. Radford concluded that despite the problems inherent
in doing educational research in the history of mathematics, it holds great potential for
understanding mathematics.
Anna Sierpinska
Sierpinska (1992) set her work within a perspective of epistemological obstacles.
Epistemological obstacles are “inherent difficulties connected with complex concepts like
function” (Selden & Selden, 1992, p. 9). These obstacles are widespread, not unique to an
individual. They frequently appear in the historical evolution of a concept and provide
difficulties for today’s students as well. The goal of her research is to identify these
obstacles.
One example of such an obstacle is “thinking in terms of equations and unknowns
to be extracted from them” (Sierpinska, 1992, pp. 37-38). Not until the time of Euler did
discussions of analysis begin with a distinction between variable and constant quantities.
For centuries before, the main discrimination in algebra (referred to as “ordinary
Analysis”) was between the known and unknown. Similarly, Sierpinska found that 16-
year-old French students had difficulty making a shift from finding the unknown to
thinking in terms of variables on an exam concerning systems of linear equations. To
answer one question, they simply needed to find the unknown (i.e., the equilibrium
point), but to answer another, they needed to think in terms of the cost of copies
23
depending upon the number of copies. Most answered the first question with ease, but
very few students answered this latter question correctly.
Another example is “regarding the order of variables as irrelevant” (Sierpinska,
1992, p. 38). Hundreds of years passed before mathematicians perceived the order of
variables as important. The concept of function came into being in the context of analytic
geometry, where relationships between different line segments frequently played a role
for the curve (e.g., diameter, axis). The order of variables is insignificant in the analytic
representation of the conics, for example. Historians credit Descartes with discriminating
between the dependent and independent variables, though Sierpinska claimed that the
role of his coordinates was symmetric, not ordered. Even Newton was not clear on the
distinction. Sierpinska found a similar disregard in one of her students who viewed a
formula as a representation of the curve, that is, the curve was there first and the formula
simply described the curve, allowing one to calculate one coordinate when the other was
given. The student did not perceive the fact that the second coordinate was always
uniquely determined by the first, but not necessarily vice-versa.
One last example of Sierpinska’s epistemological obstacles is the belief that “only
relationships describable by analytic formulae are worthy of being given the name of
function” (Sierpinska, 1992, p. 46). She noted that the definitions of function given by
Bernoulli, Euler, and Lagrange in the early 18th century all describe a function as an
algebraic expression, that is, a formula. She claimed that during the 17th and 18th
centuries, these analytic tools for describing functions became more important than the
functional relationships themselves. Functions were their algebraic expressions. This

24
belief is also common among modern-day students (Breidenbach et al., 1992; Carlson,
1998). Sierpinska (1992) suggested giving students examples of functions that can be
described by two different formulas (e.g., one recursive and one explicit) to help them
overcome this obstacle. One approach would be asking students to write short computer
programs, as did Breidenbach et al. (1992). Their results were impressive.
Historically, the need to formulate general theorems on large classes of
relationships between variables provided the impetus for extending the notion of function
beyond those expressible analytically. In particular, the solution to the partial differential
“wave” equation (concerning the motion of a vibrating string) sparked a controversy
about the type of function that should be allowed in analysis. The mathematician
d’Alembert claimed that “there exists an infinity of curves different from the elongated
cycloid (sine curve) which satisfy the problem under consideration” (Siu, 1995, p. 110).
The only restrictions on the functions were that they be periodic, odd, and everywhere
twice differentiable. Another impetus for the extension of the function concept was the
theory of trigonometric series by Fourier and the search for conditions under which the
series converges. This study led Dirichlet to formulate his general definition of function
in 1837, a definition that does not require an analytic expression.
Sierpinska (1992) closed with several implications for teaching, including
motivation (students must be interested in explaining changes and finding regularities
among these changes), introductory contexts (functions should appear as tools in
modeling situations in either science or real life), developmental contexts (using and
constructing tables), prerequisites (a certain amount of algebraic awareness is necessary),

25
representations (use multiple representations), and definitions (informal definitions are
sufficient at the introductory level). She did not, however, provide empirical evidence for
the effectiveness of her suggestions.
Anna Sfard
Anna Sfard (1991, 1992) conjectured that students’ learning of a mathematical
concept frequently parallels the historical development of the topic by mathematicians.
As does Sierpinska, she claimed that an “awareness of the long and painful processes
preceding the birth of a mathematical object may be the key to understanding some of the
difficulties experienced by so many learners” (p. 59).
Sfard’s theory, much like the theoretical framework for the present study (APOS
theory) maintains that one can conceive of abstract mathematical notions such as number
or function in two fundamentally different ways: structurally—as object-like entities, and
operationally—as computational processes (1991, p. 1). Sfard’s thesis (1992) is that
“structural conceptions develop usually out of operational” (p. 61); that is, that abstract
objects emerge from computational processes, whether one is concerned with the
historical development of a topic or its acquisition by a student.
Sfard exemplified her theory with the concept of function. She claimed that
“function’s turbulent biography can be viewed as a three-centuries long struggle for
reification” (1992, p. 62). She commented that the idea of function was the result of a
long search for mathematical models for physical phenomena and found it significant that
it flourished shortly after the development of algebraic symbolism. “No wonder then, that
the notion of function was initially connected to algebraic processes” (Sfard, 1991, p. 14).
26
She compared the concept of function and its relation to algebraic manipulation to that of
negative number for that of subtraction: “something between the product and process
itself” (p. 14).
Early definitions (Euler’s and Bernoulli’s) of function depended on the fuzzy
notion of variable. Later definitions (d’Alembert’s and Euler’s) excluded this idea and
became mainly operational in nature: “a quantity” is a function only if it depends on
another quantity “in such a way that if the latter is changed, the former undergoes change
itself” (Sfard, 1991, p. 15). Sfard (1992) claimed that the difficulty mathematicians had
with the concept of variable was due to their inability to reify it as an object:
None of them (descriptions of variable) captured the underlying process of change
and converted it into a legitimate object. In the absence of a structural conception,
there was no reasonable link between the new idea and the existing systems of
mathematical notions. (1992, p. 63)
She noted that the mathematician Frege de-legitimized the use of the term “variable,”
claiming that its use has no justification in analysis.
The graphical representation of functions did not facilitate reification either,
according to Sfard (1992). Each time an algebraic formula was modified to fit a graph, a
counterexample was found showing that a gap still existed between the two, and vice
versa. Over time, the numerous failed attempts at translating from operational to a
structural point of view led Dirichlet to his “arbitrary correspondence” idea and
eventually to Bourbaki’s definition:

27
This simple description presented function as a set of ordered pairs and made no
reference whatsoever to any kind of computational process. Bourbaki’s group
solved the time-revered problem by eliminating the “unreificable” notion of
variable and substituting it with purely structural set-theoretic concepts. Not
surprisingly, this new definition, which had very little in common with its
intuitive operational origin, evoked much criticism when first proposed. But when
at long, long last function—initially only a computational process—was
converted into a mathematical object, our schema of concept development could
repeat itself once more: on the new objects new operations could now be
performed. (Sfard, 1991, pp. 15-16)
Sfard (1991) used the above scheme, that is, the precedence of operational
thinking to structural thinking, when describing the learning process in today’s students.
She noted that each stage is a prerequisite for the next, and that one may work at more
than one level when solving a complex problem. This hierarchical aspect of her theory is
similar to that of APOS theory (Asiala, Brown, et al., 1997). In the case of function,
reification may be evidenced by the ability to solve equations in which functions are the
unknowns (e.g., differential equations), the ability to talk about processes (e.g.,
composition) performed on functions, and the recognition that sets of ordered pairs need
not be computable in order to be a function.
Sfard’s (1992) research supported her theoretical conjectures. She studied the
understandings of students who were enrolled in a traditional class where the structural
nature of functions was emphasized. She found that even in this situation, most students’
28
conception of function was operational and concluded that the operational approach is
intuitive to students. Another interesting observation was that students in the traditional
classroom setting will often develop pseudo-structural conceptions, such as identifying a
function with its representation. Her findings were not unlike those found by other
researchers studying students’ conceptions of functions: their difficulties with split
domains; their lack of understanding of the connections between representations of
functions; students giving independent solutions to basically equivalent problems when
only superficial changes in notation are introduced; and their general neglect of domain
and range. She concluded, “pseudostructural conceptions can hardly be avoided within
the usual structural way of teaching” (1992, p. 77).
She experimented with teaching functions operationally to students 22-25 years
old in a course on algorithms and computability at the Center for Pre-Academic Studies
of the Hebrew University. Her experiment had mixed results. The course was devoted to
the idea of algorithm; the concept of function was introduced as a means of dealing with
the semantics of algorithmic languages. Initially, the term function was used almost
synonymously with algorithm, later explained as being a name for the product of an
algorithm. She used many different devices to represent functions: tables, set notation,
analytic formulas, computer programs, graphs, and diagrams of different kinds.
Her findings (Sfard, 1992) were as follows.
• Students tended to identify functions with computer programs, even when
encouraged otherwise. Since they were quite capable of solving non-routine

29
problems, she interpreted this phenomenon as an indication of an operational
conception of function rather than pseudo-structural.
• Students were not able to deal formally and generally with operations on
functions.
• Students were bewildered by the idea of non-computable functions.
Though she believed that structural thinking can be externally stimulated to some degree
and quoted Dubinsky’s work with ISETL in this regard, she admitted that her own
attempts were not fully successful. Students did make progress toward that goal,
however, and students’ performance on exams was impressive. She blamed the “inherent
difficulty of reification” for her mixed results and also noted the necessity of long periods
of incubation for true understanding, as well as the need for learner determination,
stamina, and intellectual discipline.
Summary
This section looked at four theoretical perspectives for the link between history
and learning. The similarities between them are noteworthy. Piaget’s and Garcia’s claim
that the “thematization of an operation always occurs later than nonreflective use of the
same operation” (1989, p. 12) is similar to Sfard’s claim that the “structural conceptions
develop usually out of operational” (1992, p. 61). Both groups found these patterns in the
historical development of mathematical concepts and in their subjects’ cognitive
development of the same. Noteworthy also is that the theoretical framework for the
current study, APOS theory, is similar in its basic tenets to both these claims and though
not based on history, is based on the work of Piaget and widely used in research
30
documenting college students’ learning of mathematical concepts. Like Sfard, Sierpinska
(1992) found parallels between students’ difficulties learning a concept and the
difficulties in its historical development. Radford (1997) concluded that despite the
problems inherent in doing educational research in the history of mathematics, it holds
great potential for understanding mathematics. All four valued the endeavor.
Studies Using History of Mathematics as a Tool
This section reviews research studies concerning the use of history as a tool for
learning. Note that many of the studies document changes in attitude, not learning, and
that others depend solely on self-reporting of learning rather than upon objective
empirical results.
Barbin (2000b) gave an overview of nine case studies written by teachers of work
in their own classroom. Attempting to answer the question “does it work?”, the teachers
used qualitative analysis (ethnography) of the changes that occur when teachers use
history in the teaching of mathematics. Of the nine case studies, six used history
explicitly—one used problems from the history of mathematics and five used readings
and historical text; the other three used history implicitly. Three used history to bring
about change in way students view mathematics; five attempted to improve learning of
mathematics; and one used history as a means of facilitating mental construction of
mathematical concepts.
Results reported were from teacher introspection and most dealt with changes in
views of mathematics, rather than mathematics learning. One teacher group wrote:
31
The confrontation with mathematical texts changes the view of mathematics for
both teacher and student. Mathematics becomes alive, it is no longer a rigid
object. It is the object of enquiry, controversy, contains mistakes and uses
methods of trial and error. (p. 67)
Another teacher claimed that “reading old texts excites the curiosity of the students and
encourages them to question” (p. 67). And yet another: “We are no longer dealing with a
finished product but with something in continuous evolution; it is no longer a case of
accepting a discipline of divine nature, but of understanding tools, methods and concepts”
(p. 67). A student wrote, “Mathematics has for me passed from the status of a dead
science to that of a living science” (p. 67).
Barbin (2000b) claimed that history encouraged the teachers to see students as
thinking and inquiring beings and therefore look anew at student work. One of the
teachers came to a realization of the time necessary to construct a topic. As a result of
working with the history of mathematics, she has allowed her students the time they need
to construct ideas slowly and has learned to identify moments of misunderstanding. As a
result, she claimed that students are reassured and many have gained confidence in their
ability to do mathematics.
Though these results are certainly admirable and desirable, none document
changes in the actual learning of mathematics. One might argue that such attitudes result
in better learning, yet these studies fail to make a direct connection between studying the
history of mathematics and mathematics learning. Note also the dependence upon teacher
self-reporting.
32
Furinghetti (1997, 2002) reported four case studies on different uses of the history
of mathematics in teaching. The most significant dealt with using history as a different
approach to concepts at the secondary school level. The teacher focused on a topic that is
traditionally quite difficult for calculus students: the link between the derivative and
integral. An initial questionnaire given to the students of these teachers indicated that
“derivative is one of the mathematical objects to which students connect manipulation of
formulas, but not mathematical meaning” (2002, Example section, par. 1). She offered
students a re-elaboration of Isaac Newton’s work on the topic. Her intent was to help
students recover the mathematical meaning by studying the derivative as tangent.
Believing that the “pioneering period at the beginning of calculus” reveals the roots of
mathematical entities, she had students look at the various constructions of tangent line to
a curve as a first step toward constructing the derivative. Her approach is fascinating, yet
she revealed no empirical evidence about students’ growth in understanding.
In another unit, Furinghetti focused on proof, in particular the methods of analysis
and synthesis. She noted that both methods were used in mathematical research and in
teaching, commenting that the “double method of analysis and synthesis, which is good
for making discoveries, with the addition of more reasons can be employed to explain the
discovery” (1997, p. 58). She presented the method through the works of Marin Ghetaldi,
1630 and Descartes’ Discours sur la methode. The teacher maintained that presenting
proofs through analysis and synthesis is appropriate since mathematicians discovered
proofs in this manner. Textbooks, however, show only one way, leading students to
wonder how a mathematician could have ever thought of such an “artificial” or cunning
33
trick. “In this case one of the values of a historical presentation is its capacity to make the
rules explicit” (Furinghetti, 1997, p. 58). She observed that students presented with these
historical works
become free and easy in applying it to different situations, even when they are not
required by the teacher to do this. We stress that the use of history in this case
allowed the teacher to satisfy the students’ need for explicitation, which is
frustrated in the usual teaching. (p. 58)
Po-Hung Liu (2002) conducted a study whose purpose was to investigate the
relationships between a problem-based, historical approach to teaching calculus and
college students’ views of mathematical thinking. At their first class meeting of the
semester, 44 Taiwanese engineering-major college students completed a six-item
questionnaire designed to examine their views about mathematical thinking. They also
completed a mathematical biography that served as supplementary data. Nine randomly
selected students then participated in semi-structured individual interviews to validate the
written data and glean more detailed information.
During the course of the semester, the teaching of calculus followed the historical
order of its development and students learned concepts by attacking a variety of
problems. Historical problems, quite different than the usual exercises, required that
students think deeply about the concepts they were learning. Near the end of the
semester, students completed the same questionnaire and several participated in
individual interviews.
34
While responding to the question at the beginning of the semester, “In your
understanding, what is mathematical thinking?” 20 students (45%) associated
mathematical thinking with ways of deriving answers or solving problems, mainly by
following pre-determined procedures. To them “pondering on mathematics” meant
recalling and applying formulas. Twelve students (27%) viewed mathematics as a process
of logical thinking or reasoning. Even those who professed this view admitted in
interviews that they had not experienced the benefits therein.
After the 18-week historical problem-based approach, students’ views changed
somewhat. Though they still viewed mathematics as a procedure for deriving answers,
they showed an inclination to value individual creativity while solving problems and also
appreciated the need for involving concepts from other disciplines. The researcher
attributed this change to the ancient mathematicians’ imaginative approaches discussed in
class. After learning about various errors made in the development of calculus, students
were less likely to believe that mathematics is a static discipline. They also appreciated
the need for justification and logical sense. According to the researcher, “Participants’
focus seemingly shifted from mathematics as a product to mathematics as a process”
(Liu, 2002, Summary and Discussion section, par. 2). Although these results are certainly
laudable and give credence to the benefits of an historical approach, the emphasis was on
students’ views of mathematics not on students’ learning of mathematics.
Bueno and Lins (2002) claimed that a good reason for a pre-service teacher to
study the history of mathematics is to enable the understanding of the process of meaning
production for mathematics that would “allow her/him a much finer reading of the
35
learning processes in the classroom” (Abstract). They also see the history as facilitating
an understanding of the possibility of different meanings produced for the same
mathematical object. Their course began with a primary source to texts which discussed
style, interpretation, and different presentations, in the hopes of helping students develop
an awareness of the processes involved in meaning production.
Their study used a theoretical model which focuses on students’ “knowledge” and
“meaning.” “Meaning” is characterized as what a person actually says about an object in
a given situation whereas “knowledge” is characterized as a statement that a student
believes in, together with a justification for this belief (Bueno & Lins, 2002, On Meaning
Production section, par. 5). They briefly analyzed student understanding via their
language and written projects. They concluded that their experiment resulted in students’
participating reflectively in processes of meaning production and creating a first link
between interpreting a historical text and studying present-day mathematics.
Boero, Pedemonte, and Robotti (1997) conjectured that the introduction of
‘voices’ from the history of mathematics and science may mediate some important
elements of theoretical scientific knowledge. Their research goal was to develop a
theoretical framework designed to organize and analyze early student approaches to
theoretical knowledge. In deciding which aspects of theoretical knowledge to include in
designing curriculum materials, they agreed that cultural meaning and student motivation
are crucial criteria (p. 2-82) and that “priority should be given to leaps forward in the
cultural history of mankind, even if . . . these are the most difficult areas for school
study” (p. 2-82).

36
The researchers (Boero et al., 1997) designed a “voices and echoes” game and
experimented with five eighth grade classes. They described their game as follows:
Some verbal and non-verbal expressions . . . represent in a dense and
communicative way important leaps in the evolution of mathematics and science.
Each of these expressions conveys content, an organization of the discourse and
the cultural horizon of the historical leap. . . . we call these expressions voices.
Performing suitable tasks proposed by the teacher, the student may try to make
connections between the voice and his/her own conceptions, experiences . . ., and
produce an ‘echo,’ i.e., a link with the voice made explicit through a discourse.
The ‘echo’ is an original idea, intended to develop our new educational
methodology. What henceforth will be called the ‘voices and echoes’ game is a
particular educational situation aimed at activating the production of echoes by
students. (p. 2-84)
The researchers (Boero et al., 1997) hoped that by comparing a text with another
text, or with some data from everyday experience and finding congruencies or
contradictions, students would enhance the transition of their thought to a theoretical
level. They believed that the game should begin with historical voices that give a
theoretical representation close to students’ intuitions and interpretations and then
eventually overturn this theory. They performed a teaching experiment concerning the
idea of falling bodies. In the teaching experiment, teachers paraphrased and explained
new terms and information concerning the voice. Following each voice, teachers asked
students to produce echoes and then frequently discussed the echoes together as a class.
37
Five eighth grade classes of different levels and different environments were involved in
the experiment, which lasted 12 to 16 hours. Researchers collected recordings of
classroom discussions and individual student texts. The researchers noted that the
resultant learning was “better and more extensive” than those usually achieved when
eighth graders approach theoretical knowledge (p. 2-85), though they did not indicate the
instrument used to draw these conclusions, nor did they specify what they meant by
“better and more extensive.” Among the positive outcomes they noted in all classes were
the following:
• students garnered the contents, methods, and ways of organizing discourse
presented to them through the voices;
• they participated in high quality scientific debate: “ancient scientific debate
was revived and related to the present cultural and expressive horizon” (p. 2-
86).
Garuti (1997) found similar results with the ‘voices and echoes’ game, though again the
specific criteria used for drawing such conclusions, however general, was unavailable.
The purpose of the current study closely resembles the purposes of Boero et al.
(1997) and Garuti (1997) since these researchers attempted to scrutinize student thinking.
Their results are encouraging.
Arcavi (1985) developed worksheets for pre-and in-service middle school
teachers concerning the topics of negative and irrational numbers and linear and quadratic
equations. The worksheets focused on conceptual development, as the author tried to
“present the search for a formal mathematical definition of a particular concept and, thus,
38
to give a proper picture of mathematical activity, as well as to motivate the necessity for
the definition itself” (p. 16). His worksheets were based on original sources and followed
by questions, usually of one of the four following types:
• elementary understanding (One example of this type is the “dictionary,”
where students provide the modern forms [or names] for those that appear in
the original sources.)
• development and transfer
1. mathematics exercises in the style of the original text
2. completion of missing steps
• criticism, analysis, and synthesis
1. comparisons between different approaches within an original source or in
other sources
2. argument and refutation
3. summary and synthesis
The objectives of his worksheets were:
• to improve and enrich teachers’ mathematical knowledge about topics they
were/would be teaching;
• to enrich teachers’ didactics, allowing the opportunity for discussion of the
demands of didactics as opposed to the demands of pure mathematics and
comparisons with present classroom practice;
• to contribute to the creation of a reasonable image of mathematics and
mathematical activity; and

39
• to create awareness of, and positive attitude towards, the history of topics in
the curriculum. (Arcavi, 1985, p. 43)
Arcavi (1985) was one of the few to attempt to determine the learning that
occurred as a result of studying the history of a concept. Note that his work is topic
specific, that is, the pre-service teachers studied, in depth, the development of a single
topic rather than studying the history of mathematics in a general, chronological
approach. The teachers involved in his study felt that they had profited from the materials
in terms of the objectives which guided the design and felt they learned the formal
definitions of negative and irrational numbers. They believed that the materials helped
“enhance mathematical knowledge” (p. 63). Note, however, that this resultant learning of
the mathematical topic was self-reported by the teachers. Though one would like to hope
that teachers have an accurate understanding of their own conceptions, objective data
would be more convincing. Arcavi did compare the mathematical reading ability of in-
service and pre-service teachers, however, before and after completing his worksheets
and found that the materials aided in developing competence in reading mathematical
text. He also provided a short pre- and post-test on which his participants were asked to
identify irrational numbers. He found that the materials “contribute to the recognition of
irrational numbers” (p. 73) and they motivate and conceptually prepare students for the
introduction of the formal definition of negative numbers.
Van Gulik (2005) studied the value and the applicability of the history of
geometry in modern education. Sixteen to 18-year-old high school students learned about
proofs in plane geometry by working in groups on the history of non-Euclidean

40
geometry. Van Gulik collected data through questionnaires preceding and following the
series of lessons, through lesson observations, and through discussions with students and
teachers. She concluded that her participants gained a deeper insight into the origin of
geometry, the teachers found the subject challenging and inspiring, and “the ‘reinvention’
of the basic assumptions of geometry results in a more lively learning process and better
motivation” (personal communication, February 23, 2006.)
Jahnke (2000) reported about a teaching experiment on conic sections. A
secondary school teacher in Italy created 34 worksheets containing passages from a
French text and worked with 16 students aged 16 or 17 who volunteered to participate
after the regular school day. After each session he gave students a questionnaire to
evaluate their understanding of mathematical content and to check difficulties. Students
also completed a questionnaire at the conclusion of the overall experience. Among other
benefits, the teacher noticed that working directly with the text “required more careful
reflections on problems and better understanding of their meanings” (p. 312).
Not all studies reported positive results. Fleener, Reeder, Young, and Reynolds
(2002) reported the effects of a mathematics education curriculum infused with historical
explorations over three semesters in three different courses. Their study investigated the
ways that such study might help pre-service elementary education students make sense of
their own mathematics. The purpose was to critique students’ thinking about mathematics
at three points in the curriculum:
• early in their studies, that is, in a mathematics content course that was part of
a general education requirement for elementary education, early childhood,

41
and special education majors. The Mathematics Department housed this
course which students usually take during the second semester of the
sophomore year.
• at the beginning of their methods coursework, that is, in the first mathematics
education course taken after admission to the teacher education program.
Students typically take this course in their junior year. The course focuses on
problem-centered, inquiry-based approaches to mathematics teaching.
• during the last semester of coursework before their graduate-level internship.
In the mathematics content course, students selected topics from the history of
mathematics to research and prepared a short presentation. Most worked in pairs and all
submitted a two-page report of their topic. Students chose mathematical topics rather than
historical figures.
In the first methods course, students chose to study either an historical figure or a
mathematical historical topic, then prepared a formal oral report and one-page handout on
their topic. Most students again worked in pairs. They also developed inquiry activities
for elementary students using historical topics or individuals. The researchers (Fleener et
al., 2002) claimed that the topics were of a broader range than those in the mathematics
course, that is, students chose biographies, mathematics fields of study, as well as topics
in the history of mathematics.
In the last course involved in the study, students chose a mathematician from a
given list, read about this mathematician’s contributions, then role-played him or her.
42
They were tested on the contributions of the mathematicians, and throughout the entire
semester, classroom activities integrated historical topics.
At the end of the first course, students responded to the following questions on
their final exam: How has learning about the history of mathematics affected your
understanding of mathematics? How has preparing and listening to others’ final reports
of historical topics facilitated your understanding of mathematics content and
approaches to teaching mathematics? Students responded to similar questions on each of
the final exams in the other two courses.
The researchers used three categories of analysis: mathematics learning as defined
by the National Council of Teachers of Mathematics (2000); Habermas’ categories of
human interest from the point of view of learners of mathematics (technical knowing,
hermeneutical understanding, emancipatory participation); and Habermas’ categories of
interest from the point of view of future teachers. They performed a qualitative analysis
and two chi-square quantitative analyses—one to determine if there was a change in
responses from the first class to the last and the other to determine if the answers varied
from random responses. Researchers also performed a third chi-square test with collapsed
categories to determine if there was a preference for technical over other approaches to
understanding mathematics. This last Chi-square statistic was “extremely significant”
(Fleener et al., 2002, p. 77), suggesting a preference for responses in the technical
category. The researchers concluded that
These results suggest critical and historical approaches, even sustained over
several semesters of mathematics instruction, are not sufficient for students to

43
develop an emancipatory approach to their own learning and teaching of
mathematics. As suggested in much of the mathematics education literature,
students’ past experiences with mathematics and the emphasis, becoming even
more pervasive, on technical skill demonstration of competencies in K-12
mathematics, may prevent our programmatic efforts from having an impact on
students’ hermeneutical or emancipatory approaches to the study and teaching of
mathematics. (pp. 80-81)
Researchers expressed surprise at the resiliency of this technical view of
mathematics in the pre-service elementary school teachers. They expected students in the
later classes to “exhibit more hermeneutical or emancipatory ideas about mathematics”
(Fleener et al., 2002p. 80).
The Fleener et al. (2002) study was thorough, using both quantitative and
qualitative analysis. Note, however, that results depended upon analysis of student self-
reporting, rather than upon analysis of their work on mathematical tasks intended to
assess their understanding of a specific topic. Then too, the assignments (e.g., 2 page
reports) may not have been in-depth enough to reveal or require growth in understanding.
The researchers recognize that these assignments did not promote critical reflection (p.
82) and that a problem solving approach may have been more effective. They admitted
that their approach was still their “doing something TO” the students rather than enabling
students’ own drive to understand (p. 83).
One apparent critic of using history to teach mathematics is Freudenthal (1981),
who claimed that the most commonly heard argument for studying history is that
44
knowledge of the history of a subject helps in understanding the subject matter itself. He
does not believe that studying the history of mathematics helps in understanding
mathematics, however. Admitting that mathematics has a long and interesting history, he
questioned if the pre-stages of mathematics are worth remembering. “Whoever learns a
second modern language learns it in its present state doesn’t he?” Freudenthal asked.
“Maybe Hippocrates’ name will be dropped at least once in courses at medical schools,
but no examiner will expect a student to know whether anaesthesia was invented before
Christ or later” (p. 30). “Should one impose the constraint to memorise upon students
who have accepted to go the way of understanding? History, indeed, is again a thing that
is memorized, must be memorized” (p. 31).
One can readily conclude that Freudenthal thought that studying the history is a
matter of knowing dates and names, though he was really being critical of the way it was
taught. He concluded his article by stating that
The history of science [is] . . . integrated knowledge rather than items stored in
well-stocked drawers, each of them labeled and opened when the timetable
announces the history of the subject matter. . . . History is worth being studied at
the source rather than by reading and copying what others have read and copied
before. . . . Whoever is interested in the history of mathematics should study the
processes rather than the products of mathematical creativity. (p. 33)
In the present study, as in Arcavi’s (1985), the purpose is “to create a picture of
the development of a topic of relevance to the teacher, and by the way, into that picture,
enter mathematicians, dates, etc., that are subsidiary to the study of the mathematical
45
story” (Arcavi, Bruckheimer, & Ben-Zvi, 1982, p. 30). Names and dates are secondary in
a thematic approach that attempts to trace the crucial moments in a topics’ development,
that is, what caused the idea to move forward? What were some of the issues that
spawned the idea? Why was there controversy about it? What were the arguments pro
and con? How did a formal definition finally develop? What were the cultural and
intellectual settings that spawned the need for the concept? These questions get at the
meat of the topic, and can offer profound insight into the nature of the topic. Studying
history is the study of an evolution of an idea, not a study of names and dates.
As noted in the above studies, researchers are very interested in what history has
to offer to the educational process. Note that many studies document changes in attitude
not learning, however, and that others depend solely on self-reporting of learning rather
than upon objective empirical results. Van Gulik’s (2005) and Arcavi’s studies (1985;
Arcavi et al., 1982) were among the few that attempted to document student learning.
The purpose of the current study is to extend these efforts and analyze student learning
within an APOS theoretical framework.
Studies About Learning the Concept of Function
Much has been written about students’ understanding of the concept of function.
Many students think that a computational formula is a necessary condition for a function
(Breidenbach et al., 1992; Carlson, 1998; Even, 1993; Sierpinska, 1992; Wilson, 1994),
or that variables must be present to indicate input and output (Breidenbach et al., 1992).
Some students insist upon the presence of causality and many were not able to construct a
process in their minds in response to a situation (Breidenbach et al., 1992; Norman, 1992)
46
or to solve real world problems (Wilson, 1994). Breidenbach et al. (1992) noted
confusion between the requirement for being a function and the definition of a one-to-one
function. Prospective secondary school mathematics teachers in Even’s study (1993)
thought that all functions are one-to-one and onto, graphs of functions are “nice” (i.e.,
smooth and with no sharp corners), and tended to use the vertical line test as a rule for
students to follow, not necessarily with understanding. They did not consider constant
functions, those with split domains, or a function obtained by composition to be
functions. Anna Sfard (1992) noted other misconceptions and student difficulties with the
concept:
• thinking that a function is its representation;
• an inability to connect different representations of functions (graphical,
symbolic, tabular);
• giving independent solutions to basically equivalent problems when only
superficial changes in notation are introduced;
• a general neglect of domain and range;
• inability to deal formally and generally with operations on functions.
Marilyn Carlson (1998) noted others:
• believing all functions are continuous
• difficulty with function notation, that is, the role of the parenthesis in the
function representation
47
• not understanding that the “function value” refers to the y-value (assuming
conventional labeling of axes); not understanding what it means for one
quantity to be a function of another
• difficulty interpreting rate of change information from a situation
• difficulty demonstrating an awareness of the impact of change on one variable
has on the other; interpreting and graphically representing covariant aspects of
a real world situation.
Vinner (1992) studied 271 college freshmen starting calculus courses in different
science departments and noted that 56% of his subjects showed some degree of
compartmentalization concerning the concept of function. He defined
compartmentalization as the existence of incompatible pieces of knowledge in a student’s
mind without the student being aware of it. For example, though some of his subjects
defined a function as a correspondence between two sets, they claimed that a graph does
not represent a function because there was no rule to describe it.
Eisenberg (1992) noted that students have a strong tendency to think of functions
algebraically rather than visually and have difficulty connecting the graph of a function
with its analytic representation. He observed a common belief that visualization is not
central to the nature of mathematics.
Artigue’s (1992) students also exhibited this belief. Her study involved about 100
university students in a differential equations class. Her goal was to study the exact
nature of student difficulties in relating an algebraic representation to a graphic one. She
identified three “registers of interaction” (p. 114) between the algebraic setting and the
48
graphical one: the interpretation register (information is given simultaneously in the two
settings and the problem to be solved requires interaction between these two bits of
information); the prediction register (information is given in one setting only and the
problem to be solved requires a solution in the other setting); and the justification
register. Not surprisingly, she found the most difficulties in the register of justification.
She commented about the difficulty of justification but also noted that in traditional
teaching, the graphical setting is rarely, if ever, accepted as a setting for justifications.
She saw this as a pedagogical obstacle to understanding the qualitative nature of
differential equations. In the second year of the study, the graphic setting was emphasized
on the level of justification and students showed an obvious improvement (p. 128).
Other researchers note students’ tendency to trace back to each axis of a graph
rather than look at its global characteristics (A. Bell & Janvier, 1981; Monk, 1992). Monk
differentiated between a pointwise analysis of a graph and an across-time analysis. A
pointwise question asks for values of a function for a specific input value. An across-time
analysis involves asking students to describe a pattern of change in the value of a
function that results from a pattern of change in the input values. Monk (1992)
maintained that one source of difficulty students have with across-time analysis is their
incomplete understanding of relevant concepts. A. Bell and Janvier (1981) called the
same difficulty one of a “situational distractor” (p. 37). For example, a student may have
difficulty with the related concepts of speed, distance, and time. Given a graph of speed
versus time of three runners in a race, a student may be unable to completely integrate the
three. She can use the concepts for some purposes (e.g., identifying the winner by point
49
reading) but is unable to correctly differentiate change in position with change in speed.
Monk called this confusion a “blurred concept” of these ideas, “sometimes fused,
conflated, or exchanged” (p. 176). She exchanged the concept of change in position with
that of a change in speed.
An error mentioned frequently in the research is the tendency of students to treat a
graph as a literal picture of the problem situation (Bell & Janvier, 1981; Clement, 1989;
Monk, 1992; Sierpinska, 1992). Clement differentiated between a global correspondence
error and a local correspondence error (p. 83). In the latter, a visual feature of the
problem scene (e.g., the same location of the cars) is matched to a specific feature of the
graph (i.e., the point of intersection).
Goldenberg, Lewis, and O’Keefe (1992) also documented students’ difficulties
with graphs of functions. Students do not have a sound understanding of graphs of
functions because they do not really understand or even see the varying nature of the
variables. Such understanding is difficult, they claimed, due to the static nature of
traditional graphs. They developed software called DynaGraphs that enables students to
control the variation of the input values and see the effect of this change on the output
values. Rather than using the traditional xy coordinate system, they use two separate
horizontal axes. Though their initial intent was to gradually lead students to the
traditional system of graphing, the researchers found that students were experiencing a
rich understanding of rather sophisticated mathematical concepts using these dynamic
graphs. They decided that they had not exhausted the possibilities of these DynaGraphs
and also noted that some students were able to sketch the traditional graphs with ease
50
after analyzing the behavior of the function on its DynaGraph. Like Artigue (1992), these
researchers emphasized the importance of the graphical representation of functions.
Breidenbach et al. (1992) noted that college students, even those who have taken
a large number of mathematics courses, do not have a good understanding of the function
concept. They insisted that a good understanding includes having a process conception of
function, as defined by APOS theory. Since the current study analyzes students’
responses in terms of this theory, Table 1 illustrates a detailed summary of Breidenbach
et al.’s classifications.
While studying students’ understanding of derivative, Asiala, Cottrill, Dubinsky,
and Schwingendorf (1997) noted that a student’s “strong need to find or even construct
an expression in which to substitute suggests that they may not be very far advanced
beyond an action conception of function” (p. 16).
Breidenbach et al. (1992) worked with sophomore and junior mathematics majors
preparing to be high school, middle school, or elementary school mathematics teachers.
They compared student responses to the question “what is a function?” before and after
instruction. The students experienced a semester long instructional period in a computer
environment studying various topics of mathematics in the context of working with
ISETL, an interactive programming language whose syntax is very similar to standard
mathematical notation. They worked in cooperative groups. After instruction, students
were given a list of descriptions of situations and asked to decide if the situation could be
expressed by functions, and if so, to explain how. Researchers noted that of 59 students,
51
Table 1
Description of APOS Levels
Sample student responses

APOS Level Description of Level Evidence to “What is a function?”
Prefunction Student has little if any “I don’t know.

conception of function. “A mathematic equation with
variables.
“A mathematical statement
that describes something.
“A social gathering.”
Action “A repeatable mental or Emphasizes the act of “A function is something that

physical manipulation of substituting numbers evaluates an expression in
objects” (Breidenbach et in for variables and terms of x.
al., 1992, p. 251). Would calculating to obtain a
involve the ability to plug number, but not “A function is an equation in
numbers into an algebraic referring to any overall which a variable is
expression and calculate. A process. If student manipulated so that an
static conception; student does not explicitly answer is calculated using
needs to think about it one mention a beginning numbers in place of that
step at a time. or resulting object, variable.
response should be
categorized as an “A function is a combination
action. of operations used to derive
an answer.
Subject might be able “A function is an expression

to find composition of that will evaluate something
two functions by when either variables or
substituting one numbers are plugged into
formula into the other, the function.” (p. 252)
but unable to find
composition in a more
general sense, e.g., if
not given an algebraic
expressions, or if
given functions with
split domains.
(table continues)
52
Table 1 (continued)
Description of APOS Levels
Sample student responses

APOS Level Description of Level Evidence to “What is a function?”
Process Involves a “dynamic Subject can think “A function is a statement that

transformation of objects about process without when given values will
according to some having to perform it operate with these values and
repeatable means, that given and can think about return some result.
the same original object, the transformation as
will always produce the a “complete activity “A function is some sort of
same transformed object.” beginning with input being processed, a way
(p. 251) objects of some kind, to give some sort of output.
doing something to
these objects and “A function is an algorithm
obtaining new objects that maps an input into a
as the result as what designated output.
has been done.” (p.
251) “A function is an operation
that accepts a given value and
Can combine the returns a corresponding
process and even value.” (p. 252)
reverse it. Notions of
1-1 or onto functions
are accessible.
Mentions input,
transformation, and
output in general
fashion.
Object When a student can perform

an action on a process, the
authors say the student has
encapsulated the process as
an object.
53
7 appeared to start the course with strong process conceptions, 24 showed clear
progress throughout the semester that seemed to be more than one might expect
from ordinary instruction and the other students made only a small amount of
progress or their performances appeared to oscillate. None of the students got
noticeably worse. (p. 274)
Schwingendorf, Hawks, and Beineke (1992) found similar results while using this
pedagogical technique with a group of first year calculus students and claimed that it
“may produce substantial progress in the development of the students’ understanding and
‘sense’ of the function concept” (p. 147). They gave a final questionnaire six months after
instruction and noted that “our students appear to have made progress in developing a
process conception of function by constructing functions as processes in ISETL” (p. 147).
There was less confusion with notion of one-to-one and the uniqueness condition for
functions and little if any continuity restriction, that is, they did not think function had to
be continuous. These observations are indicative of process conception of function.
Confrey, Piliero, Rizzuti, and Smith (1994) developed software called Function
Probe in an effort to facilitate a constructivist learning environment. The software
allowed students to experiment, analyze, and make conjectures about functions using
multiple representations. It allowed students to build calculator buttons to generalize
numeric procedures, fill tables to organize data, and perform transformations of function
graphs. This technique is similar to that of Breidenbach et al.’s (1992) technique of
asking students to write small computer programs in ISETL.

54
Apparently, the exactness required to write a simple code enables the construction
of mathematical concepts. Previous screens were also readily accessible to students in
order to encourage student reflection about their work. The researchers concluded that the
students’ “understandings of functional relationships were strengthened and broadened
by coordinating different representations” (Confrey et al., 1994, p. 4). They also noted
that the use of contextual problems contributed to their understanding. They claimed,
however, that though the software and curriculum played important roles, the teacher’s
role was “critical in the development process” (Confrey et al., 1994, p. 4).
Dubinsky and Harel (1992) described in detail the nature of the process
conception of function. They reported on a study conducted with 22 students in a
Discrete Mathematics class. The article analyzed in depth 4 of the 13 interviews
conducted with students after they experienced instructional activities based on a
constructivist theory of learning involving computer activities in the programming
language ISETL. These activities were specifically designed to foster development of
students’ conceptions of function.
The researchers (Dubinsky & Harel, 1992) found that the process conception of
function is complex and identified four factors that evolved from their interpretations of
the observations of students’ apparent ways of thinking about function. These four factors
were
1. Restrictions that students possess about what a function is. They observed
three main restrictions: the manipulation restriction (one must be able to
perform an explicit manipulation for a function to exist), the quantity

55
restriction (inputs and outputs must be numbers), and the continuity
restriction (a graph representing a function must be continuous).
2. The severity of the restriction.
3. Ability to construct a process when none is explicit in the situation, and the
student’s autonomy in such a construction.
4. Uniqueness to the right condition, confusing the definition of function with
that of a 1-1 function.
The authors analyzed in detail the significance of each of the function situations
presented to students. A variation of these situations is included in Appendix A and was
used in the present study. The function situations deal with 7 different contexts:
1. Two finite sequences: one with integer values (#8) and one with Boolean,
true/false values (#18).
2. Two character strings (#5 and #6).
3. Three graphs: a single valued continuous curve (with respect to both axes;
#2), a discrete graph (#19), and a continuous curve (#13).
4. Three sets of ordered pairs (#1, #16, and #17).
5. One table (#12).
6. Five equations: one with a single variable (#15), three with two variables (#3,
#4, and #11), and one set of parametric equations (#14).
7. Three statements: two describing physical situations (#7 and #9) and one
specifying a set of records (#10).

56
Dubinsky and Harel’s analysis of these situations is described in Chapter 3 under the
section entitled Analysis.
In summary, many researchers have documented students’ difficulties with the
concept of function. Breidenbach et al’s (1992) didactical method of writing ISETL
programs resulted in significant growth in student understanding of the concept based on
APOS analysis. The current study attempts to discern if learning the history of the
concept of function is another avenue to such understanding.
History of the Concept of Function
This section looks at the main players in the history of the concept of function.
Much of the work described below appears in abbreviated form in the worksheets used in
the present study.
Origins
Scholars differ in their opinions about the origin of the notion of function. E. T.
Bell (1945) claimed that “it may not be too generous to credit them [ancient Babylonians]
with an instinct for functionality; for a function has been succinctly defined as a table or a
correspondence” (p. 32). The Babylonians used tables like the one for n3 + n2, n = 1, 2,
. . ., 30 suggesting that a function is a table or correspondence, between n in one column
and n3 + n2 in the other (Kennedy & Ragan, 1989). According to Sierpinska (1992), the
notion of function had its origin in
the disdained field of practical computations. It entered mathematics through the
kitchen door long before it was put on a pedestal by Felix Klein, who, in 1908 . . .
57
advocated that functional thinking should pervade all of mathematics and, at
school, students should be brought up to functional thinking. (p. 32)
In ancient Greece, the art of constructing tables, however sophisticated, did not
belong to the realm of science. It belonged to that of logistics. In the Almagest,
considered the culmination of Greek astronomy and the most influential astronomical
work from the time it was written until the 16th century, Ptolemy (c. 100 – 178 C.E.)
constructed sophisticated tables of chords, or sines, of all arcs from ½o to 180o in
intervals of ½o. Though involving the interpolation of functions of two variables, his
work did not achieve the status of true mathematics, as did Euclid’s elements and some
new geometrical theorems and plane and spherical trigonometry (Sierpinska, 1992, p.
31). O. Pederson (as cited in Siu, 1995) recognized the existence of functions in
Ptolemy’s Almagest:
But if we conceive a function, not as a formula, but as a more general relation
associating the elements of one set of numbers (viz, points of time t1, t2, t3, . . .)
with the elements of another set (for example, some angular variable in a
planetary system), it is obvious that functions abound in the Almagest. Only the
word is missing: the thing itself is there and clearly represented by the many
tables of corresponding elements of such sets. (p. 106)
On the other hand, A.F. Monna claimed that “the notion of function has no place in
Greek mathematics” (as cited in Siu, 1995 p. 106) and Youschkevitch (1976) noted that
“there was no general idea of functionality in ancient times” (p. 42). Boyer (1949)
claimed that since Greek geometry was concerned largely with form rather than variation,
58
the function concept was not developed. Apparently, one’s belief about the origin of the
function concept depends upon one’s definition of function.
The Fourteenth Through Sixteenth Centuries
A. P. Youschkevitch claimed, “The notion of function first occurred in a more
general form three centuries later [in the 14th century], in the schools of natural
philosophy at Oxford and Paris” (p. 45). In the 14th century appeared the beginning of the
notion that velocity, in particular instantaneous velocity, was measurable. The ancient
Greeks never considered velocity or acceleration as independent measurable quantities.
As mathematicians attempted to quantify Aristotle’s ideas on motion, they began to
explore the idea of representing velocity, as well as other varying quantities, by line
segments. Notions of time, distance, and length of line segments were still considered to
be continuous magnitudes, not discrete numbers. Thus, representing the abstract notion of
velocity, which was clearly continuous, as a line segment seemed reasonable. Velocities
of varying magnitudes would be represented by line segments of different lengths.
Nicole Oresme (1320-1382), a French cleric and mathematician associated with
the University of Paris, carried this idea to its logical conclusion by introducing a two-
dimensional representation of velocity changing with respect to time. He showed
geometrically that the arithmetic mean of the initial and final velocities of motion with
uniform acceleration is equivalent to motion with uniform velocity. A rectangle
corresponds to uniform velocity, a triangle to uniform acceleration.

59
E G
F
A B
AF = ½ AC
Area(ABC) = Area (ABGF)
Figure 1. Oresme’s Uniform Velocity and Uniform Acceleration
Sierpinska (1992) claimed that this geometric way of thinking about numbers
makes Oresme’s graphs qualitative models of relationship rather than true quantitative
graphs as we think of them today (p. 40). Siu (1995) disagreed, however, and maintained
that though this “dim idea of functional dependence exerted minor influence later, it
indicated: (i) quantitative laws of nature as laws of functional dependence, (ii) conscious
use of general ideas about independent/dependent variables, (iii) graphic representation
of functional dependence” (p. 107).
According to Alistair C. Crombie (1959) this idea of functional relationship was
developed only in principle, not with actual measurement. Youschkevitch (1976) noted a
lack of computational technique, claiming that these early views of function were not that
significant. Similarly, Kleiner (1989) attributed the lack of the development of the
function concept to the following:
• lack of algebraic prerequisites—the coming to terms with the continuum of
real numbers, and the development of symbolic notation;

60
• lack of motivation. Why define an abstract notion of function unless one had
many examples from which to abstract? (p. 283)
Seventeenth and Eighteenth Centuries
The study of motion by Johannes Kepler and Galileo Galilei in the early 17th
century provided the impetus for development of the function concept (Siu, 1995). Also
significant were the extension of the concept of numbers and the development of
symbolic algebra, which facilitated the invention of analytic geometry by Fermat and
Descartes, followed eventually by the creation of calculus by Newton and Leibniz. The
crucial elements in analytic geometry were the use of variables and the expression of the
relationship between variables by means of equations. These equations “provided a large
number of examples of curves (potential functions) for study and set the final stage for
the introduction of the function concept” (Kleiner, 1989, p. 283). Still missing, however,
was the discrimination between the dependent and independent variable in the equation.
Rene Descartes
Descartes went beyond the purely algebraic treatment of equations to the study of
variation in magnitudes when one variable changes through a series of values. Friedrich
Engels (1940) said:
The turning point in mathematics was Descartes’ variable magnitude. With that
came motion and hence dialectics in mathematics, and at once also of necessity
the differential and integral calculus, which moreover immediately begins, and
which on the whole was perfected by Newton and Leibniz not discovered by
them. (p. 199)

61
Newton and Leibniz
Modern students would not recognize the calculus of Newton and Leibniz. It was
not a calculus of functions, but rather one of geometric curves. Analysis of the time
involved finding tangents, areas under various curves, lengths of curves, and velocities of
points moving along curves. The variables were geometric (abscissas, ordinates,
subtangents, subnormals). In 1692, Leibniz introduced the word “function” to designate a
geometric object associated with a curve. For example, he wrote, “a tangent is a function
of a curve” (Kleiner, 1989 p. 283). Even though this study is geometric rather than
analytic, Siu (1995) maintained that it induced further study into the notion of function
simply by providing many examples of functions cloaked in various forms (e.g., the
“fluents” of Newton, the abscissas, ordinates, and subtangents studied by Leibniz, and
eventually the infinite power series of Mercator, Gregory, and Newton).
As mathematicians increasingly emphasized the formulas and equations relating
to the curves, they focused their attention of the role of the symbols in those equations,
often not even referring to the curves themselves. They studied the relationships among
these symbols. According to Kleiner (1989), Johann Bernoulli and Leibniz corresponded
(1694-1698) about the lack of a general term to represent quantities dependent on other
quantities in these formulas. They eventually introduced the use of the term “function” as
it appears in Bernoulli’s (1718) definition: “I call a function of a variable magnitude a
quantity composed in any manner whatsoever from this variable magnitude and from
constants” (Katz, 1998, p. 724).

62
Leonhard Euler
Leonhard Euler, in his book Introduction in Analysis Infinitorum written in 1748,
claimed that analysis is the science of variables and their functions. His entire approach
was algebraic, not geometric; not a single drawing appears (Kleiner, 1989, p. 184). His
definition of function was almost identical to Bernoulli’s. He used the term “analytic
expression” and though he did not define it, he stated the admissible ones: the four
algebraic operations, roots, exponentials, logarithms, trig functions, derivatives, and
integrals. He classified functions as being either algebraic or transcendental and claimed
that any function can be expanded in a power series. His work exhibited one of the
earliest treatments of trig functions as numeric ratios and was the earliest interpretation of
logarithms as exponents (Kleiner, 1989, p. 184). Interestingly, Euler’s definition of a
continuous function was one that had the same analytic expression over the entire
domain. Historian Hawkins (1975) summed up Euler’s contributions to the development
of the function concept: “Although the notion of function did not originate with Euler, it
was he who first gave it prominence by treating the calculus as a formal theory of
functions” (p. 3).
The Vibrating String Problem
The next major influence on the development of the function concept was the
controversy over the solution to the vibrating string problem. Given an elastic string with
fixed ends, which is pulled into some initial shape then released to vibrate, the problem is
to determine the function that describes the shape of a string at time t. The controversy
centered on the meaning of “function” and the type of functions which could be allowed
63
in analysis from the standpoint of a mathematician (d’Alembert), a physicist (Bernoulli),
and a mathematical physicist (Euler). In 1747, d’Alembert solved the problem by
showing that the motion of the string is governed by the so-called “wave” partial
differential equation
∂ 2y ∂2 y
=a 2 (a is constant, y(0,t) = y(L,t) = 0).
∂t 2 ∂x 2
In this representation y, the dependent variable represents the displacement from
equilibrium, x represents the distance from the origin, and t indicates time. The most
“general solution” is y(x,t) = f(x + t) + f(x – t), where f is an arbitrary function. The only
restrictions on it were that it be periodic, odd, and everywhere (twice) differentiable.
D’Alembert thought that the function describing the initial form of the string must be a
single analytic expression, that is, given by the same formula over the entire length of the
string.
In 1748 Euler solved the same problem and showed that his solution gives the
shape of the string for different values of t even when the initial shape is NOT given by a
single formula. Euler argued that the initial shape can be given by different analytic
expressions in different subintervals, or even drawn free-hand. Different pieces of the
string might be described by circular arcs of varying radii, for example. The belief at the
time, however, was that if two analytic expressions agreed on an interval, they agreed
everywhere. In other words, if a single analytic expression determined the shape of the
entire curve, looking at a single interval, no matter how small, would be sufficient to
determine behavior on the entire string. If the initial shape of the string were given by
64
several analytic expressions or drawn free-hand, it could not possibly be given by a single
expression. Thus d’Alembert’s solution could not account for all possibilities. Bernoulli
entered the debate, which lasted for several more years. The major outcome of this debate
was to extend the function concept to include:
• functions defined piecewise by analytic expressions in different intervals
• functions drawn freehand and possibly not given by any combination of
analytic expressions. (Kleiner, 1989, p. 288)
Another eventual result of this debate was the change in Euler’s definition of
function. Recall that his 1748 definition used the term “analytic expression” but the
following later one (1755) did not:
When quantities depend on others in such a way that [the former] undergo
changes themselves when [the latter] change, then [the former] are called
functions of [the latter]; this is a very comprehensive idea which includes in itself
all the ways in which one quantity can be determined by others. (Katz, 1998, p.
724)
The Nineteenth Century
Fourier Series
Fourier studied heat conduction and his work, in addition to winning him a prize
from the Institut de France in 1812, was “revolutionary in the evolution of the function
concept” (Kleiner, 1989, p. 289). He claimed that any function f(x) defined over an
interval is representable over this interval by a series of sines and cosines. Both Lagrange
and Euler, among others, had previously recognized Fourier’s results as true for certain
65
functions. Fourier’s claim that it was true for all functions was revolutionary. His
definition (1822) of function is as follows:
In general, the function f(x) represents a succession of values or ordinates each of
which is arbitrary. An infinity of values being given to the abscissa x, there is an
equal number of ordinates f(x). All have actual numerical values, either positive or
negative or null. We do not suppose these ordinates to be subject to a common
law; they succeed each other in any manner whatever, and each of them is given
as if it were a single quantity. (Katz, 1998, p. 724)
The influence of Fourier’s work is as follows:
• It disproved the belief that if two analytic expressions agreed on an interval,
they agreed everywhere. He calculated the Fourier coefficients for small
values of n for a great variety of functions and noted the close agreement in an
interval, but not outside the interval, between the initial segments of the
Fourier series and the function values of the given function (Kleiner, 1989, p.
290).
• There was a renewed emphasis on analytic expressions (Siu, 1995, p. 112).
• The recognition that an “arbitrary” function is representable by an analytical
expression (Siu).
• Reexamination of the function concept as a whole (Siu).
Both Siu (1995) and Kleiner (1989) claimed that Fourier’s work in heat
conduction set the analytic expression of a function on (at least) equal footing with its
66
geometric representation. Siu compared the vibrating string problem with the heat
conduction problem:
The shape of the string (geometry) is visible, while temperature distribution
(algebra) is not. This may explain the freeing from geometric perception of a
function and the emergence of a general notion of function in the nineteenth
century. (Siu, 1995, p. 111)
He also gave profound insight into the problem of accepting a new concept, both
historically and by students:
Why was the “Eulerian” concept of function maintained so long after the
realization that is was inadequate? What lesson do we learn from this experience?
(If only a particular form is used, students unconsciously accept that particular
form as the definition. We witness the same psychological effect in
mathematicians of the seventeenth/eighteenth centuries. A new concept receives
recognition only when it is relevant to current usage. This is as true in research as
it is in teaching.) (p. 113)
Cauchy and Dirichlet
Cauchy was one of the first mathematicians to usher in a “spirit of rigor” in
mathematical analysis and defined the concepts of continuity, differentiability, and
integrability of a function in terms of limits (Kleiner, 1989, p. 291). His definition of
function in 1821 was not that different from his predecessors:
When the variable quantities are linked together in such a way that, when the
value of one of them is given, we can infer the values of all the others, we
67
ordinarily conceive that these various quantities are expressed by means of one of
them which then takes the name of independent variable; and the remaining
quantities, expressed by means of the independent variable, are those which one
calls functions of this variable. (p. 291)
Another main player of the time was Dirichlet. He questioned Fourier’s result and
in 1829, gave sufficient conditions for Fourier’s representability of functions, that is, that
the function needs to have only finitely many discontinuities and finitely many maxima
and minima on the interval. Dirichlet was the first to take seriously the notion of function
as an arbitrary correspondence (Siu). He is known for the celebrated “Dirichlet
function:”
⎧ c, x is rational
D(x) = ⎨
⎩d , x is irrational
This function was the first explicit example(s) of one not given by analytic expressions,
nor was it a curve drawn freehand. It was also the first example of a function
discontinuous everywhere. It also illustrated the concept of a function as an arbitrary
pairing (Kleiner, 1989, p. 292).
Dirichlet’s definition of function follows:
y is a function of a variable x, defined on the interval a < x < b, if to every value
of the variable x in this interval there corresponds a definite value of the variable
y. Also, it is irrelevant in what way this correspondence is established. (Kleiner,
1989, p. 291)
68
Noteworthy in this definition was the explicit restriction of the domain to an interval . . .
also a first.
Riemann and Weierstrass
In 1854, Riemann dealt with the representations of functions in Fourier series.
The coefficients of a Fourier series are given by integrals. Cauchy had developed his
integral only for continuous functions, but Riemann extended this concept to
discontinuous functions. As a result, he enlarged the class of functions representable by
Fourier series. Both Riemann and Weierstrass delighted in studying irregularities and
discovering exceptions in analysis (Kleiner, 1989, p. 292). Weierstrass found an example
of a continuous, nowhere-differentiable function and thus began the “disengagement of
the continuous from the differentiable” in analysis (Kleiner, 1989, p. 293).
These “pathological” examples of functions characterized the change in emphasis
in analysis during the late 19th century:
The main difference between methods of studying functions within the framework
of mathematical analysis and the theory of functions is that classical analysis
deduces properties of any function starting from the properties of those analytical
expressions and formulae by which it is defined, while the theory of functions
determines the properties of function starting from that property which a priori
distinguishes the class of functions considered. (Luzin, as cited in Siu, 1995, pp.
114-115)
69
Function as Correspondence
With Cantor’s development of set theory and progress made in algebra, the notion
of function as mapping dominated towards the end of the 19th century (Siu, 1995). A
sampling of such definitions follows. Note that these definitions have their basis in set
theory:
Dedekind (1888): “A function φ on a set S is a law according to which every
determinate element s of S there belongs a determinate thing which is called the
transform of s and denoted by φ (s)” (Katz, 1998, p. 724).
Peano (1911):
The function is a special relation, by which to each value of the variable there
corresponds a unique value…a function is a relation u such that, if two pairs y;x
and z;x, having the same second element, satisfy the relation u, it necessarily
follows that y = z whatever x, y, z may be. (Siu, 1995, p. 116)
Bourbaki (1939):
Let E and F be two sets, which may or may not be distinct. A relation between a
variable element x of E and a variable element y of F is called a functional
relation in y if, for all x ∈ E, there exists a unique y ∈ F which is in the given
relation with x.
We give the name function to the operation which in this way associates with
every element x ∈ E the element y ∈ F which is in the given relation with x; y is

said to be the value of the function at the element x, and the function is said to be
70
determined by the given functional relation. Two equivalent functional relations
determine the same function. (Siu, 1995, p. 116)
Conclusion
Kleiner (1989) summarized the development of the function concept succinctly.
He claimed that implicit manifestations of the function concept date as far back as 2000
B.C., though its explicit form did not emerge until the beginning of the 18th century. He
viewed the evolution of the function concept as “a tug of war between two elements, two
mental images: the geometric (expressed in the form of a curve) and the algebraic
(expressed as a formula—first finite and later allowing infinitely many terms)” (p. 282).
Later, the “logical” definition of function appears and the geometric conception is
gradually abandoned.
Summary
This chapter described four theoretical frameworks for analyzing the link between
the history of mathematics and the learning of mathematics. This research confirmed
significant links between history and student learning, but did not connect individual
study of history to learning a concept. Piaget and Garcia’s (1989) triad is similar to the
components in Asiala, Brown, et al.’s (1997) APOS theory and APOS provides the
theoretical framework for the current study. This link suggests possible student benefits
from studying history. Students may recognize their own conceptions in the history of a
concept and note how these conceptions were found to be inadequate. Once these
inadequacies were discovered, the concept developed historically. Will similar results
occur cognitively in students?

71
The second group of studies reviewed concerned the effectiveness of using the
history of mathematics as a tool to learn the mathematics. Most studies in this area dealt
with student attitude towards mathematics or views of mathematics. Several involved
self-reporting of participants and lacked the objectivity for convincing empirical results.
The recent dissertation by Van Gulik (2005) is one exception, however. The current study
attempts to more closely investigate the learning that occurs as a result of studying
history.
The third section reviewed the literature concerning students’ understanding of
the concept of function and provides much of the structure to the current study. In
particular, Harel and Dubinsky’s (1992) and Breidenbach et al.’s (1992) APOS analyses
provide very specific guidelines for analysis. The closing section on the history of
function gives historical background to the reader, so that he may understand the context
of this study.
CHAPTER III
METHODOLOGY
Introduction
Asiala, Brown, et al. (1997, p. 38) referred to Thomas Kuhn’s work (1970) as
they described a paradigm to be a
Collection of understandings (explicit or implicit) on the part of an individual or
group of individuals about the kinds of things one does when conducting research
in a particular field, the types of questions that are to be asked, the sorts of
answers that are to be expected, and the methods that are to be employed in
searching for these answers.
A paradigm shift generally occurs gradually (Asiala, Brown, et al.), as a result of
dissatisfaction with the status quo over an extended period of time. A change then takes
place and a new paradigm emerges. Such a paradigm shift occurred in research models in
mathematics education. The statistical model, comparing an experimental and a control
group, caused consternation among researchers since controlling enough variables to get
an appropriate control group is difficult, if not impossible. Mathematics education
researchers rarely consider two nearly identical classes and subject them to a single small
change in curriculum or pedagogy. Also, statistical information rarely provided enough
information about students’ thinking or understandings (Selden & Selden, 1992).
72
73
Design
The current study attempts to determine the impact of historical studies on
students’ thinking and learning. The hope is to describe the changes in student
understanding that result from studying the history of a concept and demonstrate that
some students are able to move from one level of understanding to another in part due to
their exposure to historical readings. In the recently published ICMI study, E. Barbin
(2000a) commented that
The question of judging the effectiveness of integrating historical resources into
mathematics teaching may not be susceptible to the research techniques of the
quantitative experimental scientist. It is better handled through qualitative research
paradigms such as those developed by anthropologists. (p. 63)
Constructivist research is qualitative. Its purpose is to study the construction of
mathematical concepts in the minds of students. Constructivist researchers view
mathematics teaching and learning from three perspectives: the experiential perspective
from which they attempt to understand what students’ mathematical worlds are like, a
cognitive perspective from which they attempt to determine what mental operations give
rise to these mathematical worlds of students, and a social/cultural perspective from
which they attempt to describe the social and cultural factors which affect students’
construction of meaning (Battista, 1999a). Information from these three perspectives
gives a well-rounded view of student thinking and learning.
A teaching experiment is one constructivist research methodology. It is “primarily
an exploratory tool . . . aimed at what goes on in the student’s head. To this it adds
74
experimentation with ways and means of modifying the student’s operating” (von
Glasersfeld, 1987, p. 12). A teaching experiment is more than an interview aimed at
understanding where the child is in his understanding. It also aims at moving the child
along toward adult competence (Battista, 1999a). In order to do so, however, “the
experimenter/teacher must not only have a model of the student’s present conceptual
structures but also an analytical model of the adult conceptualizations towards which his
guidance is to lead” (von Glasersfeld, p. 13). “The ‘guidance’ must take the form of
either questions or of changes in the experiential field that leads the child into situations
where her present way of operating runs into obstacles and contradiction” (p. 14).
This emphasis on understanding student thinking and moving the student toward
adult competence makes the teaching experiment a particularly good methodology for
answering the research questions of this study:
school teachers?
understanding to a process level understanding as described by APOS theory
(Action, Process, Object, and Schema)?

75
understanding?
According to Battista (1999a), the components of teaching-experiment-based
research are (a) preliminary interviews that allow for the researcher to make initial
distinctions in students’ conceptualizations, (b) Model building during which the
researcher formulates working models of student construction of ideas, (c) Retrospective
analysis to determine student’s ways of thinking, and (d) Scientific model building which
“provides an account of students’ mathematical concepts and underlying mental
processes at particular points in time as well as changes in those concepts and processes
caused by instruction and interaction with others” (p. 13).
In the current study, an initial questionnaire and interview provide information
about current student thinking, historical studies provide the potential perturbations that
move students along toward adult competence, and APOS theory provides the framework
for analysis of growth.
Sample
During spring semester 2006, students from a junior-senior level History of
Mathematics class at a large Midwestern university were chosen for this study. A week
before the five-week unit on the history of functions was to begin, the entire class of 17
students completed an extensive initial written questionnaire. Based on the responses to
this questionnaire, 7 students were asked to participate in the study. In order to obtain a
heterogeneous sample, the researcher asked the following to participate: the one student
76
whose responses suggested a process conception, three whose responses suggested an
action—or perhaps even a pre function—conception, and three who evidenced some
characteristics of an action conception and some of a process conception. The researcher
categorized these students as having an “emerging process” conception. Four of the
participants were juniors, two were seniors, and one was a post-undergraduate. Of the
seven, five were studying to be secondary school mathematics teachers. One of the
junior-level pre-service teachers became ill with mononucleosis during the unit on the
history of functions and dropped out of the study. One student missed his initial interview
because of a schedule conflict but chose to remain in the study.
Procedure
Students enrolled in the History of Mathematics course participated in a five-
week unit on the history of the concept of function, as described by the worksheets in the
Appendices. These learning materials are nontraditional in that they did not follow a
chronological approach of multiple topics typically used in history of mathematics texts.
They focused on the development of a single concept and thus did not fit into the
curriculum as the course is usually taught. Only one section of the course is taught per
semester. Hence, the researcher was the instructor of the course. The culture of the
classroom was inquiry-based, with students working collaboratively.
All students in the class completed questionnaires before and after the
instructional program to ascertain their conception of function. These questionnaires are
in Appendices A and B and most of the questions on them are adapted from the work of
Dubinsky and Harel (1992). Based on student responses to the initial questionnaire, the
77
researcher chose seven students to interview. The initial interviews consisted of
questioning students about responses to the initial assessment. In particular, the
researcher asked them to explain their thinking as they responded to each question. Each
of these interviews was audio taped. Each interview lasted between 60 and 90 minutes.
All tapes were transcribed.
All students in the course worked in groups both in class and outside of class to
complete the readings and worksheets in the Appendices. After each reading or
worksheet, each student wrote a one page summary reflection indicating the following:
• her understanding of the concept of function;
• if and how the worksheet led to new insights concerning the concept of
function.
The researcher conducted another individual interview with each of the 6 participants at
the conclusion of the unit. These interviews were also between 60 and 90 minutes in
length and focused on students’ responses on both questionnaires. In particular, the
researcher asked the participants again to explain their thinking in detail as they worked
through the second questionnaire and asked if they would change any of their answers on
the initial questionnaire. These interviews, the completed questionnaires, the student
worksheets, and the individual reflections comprised the data for this study. This plan is
summarized in Table 2.
Triangulation
The researcher was careful to gather and analyze data from different sources. The
written responses on the initial questionnaire were further explored in the interviews.
78
Students’ written reflections provided insight in a way that interviews and questionnaires
could not since the students wrote these immediately following their work on each
worksheet. Their thinking about the history was fresh in their minds and the questions
were open ended. The students’ completed worksheets were another data source.
Inter-Rater Reliability
Qualitative researchers have developed standard techniques for dealing with
issues of reliability of their research. One is the training of other observers who then can
work through the same body of data independently and compare notes afterwards. If all
goes well, the observers will code student responses similarly. This technique is
“interrater reliability” (Schoenfeld, 2002, p. 463) and identifies the degree to which
independent researchers assign a body of data the same coding.
The researcher of the current study conferred with a colleague concerning the
interpretation of data. Ed Dubinsky, the original developer of APOS Theory, was a co-
director for this research and a consultant on all data analysis. As a prolific researcher and
developer of the theory, he provided a level of confidence in the analysis that is
unparalleled. He and a group of researchers had developed the theory over several years
Table 2
Summary of Research Design
Research question Purpose Data tools Details
What is the function conception of Create a baseline of Function Given to all students in class on March 23. Researcher
junior and senior level pre-service high information about questionnaire in analyzed and coded questionnaire responses and chose 7
school teachers? student understanding Appendix A. students to interview in depth about their responses.
so that accurate Students were chosen on the basis of their responses, so
assessment of growth In depth interviews that the result was a heterogeneous sample. These students
can occur. then became the primary participants.
Does studying the history of the concept Gather information Student worksheets; Students were assigned a series readings and worksheets on
of function deepen a student’s about student thought the history of the concept of function. These are in the
understanding of the concept in any processes. Written individual appendices. Students worked in cooperative groups in class
way and if so, in what way? reflections and outside of class on these worksheets. Class discussion
Glean if and when concerning the ensued.
In particular, does studying the history growth occurred and understanding of the
facilitate his or her move from an action what caused it. concept of function. After each worksheet, each student reflected upon their
level understanding to a process level understanding of the concept of function in writing: What
understanding or from a process level to Evaluate overall Written second insights did the reading or worksheet facilitate? What
an object level as described by APOS learning as result of questionnaire and questions arose? What were you confused about?
Theory? entire project. second interview
The second questionnaire is in Appendix B.
Interviews were based on student responses to initial

questionnaires (i.e., explain your answer. Would you
change your answer now?) and final questionnaires.
(table continues)
79
Table 2 (continued)
Summary of Research Design
Research question Purpose Data tools Details
In what ways can studying the history of a Determine the most All of the above. In analyzing the different types of data collected
mathematical concept be used to deepen a effective ways of (written, discussion, readings, problems) the researcher
student’s understanding of the concept? using history. attempted to discern exactly which pieces facilitated
growth.
Does a student’s studying the history of the Evaluate overall Written second The second questionnaire is in Appendix B.
concept of function facilitate his or her learning as result of questionnaire and Interviews were based on student responses to initial
move from a process level understanding entire project second interview questionnaires (i.e., explain your answer. Would you
to an object level? change your answer now?) and final questionnaires
80
81
and have applied it to various topics in collegiate mathematics (Asiala, Brown, et al.,
1997) as early as 1992 (Breidenbach et al., 1992). After the researcher analyzed each
interview transcript, she sent the transcript and analysis it to him for his comments and
interpretations. On two occasions, disagreements occurred. A discussion ensued to
resolve the minor differences. In both cases, the researcher deferred to Dr. Dubinsky’s
more experienced judgment. The detailed sample student responses and interpretation
provided in the Dubinsky and Harel article (1992) further ensured reliability since the
questionnaires used in the current study closely mimicked those described in the article.
Thus, by analyzing students’ thinking from several viewpoints—responses on
questionnaires, comments during interviews, written reflections, and a colleague’s
analysis—triangulation and inter-rater reliability were achieved, thereby strengthening
confidence in the findings.
Data Analysis
The theoretical framework for this study was APOS theory, described in Chapter
1. Analysis closely followed that in the articles by Breidenbach et al. (1992) and by
Dubinsky and Harel (1992). Since the questions for the current study are adapted from
the Dubinsky and Harel study, their detailed analysis of the situations presented follows.
The problem numbers are in parentheses following the name of the problem type. These
situations are in Part 2 of the questionnaire in Appendix A.
Finite Sequences (#8, #18)
For a student to identify a function from a sequence, she must have thought in
terms of a first term, second term, and so forth, something not given to her in the
82
situation. The authors claimed that if a student can accept a positive integer as the ordinal
position of one of the quantities in the sequence, and take that quantity as the output, then
that student is using a process conception. However, in these particular questions, the
format of the problem strongly suggests that construction (Dubinsky & Harel, 1992).
Therefore, a student’s success in dealing with these questions might only indicate an
action conception. Since #18 involves Boolean values for output, though, it “does provide
us with a context for suggesting the possibility that those who were successful were
capable of more than an action conception of function” (p. 92).
Character Strings (#5, #6)
The mental constructions required for strings are mathematically equivalent to
those required for sequences. They pose more psychologically difficult for students,
however, since the outputs are characters rather than numbers, and no suggestion of the
construction is evident (Dubinsky & Harel, 1992). All we see is the result.
Graphs (#2, #13, #19)
The authors believed that these examples have potential to provide indication of
one’s ability to use a process conception of function, particularly the one that required
use of values on the vertical axis as the domain (#13). Also #19 is a good indicator of a
process conception since a student with a process conception of function would see the
function process even if the domain was very small. The researchers noted student
difficulty with this type of question, reminiscent of Carlson’s (1998) finding that students
often believe that a function had to be continuous.

83
Sets of Ordered Pairs (#1, #16, #17)
Dubinsky and Harel (1992) noted significant student difficulty with this
representation of function. Students often confused the process of constructing the set of
ordered pairs with the function itself. However, the function can only be constructed if
the ordered pairs are already there by identifying the domain as the first element and
taking the second element as the result of the function process. The ordered pair
representation does not suggest this construction, so it must come from the student
himself.
The researchers noted the common student confusion of the uniqueness condition
with the notion of one-to-one, results also consistent with Vinner’s (1989) study. Because
of this confusion and the necessity for the construction coming from within the student,
they claimed that “the set of ordered pairs is a bellweather type of situation for detecting
the presence and strength of a process conception of function” (Dubinsky & Harel, 1992,
p. 93).
Tables (#12)
Dubinsky and Harel (1992) claimed that, for the purpose of their analysis, tables
are similar to ordered pairs. A student who insisted upon a rule relating the first number
to the second and/or cannot construct a process as the act of going from an item in the
first column to one in the second, is “probably displaying an action conception of
function” (p. 93).

84
Equations (#3, #4, #11, #14, #15)
If a student insisted upon solving an equation in two variables for one in terms of
the other, she probably displayed an action conception of function. If she can describe the
process without actually doing it, she “probably” exhibited “at least the beginning of” a
process conception (Dubinsky & Harel, 1992, p. 93).
With equations involving one or more variables, one can view as input any
numerical value(s) and as output a Boolean value, True or False. Since such a
construction must come from the student himself, it is evidence of a process conception
of function.
Statements (#7, #9, #10)
These are the most open-ended in Dubinsky and Harel’s (1992) list of situations.
They included these simply to observe what type of functions the participants would
construct, given a situation with very little structure.
One consistent theme for evidence of a process conception of function is that the
subject did the construction himself, that is, the representation does not suggest the
construction.
Pilot Study
During the spring semester 2005, two colleagues ran a pilot study with the
researcher and used a similar, but scaled down research design. Data consisted of student
written responses on the initial and second questionnaires and student work on the
worksheets in the Appendices. The students in the study were from two groups. One
group of 10 students came from a lower level statistics course. These students were either
85
beginning calculus students or had no prior experience at the calculus level. The other
group consisted of 4 students who were enrolled in an upper level history of mathematics
course. Their previous mathematics experience was beyond the calculus III level.
In both groups, the researchers assessed levels of student understanding of the
concept of function with an initial questionnaire very similar to the one in Appendix A. It
included open-ended questions and situations which might be described with functions.
Students were asked to find a function in each of the situations. They then completed a
worksheet based on readings from the works of Nicole Oresme (1320-1382), a French
cleric and mathematician who first introduced two-dimensional graphs of varying
quantities. In particular, students read about his representation of velocity changing with
respect to time. The researchers tested student understanding of the readings with focused
questions that made connections to modern day function graphs and also asked students
to explain their understanding in journal entries.
Three of the students in the upper level group also completed a series of 6 other
worksheets over a 4-week period of time, based on the history of the concept of function.
These worksheets focused on the works of Fermat, Descartes, Leibniz, Euler, and
Fourier, and on the changing definition of function over the years and are found in the
Appendices. Three of these worksheets were from the Mathematical Association of
America’s CD, Historical Modules for the Teaching and Learning of Mathematics. The
others were this researcher’s design. The fourth student in this upper level group chose to
write a term paper. Her term paper encompassed the history of the concept of function
86
through the time of Euler. Because of her term paper work, she did not complete any of
the worksheets.
Three of these four students completed a questionnaire at the end of the semester
to discern their growth of understanding of the function concept. This second
questionnaire was similar in design to the initial questionnaire with open ended questions
and function situations (see Appendix B). Further, on the final exam the researchers
asked the entire upper level class of 25 students to comment on the benefits of studying
the history of mathematics in general, not necessarily relating to the concept of function.
Of the four students in the upper level group, only one of them correctly sketched
a graph of the height of water in a bulb-shaped bottle as a function of the amount of water
in the bottle on the initial questionnaire. On the second questionnaire, all three who
completed it correctly sketched the shape of a bottle when given a graph of the height of
water in the bottle versus its volume.
Other responses showed growth in understanding, as well. While answering the
question, ‘What is a function?’ on the initial questionnaire, student A made the common
mistake of confounding the function definition with the definition of a one-to-one
function. Her answer to the same question on the second questionnaire, however, was
concisely correct. Student B gave what appeared to be a correct textbook definition on
the initial questionnaire, but he implied the necessity of an algebraic formula. On the
second questionnaire, he emphasized the fact that “functions are not necessarily analytic
expressions,” and quoted Euler in his definition. As this student was working through an
activity critiquing different definitions of functions, he struggled with the idea that a
87
function need not require a formula. He was impressed, too, with “Euler’s evolution of
the definition of function.” He had a real epiphany, not only recognizing the generality of
the function concept but also its dynamic nature. In fact, we had a very stimulating
conversation about the dynamic nature of functions. Student C had provided a correct
definition on both instruments though claimed that her greatest insight was “how much
better I understand the ideas after looking at them from a different perspective.”
Another interesting development was the students’ ability to see functions in real-
world scenarios. On the initial questionnaire, none of the three students correctly
recognized a function in a list of student names paired with club dues owed. Student A
claimed, “there is no correlation between the data.” Student B claimed that it “can be a
function, just don’t know how to write it.” These comments suggest a conception of
function as needing some definable formula of correspondence. Student C left this
question blank on the pretest. The responses on the second questionnaire were, however,
much improved. On a question listing students’ names with a test scores, all three
recognized the function. Student C, who left 10 of 16 function existence questions blank
on the initial questionnaire, left only 1 of 8 such questions unanswered on the second.
There was no class discussion per se concerning the two questionnaires since only four
students worked on this project. Thus, one can reasonably conclude that her working
through the historical worksheets provided her the insight and perhaps extra confidence
to work more correctly and completely with functions.

88
Though the pilot study involved no interviews with students, there were enough
differences in student answers on the two questionnaires to suggest some growth in
student understanding about the concept of function. A more detailed study was in order.
Using a teaching experiment methodology, this study describes the changes in
thinking about functions that occurred as a result of studying the history of the concept in
each of six History of Mathematics students. Through examination of the various data—
written questionnaires, interviews, student worksheets, and written reflections—this
researcher uncovered what these six students knew about functions initially and how their
thinking differed after instruction.

CHAPTER IV
DATA
Introduction
This chapter contains the analysis and interpretation of data collected during
spring 2006 semester. The first section categorizes each of the participants’ conception of
function before studying the history of the functions. This section consists of two
subsections: understanding functions in general and understanding graphical
representations of functions. It answers the question: What is the function conception of
junior and senior level pre-service high school teachers? The second characterizes the
participants’ understanding of graphical representations using APOS terminology. With
Dr. Dubinsky’s approval, the researcher extended APOS analysis to include responses to
graphical representations of functions. This section describes and analyzes in general
student comments that characterize each of the APOS levels and then categorizes student
response for particular graphing tasks. The third section categorizes and discusses each of
the participants’ conception of function after studying the history of functions. This
section deals with the questions:
understanding of the concept in any way and if so, in which way In particular,
understanding to a process level understanding as described by APOS theory?
89
90
deepen student’s understanding of the concept?
understanding?
Initial Conceptions of Functions
Participant 1: DB
General Function Conception
DB’s initial conception of function is an emerging process one. The following
analysis is based on her responses on the initial questionnaire and her comments in the
initial interview.
Evidence of a process conception. The reader may find the description of APOS
levels in Chapter 2 of this study helpful. When asked what a function was, DB replied, “a
function is almost like a machine or some kind of process where you input a number or
something and then you get an output, a result.” When discussing Task 2, she put her
finger on the horizontal axis and pointed to the different values on the vertical axis, then
said, “You put in a value for x and the y value is . . . it’s increasingly larger.” These
comments indicate that DB can think about the process without actually having to
perform it.
In a typical mathematical sequence, a participant needs to think in terms of a first
term, a second term, and so forth, something not usually given in the situation. Dubinsky
and Harel (1992) claimed that if a participant can accept a positive integer as the position
91
of one of the quantities in the sequence and consider this integer as the output, then she is
using a process conception. Note DB’s comments in the excerpt below, where the letter I
indicates the interviewer’s comments.
TASK 8: If possible, describe the following situations using one or more
functions. If not, explain why not: {3n + n 2 : n in[1...100]} .
I: I see you have something crossed off there. Do you want to expand upon
your answer a little bit please?
DB: Okay, I wrote, “this is a function that works for values of n being 1-100,
which that is given, and as a function what you are doing you would be just
plugging in values for n and then you are using multiplication and addition
to come up with a result.”
These comments again suggest that DB was able to think about the process
without having to perform it. Dubinsky and Harel (1992) claimed, however, that this
particular representation itself suggests the function since the inputs are already given.
Thus one cannot be sure if DB would be able to construct the notion without the
suggested input values. Is her response coming from within her or from the representation
of the problem? When asked what the actual functions are, she replied, “Well I guess the
functions are multiplication and addition.” She saw a function as an operation, though did
again reveal her ability to discuss the process without actually performing it with
numbers: “I don’t know I guess I almost see two functions. You are taking the value of n
times something, and then you are taking that value of n times itself and then you are
adding the two results together.”

92
Task 18 involves Boolean values for outputs (true or false) and DB’s responses
are enlightening.
functions. If not, explain why not: {2n > n2 + 3n: n in [1 . . . 100]}.
DB: 2n is greater than n2 + 3n in a set of 1-100. Um . . . [Pause] Yeah, you would
plug in values for n from this set and then you would get a result.
I: What would your result be?
DB: Um . . . [Pause] I think it would be either true or false. You would find a
value of n that makes that statement true.
I: So your input would be?
DB: Um . . . any number 1 through a hundred.
I: And your output would be?
DB: Um . . . an inequality . . .?
I: So the output is the inequality, is that what you mean?
DB: Either true of false.
Though uncertain, DB did consider Boolean values to be viable function outputs.
Dubinsky and Harel (1992) claimed that this particular task “does provide us with a
context for suggesting the possibility that those who were successful were capable of
more than an action conception of function” (p. 92).
Dubinsky and Harel (1992) believed that Tasks 13 and 19 have potential to reveal
one’s process conception of function. Task 13 in particular is helpful since it requires the
use of values on the vertical axis as the domain. DB was weak in this regard. On the
93
written questionnaire, she indicated that you “plug in values for x and determine the
resulting y values.” She appeared to understand the notion of inputs and outputs but was
ignoring the uniqueness output criterion for functions. This criterion stated that in order
for a relation to be a function, each input must have a unique output. During the
interview, which due to scheduling difficulties occurred after she completed the
worksheet on graphs of functions, she exhibited confusion concerning the dependent
versus independent variable:
functions. If not, explain why not:
Figure 2. Graph for Task 13
I: Can you explain your answer please and tell me what you meant by that and
what you were thinking?
DB: Well, I mean we’re given a graph that I don’t necessarily recognize as being
a graph that I know. It doesn’t look like anything . . . [inaudible]. Um . . .
and all I thought was when you would plug in the results for x and then you
would find out what the y . . . the independent variable is. Um . . . so I

94
suppose that is a function because you put in a result and get a different
result.
DB’s notion of “getting a different result” is not clear here since any given x in the
domain is matched with more than one output.
Though DB left Task 19 blank on the written questionnaire, during the interview
she determined that the plotted points could represent a function.
I: Number 19? I noticed you left it blank on the written questionnaire. Do you
remember what you were thinking about it when you were doing it?
DB: It is just a graph . . . with some scatter plots. And I . . . I guess it would be a
function if I’m assuming like with the other graphs you would pick a point
. . . [pause] and find out the value for x and see what y, the result of y is.
I: So you changed your mind, so you think it is a function?
DB: I think so, yes.

95
Please note, however, since the interview occurred after she completed the worksheet on
graphs of functions, it is hard to determine if she had this conception originally or if the
worksheet facilitated her insight.
Weaknesses in her process conception. When asked if one needs a formula in
order to have a function, DB replied, “Yes. I think so.” The researcher was curious about
her response if given both a formula and a graph. Would she understand that each could
represent the same function, or would she mistakenly think that either the graph or the
formula was the function rather than just a representation?
I: So is the graph the function or is the formula the function or what?
DB: [Pause] . . . I believe that the formula is the function.
I: Now if you felt uncomfortable and you couldn’t find a formula for that,
would you still think it was a function?
DB: [Pauses to think] Yes. You mean even if I couldn’t find the specific
function?
I: Yes . . . Would there still be a function there?
DB: [Pause] Maybe not. And I would just have the graph?
I: Yes. And with the graph, there didn’t seem to be any function that you were
familiar with or maybe you couldn’t come up with one. Would it still
represent a function?
DB: Hm . . . [Pause] I don’t know.

96
She apparently did not understand that a function is a process and a formula or a
graph is simply one representation of that process. A function, therefore, may have more
than one representation. To DB, though, the representation itself was the function.
Recall that finite sequences can reveal whether or not a student has a process
conception. Dubinsky and Harel (1992) claimed that finding functions in Tasks 5 and 6 is
difficult for students since the outputs are characters rather than numbers; no suggestion
of the construction is evident. DB claimed that since there was no obvious dependency
relationship in Task 5, it was a statement rather than a function. The situation in Task 6
was not even a statement according to her definition. DB was unable to construct, in her
own mind, a set of inputs and corresponding outputs in these situations.
In another example, she was unable to find a function in a table, where a formula
was not evident.
TASK 12: If possible, describe the following situation using one or more
functions. If not, explain why not. The club members’ dues status:
Name Owed
John $15
Julie 10
Jen 0
Dave 7.50
Brittany 10
Alex 5
Mary Ann 25
Sam 20
Sally 17
Figure 4. Table for Task 12 giving club members’ names and the dues they owe
97
I: What about this one? The club members’ dues status. You have two
columns: one with the name and one with the amount owed, and you said,
“This involves a function of subtraction, each members’ dues are viewed
and the amount they have paid is subtracted from the total due. The amount
left is the amount due.” And how does the function fit into that?
DB: The function is . . . I guess . . . you would input . . . the independent
variable, then would be the amount of their dues and you are inputting how
much they have paid and when you apply the function of subtraction they
would give the amount they owe. So that would be the result.
She appeared to be thinking that a function is a mathematical operation. Inputs must be
numbers, according to DB. She exhibited a rather elementary view of functions. In Task
16, she again was unable to consider functions as arbitrary pairings:
functions. If not, explain why not:{(x,y): x, y in the set of all rational numbers}.
DB: [Pause] Um . . . [Pause] I wrote all possible values for x and y give rational
numbers. Which could go on forever. Um . . . I don’t really see a function
there. I mean a function is that they’re rational numbers. But they’re, I mean
you would plug in but you don’t get a different result so I don’t see it.
Finally, consider her responses to Tasks 3 and 4 on the initial questionnaire:
TASK 3: If possible, describe the following situation using one or more functions.
If not, explain why not: y4 = x2.
TASK 4: y4 = x3
98
I: In number three I noticed you solved for x. Is there any reason you solved
for x and not for y? And I noticed you did that in number 4 as well. Any
particular reason you did that?
DB: Well the first one I think I solved for x because it’s to me easier to find the
square root of something. Whereas if I solved for y you would have to
find the quadratic root [DB is exhibiting an error here, confusing fourth root
with quadratic root.] That is why I chose to solve for x. And I guess I still
find a cubic root would be easier again. Maybe not.
I: So how would you describe these responses then?
DB: Functions when the two variables are dependent on each other.
I: So each one is dependent on the other, or is one dependent on the other and
not the other way around, or doesn’t it matter or what?
DB: I would think one is dependent on the other. I would say y really should be
the independent variable in both of them.
DB exhibited some flexibility here, though she still ignored the uniqueness
criterion for functions. Noteworthy also is the fact that she actually solved the equations
for x. Dubinsky and Harel (1992) claimed that if a student actually needs to solve the
equation, the concept is at the action level.
As evidenced above, DB’s general understanding of the concept of function
appears to be an early process level. What follows is an analysis of her understanding of
graphical representations of functions.

99
Understanding of Graphical Representations
Note: Because of a scheduling conflict, DB’s first interview occurred AFTER she
had completed the first worksheet, which concerned the first known graphs of two
varying quantities. The following analysis, then, is based on her answers on the written
questionnaire (completed before the unit on functions) and the interview (completed
shortly after her work on Worksheet 1). I had asked her to “say what you were thinking
right off the bat, or why you left it blank and then tell me if you had any additional
insights after you had worked through the worksheet.”
DB began the unit with a very shaky understanding of graphs of varying
quantities. In two of the five scenarios in the graphing section on the initial questionnaire,
she interpreted the graph as an actual picture of the scenario, a misconception frequently
noted in the literature (Bell & Janvier, 1981; Clement, 1989; Monk 1992; Sierpinska,
1992). She answered only one of the five correctly.
GRAPHING TASK 3a: The graph below represents speed versus time for two
cars. (Assume the cars start from the same position and are traveling in the same
direction.) State the relationship between the position of car A and car B at t = 1
hr. Explain.
100
Speed
t=0 t=1
Time in hours
Figure 5. Graphing Task 3 showing speed vs. time for two cars
Note. From “An Investigation of the Function Concept” by M. Carlson (1998).
Research in Collegiate Mathematics Education, III, 7, 147. Copyright 1998 by American
Mathematical Society. Reprinted with permission.
DB indicated that the “two cars meet at the same point at t = 1 hr. because this is
where their paths cross.” When asked about the relative position of the two cars during
the time interval between t = .75 hr and t = 1 hr. (i.e., is one car pulling away from the
other?), she answered, “At t = .75 hr. car A is driving towards car B and car B is also
driving towards car A.” Interestingly, she also noted that “the speed of the two cars at t =
1 hr. is also equal because both paths correspond to the same speed.” She appeared to be
confounding the notions of speed and position.
Also interesting is her response to Graphing Task 5.
GRAPHING TASK 5: Can you indicate, from the graph below, how many bends
there are along the track on which the car was driven?
101
Speed of a racing car along a 3-km track (during second lap)
Speed (km/hr)
200
180
160
140
120
100
80
60
40
20
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3
Distance along the track

(km)
Figure 6. Graphing Task 5 showing speed vs. distance along a race track
From “Use of Situations in Mathematics Education” by C. Janvier (1981).
Educational Studies in Mathematics, 12(1), 115, graph 1.1. Copyright 1981 By D. Reidel.
Reprinted with kind permission from Springer Science and Business Media.
DB replied “9.” She counted the number of times the graph itself bent, thus
interpreting the graph as a picture of the actual track. In actuality, the car slowed down
only three times and thus there are three bends in the track.
DB had difficulty with other graphing situations as well. She left Graphing Task 1
blank. It showed a position versus time graph of two objects moving at different, but
constant rates and asked students to compare the speeds of the two objects at a given
instant.
GRAPHING TASK 1: Consider the graph below. At the instant t = 2 seconds, is
the speed of object A greater than, less than, or equal to the speed of object B?
102
position (cm) B
2 4 6 8 time (sec)
Figure 7. Graphing Task 1 showing position vs. time of two objects
Note. From “Students’ Understanding of a Function Given by a Physical Model”
by S. Monk (1992). The Concept of Function. Aspects of Epistemology and Pedagogy.
MAA Notes, 25, 175. Copyright 1992 by the Mathematical Association of America.
Adapted with permission.
The concept of slope as rate of change did not occur to her. When asked about it
during the interview, she said, “And I had a hard time understanding . . . when I first
looked at it what . . . it was actually showing. [Pause] Um . . . Which is why I left it
blank.”
In Graphing Task 4, when asked to make a sketch of height as a function of the
amount of water in a bulb-shaped bottle, she drew a straight line. She also incorrectly
answered Graphing Task 2c.
GRAPHING TASK 2c: This graph shows the speed in meters per second of a
cyclist over a 10-minute period. Does the cyclist travel further during the first five
minutes or during the last 5 minutes? Please explain your answer.

103
Speed4
2 4 6 8 10
Time (min)
Figure 8. Graphing Task 2 showing speed vs. time of a cyclist
Mathematical Society. Adapted with permission.
DB stated that the cyclist “travels further in the last 5 minutes because he is
traveling at a more constant rate of speed.”
Conjecturing why a post-undergraduate mathematics student has such an
elementary view of graphs of functions is difficult. The time lapse between her History of
Mathematics course and previous mathematics courses was small. She began during Fall
2003 semester and was a student in the History of Mathematics class during Spring 2006.
She took this class near the END of her undergraduate mathematics requirements before
entering a master’s program to earn her teaching license. Perhaps she rushed through the
assignment and/or had not thought about graphs for a while. Perhaps she was having a
bad day. Nonetheless, the data is there. Her responses indicate a very low level of
understanding.
104
Participant 2: CW
CW’s initial conception of function was at the action level. The following
analysis is based on her responses on the initial questionnaire and her comments in the
initial interview.
When asked, “What is a function?” she replied, “A function is an equation that
relates variables.” When replying to Task 1 in Part 2, she replied, “this function has
certain values (the set of all integers) to be inputted for x, and then that x is put into
3x + 2 and an answer is given, that is, x = 2 gives 3(2) + 2 = 8 .” In the initial interview, she
claimed
Functions usually are in the form f of x equals something or f of y usually and then
there’s another variable like f of x equals x squared plus 2 and then you have to
substitute in for that variable. So I guess . . . I usually work with functions as an
equation.
Dubinsky and Harel (1992) claimed that this emphasis on equations, numbers, and
evaluating expressions is typical of an action conception. While discussing Task 8, one of
the finite sequence questions, CW used function notation, though note again her
insistence on the need for an equation.
functions. If not, explain why not: {3n + n 2 : n in [1...100]} .

105
CW: Um . . . Yeah, I was thinking the same thing as in number 1 . . . or whatever
that one was where you had an equation and you could put in values for n
. . . And here it says your equation is 3n 2 + n and your n’s can be from 1 to
100. So I was just saying that f (n) = 3n2 + n and whatever n you put in
would give you your f (n) . So I said it was a function.
Her response to Task 5 also revealed her emphasis on function as equation.
functions. If not, explain why not: “Kent State’s Men’s basketball team makes it
to the Sweet Sixteen.”
CW claimed that this scenario is “not a function because it is just a statement.
There is no equation here.” She left the two other sequence problems, Tasks 6 and 18,
blank on the initial questionnaire. Recall that successful answers to these suggest a
process conception.
Dubinsky and Harel (1992) claimed that the graphs in Tasks 2, 13, and 19 have
potential to provide indications of one’s ability to use a process conception of function,
particularly Task 13, which requires the use of values on the vertical axis as the domain.
CW’s responses are revealing. According to her, the graph in Task 2 “could be the
function f ( x) = e x because it is exponential.” She sketched in a vertical line. Though this
is indeed a correct answer, she again showed her reliance on a known function formula,
an equation, to express the relationship. She was unable to find a function in Task 13,
claiming that it is “not a function because it does not pass the vertical line test; functions
106
have a 1-1 correspondence. So for every x there should be one y, but here, there are more
than one y value for each x.” Though she correctly used the vertical line test, she
appeared to be confusing the idea of one-to-one correspondence with the uniqueness
criterion for functions. She also did not have enough flexibility to consider the values on
the vertical axis as input. For Task 19, she wrote that these are “just points?”
When questioned further, she responded,
CW: Yeah. I would say just plotting points. Again, like I said there is no
relationship between them unless you connect them by a line.
I: But if I connected them by a line, I could make a function of them?
CW: Yeah. Because you could use the vertical line test. You could draw a
vertical line and it would not intersect that graph in more than one point.
I: But if I just have the discrete points like this I couldn’t do that?
CW: Yeah.
I: That’s not a function.

107
CW: Right, because you’re not . . . that’s just plotting points. You’re not . . . you
can’t really see an actual relationship between any of those points. It’s
almost like having a list. We’re just saying, “okay put them on a graph.”
And . . . like I said before, you have to have a pattern or some sort of
relationship. So I think it’s not a function.
CW’s insistence on requiring an equation, a pattern, or a relationship for a
function is consistent throughout her work on the initial questionnaire and in the initial
interview. Note her response to Task 16: “No, not a function because there is no equation
relating x and y, you are just inputting values and there is the possibility of (3, 2) and
(3,7) which breaks the 1-1 correspondence rule.” In Task 17, she was unable to see the
output as a point, claiming there was no function there. Finding a function in this
situation requires sophisticated and flexible thinking about functions; CW apparently
never had the opportunity to experience anything other than function as formula.
CW’s answer to Task 12 further confirmed her action view of functions: “again
just a list with no pattern. So cannot write a function to represent total owed or anything.”
Recall that Task 12 gives a table pairing names with the amount of club dues owed.
functions. If not, explain why not: The Club Members Dues Status.
CW: Yeah, . . . if you’d have the previous one plus five, you would have a
function. You could put it in the previous one to get out the new value. So
you’re putting in for an x, get out a y. That . . . then it could be a function.
But here it’s just a list.

108
I: There has to be . . . again you are saying there has to be a pattern?
CW: Yeah.
Her discussion about Task 10 is similar.
functions. If not, explain why not: A record of all MAC women’s basketball
teams giving, for the 2003-2004 season each team’s field goal shooting
percentage at home and its field goal shooting percentage away.
I: Could you just explain your answer?
CW: Okay, well if you are just are getting the data, it’s a list, it’s not a function.
You’re not looking . . . but you can look closer at the data and see that
there’s a pattern and you can use a function to represent that pattern. But
this statement is general, it’s just a list.
I: What happens if there was a list and there wasn’t a pattern?
CW: Then you couldn’t use a function to represent it.
Again and again, CW insisted upon a recognizable pattern or equation.
According to Dubinsky and Harel (1992), if a student insists upon solving an
equation in two variables for one in terms of the other, she is probably displaying an
action conception of function. CW did this in Tasks 3 and 4, but ignored the necessity of
restricting the co-domain. In Task 11, she tried to solve for y but could not, so she simply
stated that this equation “can be solved for x and y to form a function (need variables on
opposite sides: y = something( x) ).” Here she showed some flexibility since she was able to
discuss the process even though she was unable to actually perform it.
109
In conclusion, the evidence suggests that CW was operating at an action level,
while exhibiting some instances of an early process conception of function. She was
confused about the notion of one-to-one correspondence and how it relates to the
definition of functions, and held an apparent rigid view of function as formula.
CW appeared to have a fair understanding of graphs of functions, as evidenced by
her correct answers to most of the graphing tasks. She interpreted slope of a line in a
position versus time graph as speed in Graphing Task 1 and correctly identified the
interval of longest distance in Graphing Task 2. Interesting to note in this response,
however, was her inability to consider area under the curve as total distance traveled.
GRAPHING TASK 2: This graph shows the speed in meters per second of a
cyclist over a 10-minute period. Does the cyclist travel further during the first 5
Speed 4
2 4 6 8 10
Time (min)

110
I said because they were going the fastest between those times so I said they
traveled the farthest because for a longer time period they’re going slower. And
it’s only between 2 and 4. I mean, they’re slowing down like drastically between
2 and 4 seconds and they’re going a lot slower even between 4 and 10 seconds. So
between 0 and . . . like 2 seconds they’re going really fast so I . . . I was thinking
that they were going the farthest between that time period.
This response was consistent with that in Graphing Task 3. Her inability to consider area
under the curve as distance was again evident, as was her confusion with determining the
relationship of car A and car B at t = 1.
GRAPHING TASK 3: The graph below represents speed versus time for two
direction.)
Speed
t=0 t=1
Time in hours

111
CW: Um . . . I looked . . . The first time I looked at this I was thinking . . . the
same position. Then I realized that here is more . . . um . . . It depends . . .
well B is . . . it depends on what their slopes are. Well, if . . . I was confused
on this part. I know they’re not in the same position there. So . . . I think . . .
[pause].
I: How do you know that?
CW: Well it depends . . . if A and B have the same slope then yes, they’re at the
same position. But, I couldn’t tell from what this picture was whether or not
they have the same slope. Like A has a greater slope than B then A is going
to be farther along than B at t = 1 hour.
I: Uh Huh . . .
CW: But if B has a . . . then vice versa. But here I was just getting myself
confused on whether or not they have the same slope or not.
I: So are you talking about, if we drew, like a tangent line at that point and
you want to compare those slopes? Is that what you are saying?
CW: Yeah. You can do that.
I: And we did that . . . we did that . . . [pointing to tangent lines]
CW: Then we know that B has a greater slope than A from that picture so that B
is going to be farther along.
I: B will be farther along.
CW: Yeah. Yeah.

112
After further discussion, CW still believed car B will be further along, but is not
confident in her answer. It is as if she knows there is something more, but just cannot put
her finger on it. When asked to compare the relative positions of the two cars during the
time interval t = .75 and t = 1 , she left it blank on the written questionnaire, but in the
interview claimed, “Is one car pulling away from the other? . . . I’d say B is because its
slope is greater in that time interval.” This, of course, is incorrect.
She left the bottle problem blank on the initial questionnaire, but had no difficulty
correctly identifying the 3 bends in the track in Graphing Task 9.
Participant 3: MJ
Evidence suggests that MJ had an emerging process conception of function prior
to the unit on the history of functions. His understanding of graphs was sound, but had
some holes. The following analysis is based on his work on the initial questionnaire and
the interview during which he discussed his reasoning.
Evidence of a process conception. MJ showed evidence of a process conception
of function, as evidenced by the excerpts that follow.
functions. If not, explain why not: The string “ABCDEFG.”
MJ readily found a function in this task. Recall that this problem, in itself, does
not suggest that construction, so it must therefore come within the student (Dubinsky &
Harel, 1992). He wrote, “Yes, each letter can be given a position (1,2,3,4,5,6) and each
position has a unique letter ∴ one to one and onto.” He explained his thinking as follows.
113
MJ: Again, I was assigning each variable, each letter a number and then that
number corresponds to a position. Therefore, you can have a unique
relationship between the numbers and the letters.
I: OK, so the input would be?
MJ: If I gave you a number and then the uh . . . the thing would be . . . that’s the
position it’s at and it’s going to point at a letter.
I: So what happens if I give you this string? So, if I gave you, if this was a
different string, X, Y, Z, P, Q could you identify a function there? Or is it
the same function or a different function or is there not a function?
MJ: I’m not sure. Um . . . I would think that really you could do that. It really
wouldn’t matter what the letters are. Um . . . [pause]. I mean, that’s what I
think. I’m kind of just going off like a computer program that can generate
out a relationship between those numbers and assign them, or those letters,
and assign them a value and you could always determine what you’re
working with.
I: Sure.
MJ: I mean . . . I was looking at the computer.
His reference to computers is interesting. MJ had an associate degree in electronic
technology and had worked as an engineering assistant before returning to school,
troubleshooting problems with circuit boards at the component level. His previous work
with computers may have enriched his notion of function. Research supports this
phenomenon (Breidenbach et al., 1992).

114
A participant’s ability to discuss the manipulation of an algebraic expression
without actually performing those manipulations is evidence of a process conception. MJ
illustrated this ability in his response to Task 11, albeit after some struggling.
functions. If not, explain why not.
I: What about #11? I noticed you tried to start doing something there. I’m not
so sure what you were doing.
MJ: What I was . . . my thought process on it was, I could solve for y and uh . . .
to me from um just the calculus classes I know that what I am working in is
a uh three dimensional thing, with separate xy’s. So I mean, it’s a . . . it’s a
. . . [pause]. I mean I know that you have to solve that by maybe doing the
derivative of x and then do the derivative with respect to y or pull out a y.
Just looking at it right now I kinda noticed like oh that’s the log of y, that’s
what kind of messed me up here. [Laughs] It isn’t like this is going to be the
easiest thing to separate into two variables.
I: So you are just trying to separate the y by itself, is that what you’re trying to
do?
MJ: Yeah, but you might be able to deal with it somehow with it in there, uh . . .
I’m not sure how you would actually do that.
I: But looking at it as it is now do you think there is a function there?
MJ: Yeah, I think it’s a function.
I: And what would be the input, would you say?

115
MJ: Um . . . well the only thing that would really limit your input would be since
you have a . . . a . . . square root of x that would limit your . . . your domain.
And if the x is your independent to your positive numbers then a y could
still go . . . whatever, right?
Weaknesses in his process conception. Though initially indicating that a function
is a mapping, in the interview he concluded that a function is different than a mapping; a
function has an algebraic expression, a formula, whereas a mapping is a generalized
procedure:
MJ: Okay, I said a function is to map each item in a set A to a unique item in set
B and my, uh, thought process there is that, um, you have two separate sets
that may contain different elements but there is a rule that you would follow
that would allow you to select an item in the first set and following that set
of rules it would point to a unique item in set B.
I: Now you mentioned the rule in your discussion now and um . . . what type
of rule are you thinking? Does there have to be a set rule in order for this
function to occur?
MJ: Um . . . Yes, it has to be a . . . [pause], my thought of it, a . . . my thought of
it as a . . . uh, basically a procedure . . . um either numerically or, um . . .
algebraically that would be written. That when you implemented that
procedure it would carry you to an element in set B.
Notice that he suggested an algebraic procedure, an indication of an action conception.
When questioned further, he concluded that functions are different than mappings.
116
I: Okay, so there has to be some type of a rule or formula . . . is that what
you’re thinking?
MJ: Yeah, a formula, maybe a . . . a formula . . . that’s what I’m thinking on
that.
I: But yet you call that a mapping opposed to a formula is there any . . .
MJ: Okay . . . a function . . . [pause], I mean I say mapping because I see
mapping as a . . . more of a procedural approach to following how to get
from set A, from an element in set A, and put it into a unique element into B.
But um . . . A function I see as more of algebraic expression.
When pressed further, however, MJ admitted that a function might exist even if a
formula was not present, though he was unable to come up with his own example. When
the interviewer developed one, he did recognize a function in the situation.
I: Okay and you have here . . . Is there anything else you were thinking, were
you thinking of anything else in number two describe the ways a function
can be represented. Do you have anything else to say about that?
MJ: Well, I know it could be represented graphically and in tables, but I also
know that it could be like a description that relates how two items are
related. So you could have, um . . . I can’t think of one right now, but . . . it
could be a . . . [pause]. I don’t know just two, maybe a sentence or
something that tells you the relationship between two sets of items.
I: Uh huh . . . A verbal description?
MJ: Yeah verbal, a verbal . . .

117
I: So that would be a function then?
MJ: Yeah, that’s like saying here, mapping is more like a procedural thing, it
shows you the relationship—it could be a verbal.
I: So if I said something like here’s a bunch of people and here’s a bunch of
people and I say “is the brother of,” is that the type of verbal thing you
mean?
MJ: Yes.
I: Or are you talking about an algebraic formula type of . . .?
MJ: No I’m saying like that, that could be a function. If there’s a unique
relationship that’s pointing from one brother that characterizes their
relationship to another individual, a parent being a daughter . . ., there’s a
unique relationship between the two.
I: So it really doesn’t have to be algebraic formula, just so there’s a regular
relationship?
MJ: Yes.
MJ’s conception of function, then, does not require a formula, only a regular
relationship. When identifying a function in Task 2, however, he immediately suggested
a formula, not a generalized “input-unique output” relationship.

118
Figure 12. Graph for Task 2 showing an arbitrary graph with no labels
I: What about number two, I noticed you have its exponential, and um . . . do
you remember what you were thinking as you were answering that
question?
MJ: Um, I was kind of . . . I was thinking . . . I was debating between it being a
logarithm or uh . . ., because I’m not sure how they sweep.
I: Is there a way you can figure that out?
MJ: Um . . . Well I could actually do a plot, 2 x and just say . . . it’s zero and then
see how it’s actually climbing in effect with zero. I . . . it would actually go
negative so actually that might be a logarithm. [Laughs] Now that I’m
thinking about it. [Laughs more]
I: [Laughs too] Okay, but you were just starting to say how you decided . . .
what made you decide to put the answer that you did? Do you remember?
MJ: I did then actually try to graph and see if I could get the points that would
be in the relationship to what was there. Now I’m saying if you put a
negative x here it just goes smaller and smaller and approach zero.
119
His insistence on a definable relationship was evidenced in his response to Task 10 as
well.
functions. If not, explain why not: A record of all MAC women’s basketball
teams giving, for the 2003-2004 season each team’s field goal shooting
percentage at home and its field goal shooting percentage away.
MJ: Well . . . I just said because I mean for some reason I said that there is no
relationship they could really uh develop because their distinct events that
don’t . . . [pause] they are not really related to each other. [Pause]
I: So . . . nothing there, huh?
MJ: Not for me. [Laughs] Maybe there is, but not for me. [Laughs]
He was similarly unable to arbitrarily match names with numbers in Task 12 on the initial
questionnaire, the club members’ dues status problem. Recall that Dubinsky and Harel
(1992) claimed that a student who insists upon a rule relating the first number to the
second and/or cannot construct a process as the act of going from an item in the first
column to one in the second is “probably displaying an action conception of function” (p.
93).
functions. If not, explain why not: The club members’ dues status.
I: Oh, I noticed you left number 12 blank, any particular reason? Do you
remember what you were thinking there?
MJ: [Laughs] No . . .
120
I: It looks like you started to write something, and then I can’t tell what you . . .
MJ: What I was trying to think is that any little, a . . . [pause]. I’m just not
catching a . . . a relationship, I think well can you really relate the names to
these numbers, and it’s and um . . . I was starting to try to graph it a little bit
and see if there was any kind of . . . relationship and I didn’t see anything.
So I said it’s not . . . I just don’t think it’s a function or if there is I’m not
aware of it.
That he is unable to find a function in this situation, but found one easily in the string
problem [Task 6] is interesting to this researcher. Perhaps he worked with arbitrary
sequences on compute—as inputs and outputs, but nothing in his computer experience
related to tables as inputs and outputs.
MJ was also unable to consider a Boolean function. Note his response to Task 15.
functions. If not, explain why not: x 2 + 3x+ 2 = 0 .
I: What about 15?
MJ: That’s a function.
I: Okay, you did not have anything written there. I was wondering what you
were thinking.
MJ: Well actually that would just be a uh . . . just two numbers, so I mean you
would have an x and then you get out a specific number. I mean I guess
there is only going to be maybe two numbers that could actually solve that.
I: So do you think you would have a function, or not then?

121
MJ: But it’s only one. [Laughs]
I: [Laughs too] It’s getting late in the day . . .
MJ: Um . . . No I don’t think it’s a function. There is only one solution to it, you
know what I’m saying, it’s not . . . it’s not really getting you . . . it’s just
telling you that there’s a . . . that just might be a parabola. [Pause] Now I’m
trying to think . . .
I: Take your time.
MJ: (Long Pause) No, I don’t think . . . I think its just one solution to it.
I: So since there is one solution to it there is no function at all?
MJ: That’s what I’m saying, yeah.
Notice in the following excerpt, MJ was unable to consider Boolean outputs, even
when the interviewer strongly suggested it.
I: Okay. Eighteen—you left blank any thoughts about that one?
MJ: Um . . . [pause]. No, not really. [Long pause] Then you would be having a
number. I mean to actually make that . . . I guess it would be just a unique
set of numbers that could actually, you’d have to go above a certain a . . . to
point at a certain set of numbers . . . I guess it would point at a certain set of
numbers. [Pause]
I: So it would point at a certain set of numbers when this was true, is that what
you’re saying?
122
MJ: Yeah
I: And then, what would be the function there?
MJ: [Laughs] 2 n − n 2 − 3n is greater than uh . . . zero.
I: So when that is true, then there is a function and the function is that
formula; is that what you are saying?
MJ: Yes. That is what I would think the function would do.
I: Anything more about that one?
MJ: No.
He appeared to be unaware of the possibility of Boolean outputs. His response to Task 16
provides another example of his inability to make an arbitrary pairing.
functions. If not, explain why not: {(x,y): x, y in the set of all rational numbers}.
I: Um . . . Sixteen says, the set of all x’s and y’s such that x and y is in a set of
all rational numbers, could there be a function there?
MJ: Again, I just looked at it as being a set of coordinates . . . in . . . in the
coordinate plane so um . . . I don’t think that there would really would be a
function, I’m not sure how . . . how they relate . . . how they’re related. I
mean you just pick arbitrary sets of numbers you could put it in there but
without a third variable, particularly a number, I just didn’t see anything,
what the relationship would be.
He continued to insist upon a regular relationship.

123
Recall that Dubinsky and Harel (1992) claimed that one’s ability to find a
function in a discrete set of points indicates a process conception. MJ is unable to do that,
as evidenced below.
What I see, that you get a series of different points, and I really didn’t see that it
could be a function or anything because without, and this could just be arbitrary
points in a plane, so I didn’t see any type of reason why there would be a
relationship in the interval.
In conclusion, though MJ was able to find functions in strings and discuss the
manipulation of an algebraic expression without actually performing those manipulations,
his process conception is weak. His “function as formula” conception was strong. He was
unable to find a function in either a table or a discrete set of points, and was unable to
consider Boolean outputs. His conception of function appeared to be an early process
one.
124
MJ appears to have a good though not perfect understanding of graphs as
indicated by his written explanations on the initial questionnaire. Because of his time
constraints, MJ was unable to verbally explain his answers on the graphing portion of the
questionnaire during the initial interview.
As evidenced by his answers to Graphing Tasks 2 and 3a, he understood that
given a speed versus time graph, the area under the curve represents the distance traveled.
Speed4
2 4 6 8 10
Time (min)

125
When asked if the cyclist travels further during the first five minutes or during the
last, he indicated the first “because the area under the curve is greater during that time
period.” He was similarly accurate in his response to Graphing Task 3a.
GRAPHING TASK 3a: The graph below represents speed vs. time for two cars.
(Assume the cars start from the same position and are traveling in the same
hr. Explain.
Speed
A
B
t=0 t=1
Time in hours
While answering this question, he wrote, “A is ahead of B because the area under
the curve is greater.” He also had no trouble comparing speeds of two objects or of a
single object at two different times on the same graph. Similarly, he accurately compared
acceleration of the two cars: “B is accelerating faster than A because the slope of the
tangent line is greater.” He also recognized changes in rates of change, as evidenced by

126
the changes in concavity he sketched while making a graph of height as a function of the
amount of water in a bottle.
MJ exhibited some misconceptions on his initial questionnaire, however. When
asked in Graphing Task 3 above about the relative position of the two cars during the
time interval between t = .75 hr. and t = 1 hr., he said, “B is accelerating faster than A
therefore B is pulling closer to A.” Though he previously demonstrated an understanding
of the area curve as representing distance, he did not apply that knowledge in this
instance. He interpreted a larger acceleration as a faster speed. He also interpreted the
height of a line, that is, its output value, rather than its slope, as determining speed in a
position versus time graph.
GRAPHING TASK 1: Consider the graph below. At the instant t = 2 seconds, is the
speed of object A greater than, less than, or equal to the speed of object B? Explain.
position (cm) B
2 4 6 8 time (sec)
Figure 16. Graphing Task 1 showing position vs. time for two objects
Note. From “Students’ Understanding of a Function Given by a Physical Model
by S. Monk (1992). The concept of function. Aspects of epistemology and pedagogy,

127
He sketched the points (2, y1 ) on line B and (2, y2 ) on line A, then claimed “A is greater
than B because y1 < y2 ,”
In conclusion, though MJ exhibited strong understanding of graphical
representations in some instances, his understanding was weak in others. The error in
Graphing Task 1 may have simply been a careless oversight, with MJ not realizing that
the dependent variable was position, not speed. The understanding needed to answer
Graphing Task 3d correctly is more subtle, however, and unlikely due to oversight.
Participant 4: BG
Evidence suggests that BG had a strong process conception of function as he
began the unit on the history of the concept of function. His understanding of graphs was
well-developed; he showed none of the common misconceptions either in his written
responses or his verbal explanations.
Evidence of a process conception. BG’s written response to the question, “What is
a function?” stated that “a function is an equation in terms of one or more variables such
that any given set of variables will yield a specific output.” When prompted to further
explain his thinking in the interview, however, he said,
Okay, uh . . . well . . . this question threw me for a loop; actually, uh . . . because I
just never had to . . . really think of what a function is, just . . . going back to 11th
grade mathematics class. I took Elementary Functions and it explained some
things. The equation thing I realized later even when I was taking the, . . . the uh
128
. . . the worksheet that probably, it doesn’t necessarily need to be an equation, um
. . . but it is whenever you have, uh . . . [pause] one thing leading to another, like,
uh . . . how should I say this? [Pause] For a given . . . situation there is only going
to be one outcome. Uh . . . so if, if uh . . . [pause] if, if you have, a set of a . . .
parameters I guess, you’re not gonna have more than one solution to that. It is not
going to lead you to this or this or anything, it’s . . . these things lead to one
specific answer. So that is what I was trying to get at, I . . . that’s what I meant by
specific result, leading to one solution to your problem.
He is discussing a generalized process, realizing that an equation is not necessary. When
discussing the graph in Task 2, he wrote f (x ) = e x , but in the interview, he reiterated that
a formula is not necessary.
If not explain why not.
I: And about the graph, . . . if you didn’t recognize so quickly that there was
an equation for that graph, if I had an arbitrary graph whose formula was
not so evident, would there . . . could there still be a function represented?

129
BG: Oh yeah, there could be a function represented. I would have described it
um . . . oh gosh . . . I would have described it in terms of, well for the values
uh . . . [pause], along the x-axis if it comes out to be the other . . . on the y-
axis there. So I’d just, yeah . . . [pause].
I: So it is just sort of, well it’s nice if there’s a formula but it’s not necessary.
BG: Right, it doesn’t have to be a formula.
His responses and explanations to several of the other questions further indicate
the strength of his process conception. Though he did not find a function in Task 3—he
did not think of restricting the domain—he showed flexibility in his thinking. He
considered both x and y to be inputs, something his peers did not think to do. He found a
function in Task 4 by considering y as the input—not the convention.
BG recognized a function in Task 6, the string “ABCDEFG.” Recall that this
representation does not suggest a function, so the construction came from BG himself.
If not, explain why not: The string “ABCDEFG.”
I: What about number 6?
BG: Oh . . . the string ABCDE . . . uh . . . I have something here, but I’m not
sure if the terminology is correct though.
I: That’s okay. That’s really why I want to talk to you about it so that I
understand what you meant by your terminology.
BG: Uh. I was . . . if . . . if . . . the way that I visualize this: I had this graph.
Let’s say the horizontal axis was a . . . let me make sure I understand . . .
130
[pause] now say that uh . . ., the horizontal axis for this was . . . a to the . . .
or . . . rather a sub n [an ] and the uh, . . . vertical axis was the alphabet, . . .
I: Oh I see . . .
BG: Um, you have . . . you know the first . . . n . . . n sub . . ., a sub 1 [a1], would
be A, a sub 2 would be B, a sub 3 would be C, and so on and so on.
I: Now would it matter if I did something like this? [writes “PQRADF”]
BG: Uhhh . . . No actually . . . I uh . . . you could . . . because they’re all the
same. Now if you have like, p, q, p . . . oh no you can have it. I think it was
just because of the pattern . . . [inaudible].
I: Could you do it?
BG: Yes you could do the same thing. [Pause] Now if you had uh like . . . “p or
a” here, and then q, p, then it wouldn’t be.
Recall that if a student needs to actually solve one equation in terms of another,
she is exhibiting an action conception of function. While discussing Task 11 on the initial
questionnaire, BG attempted to do this; yet when unable to do so, he still recognized the
existence of a function.
functions. If not, explain why not: 2 x 3 y − x log y = 2 .
I: What about this one, number 11?
BG: [Laughs] I . . . I tried for a little while to solve for one of the variables and I
couldn’t do it to my satisfaction. Looking at it now I probably could. [Long
pause] I’m not sure if I . . . well, hold on. I could’ve, I could have done it.
131
I: How would you do it?
BG: [Long pause] Wait a minute, never mind, I don’t think I could. [Inaudible]
I: So does that mean that there’s not a function there now?
BG: Not necessarily. There could be. I mean if you plotted that out on a graph,
um . . . it might come out so that one of the variables doesn’t duplicate. Or
. . . it could be a circle . . . I don’t know . . . All sorts of things it could be. I
am going to guess that it is a function.
I: Which one would be, like you were talking that the input would be n here.
BG: I am going to guess that it is a function of x . . . [Pause]. Yeah . . . that’s it.
BG similarly had no difficulty identifying a function in Task 12, the club members’ dues
status.
functions. If not, explain why not: The club member’s dues status.
I: Number 12 is a chart with the club member’s due status and can you
describe your function there please? And why it is a function?
BG: (Laughs). Uh . . . Oh, oh . . . It’s . . . it’s a function, um . . . when you plug
in a name and one particular amount that they owe comes out. So, . . . that’s
as simple as it gets, I think. You have a chart outlining the function, right
there. See? One amount owed.
Note also the flexibility in his thinking when discovering a function in Task 13.

132
BG: 13. I see there that you’ve got a function of y. you have to turn the page a
little bit, 90 degrees. Turn this 90 degrees . . . and then you’ve got a
function.
I: In what way?
BG: Well, I’m not sure how . . . The only reason I was worried about that one is
because I know uh . . . convention has that you put the uh . . . the
independent variable on the horizontal axis, and that was the only thing that
I was worried about but I decided to . . . push my . . . luck. Turn it and see
what happens.
Though BG initially left Task 18 blank, he constructed a Boolean function during the
interview.
I: What about 18? Do you remember what you are thinking as you were
thinking about 18?

133
BG: I think that the inequality threw me off again. I didn’t quite know what to
make of it. Um . . . I couldn’t figure out which numbers actually satisfy it.
You know, for like for n in from 1-100, 1 is . . . I guess, you know . . . what
the outcome could be is just whether or not it is true or false . . . So, right
. . . going back several . . .
I: That’s okay.
BG: Going back several uh . . . to the sentence about Kent State Basketball. That
just comes out as false. So the input is that [pointing to the sentence].
I: Oh I see.
BG: And the output is just false. But I can do that, I can say values of n that
satisfy that equation, uh . . . is true, for values that don’t satisfy the
inequality it would be false. And that would be your output.
I: So that would be a function then?
BG: That would be a function, there we go.
Finding a function in Task 19 was similarly easy for him. Recall that Dubinsky and Harel
(1992) considered this problem to be a good indicator of a process conception of
function.
functions. If not, explain why not. [The graph of discrete points is given.]
I: What about 19?
BG: [Laughs]
134
I: Okay, so you sort of explained it, I mean if you could just sort of explain
what you were thinking there. There is nothing wrong with what you did.
BG: This is a function of x, such that when x = 0, y = 0. I saw that point there.
For the rest of them I just kind of . . . tried to label where they were on the
graph and then for any value of x, there is only one value of y on this graph
so . . . I just kind of plugged them in.
Weaknesses in his process conception. BG was unable to discern a function in
Task 15, which could be considered a function with Boolean outputs.
I: What about 15? What were you thinking there?
BG: Well . . . it’s . . . it’s an equation. There is only one variable. You’ve got
x = 2 or 1 . Uh . . . so it’s not, I mean, I guess you could say, for some
reason I was thinking, it’s just a sentence, it’s not . . . you’re not really
solving for anything. You just have two values for x.
I: Okay
BG: I don’t know if I’m still having a brain freeze . . . When I saw that, there
were two solutions for x, it’s not . . .; it’s the only variable that I see in
there, unless the whole thing can be it. And say f ( x) = x 2 + 3 x + 2 . But that
is not what it says, so I am just going to stick by that.

135
In conclusion, BG was able to discern a function in almost all of the situations
given in Part 2 of the initial questionnaire, including those that Dubinsky and Harel
(1992) considered significant indicators of a process conception of function. There was
little hesitation. He was a confident and flexible mathematical thinker.
BG’s understanding of graphs was similarly sound. He immediately recognized
the meaning of slope in Graphing Task 1.
This is Part Three now and considering the graph below it, at the instant t = 2
seconds is the speed of object A greater than, less than, or equal to the speed of
object B? And you wrote, “The speed of A is less than the speed of B for the
duration of the exercise. The speed is represented by the slope.” Do you have
anything else maybe that you want to add?
position (cm) B
2 4 6 8 time (sec)

136
BG: Let’s check it out. Yeah. [Long pause] You’ve got the change in position
which is less than the change in position for the same amount of time. So,
yeah, B’s going faster.
He understood the meaning of area under a curve, as evidenced by his response to
Graphing Task 2.
cyclist over a 10-minute period.
Speed4
2 4 6 8 10
Time (min)
I: Does the cyclist travel further during the first five minutes or during the last
5 minutes?
BG: The distance traveled would be the area under the curve. During the first
five minutes a much greater distance covered by the cyclist.
I: Is there anything else you want to say about that?

137
BG: No, I’m pretty satisfied with that answer.
When answering Graphing Task 3d, he was one of the few who gave an accurate
interpretation.
GRAPHING TASK 3d: The graph below represents speed versus time for two
direction.)
Speed
A
B
t=0 t=1
Time in hours
I: What about the relative position between the points t = .75 and t = 1 hour?
BG: Between t = .75 hours t = 1 hr. Oh yeah . . . Car A is continuing to pull
away from car B up until one hour. Uh . . . it’s going faster, it’s continuing
to go faster than car B so it’s gonna continue pulling away. After the hour
it’s gonna start catching up.

138
He drew an accurate sketch in Graphing Task 4 and correctly identified the three bends in
the track in Graphing Task 9.
Thus, BG had a sound grasp of the function concept and of graphs before
studying the history of functions. The above evidence suggests that his conception of
function was a strong process one.
Participant 5: MS
Evidence suggests that MS began the unit with an emerging process conception of
function, but a strong understanding of graphical representations of functions. The
following analysis is based on his responses to the initial questionnaire and his comments
during the interview.
Evidence of a process conception. MS’s responses to several tasks suggest that he
possesses at least the beginnings of a process conception.
TASK: What is a function?
I: Okay the first is “what is a function?” And if you can tell me what you
meant by the fact “the output then gives a clear result” [written on
questionnaire].
MS: Um . . . I started out in computer science so I was used to writing programs.
Usually you put some data in then you get something out. You get an
answer.
I: Interesting. So when you say clear result, um . . .
MS: You get something.

139
I: You get something? A specific answer..
MS: Usually. Usually it will be defined, unless it’s not…
I: And if you ran the program again, what would happen?
MS: You’d get the same result. If you used the same input.
MS was clearly discussing a process in a general way. He was not dependent
upon specific numbers to discuss the concept. He talked freely and easily, readily
comparing a function to a computer program. His comments brought to mind the research
of Breidenbach et al. (1992), which showed that writing computer programs facilitates a
process conception of function. Note in the following excerpt, he discussed function as
mapping, also in a general way.
TASK: Discuss the ways a function can be represented.
I: What were you thinking when you did that?
MS: Hmm . . . I was in topology and he just showed us that so I thought I’d use
that x goes to y and you can map it onto a different variable.
I: So you could map it onto different variables. Is that a representation of a
function?
MS: Sort of like a number line, it shows . . . [pause].
I: Like what . . .? Like what you have here on one of your later questions?
Like number . . . like that one, or this one here? Number 1 or number 7?
MS: Yeah.
140
Figure 22. MS’s sketches showing function as mapping
I: So you think a function can be described as a mapping.
MS: Yeah. If it’s got two variables. You put in one so that you can see the
numbers . . . you don’t need . . . the graph.
I: Okay. Any other examples of functions? I notice you have the f(x) here . . .
Any other examples you can think of? What does that f(x) mean?
MS: It pretty much anything that manipulates . . . something.
I: It manipulates the number?
MS: Like sine, cosine . . .
In the following scenario, MS again discussed the process in general, without
solving for x or y.
functions. If not, explain why not: 2 x3 y − x log y = 2 .
I: Number 11. I notice that you started writing something and then you left it
blank. What were you thinking when you wrote –

141
MS: Well, I was going to try and isolate y again, but then it just looked like a
pain, so I didn't do it. Guess that would probably be the way to do it . . . So I
was trying to do it – it's usually not so complicated.
I: So I'm not clear if you saw the function there or not.
MS: Uh . . . It could be, but if you wanted to do it with one variable, you would
just have to give a value to one of them. So I guess it could be a two
variable . . . you’d just have to put it in and solve . . . so it equals . . .
I: solve for . . .
MS: You have to find an x and y so it would work, and then that would be . . .
MS was convinced that a function existed here, even though he was unable to solve for
one variable in terms of the other. This ability is one indicator of a process conception.
While discussing the club member’s dues status problem, he also had no trouble
finding a function. Note in the following discussion he again used computer terminology
to discuss the function concept.
functions. If not, explain why not: The club members’ dues status. [A chart with
names and amount of dues owed is given.]
I: Okay 12. The club members’ dues status. What were you thinking here and
would you explain about it?
MS: Well, if you had a database, say, with all these names, they would have a
number attached. So if you entered a name, you would get a number.
I: So is there a function there?

142
MS: Yes.
His reference to a computer database was interesting. MS had been considering a major
in computer science and had taken programming courses. He often referred to his
programming experience as he discussed functions.
MS answered Task 13 readily, showing flexibility in thinking by easily
interchanging conventional inputs and outputs.
MS: That just looks like a cosine function. But it’s turned on its side. So if you
turn the paper, it would look correct.
I: Would you say that again?
MS: So if I turned the paper, it would look correct.
I: So if I turn the paper this way, then you would say there's a function. What
would the input be?
MS: Whatever. Whatever you want. [Long pause]

143
In the following task, MS was able to consider a number as an input and a point as an
output. He did not make the common mistake of invoking the vertical line test.
functions. If not, explain why not: {(1, 2 x) : x in [1...100]} .
MS: So for this one, I would assume that this is looking at a set of points where
you take x and put 1 and then 2 or 3 up to 100. So if you have a point 1 and
2x – there's no x in the first one so it's going to be always x = 1, so you’ll
basically have a line going from y = 2 to y = 200, . . . a set of points.
I: And that's the function, then?
MS: Sure. When you're inputting 1 through 100, and they would get a point.
I: You get a point or you get . . . okay.
MS: Yeah, so you get the coordinates.
Weaknesses in his process conception. Some evidence exhibits the weaknesses of
MS’s process conception, however. He was unable to consider a pair of parametric
equations as a single function of t; he often searched for a formula when discussing
scenarios, and suggested that numbers are needed as inputs and outputs. Then too, he was
unable to consider Boolean (true or false) outputs, even when he considered the truth or
falsity of a mathematical statement.
The following two excerpts show his inability to consider Boolean outputs.
functions. If not, explain why not: x 2 + 3 x + 2 = 0
I: What were you thinking when you answered 15?

144
MS: It was a quadratic and I factored it, and then I found the two solutions.
I: So is there a function there?
MS: Yes, but it only would be defined at those two numbers, −1 and . . .
I: Oh.
MS: Cuz otherwise it would be false.
I: I see. So you're setting the function equal to the left hand side, and then the
only thing you allow for input is negative 1 and negative 2? Is that what
you're saying?
MS: Right.
functions. If not, explain why not: {2n > n2 + 3n : n in [1...100]} .
MS: Um . . . [long pause] . . . It didn't state . . . it isn't true, as far as I can tell.
Because we take say, 2 2 it’s not greater than 22 + 6 . So it didn’t make any
sense . . . So it's not a function. It's never true when you’re . . . [pause].
I: If it were true would it be a function?
MS: Sure. Cuz then it would work. But it doesn't make sense. Because right now
it says the smaller number is bigger than a bigger number.
I: But there is no function there as it stands.
MS: Right.
MS seemed unable to find a Boolean function in a situation. The ability to do so is one
indicator of a process conception.

145
While discussing Task 5, MS came close to finding a function there, but was
unable to clarify his thoughts.
I: What were you thinking about #5? “Kent State’s Men’s basketball team
makes it to the Sweet Sixteen.” When I asked you to find a function there.
MS: Well, the only way there could do that is if they won, so if they did . . . IF
they won then could come back. But I wasn’t really sure what Kent State
...
I: You’re not really sure if Kent State was winning?
MS: Yeah. Once again I just made something up.
I: But how does “If x then y” [reading paper] relate to that statement?
MS: If the team wins, then they go . . .
I: Oh, so x would be . . .
MS: So they would have had to do something for x.
I: So like x would be “if the team wins” then they’d go to the Sweet 16, is that
what you’re saying?
MS: Right.
I: And what would be the function there?
MS: Uh . . . this wouldn’t really be a function but it would just be . . . [pause] not
sure how to say it . . . It would just be an if statement.
I: It would be an if-then statement?
MS: Right.
146
MS: [Pause] I guess it could be a function, but I’d come up with some
numbering system for the whole.
I: Like what type of numbering system?
MS: No idea.
I: No idea? [Laughs] So if I had to ask you, Can you describe it in general? A
numbering system . . . that would, would . . . satisfy your definition of
function.
MS: Well you could say a win is 1 a loss is 0.
I: And how would that be a function, then?
MS: If x is bigger than a certain number . . . then maybe you could . . .
something. So I guess if it’s single elimination . . . look at zeros and ones
. . . [pause].
I: This is interesting.
MS: I don’t know if this makes any sense.
I: Well, if the input is 1, then what would the output be then?
MS: Then, whatever you want, say, then go to the next bracket [inaudible]. It
would just be a [inaudible]. So it would probably wouldn’t be any easier.
I: So if I have to ask you now, is there a function in this situation?
MS: Probably not.
When discussing the Task 19, MS’s first tendency was to connect the dots and find a
formula.
147
I: I notice that you've connected the dots in No. 19. What were you thinking
when you answered that question?
MS: It looked like it could be minimum to maximum on some curve, so I just
made up a curve.
I: But if you weren't allowed to connect the dots—that was fine that you did,
but if you didn't connect the dots, would the dots themselves be—
MS: Yes, you could make a table then, just be a scatter plot–
I: But would it be a function if you did that?
MS: Sure, because you could probably make a function from the data.
I: How would you make a function from the data? What do you mean by that?
MS: You'd have to map it. I'm not even sure how you – [Pause] . . .
I: What do you mean by map? I mean, do I have to have something that says
do this, or can I just list them. What do you think?

148
MS: Well, you have the coordinates, so you could pick one of the axes and use
that as the independent variable. And see how the other one . . . changes . . .,
I mean . . . Like this point would be going up on the y plane. Through the y
axis. So you could probably find a pattern of how to change it, but it's not
necessarily true.
I: Okay. But do you need a pattern for them to have a function then, is that
what you're saying?
MS: Well, it could be predictable, because when you have a function, you know
all the outputs.
He appeared to be insisting on a regular relationship, almost a formula, here, though he
did acknowledge the existence of a function from a table. Recall that this problem is a
good indicator of a process conception of function. His ability to find a function here
would indicate that he has that conception, but his tendency to connect the dots and look
for a formula suggests that it may be weak.
In conclusion, evidence suggests that MS exhibited a process conception of
function before the beginning of the unit on the history of functions. A few minor
weaknesses exist.
MS’s understanding of graphical representations of functions is strong, as
evidenced by his insightful responses to the graphing questions both on the written
questionnaire and during the interview. None of the common misconceptions were
evident.
149
Participant 6: CS
CS’s initial conception of function appeared to be at the action level. The
following analysis is based on his responses on the initial questionnaire and his
comments during the second interview. Due to a scheduling conflict, he did not
participate in an initial interview. The reader might note the similarity between CS’s
responses and those of CW, who also evidenced an action conception of function. In
particular, note the repeated references to the need for a formula or equation.
On the initial questionnaire, he defined a function as “an equation that can have
numbers plugged in to form a graph.” He was unable to discuss alternate representations
of functions. For example, he left Task 12 (the club members’ dues status problem)
blank. He was unable to find a function in Task 13, which required a reversal of
conventional inputs and outputs on a coordinate system. He was also unable to consider a
string as a function on the initial questionnaire, leaving Task 6 blank. His response to
Task 7 was revealing, as were his comments about it during the second interview.
If not, explain why not: A swimmer starts from shore and swims to the other side
of Walden Pond.
I: On your original questionnaire, this was #7, um . . . you said you can’t
because you don’t know if the swimmer takes a straight line to the other
side. Or if he takes another route.

150
CS: Uh . . . I guess . . . like what I was thinking there . . . is that you, you
couldn’t make a function out of it because you have to know like, exactly
what type of direction he’s gonna go with it. Um . . . like, I was just
thinking that you . . . you couldn’t have an input and output because you
don’t know . . . what your input and output would be.
His comment, “ . . . what your input and output would be,” suggests a need for a specific
rule for determining inputs and outputs. He similarly referenced the need for an equation
in his response to Task 5.
functions. If not, explain why not: “Kent State’s Men’s basketball team makes it
to the Sweet Sixteen.”
I: Do you remember what you were thinking when you tried to answer this
one?
CS: Again, that seems like such an abstract idea for a function that . . . I couldn’t
come up with an equation that seems like it would make any sense for that.
Like . . . it’s really not giving any relationship.
If no clear relationship exists, no function exists, according to CS.
His discussion about Task 8, which he left blank on the initial questionnaire, is
strikingly similar.

151
CS responded, “I’m not 100% sure what I did there. Uh . . . I think I tried to make an
equation out of it.” Likewise, his response to Task 10 included an attempt to create two
equations giving percentages of shots made or shots missed.
In conclusion, CS showed dependence upon the notion of function as equation.
He left seven tasks blank on the initial questionnaire and those he did answer exhibited a
tendency to write an equation. This evidence suggests that CS has an action conception of
function.
CS’s understanding of graphical representations was similarly weak. Though he
interpreted some graphs correctly (Graphing Tasks 1, 2a, 2b, 3a, 3b, 3c), he showed
serious misconceptions in others. Interestingly, in Graphing Task 3a, he answered, “Car
A would be ahead of Car B because he has covered more area,” yet he is unable to apply
the notion of area under the curve to Graphing Task 2c.
GRAPHING TASK 2C: This graph shows the speed in meters per second of a
His response was, “The last five minutes because it takes him time to get up to speed then
he quickly slows; however the last five he is already at a speed of about 2 and increases.”
Similarly, he ignored the concept of area under the curve when answering Graphing Task
3d.
GRAPHING TASK 3D: The graph below represents speed vs. time for two cars.
152
direction.) What is the relative position of the two cars during the time interval
between t = .75 hr. and t = 1 hr. (i.e., is one car pulling away from the other?).
Explain.
Speed
A
B
t=0 t=1
Time in hours
CS thought that “Car B is pulling ahead because it is exponentially growing while the
other is leveling off.”
While answering Graphing Task 5, he made the classic error of mistaking a graph
as a picture of the scenario.
153
Speed (km/hr)
200
180
160
140
120
100
80
60
40
20
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3

(km)
Figure 26. Graphing Task 5 showing speed vs. distance along a race track
From “Use of Situations in Mathematics Education” by C. Janvier (1981).
Educational Studies in Mathematics, 12(1), 115, graph 1.1. Copyright 1981 By D. Reidel.
Like DB, CS replied “9.” He counted the number of times the graph itself bent, thus
interpreting the graph as a picture of the actual track. In actuality, the car slowed down
only three times and thus there are three bends in the track.
His sketch of the height of water versus amount in Graphing Task 4, the bottle
problem, showed no change in concavity. His initial concavity is incorrect and he showed
no indication of a constant rate of change for the neck of the bottle.

154
Figure 27. CS’s sketch for graphing Task 4
Thus CS exhibited a relatively weak conception of the notion of area under a curve and of
graphical representations of rate of change.
Summary of Participants’ Initial Conceptions
Quite a disparity exists between the initial function conceptions of the six
participants of this study. Using APOS analysis as a theoretical framework for describing
these differences, CW on the one hand held an apparent Action conception, depending on
equations and needing to perform calculations on numbers while discussing functions. At
the other extreme was BG who appeared to have a strong process conception as
evidenced by his ability to discern a function with Boolean (true-false) outputs before
beginning the unit on the history of the concept of function. The function conceptions of
the six participants are summarized Table 3.

155
Table 3
APOS Levels of Participants Before Studying the History of Functions
Action Emerging Process Process Object
DB *
CW *
MJ *
BG *
MS *
CS *
APOS Theory Applied to Understanding Graphs
The purpose of this section is to characterize a student’s understanding of graphs
of functions using APOS terminology. The researcher proposes the descriptions and
analysis of student answers that characterize each of the APOS levels. This
characterization is an extension of APOS theory and is the researcher’s innovation. The
section begins with a brief overview of the characterizations of other researchers.
Brief Summary of the Literature
Bell and Janvier (1981) identified several aspects of understanding found to be
significant in their work of students’ conceptions of graphs. Among these are:
1. The recognition of global features by a progression from point reading to
interval and to gradient
2. The measurement and comparison of intervals or gradients

156
3. The existence of various distracters. One type is the pictorial distracter,
which occurs when the student confuses the shape of the graph with that of
the hill being climbed or the race track being traversed. Another type is a
situational distracter, which occurs when the experience of the situation
interferes with attention to the meanings of the abstract features of the graphs.
(p. 37)
Monk (1992) differentiated between a pointwise analysis of a graph and an across-time
analysis. A pointwise question asks for values of a function for a specific input value. An
across-time analysis involves asking students to describe a pattern of change in the value
of a function that results from a pattern of change in the input values. Monk maintained
that students find across time questions more difficult than pointwise questions.
APOS Characterization
An APOS characterization of student understanding of graphical representations
builds upon the work of these researchers. The researcher proposes following
characterization, summarized in Table 4.

157
Table 4
APOS Characterizations for Understanding Graphical Representations of Functions
APOS Level Description of Level Evidence
Prefunction Student has little if any conception Is unable to see graph as a representation of a
of graphs of functions relationship between variables;
Sees a graph a picture of the event.
Prefunction Student has little if any conception Is unable to see graph as a representation of a
of graphs of functions relationship between variables;
Sees a graph a picture of the event.
Action “A repeatable mental or physical Emphasizes a reference to specific outputs and

manipulation of objects” (Dubinsky corresponding inputs on the graph. No reference to
& Harel, 1992, p. 251). A static global features of a graph.
conception; student needs to think Must measure intervals in order to compare them.
about it one step at a time.
Object When a student can perform an Subject considers area under a curve as significant.
action on a process, the authors say The act of finding the area is an action, which she
the student has encapsulated the is performing on the process represented by the
process as an object. graph. Can consider area under a graphical
representation of speed vs. time as distance
traveled, for example.
Recognizes a graph as one of several

representations of a function.
The Pre-Function Construct
A student holding this construct of graphs has little or no understanding of a graph
as a relationship between two varying quantities and will frequently mistake a graph as a
picture of an event. Student DB exemplified this conception.
GRAPHING TASK 3: The graph below represents speed vs. time for two cars.
158
hr. Explain.
Speed
A
B
t=0 t=1
Time in hours
In response to this task, she indicated that the “two cars meet at the same point at t = 1 hr.
because this is where their paths cross.” When asked about the relative position of the
two cars during the time interval between t = .75 hr and t = 1 hr. (i.e., is one car pulling
away from the other?), she answered, “At t = .75 hr. car A is driving towards car B and
car B is also driving towards car A.” Interestingly, she also noted that “the speed of the
two cars at t = 1 hr. is also equal because both paths correspond to the same speed.” She
appeared to be confounding the notions of speed and position.

159
there are along the track on which the car was driven? [The graph of speed vs.
distance along the racetrack is given.]
Her response to this problem was “9.” She counted the number of times the graph itself
bent, thus interpreting the graph as a picture of the actual track. In actuality, the car
slowed down only three times and thus there are three bends in the track.
We classified any misinterpretation of a graph as a pre-function conception of a
graphical representation.
The Action Construct
In APOS theory, a student who has an action understanding of functions sees an
algebraic expression as a command to calculate. Such a student can carry out a
transformation only by reacting to external cues (textbook directions, teacher suggestion,
etc.) that give exact details on what to do. This conception is like a recipe and they must
apply it to some number before it will produce anything. A student with an action
conception of graphical representations of functions refers to specific outputs and
corresponding inputs on the graph. She does not refer to the global features of a graph.
This construct is similar to Monk’s (1992) characterization of a point-wise analysis of
graphs. When trying to compare intervals on a graph, a student with this conception
would measure the intervals, again referring to specific inputs and outputs.
One example of this conception of graphs is MJ’s response to Task 1 in the
graphing section of the initial questionnaire.

160
position (cm) B
2 4 6 8 time (sec)
MJ interpreted the height of a line, that is, its output value, rather than its slope, as
determining speed in a position versus time graph. He sketched in the points (2, y1 ) on
line B and (2, y2 ) on line A, then claimed, “A is greater than B because y1 < y2 .” It is
interesting to note that though MJ exhibited a significant misconception here, his
discussions of other graphing tasks revealed a much better understanding of graphs.
The Process Construct
A process is an internal construction that performs the same transformation as the
action, but it is internal and hence under the control of the individual. She no longer
needs the external stimuli, no longer needs to actually evaluate an expression to think of
161
its result. A student with this conception of graphs of functions refers to the global
features of a graph, discussing its overall qualitative nature. She does not need to refer to
specific inputs or outputs for his analysis. CW’s response to Graphing Task 2 exemplified
this phenomenon.
minutes or during the last 5 minutes?
Speed4
2 4 6 8 10
Time (min)
CW: I said because they were going the fastest between those times so I said they
traveled the farthest . . . they’re slowing down like drastically between 2
and 4 seconds and they’re going a lot slower even between 4 and 10
seconds. Between 0 and . . . like 2 seconds they’re going really fast so I . . .
I was thinking that they were going the farthest between that time period.
162
Note that even though she referred to specific input intervals, she readily discussed the
qualitative features of the graph without referring to specific outputs.
Similarly, in the brief excerpts below BG, MS, and CS considered the qualitative
features of the graph without referring to specific inputs and outputs.
[Please see the graph for Graphing Task 1 on the previous page.]
BG: Let’s check it out. Yeah. [Long pause] You’ve got the change in position
which is less than the change in position for the same amount of time. So,
yeah, B’s going faster.
MS: A is traveling slower, its graph has a lower slope.
CS: Less because B is rising much faster than A.
The Object Construct
When one tries to perform an action on a process, APOS theorists say she has
encapsulated the process to an object. The recognition of area in a speed versus time
graph as distance traveled is applying an action (finding area) on the process (relating
speed to a given time) represented by that graph. Therefore, participants who are able to
compare distances traveled by comparing areas under the curves exemplify an object
conception of the graphical representation.
Two participants provided such evidence when discussing Graphing Tasks 2c and
3a on the initial questionnaire. Note the references to area in the following excerpts.
163
minutes or during the last 5 minutes?
Speed4
2 4 6 8 10
Time (min)
MJ: Because the area under the curve is greater during that time period.
BG: The distance traveled would be the area under the curve. During the first
five minutes a much greater distance covered by the cyclist.
Similarly, many respondents referred to area when discussing Graphing Task 3a
on the initial questionnaire.
GRAPHING TASK 3a: The graph below represents speed vs. time for two cars.
hr. Explain.
164
Speed
A
B
t=0 t=1
Time in hours
BG: Car A will be ahead of Car B at t = 1 hr. Up until that point, car A had been
traveling faster than Car B, so it would be ahead. The areas under the curve
represent distance traveled.
CS was initially not quite so certain, but appeared to understand the connection between
distance and area under the curve.
CS: . . . to find t . . . I . . . can’t remember really what I was thinking here, but it
looks like I started out thinking that the cars would be pretty much at the
same spot at that time . . . and then for some reason I started to think that car
A would be ahead of car B cuz it’s covering more area . . .
I: How could you tell if it’s covered more area?

165
CS: Uh . . . because it’s went at a greater speed for . . . um . . . for uh . . . the
same amount of time car B has went at a slower speed. So yeah, I guess I
would still stick with that answer.
I: Now when you say covering more area, what do you mean?
CS: Um, distance-wise, it would have traveled farther so they wouldn’t . . .
position-wise they wouldn’t be at the same position, because car A had been
going at a greater speed for the hour and car B was going at a slower speed.
I: So when you say area, you’re talking about the distance he traveled.
CS: Yeah . . . the distance . . . that they traveled.
MJ was confident in his speedy response:
MJ: A is ahead of B because the area under the curve is greater.
Table 5 is a summary of the levels of responses indicated by each participant for
particular graphical representations before studying the history of functions. The
questions included in the summary are those that required a significant level of graphical
understanding and those that exhibited the most variation in student responses.
166
Table 5
Conceptions of Graphical Representations Before the Unit on History of Functions
Task
1 2c 3a 3d 5
Student
NA
DB (no answer) Pre-Function Pre-Function Pre-Function Pre-Function
CW Process Process Pre-Function Pre-Function Process
MJ Action Object Object Pre-Function Process
BG Process Object Object Process Process
MS Process Object Process Process Process
CS Process Pre-Function Object Pre-Function Pre-Function
Conceptions of Function After the Unit on History of Functions
Participant 1: DB
This section begins with an analysis of DB’s growth in understanding graphs of
functions, since her development is most profound in this area.
The unit on the history of functions began after the midpoint of the semester.
DB’s work habits improved and she put a great deal of effort in the first worksheet,
created after reading a translation of Oresme’s original work. The purpose of the
worksheet was to exemplify an early graphing technique, then determine if, how, and
why studying the technique affected students’ understanding of graphs. Oresme sketched
vertical lines of varying lengths to measure the “intensity” of a varying quantity. Intensity
refers to the degree to which the characteristic is present. The worksheet began with short
167
readings of Oresme’s work, and then asked various questions about them: “Why do you
think Oresme felt this way?” “Is Oresme’s technique more intuitive than modern graphs?
Why or why not?” “Do you think his method would make sense for young students just
learning about graphs?” Students explained Oresme’s terminology and represented
several scenarios using modern-day graphs and graphs using Oresme’s technique. Each
student wrote a summary and reflection paper and then reworked the graphing questions
on the pre-test, making any changes they deemed necessary. The reader may refer to
Appendix C to view the worksheet.
DB reworked almost all her answers from Part 3 of the initial questionnaire. Her
answer to problem 1 became, “The speed of object B is greater than the speed of object A
at time 2 because the slope of B is greater than the slope of A.” Her revised response to
the area question in problem 2c was also correct and showed her thinking about the area
under the curve as representing speed times time: “The first 5 minutes has the greater
area under the curve which correlates to the distance, therefore the cyclist traveled further
during the first five minutes.” One can reasonably conjecture that working through
Activity 5 on the worksheet prompted this epiphany.
Similarly, her revised response to problem 3 exhibited growth: “Car A has
traveled further than car B at t = 1 hr. because the area under the curve for A is greater
than the area under the curve for B.” In the interview, she stated:
DB: I wrote that the two cars meet at the same point at t = 1 hour because this is
where the paths cross. Now I know it is different, but I think at the time I
168
was looking at it I saw, okay the lines cross then the cars must cross. But I
don’t think that is necessarily the case now.
I: If you want to say what you are thinking now that is ok.
DB: I guess I am looking at the distance that car A is traveling is much greater at
that point the area under that graph of A and then the area under the graph
of B is a lot less. So I know that would be at a different position. Their
meeting signifies how I looked at my worksheet. Now I don’t think they are
at the same point.
Note her revised response to problem 3d, “Car A is pulling away from car B. Car B is
still trying to catch up to A.” Her sketch to Task 4 was correct, as was her response to the
number of bends in the race track.
On the second questionnaire, she answered both graphing questions correctly. Her
comments confirmed her new-found understanding. One may note the evidence of her
inclination to think of a graph as a picture, but she was able to correct the error in her
thinking. (The notation “Q2” indicates that this question is from the second
questionnaire.)
Q2 TASK 12: Make a sketch of the relationship between velocity and time of a
bicyclist as that bicyclist is approaching the top of a hill, then goes over the top
and down.
DB: So . . . initially I thought . . . okay this velocity is going to be increasing I
just kind of drew him going up a hill, then down. Then I thought oh no . . .
that’s not right. Because if he’s going up the hill, I would think his velocity
169
is decreasing. He’s going to be going slower and slower and especially
when he gets close to the top, he may be barely cycling along. As he gets to
the top . . . yeah. That’s why I put his velocity kind of decreasing as he gets
to the summit cuz when he gets to the top; his velocity is going to shoot
right up. He’ll be going down the hill.
I: So point to the summit for me.
DB: Right there. [Points to the relative min part of curve.]
Figure 33. DB’s sketches for Q2 Task 12
Recall DB’s difficulty with the racetrack problem, which is similar to this one, on the
initial questionnaire.
Such significant growth is puzzling to the researcher. Surely working through one
worksheet could not have this profound effect. Did she really have this understanding
initially yet not exhibit it? Asiala, Brown, et al. (1997) found that an inability to
immediately access mathematical knowledge one has is quite common in students. Did
discussion with her group members prompt this understanding? She did come to the
researcher’s office frequently, asking questions, as she grappled with the ideas discussed
in the worksheet. Perhaps this guided discussion facilitated her growth as well.
170
DB’s comments on the worksheet and reflection paper lend significant insight to
her thinking and subsequent growth. When asked why Oresme may have thought that
lines are good representations of varying quantities, DB wrote, “because lines are
physically measurable representations of magnitudes.” In her reflection paper, she
claimed, “Oresme’s way of graphing offers a visual interpretation that makes sense in an
elementary way. By measuring the intensities of objects by lines, one has physical
representation that is measurable.” She admitted that this approach enriched her own
understanding.
For example, looking at the data comparing years of marriage with levels of
satisfaction in a graph the way Oresme would have graphed allowed me to see an
actual line representing the level of satisfaction. In comparing lines with each
other it was easy to see which levels were higher and which were lower. However
when comparing Oresme’s graph with our modern-day graph, they looked
strikingly similar. These comparisons helped me to realize that height represents
the dependent variable.
After correcting her work on the bottle problem (Graphing Task 4 on the initial
questionnaire), she made the comments below.

171
Figure 34. DB’s sketch for graphing Task 4 after unit on history of functions
I: That is quite a difference from your original answer, what prompted that
difference, would you say?
DB: Definitely after working on the Oresme’s—the function sheet and talking
with group members, we talked about the bottle being concave and that . . .
I was imagining the water going in the bottle and that the widest part is
going to fill up a lot slower than as the bottle gets narrower, . . . it will fill
up faster, and then as you have the neck of the bottle that will fill up very
fast.
This section considers the growth in DB’s understanding of functions in general,
and her increased ability to identify functions in situations. Her definition of function had
not changed that much, but her confidence had.

172
I think the biggest thing I gained was confidence in what my understanding of
basic functions is. Um . . . before I was a little fuzzy on defining a function and
now I feel that a function simply stated is something that takes an input and
produces an output.
On the initial questionnaire, DB ignored the uniqueness criterion for functions, perhaps
not realizing its necessity. In the second interview she commented about it:
DB: I think I understood that originally. That function was something that takes
an input and produces an output . . . [Pause] Um, but it was a little more
vague. I said an input takes numbers and an output is given as a result . . .
[Pause] Um, and I wasn’t clear on that the output is going to be a unique
output . . . [pause].
I: So, that was something that came up in your work?
DB: Yeah.
While reflecting on her work on the worksheet concerning Euler, she commented,
“The big difference I see between Euler’s use of the word function and our use today, is
the [that] Euler allowed for multi-valued functions. Our modern definition states that
functions have only one output for each input.” She commented that while working on
the Euler worksheet, she checked the modern definition of function in her calculus book.
Apparently, this particular worksheet prompted her to think about our current definition
and brought the uniqueness criterion to light for her.
When asked whether a list pairing each student’s first name with that student’s
score on a test could represent a function, she replied, very confidently: “Um yeah. That
173
could represent a function. The input would be the students’ name and the output is the
score. The output is dependent on the input.” Her confidence level was significant.
Though not evident from the transcription, she replied without hesitation. This question is
comparable to the club members’ dues status scenario on the initial questionnaire. Recall
that she was unable to make an arbitrary pairing in that scenario. Note DB’s own
comments about her change in understanding.
I: I’m trying to compare it now to your answer on #12 on the original
evaluation, which as a list with the students’ name and amount owed. You
thought that was a function as well. You described it a little differently. Um
. . . you wrote that there was a subtraction there. So that’s different.
DB: Yeah. I was looking at the actual function being different. I think originally
I was thinking that the function would be . . . um . . . you would take the
member’s amount, then subtract what they already paid to get the amount
owed. I wasn’t seeing a function as basic as inputting the name and getting
what they owed out.
I: So there is a difference.
DB: I think I was originally thinking that a function involved, kind of, . . .
[pause] mathematical operation. But it doesn’t have to . . . now.
One may wonder what particular aspect of the function project spurred this growth. Her
further comments concerning the Euler worksheet are pertinent.
Working through the Euler worksheet I have a clearer understanding of a
function. Euler starts off by saying that a function is an analytic expression.

174
Today we know that a function can be represented by a graph, by words, or by
different formulas. If the output is the same in each representation, we have the
same function.
The following excerpt illustrates her new-found understanding of the necessity of
the uniqueness criterion and her increased ability to think creatively to find a function.
She was discussing Task 13 on the initial questionnaire.
DB: Because you would put in a value for x and get a y. But now looking at it I
realize you put in value for x and you can get . . . you would get two y’s . . .
[pause] in a couple spots. So today, by our definition of a function, it’s not.
Y would not be a function of x. But looking at it I would see that you could
make x a function of y. Because you could put in a value for y and get one x.
On the second questionnaire, DB also was able to find a function in Task 7.

175
Q2 TASK 7: One can sometimes tell a great deal about a text (its author, when it
was written, etc.) by various characteristics (e.g., word length, frequency of
occurrence of certain words, etc.). For example, one might use the number of
occurrences of each vowel, a,e,i,o,u in a text. Suppose that you were trying to
analyze the following text (of course, in actual practice, one uses a much larger
amount of material) by studying the number of occurrences of each vowel.
The quick fox jumped over Farmer Brown’s lazy dog.
Explain how you might use a function to organize this information. Write
whatever you think is necessary about your function to specify it. Evaluate your
function once. (Dubinsky & Harel, 1992, p. 282)
She considered the function to be “something that would count . . ., that would
count the number of occurrences of each vowel. So that I wrote the vowel a, e, i, o, u and
then I wrote how many times each one occurred.” She also found a function in Task 9, a
string with nonnumeric outputs. She commented that though she worked with sequences
in her Discrete Mathematics class, she never really thought of them as being functions.
Her response to problem 9 on the second questionnaire is enlightening.
Q2 TASK 9: If possible, describe the following situations using one or more
functions. If not, explain why not: The string “ZYXWVUT.”
DB: At first I thought this was not a function but the more I think about it there
might be a function here. If you input a position, say a third, you would get
an output, the letter X. Therefore the string would be a function of position.

176
She confidently commented in the interview that “I . . . I think that seems . . . right to me.
Input would be a letter . . . a position, then you get a letter as the output. So that’s a
function.”
After this discussion, we returned to the initial instrument, Task 6, and she was
able to find a function there as well.
functions. If not, explain why not: The string “ABCDEFG.”
I: Would you change your answer to number 6 now on the original one?
DB: Yes. And . . . I just wrote “this is not a statement.” I . . . maybe it’s not a
statement. It’s just saying a string ABCDEFG, but . . . it is a function,
because you pick a position and you get a . . . a letter.
The worksheet about the Fibonacci sequence apparently facilitated this understanding,
according to DB:
At first I did not see the Fibonacci sequence as a function. However after working
through this sheet, it seems apparent that as long as the sequence can be evaluated
at the nth term, a formula can be written to represent the function.
Another growth was DB’s understanding the meaning of a series approximation
to a function. Judging from the following comments, she did not have this understanding
before beginning work on the project:
It is interesting that as more terms were added in the summation the graph became
closer to the original graph of Y. The claim of Fourier was that any function y =
177
f(x) can be represented as an infinite trigonometric series . . . It is quite amazing
how the graphs would become more like the original function as n got large.
She appeared confident with her new-found understandings. She identified a
function with Boolean outputs in Q2 Task 11 and, with probing, was able to find a
function in the caterpillar problem (Q2 Task 6).
In summary (see Table 6), DB has shown growth in understanding functions over
the course of the project. Her responses on the second questionnaire and her revisions
during the second interview of her answers on the initial questionnaire definitely suggest
a process conception of function. One can reasonably conclude that her work on this
project facilitated learning. In DB’s own words:
I think in this class we have been forced to talk about it and analyze it. And I
guess through the function projects, working with other students was helpful
because when they would tell me their insights and I . . . so, yeah, I would say
talking . . . about it . . . has really . . . changed my understanding.

178
Table 6
Summary of DB’s Growth in Understanding Functions
Growth Noted Apparent Impetus Evidence
Significant growth in ability to Worksheet on Oresme, small group Initial Interview,

interpret graphs of functions discussion written reflection
Ability to see a sequence as Fibonacci Worksheet Written reflection

function, even when outputs are
non-numeric
Ability to consider arbitrary Class and small group discussion; occurred Written reflection
correspondences as functions between work on the Descartes and Euler
worksheets
Recognizing uniqueness criterion Euler worksheet Written reflection

for functions
Understanding series Fourier series worksheet Written reflection

approximation for function
Participant 2: CW
This section begins with an analysis of CW’s growth in understanding of
functions in general. The growth in CW’s understanding of function is not nearly as
profound as that of DB.
CW appeared to be moving toward a process conception of function. On Q2 Task
5 (the table pairing students’ names and test scores), Q2 Task 6 (the caterpillar problem),
Q2 Task 7 (the quick fox problem), both problems listed as Q2 Task 8, and even came to
a recognition of the Boolean function in Q2 Task 11. Upon revisiting the initial
179
questionnaire, CW found a function in Tasks 5, 12, 13, 14, 17, and 19. In her reflection
after the “Definition of Functions” worksheet, she wrote,
All in all, my understanding of a function is that it is a relationship between
variables in which one input produces exactly one output, and the relationship can
be shown in a number of ways, not just an analytic expression. This worksheet
helped to solidify my understanding of function as a whole.
Similarly, when asked about her definition of functions on the second
questionnaire, she replied, “Let’s see. Definitely that a function is not . . . can be
represented in other ways. It’s not just an equation . . . It’s not an equation it’s a
relationship that relates variables.” Though she still insisted on the use of variables, she
recognized other inputs and outputs, as the following transcript about the caterpillar
problem indicates. Note her initial confusion; a perturbation eventually came.
Q2 TASK 6: A caterpillar is crawling around on a piece of [coordinatized] graph
paper, as represented below. If we were to determine the creature’s location on
the paper with respect to time, would this location be a function of time? Why or
why not? Can time be described as a function of its location? Explain.
a) b)
Figure 36. Graphs for Q2 Task 6 showing a caterpillar’s path

180
CW: Well, if say your . . . if your input is a time and your output’s a location,
then you are going to . . . for every input there’s going to be . . . one output,
so it would work that way but like the way this graph is situated . . . you
can’t do the vertical line test. You’re going to have like two different, you
have two different times a specific location . . .? where that’s not a function
. . . So say you had a specific location and you have two times that it was at,
then that’s not a function. But if you had like a specific time . . . It’s well
. . . I’d say it’s not a function either way cuz you’re gonna, from like the
chart you can see that the caterpillar is like at the same place at . . .
I: Point to me where the caterpillar is at the same place.
CW: Well, say like this is time.
I: So you’re saying . . . okay, CW is pointing to the horizontal axis.
CW: So this is time. And then this would be location. It’s at like . . . two different
locations at one time, like that doesn’t make sense.
I: And what would the, the actual description of the vertical axis be?
CW: Location.
I: Like what location? Like what? What do mean by location?
CW: Like a point in space, like I don’t know, like his point, well, on the graph
. . . crawl around. I don’t know if that could even be . . . that would even be
a graph of it. Because it’s just like . . . the caterpillar’s path. It’s not even a
graph. That’s just a picture.
Then came the realization that a function can indeed represent the scenario.
181
CW: So, I mean it would just. I don’t think . . . that wouldn’t be the picture of the
time versus his location that’s just him crawling around on little . . . table
. . . [laughs a little].
I: [Laughs a little]
CW: So that wouldn’t, I mean . . . I guess . . . yeah, you could represent it as a
function. I was thinking of it as . . . I was thinking of that being the graph of
. . . so if you did . . . that would be a function because at each time, at each
point in time he’s going to be at a different location. So you’re going to
have an input of time and an output of a single location. So . . . um . . . then
it would be a function. But if you just . . . I was just looking at these graphs
as . . . the function, so . . .
I: uh hum . . .
CW: Yes it is a function now. [Laughs]
Her comments during the second interview while discussing Task 13 on the initial
questionnaire displayed increased flexibility in finding functions in situations. She
originally left this problem blank on the initial questionnaire. Recall that Task 13 showed
an arbitrary unlabeled graph that would not pass the traditional vertical line test.
CW: The other way around . . . it is . . . not . . . or it is a function. It could be just
the inverse that we were talking about. That’s one-to-one correspondence,
right? If it’s inverse and it’s . . . so it’s not . . . it does not have one-to-one
correspondence here, but it is a function. This graph exactly is a function . .
. or is not a function, but if it’s . . . it’s inverse is a function [laughs]. So this

182
graph’s not, but its inverse is, so it does not have one-to-one
correspondence.
One might reasonably conjecture that her work on the Euler worksheet prompted this
discovery. Though she was using the term “one-to-one correspondence” incorrectly, her
ideas were correct.
Similarly, her comments about the “quick fox” problem exhibited her newly
found ability to find functions in situations.
CW: Say you have like, a,e,i,o,u. It’s not exactly a graph but . . . I don’t know
why you would use it like this . . . Let’s see, 1,2,3,4 . . . [making sketch]. So
a occurs . . . it would be like a bar graph, maybe . . .
I: Uh huh.
CW: a occurs uh . . . 2 times, so it would be here, e occurs 4 times, i one, o four,
and u two. And that could just . . . It’s not exactly, it’s not really a graph,
it’s more like a set of points. Something like that would just help you to . . .
Like if you wanted to compare how many . . . which vowels occurred the
same amount of times or something. But like . . . I would do it as a table.
And like . . . for a function, if you wanted to show it as a function, then the
table would work, more than the graph.
Recall that strings, particularly those whose representations do not suggest functions,
offer evidence of a process conception. CW commented about Q2 Task 9.

183
CW: Well, yeah . . . if you looked at it the opposite way where your inputs were
the position and your outputs were the letters then it would be a function
because for every position you’re gonna have one letter. You’re not gonna
have more than one letter.
I: Would that satisfy the definition of function?
CW: Well, yeah, that would but this one . . . [pointing to “ZYXXWVUT”] the
other way around doesn’t. So if your input’s the position and your output’s
the letter, that is a function, but if your input’s a letter and your output’s a
position, it depends on the string but it won’t be a function.
Note also her ability to find a Boolean function in Q2 Task 11. She was a bit
uncertain, but did consider it without prompting.
functions. If not, explain why not: {2n > n 2 + 3n : n in [1..100]} .
I: What would your input be?
CW: Input’s your n.
I: And what would your output be?
CW: And the output is the inequality, regardless of its truth value.
I: Okay, this would be the output then [pointing to inequality].
CW: Yeah. The entire inequality. You could go further with that and say . . . Is
this output true or false? And then you’d have to put another output which
184
is, which is not really a function . . . You’re just answering questions from
there on. It’s not really like function.
I: Uh huh . . .
CW: You could have something . . . probably it would be more appropriate to
have like, you’re evaluating that and your output would be whether or not
that’s true. So, your input would be the numbers . . . for n, and your output
would be, say if this works at all times which it doesn’t. So I mean I guess
it’s not a true function, but the output would always be false. Like, well, in
n: 1 – 100 . . .
Evident in the above transcription is CW’s uncertainty about the output of a
function. She initially claimed that the output is the entire inequality. She showed this
tendency in other scenarios as well. When discussing #5 on the second questionnaire, she
said,
I: So if I had . . . If I listed everybody’s scores on the final exam for this class
...
CW: Uh hum . . .
I: Um . . . then what would the input be?
CW: The input would be our . . . our scores for each of our names and then the
output is the whole table, the overall result.
Recall that Task 12 (the club members’ dues status table) on the initial questionnaire is
similar to this one and her response in the second interview is similar to the above.
185
Um again, now that is a function um because you have inputs and your output’s
the entire table. Yeah, so I would say. You’re showing relationship between
names and their . . . money that they owed and your entire table is the whole
function.
Her notion of the output being the entire table was troublesome. On the one hand, she
claimed that the table is showing the relationship between names and money, so the
“entire table is the whole function.” Thus she appeared to have the input-output idea. Yet
she also claimed that the entire table is the output.
When discussing the graph of discrete points in Task 19, she claimed, “For all
your inputs which are gonna be the points, you’re gonna have your output which is gonna
be the entire graph of the points.” Even when discussing the fox problem, she eventually
said, “So, you have your inputs are the vowels and the times they occurred and your
output the entire table.” She made similar comments about the string problem (Q2 Task
9) and Task 5 on the initial questionnaire.
These comments indicated the weakness of her conception—an uncertainty about
function outputs. One might conjecture that this error was comparable to identifying a
function with its formula or its graph. At any rate, her comments indicated that though
she was beginning to think about a process when discussing functions, she was still
confused about the fundamental notion. She needed to reflect more on the process behind
the inputs and outputs that she identified.
CW’s ability to interpret graphs improved as well. In her own words,

186
For number 3, I realized I made a mistake in part (a) and (c). At first I thought
their positions would be the same, but now I realize that this graph represents
speed and that their speeds are what is the same at t = 1 hr, instead of their
position and acceleration. I think I could answer number 3 much better now that I
really understand what is going on with this graph . . .
Before I hadn’t looked at the graph as closely, but after working through
this worksheet, I think I would look harder at the graph and be able to answer the
questions in a more correct fashion. I wasn’t even sure how to do number 4 before
this worksheet. Now I feel confident after reviewing so much about graphing
through this worksheet that I could attempt to sketch a graph this time.
During the second interview, she sketched the following graph for Task 4, the
bottle problem, #4 on the initial questionnaire.
GRAPHING TASK 4: Imagine this bottle filling with water. Sketch a graph of the
height as a function of the amount of water that’s in the bottle.
Figure 37. CW’s sketch for graphing Task 4 after unit on history of functions
187
Though she was leaning toward a piecewise linear function, her change in concavity was
accurate. In Q2 Task 13, she accurately sketched a cylindrical bottle and made an
accurate sketch for Q2 Task 12. Noteworthy, however, is the fact that even on this
problem, she initially confounded the graph as a picture, sketching part of the hill, but
then caught herself and drew it correctly.
In conclusion, evidence suggests that CW was moving toward a process
conception of function, but was not there yet, even after her work on the history of
functions unit. She did make significant progress, however. She no longer thought that an
equation was a necessary condition for functions, improved her ability to find functions
in situations, increased her understanding of graphs, and eventually came to realize that
her use of the term “one-to-one correspondence” was incorrect. Some weaknesses
remained, but significant growth was evident.
Table 7
Summary of CW’s Growth in Understanding Functions
Growth noted Apparent Impetus Evidence
Ability to consider arbitrary correspondences as Definition of Functions Written reflection

functions Worksheet
Ability to recognize a Boolean function Unknown Second interview
Ability to see a string as function Unknown Second interview
Clarified difference between one-to-one Classroom discussion Written reflection

correspondence and uniqueness criterion for
functions
Growth in ability to interpret graphs of functions Oresme worksheet Second interview, written
reflection
188
Participant 3: MJ
Following the unit on the history of the concept of function, MJ exhibited a strong
process conception of function. Though his understanding of graphs on the initial
questionnaire was good, his comments during the second interview suggest a deeper,
more integrated view. The following analysis is based on his work on the second
questionnaire and the interview during which he discussed his reasoning.
Evidence of a strong process conception. When asked to describe any insights he
gained concerning the concept of function over the course of studying its history, he
wrote, “That the mapping can be arbitrary, that is, as long as the mapping explains how
individual elements in two different sets are related you have a function.” Recall that in
the initial interview, he insisted on the existence of an algebraic formula. He elaborated
during the interview:
MJ: Um . . . Over the last few weeks, I think that the main thing that kinda
broadened in my definition and understanding of a function was the
characteristic of mapping and I’m starting to . . . I think what I’m
understanding more clearly is that the mapping is a lot more of a
generalized notion than what I originally um . . . conceived it to be. I’m
getting a broader notion of what the actual meaning of mapping means in
the definition.
I: Okay. And we talked also about your definition of function as a mapping,
now you say, one set to a unique element in another set. And . . . I notice
189
that even on your first one [initial questionnaire] you said a function maps
each item in a second set.
MJ: Yeah, um . . . [pause] the . . . [pause] I think that the uh, main difference, I
mean really I think they’re about the same. I really haven’t changed my
overall notion of what a function is. I think they’re fundamentally the same.
The overall idea. I think it’s just the actual . . . my understanding of
mapping, or . . . the actual process that you use to take one element and
relate it to another element in another set. That . . . I think that that’s the
idea . . . it’s more of a relationship that’s developing than a strict necessary,
um . . . a um . . . algebraic expression that you have to [italics added] use to
make the connection between the two sets.
I: So, it’s a little more broad.
MJ: Yes, I would say it’s more of a broader notion.
Even though he claimed that his understanding had not changed much, evidence
suggested the contrary. On the written second questionnaire, he was able to identify a
function in a pairing of students’ names with scores on a test. Recall that he was unable to
find a function in a similar problem on the initial questionnaire, even with probing during
the interview. In the following excerpt he exhibited his new-found ability to discern a
function as an arbitrary pairing.
Q2 TASK 5: Suppose you were to make a list pairing each student’s first name
with that student’s score on a test. Could this pairing represent a function?
Explain why or why not.

190
MJ: And I said yes. The reason why I said yes goes into the fact that because
each student gets a unique grade. They can only have one grade, then it’s
. . . uh . . . a relationship. You got a relationship between the student and an
individual grade. And that also I think is the idea where I developed the idea
of mapping as getting more broad that it doesn’t have to be, like I said, an
algebraic expression. This just shows a unique relationship between two
sets.
I: Uh huh. I noticed on the initial questionnaire you left #12 blank, which was
the list of names of students with their dues owed. Would you see this
differently now? Would you change your . . .?
MJ: Yeah I would relate that to more of what 5’s doing. The names are your
input and each student is paired with a unique . . . amount owed, so then
you could have a function there.
In the same scenario, he was beginning to apply characteristics of functions to this
arbitrary pairing:
MJ: Um . . . the only thing that I really don’t understand about those types of
functions are, um . . . maybe the properties, like . . . usually when we, uh . . .
um . . . look at functions we look at all the properties of a specific type of
function.
I: Uh huh.
MJ: And I’m not really seeing, understanding those properties. You have one-to-
one, onto . . . I mean I could say that that function isn’t onto . . . I mean one-
191
to-one, um . . . because you got two values there 10 that would point to two
different names going . . . back into the domain.
I: Uh huh . . . uh huh . . .
MJ: Now, I’m not seeing how . . .
I: You’re not sure if it’s one to one? Is that what you’re . . . You’re not seeing
how . . . What other characteristics?
MJ: That’s what I’m saying. Usually you have a whole, I mean what would the
actual range be? What would the actual domain be? Or, I mean, I’m not sure
how you . . . specify that type of idea when you have names, or you don’t
have numerical values . . . representing your input and output.
I: What could be the domain here? What could we identify as the domain?
We’re looking at #12 on the initial questionnaire [club members’ dues
status].
MJ: I mean on that one I would just say it’s every name. I mean, but if you
actually have a double name, if you had two . . . let’s say if we had two
Johns there and then all of a sudden it wouldn’t be a function because John
could be associated with two different . . .
I: Uh huh . . .
MJ: . . . so it had to be a unique names that you’re pairing with . . .
I: If we did that, would that help . . . [wrote “son” by John’s name]
MJ: Junior?
I: Or Jones or Smith or something like that?

192
MJ: Yeah. You’d actually have to have individual names. But that would be,
you’d have to list each one as your domain.
I: Uh huh . . .
MJ: You couldn’t . . . you just couldn’t . . . you don’t have like, real numbers.
unless you assigned a numerical value to each name, somehow.
I: Um . . . which we . . . which people do . . . the government does.
MJ: Yeah.
I: Right?
MJ: It’s a Social Security number.
I: Yeah. Yeah. Is that an example of one-to-one?
MJ: Yeah. There you go. [Laughs].
I: Is it?
MJ: Yeah. Every person has a unique . . . That would be one-to-one.
Was MJ confusing the notion of one-to-one function with the uniqueness criterion here?
Probably not, since his previous discussions exhibited a clear understanding of the
uniqueness criterion.
I: Okay. And then the range . . . do you want to talk about that more? The
range . . . that’s kind of an interesting idea . . . to pursue . . . Can you relate
all the different function characteristics for this? What would be the range?
MJ: Um . . . the range is all the uh . . . the values . . . the unique different values
you have that the function actually . . . um . . . has, the set.
I: Uh . . . huh.
193
MJ: Um . . . now . . . since you have two values of 10, that means that 10 is
pointing back so you wouldn’t have an inverse function . . . because you’d
actually have two people that owe that amount.
I: So you’re using function terminology in this set-up. I know we talked about
domain, we talked about range, and we talked about inverses . . .
MJ: I guess . . . I mean I guess that is true. I guess, yeah, um . . . You just don’t
have to look at it algebraically; you can look at it visually. Maybe that’s
why it doesn’t matter . . .
I: Uh huh . . . But are the . . . the fundamental ideas . . . are able to be defined.
MJ: Yeah.
MJ apparently had an insight here, realizing that he could apply characteristics of
functions he usually associated with algebraic representations to functions of arbitrary
pairings.
MJ displayed flexibility in answering Q2 Task 6, the caterpillar problem, a
problem requiring significant insight.
Q2 TASK 6: A caterpillar is crawling around on a piece of coordinatized graph
why not?
MJ: If . . . I initially just looked at the thing as a um, an xy plot, and looking at
the figure itself I’d say okay you don’t have a function because it doesn’t
194
pass your vertical line test. Each . . . you could have one x input and getting
several y outputs, so . . .
I: Uh huh . . .
MJ: Then I looked at it and said well, what if I had time as my input in each, um
. . . point . . . on the graph as a coordinate output. Then, for each t, I’d have
a unique . . . coordinate output.
I: Uh huh . . .
MJ: So therefore now I have a function now and . . .
I: Uh huh . . . So the input would be . . .
MJ: time.
I: And the output would be . . .
MJ: um . . . a coordinate, or a . . . what do you call that . . . ordered pair.
His understanding of the uniqueness criterion was clear as he explained his
thinking for part (b) of the same problem:
MJ: And now on part (b), you could follow the same notion, but since there’s an
intersection, we decided that um . . . if you have a time interval, you could
have unique x, y, but at that one spot, you actually . . . it couldn’t be a
function now because you’re having t point at the same, um . . . two
different ts . . . well that could be a function that way . . . You couldn’t have
it come back.
I: Uh huh . . . uh . . . huh . . .
MJ: It wouldn’t be onto . . .

195
I: Uh huh . . .
MJ: . . . because you have one point pointing at two different values of time.
I: One-to-one you mean.
MJ: Yeah . . . one-to-one, going back. You couldn’t have an inverse function.
I: Okay. Could you have an inverse function here? [pointing to the first one].
On the first one.
MJ: Yeah. On that one. Yes. Because you don’t have actual crossing over top of
itself.
Though MJ was showing good insight to the nature of the functions in these
scenarios, he was perhaps not showing much growth since he was able to answer
similarly the parametric function in Task 14 on the initial questionnaire. He was able to
apply the notions of one-to-one and inverse functions, however, building on his earlier
insight.
In Task 19 on the initial questionnaire, though, he showed significant progress.
Recall that Dubinsky and Harel (1992) believed that this problem is likely to provide
indication of one’s ability to use a process conception of function.
I: All right . . . I think um . . . that’s it for those. Oh, here’s another graph . . .
in #19. Now you didn’t mark anything in #19 on the first questionnaire
originally. Would you change anything about that now?
MJ: I don’t know what to do with that one . . . [pause]. Each x corresponds to a
unique y . . . for each point. So you do have a function . . . [laughs]. I don’t
know how to [pause] . . . if you . . .

196
I: Okay . . . I mean if you want to go back. Are you questioning like domain
and range type questions. Is that what you’re wondering about? Or how
would you write it out?
MJ: Yeah, um . . . um . . . this is where I mean, you could visually say, you have
for each x you have a unique y. Well actually, well . .. yeah for each x you’ll
have a unique y. But . . . my problem is . . . how do you explain how you get
from . . . without making a table . . . I mean the only way you could do it is
make a table.
Note that he was struggling, but had the idea. As he continued, he was quite
certain of the existence of a function, but appeared unable to discern its characteristics.
MJ: And say okay I’ll make a table of x’s and y’s and so therefore I got a
function, but is there . . . I mean is that the only way you could do that? Am
I allowed to take an algebraic expression and make it hit all those points and
say . . . plug in that point and you’ll get that number?
I: I guess there’s probably more than one way of doing . . . you know. It’s
whatever you think, really.
MJ: I mean I kinda see it could . . . it’s a function but . . . I don’t know really . . .
the nature of it . . . I guess.
Recall that Dubinsky and Harel (1992) claimed that the Boolean function
problems suggest a strong process conception, since the representation itself does not
suggest that construction. Recall also that MJ was unable to find a function in Task 18 on
197
the initial questionnaire. In the following excerpt, his understanding of the notions of
one-to-one and onto was accurate.
Q2 TASK 11: If possible, describe the following situation using one or more
functions. If not, explain why not: { 2n > n 2 + 3n : n in [1..100] } .
MJ: W ell what I thought on that one what you have is an inequality and it’s either
true or false, depending upon what n you put in.
I: Uh huh . . .
MJ: So every n is gonna have one result . . . true or false, so therefore it’s a
function. Now, the characteristics of the function . . . [laughs] . . . I’d have
uh . . . let’s say that each n would be your input.
I: Okay.
MJ: And then a true or false for your output. So my um . . . my domain would be
1 to a hundred . . . and . . .
I: uh huh
MJ: My range is just true or false . . . or my range is true or false.
I: Is it one-to-one?
MJ: Uh . . . [pause] no . . . because 1 T or 1 False could come back to 2 different
. . . you’re just going back the other way.
I: Would you say it’s onto?

198
MJ: Um . . . [long pause]. Now that’s a hard one. My idea of onto is that every
um . . . element in the range is used, which it is. Cuz that’s how I defined it
as. [Laughs]
I: Okay. I notice on this one, though [Task 18 on the initial questionnaire],
you left that one blank. So, yet on this one [second questionnaire] you were
able to come up with a mapping. Maybe it goes back to that idea of . . . it’s
a little bit broader view of what a . . .
MJ: Yeah. That’s exactly what it is.
MJ’s reflection written after the Definition of Functions worksheet revealed his
struggle to come to terms with a new, expanded conception of function.
My understanding of what a function is, often gets confused with the properties of
a given type of function, such as one-to-one, onto. I still view a function as a
relationship between two sets where a rule defines how to determine one element
in a given set when given an element from a second set. Does the rule have to be
an algebraic expression or are algebraic expressions used to define only a special
kind of function, while a relation is used to define another set of functions and
where a verbal description is used to define yet another set of functions. I am still
fuzzy about the word “rule.”
He was impressed by and made note of mathematician Richard Courant’s
comments, “It’s important to recognize that the definition of a function is NOT stagnant,
that its interpretation is open to new insight in mathematics.” History, indeed, had given
MJ food for thought.

199
In conclusion, MJ apparently had strengthened his emerging process conception
of function by studying the history of the concept. He was able to lend insight to several
bellwether type problems which he was unable to analyze on the initial questionnaire.
These include an ability to recognize a Boolean function in a situation, the ability to find
a function in a graph of a discrete set of points, and the ability to find a function in a table
of values not related by an algebraic expression. He was able to apply characteristics of
functions (one-to-one, onto, invertability) to functions with representations other than
algebraic expressions. By his own admission and by the evidence, the study of history has
broadened his conception of functions.
MJ’s understanding of graphs exhibited on the initial questionnaire was good,
though some misconceptions were evident. The following analysis is based on MJ’s
written changes to his answers on the initial questionnaire which were part of his
reflection after working on the Oresme worksheet, his answers on the second
questionnaire, and the follow-up interview.
After working through the worksheet about Oresme, “Graphs of Functions,” MJ
made four changes to his initial responses. His revised answer to Graph Task 1 was
“Speed is the slope of the line ∴ [therefore] B always will have a greater speed.” Recall
that in the initial questionnaire, he erroneously interpreted the larger function value as
indicating a higher speed. Significant also was his new answer to Graph Task 3d. When
asked about the relative position of the two cars during a given time interval, he wrote:
“A is still pulling away from car B because it is covering more distance.” Recall that
200
previously he was unable to apply his understanding of area under a curve to this
scenario. He changed his sketch of the bottle problem as well. Though his initial sketch
showed changes in concavity, it erroneously showed a vertical line to indicate a constant
rate of change. In his new sketch, he corrected this error.
Though he probably initially made a careless error, MJ readily corrected his
mistake on Graph Task 5 on the initial questionnaire. When asked in the second interview
about his initial response to the racetrack problem, he replied,
MJ: I looked at it the wrong way. [Laughs] Um . . . My process of thinking is
that um . . . what happens when you’re racing a car. You’re speeding down
a track and you see a curve so you let up the . . . uh . . . gas pedal. Your car
slows down. You go through the curve and then you start to accelerate.
I: Uh . . . huh.
MJ: S o . . . knowing that, now I have to interpret my graph.
I: Okay.
MJ: So . . . and what I . . . I think on the first one, what I did is I um . . . didn’t
take into account the um . . . I don’t know what I did. I think I actually
counted the top.
I: Oh . . .
MJ: As the turn.
I: Uh huh . . .
201
MJ: But I’m not sure. But what I know now is okay, he . . . here is speeding
down the track, he’s accelerating. He sees the curve coming up and at the
top of the um . . . curve, at its highest point, he lets off his gas pedal . . .
I: Uh huh . . .
MJ: . . . to come in to the curve. And his car starts slowing down.
I: Uh huh . . .
MJ: He goes into the curve, probably even brakes um . . . and then he starts
accelerating out of the curve . . .
I: Uh huh . . .
MJ: . . . and he stays long.
I: Uh huh . . . So how many turns would there be?
MJ: You know you have the three at the bottom.
I: You’d have three?
MJ: Yeah.
It is quite reasonable to conclude that MJ’s understanding of graphs increased as a
result of working on the Oresme worksheet since he corrected all his erroneous answers
after working on it. While responding to the question, “Has Oresme’s approach enriched
your understanding of graphs in any way,” he replied, “Interpreting graphs have always
been a challenge for me, and breaking the graph into isolated events and comparing data
by illustrating it as a line helps to clarify what relationships you are referring to in the
graph.” His response to: “Have the above exercises enriched your understanding of the
nature of graphs in any way,” he wrote,

202
I did not understand the underlining concept of integration and why it was
necessary to find the area under the curve to find the total distance, or average
velocity. Problem 14 [on the worksheet] clarified how the area under the curve
represents these quantities for the given graph.
Clearly, working on the Oresme worksheet helped him make a connection between area
under a curve and integration.
On the second questionnaire, MJ personalized the scenario concerning speed
versus time of a bicyclist approaching the top of a hill. Reed (2004) noted that students
were better able to interpret graphs if the scenario is familiar to them. MJ’s ease of
response supported this claim.
Q2 TASK 12: Make a sketch of the relationship between velocity and time of a
bicyclist as that bicyclist is approaching the top of a hill, then goes over the top
and down.
I: Would you just explain how you did that please?
MJ: Again what I did was I visualized a . . . me pedaling a bicycle up a hill. I
realized [laughs] as I came close to the top [laughs], I’m slowing down.
I: Me too. [Laughs, too]
MJ: So I knew my velocity as I came to the top of the hill would be decreasing
and then as I went over the top, I’d start picking up speed again. And what I
did on my curve is I started a high velocity and I kinda let it drop a little bit
...
I: Uh huh . . .
203
MJ: Made like a U.
I: Uh huh . . .
MJ: Representing going over the top of the hill.
In conclusion, a deepening of understanding of graphical representations is
apparent in MJ after he worked on the Oresme worksheet. At the very least, he fine-tuned
his appreciation for integration, was better able to interpret slope as rate of change,
clarified some confusion between speed and acceleration, and strengthened his ability to
interpret area under a curve.
Table 8
Summary of MJ’s Growth in Understanding Functions
In general, strengthened his Unknown Second questionnaire and second

emerging process conception interview
Recognize a Boolean function in Unknown Second questionnaire and second

a situation interview
Ability to recognize graph of Unknown Second interview

discrete points as function
Ability to find a function in a Unknown Second questionnaire and second

table of values not related by an interview
algebraic expression
Improved ability to interpret Oresme worksheet Second interview, second

graphs; slope as rate of change, questionnaire, written reflection
area under curve, integration as
area
204
Participant 4: BG
BG exhibited a process conception of function on the initial questionnaire. On the
second questionnaire he exhibited the same. During the second interview, however, he
exhibited evidence of an object conception of function.
On the second questionnaire, BG’s stated definition of function was more general
than that on his initial questionnaire. Recall that his initial definition referred to an
equation, whereas in the following he does not.
I: . . . I noticed in #2 you have your definition of function is “. . . a situation
where there is a unique output for any given input.” Um . . . On your first
questionnaire, you referred to a function as being “. . . an equation in terms
of one of more variables . . .” [Pause] How is that different? I mean, do you
see that there’s a change from your first one to your second one?
BG: Yeah . . . There is. I . . . I . . . was thinking of a function more in a . . . in the
first one I was thinking more in terms of mathematical equations . . .
I: Uh huh.
BG: . . . written out. I guess you could say more, more like, Euler’s first
definition..
I: Uh huh.
BG: Um . . . [pause]. Anyway, I was thinking of things numerically. And, and
. . . I guess through this I . . . I have kinda opened, opened the door to . . .
non-numeric variables.
205
By his own admission, he more readily considers non-numeric inputs and outputs.
Interesting to note is the fact that though his initial stated definition referred to function
as an equation, he was able to identify functions in several situations with non-numeric
inputs and output. Vinner and Dreyfus (1989) found similar disparities between a
student’s concept definition and concept image. BG was quoting a definition that he
perhaps learned in class, yet had a deeper conception of function than his definition
reveals. Nonetheless, after the unit on the history of function, his concept definition was
more in line with his concept image.
BG readily found a function in each of the situations given on the second
questionnaire. Not surprisingly, finding a function in a list pairing students’ names with
test scores came easy to him. The ease with which he found a function in Q2 Task 6 was
impressive.
paper, as represented below. [See sketches in Appendix B, Task 6.] If we were to
determine the creature’s location on the paper with respect to time, would this
location be a function of time? Why or why not? Can time be described as a
function of location? Explain.
I: What were you thinking as you worked through that?
BG: Uh . . . alright, well the . . . [pause, then laughs], the little worm’s path. I . . .
I realized later that uh . . . when . . . all right, you can . . . you can say is a
function of time with position. Yeah, there is a unique position for any . . .
uh . . . point in time as it crawls around. Ah . . . I realized later that . . . that

206
when we’re talking about the . . . time being described as function of
location, I should have said . . . it’s a pretty convoluted domain, but . . . you
can do it.
I: Uh huh . . . uh huh . . .
BG: Uh . . .
I: in both of them . . . or . . .
BG: No . . . not in B.
I: . . . in part A?
BG: Yeah, in part A. In part B, I . . . I said that . . . that it would not be, you
could not say . . . time is a function of location because there are certain
locations which have two values for time,
I: Uh huh . . .
BG: Ummm . . . which, which, we say that that would not be a function.
I: And you’re pointing to point A and Point B which are on the graph.
BG: Point A . . .
I: Yeah. Um . . . When you saw that right off the bat, is that what you thought
of right away? Or did you think . . . did you think anything else right away?
Some students look at this and think . . . oh . . . vertical line test. Did you
. . .?
BG: Oh no. I . . . I saw that you had put the graph in there and thought, “Oh
she’s trying to trick me . . . ”
I: [Laughs]
207
BG: [Laughs too] Well I mean, . . . When I just looked at the graphs by
themselves I . . . I . . . that’s what I thought, but when I read it I thought,
“Oh . . . she’s trying to trick me.”
I: Oh . . . Well . . . [laughs] okay.
BG: [Laughs]
I: You picked up on that. Not that I was trying to trick you, but you certainly
did find a function pretty quickly. I was just wondering if it was immediate
. . . so as soon as you read it . . . you sort of . . . you sort of realized that
there was a function there.
BG: Right.
Note also the ease with which he found a function in the scenario in Q2 TASK 7,
even though he struggled with the notation.
was written, etc.) by various characteristics. For example, one might use the
number of occurrences of each vowel. Suppose that you were trying to analyze
the following (one uses a much larger . . . of course in practice) by studying the
number of occurrences of each vowel. Explain how you might use a function to
organize this information. Write whatever you think is necessary about your
function to specify it. Evaluate your function once.
I: Let’s just expand upon what you wrote.
BG: All right. The domain would be the sentence itself that we’re looking . . .
we’re limiting our look at this one sentence. Uh . . . the input of the
208
independent variable would be . . . [pause] . . . each vowel, I guess and the
dependent variable would be the number of vowels. I think that’s what
you’re asking for. The number of vowels, right?
I: [pause] . . . the number of occurrences . . .
BG: . . . the number of occurrences for each vowel. So if I . . . I guess really the
domain would be the output . . . or the . . . the vowels in the output, not
necessarily the sentence.
I: Okay . . .
BG: Uh . . . [pause]. I really struggled with how to write this out . . . [laughs a
bit].
I: Okay . . . That’s okay . . .
BG: Yeah. A would be . . . 2, E would be 3, 4 . . . 4. I would be 1 . . .
I: So the input is . . .
BG: Yeah. I mean . . . Input . . . input’s a vowel, output’s the number of vowels
in that sentence.
One area of growth is evident in BG’s response to Q2 Task 8, which asked him to
describe a sequence using one or more functions.
functions. If not, explain why not: { 1 1 1 1

}
, , , ,... .
2 4 8 16
I: Okay, and um . . . for #8, any comments about that?

209
BG: No . . . I . . . I do have a question. Um . . . I . . . I mixed up the, the . . . in my
final test . . . what is . . . when I say “for n within N.” Like this [pointing to
#8 on second questionnaire]. What is the symbol there that can be used?
I: Um . . . Is an element of . . .?
BG: Okay . . . All right, I think I used it backwards.
I: A backwards one means “such that.”
BG: Yeah. That’s what I thought. All right. All right.
I: Yeah. Um . . . that’s fine. So, was that . . . Have you been in the habit of
thinking of sequences as functions or did that . . . what do you think of that?
BG: Uh . . . When I took . . . I hadn’t really thought of them as functions until
. . . we looked at them as functions here. I . . . I . . . [pause] I guess I had
converted sequences into functions in calculus II.
Q: Uh huh.
BG: Uh . . . I took Honors Calculus II at Miami University and the professor
kind of did the sideline thing for me.
I: Uh huh.
BG: Looked at how to convert things.
Q: Uh huh.
BG: Um . . . But, uh . . . Yeah . . . I hadn’t really thought that well a sequence is
a function until we . . . [pause]
I: . . . with the domain being the natural numbers as you have indicated here
[points to Q2 Task 8].

210
BG: Right.
I: Well that’s something.
BG: Yeah. That’s . . . that’s great.
It is reasonable to conclude that BG came to this realization as a result of working on the
Fibonacci worksheets, since sequences were the main topic of study on these sheets and
were not discussed on the others.
BG showed creativity while describing a function for the string “ZYXWVUT,”
given in Q2 Task 9. Recall that he readily found a function in a similar problem on the
initial questionnaire (Task 6). On the second questionnaire he did not want to simply
replicate his answer from the initial questionnaire; he searched for a new approach.
I: Okay. Would you explain #9? I wasn’t really clear on what you meant by
#9. This is the problem with the string “ZYXWVUT.”
BG: [Laughs] I was trying to be more creative than I was in the last one. There’s
a string of letters, but they weren’t in any order.
I: Yes, yes. Okay.
BG: Uh . . . so I . . . I thought all right if we’re gonna . . . if we’re going to say
that uh . . . a1 = capital letter A, a2 is capital letter B, then a26 would be the
capital letter Z, and so we have a string from a26 to a . . . 20 . . .?
I: Okay.
211
BG: Yeah . . . Um . . . and so I said, well all right, this one . . . I didn’t write this
out anywhere correctly, but the sequence a (27 – n) for n = 1 through 7
would give uh a26, 25, 24 . . . and so on . . .
I: So the input is [reading from Q2 Task 9] the natural numbers from 1 to 7.
BG: Yeah.
I: And the output is . . .
BG: Uh . . . these capital letters. Yeah . . . That’s fine. That’s what I was trying
to write.
I: Does it matter that these are in order? I mean, I don’t . . .
BG: Well yeah . . . from 1 to 7.
I: How would you have written it if um . . . these weren’t in that order? Would
it have mattered?
BG: Uh, probably wouldn’t.
BG: I’d probably have done it the same way as I did over here.
I: On the first questionnaire?
BG: Yeah.
I: Okay. Pretty creative there.
BG: Why, thanks.
BG readily found a Boolean function (true-false outputs) in Task 12 and answered
the two graphing questions correctly.

212
functions. If not, explain why not: { 2n > n 2 + 3, n in [1..100]} .
I: Now when you first did the first questionnaire, you didn’t have one [find a
function], but then in our conversation, it came up and I don’t think I
prompted you but you found that there was a function.
BG: Right.
I: Remember in the conversation? I remember you doing that. So, and I notice
you just got it right away, right here. So that was . . .
BG: Right.
I: Were you thinking anything different? Or you just saw that and said oh, you
remembered from last time.
BG: I kind of remembered it from last time. But I had . . . I remembered in our
conversation last time getting the true false thing and then I thought,
“Alright, well this is a true-false situation.”
I: We actually call that a Boolean function.
BG: Right.
Conclusion
Since BG exhibited a strong process conception initially, one may wonder what
benefits he obtained from studying the history of functions. His concept definition
became aligned with his concept image, and he realized that a sequence is indeed a
function with a domain of natural numbers. He more readily identified a Boolean

213
function and exhibited creativity while finding functions in non-numeric situations. His
process conception remained strong. In his own words, “The study of functions continues
to broaden my horizons. Some of these aspects I already knew, but the exercises have
simply brought them to mind again, but once again, that is what much of learning is.”
The researcher was curious if BG possessed an object conception of function. An
object conception results when a student tries to perform an action on a process. Though
the researcher had no questions prepared, she attempted to probe the notion of function as
object with BG.
I: Okay so you’re talking about maybe taking differential equations or
something. What other . . . where else in calculus are you familiar with
functions? I mean when you . . . how else do you think of functions in
calculus?
BG: Um . . . what do you mean?
I: Like, I mean . . . um . . . like when you take an integral, what are you doing?
BG: Uh . . . well . . . generally it’s taking the area under a curve or volume or
about finding how much space there is . . . contained in the . . .
I: And what’s your result? Geometrically it’s an area, but algebraically, what
is the result. I mean you’re taking an integral of the function, what are you
getting?
BG: Oh it’s a . . . You’re getting another function.
I: Uh huh.
214
BG: Either of one power lower. You know, the transcendentals have different
things that they come out to be.
I: And what about when you use the chain rule . . . or taking derivatives in
general?
BG: Right. You get another function again.
I: Yeah. What’s your interpretation of the chain rule? Does that . . . I don’t
have these questions prepared. [Laughs]
BG: [Laughs] Let’s see if I can remember the chain rule. The chain rule is if
you’ve got have f(g(x)) and then you get f(x) . . . (pause) .f ’(x) . . . (pause).
I: . . . then you take the derivative of what’s on the inside.
BG: Right. Yeah. There we go. [Laughs] Yeah that’s right. f '( g ( x)) g '( x)
I: I guess I’m driving at composition of functions, basically, because that’s a
composition of functions.
BG: Right.
I: You work with one function then you do the other function . . . and your
result is again a function. That’s what I was driving at.
BG: Right.
BG: Um . . . Yeah. It is interesting that the functions can be broken up into basic
functions that you . . . have f ’(x) and g’(x). That you break it up into as
small a function as you can.
I: So the result is still a function, then?
BG: Yeah.
215
Though the above dialogue skimmed the surface of BG’s thinking about functions
as objects, he appeared to have at least the beginnings of an object conception. He was
able to discuss in general terms the acts of integration and differentiation, acknowledging
functions as outputs. Dubinsky and Harel (1992) claimed that one’s ability to perform an
action on a process is evidence of an object conception. Whether or not BG had this
conception before the history of functions unit was unknown since the researcher did not
have this conversation with BG previously.
Table 9
Summary of BG’s Growth in Understanding Functions
Concept image more in line with Unknown Second interview

concept definition; able to
verbalize concept image
Exhibits object conception of Unknown Second interview

function
Participant 5: MS
MS was not a particularly conscientious student. He did not complete assignments
on time and had low grades on most of them. He did not interact with others, despite
continued encouragement to do so. He scored a 66% on the functions project, with a
score of 13 out of a possible 20 on the Oresme (graphing) assignment. He ended the
course with a grade of D.

216
Not surprisingly, in most areas he showed little change in his conception of
function. His ability to interpret graphs was strong at the beginning and remained so at
the end. He was unable to identify a Boolean (true-false outputs) function to begin with
and was similarly unable to do so at the end, even with prompting from the interviewer.
functions. If not, explain why not: {2 n > n 2 + 3n : n in [1..100]}.
I: What about #11.
MS: Uhhhh . . . it’s not a true statement. Same as last time.
I: Same as last time? What would some possible inputs be?
MS: Uh . . . maybe negative numbers . . . no that wouldn’t work. [Pause] But as
it stands, that’s never true, I don’t think.
I: That’s never true . . .? Do you want to try some numbers?
MS: Well, if you take 1,
I: n is one.
MS: 2 is never greater than 1 squared plus 3.
I: . . . if n is one. Okay so . . .
MS: And then 4 is not greater than 4 plus 6.
I: Okay. Do you want to pick another number?
MS: 3 . . . 8 that wouldn’t be greater than . . . 9.
I: Pick another number?
MS: 4 . . . 8, 16 isn’t greater than 16 . . .

217
I: Pick another number . . .?
MS: 32 . . . would be 25 plus 15. Doesn’t work. Maybe it doesn’t actually.
I: I don’t know—pick a huge number. Up to 10, you don’t know what 10 to
the tenth is.
MS: Well, . . . it would be 64 . . . 36 plus . . . okay, I guess it does work for this
one . . . wait . . . well . . . it doesn’t work for all cases, so it doesn’t really
count.
I: So, it doesn’t really count? So there’s no function there?
MS: No. You would have to change the range . . . or whatever that is . . . domain.
Conditions to put it.
I: Oh . . . for it to be . . . how would that make it . . .
MS: Because it’s not true, but still you could graph it as an area, I guess . . . since
it’s an inequality.
I: You can graph what as an area? Like what would the area be? [Long pause]
I mean, what are you going to graph? I’m not sure what you’re going to
graph?
MS: I don’t even know how you would set this up. I guess you could hm . . .
well . . . I don’t like this one. [Laughs]
I: [Laughs] You don’t have to like it, it’s okay. It doesn’t work? Because it’s
not always . . .
MS: It’s just a relationship. It isn’t any . . .
I: So if your input is one of the numbers . . . there’s no real . . . uh . . .

218
MS: You just get another relationship. It doesn’t . . . mean anything.
I: It doesn’t mean anything, okay.
MS: Well, it does, but . . .
I: What does it mean?
MS: It just gives you two numbers that are either greater than or less than.
I: Okay, so what might an output be?
MS: I guess you could have one number . . . like one number on the left could be
the input or vice versa.
I: Uh huh. Okay.
MS: It still doesn’t work . . . for the first . . . five.
I: Yeah. Any other outputs? Any other possible outputs?
MS: You could do it the other way around. The right side is the input. You could
switch them.
I: Oh. And the output would be . . .
MS: 2n
I: Okay.
MS: I still don’t like it.
I: Still don’t like it.
MS: So I guess you could say that one side’s one x and this one side’s the other.
I: Okay. Would there be a function there?
MS: [Laughs] No because . . . [inaudible].

219
Similarly, when revisiting Task 15 from the initial questionnaire, he was unable to
consider a Boolean function.
I: Do you remember what you said about 15?
MS: No. . . . but it looks like I just factored it. I found the solutions.
I: Do you think there’s a function there?
MS: Yeah, but it would only give . . . you’d only have two inputs . . . that are
possible.
I: So what would be the input, then . . . the input would be . . .
MS: -1 and -2.
I: And what would your output be?
MS: Always 0?
I: Okay. Anything else about that? I mean could there be a different type of
function there?
MS: You could, if you didn’t have 0 . . .
MS is unable to consider the possibility of “true” or “false” as outputs. He
similarly had a weak understanding of parametric functions, something also evident on
his initial questionnaire and interview. On the second questionnaire and during the
second interview, he continually referred to the need either for two graphs or for three
dimensions when discussing the caterpillar problem, Q2 Task 6.

220
why not? Can time be described as a function of its location? Explain.
a) b)
Figure 38. Graphs for Q2 Task 6 showing a caterpillar’s path
MS: Uh . . . well you could, but I don’t think I would want to use a graph that
looks like that.
I: Okay, would you, could you do a graph that would be different?
MS: I’d probably do . . . some sort of parametric . . . with two graphs. Like each
one would have time . . . each one would have time on the axis and one
would have the x and one would have the y. That’s the way I would do it.
Apparently, he had a vague notion of parametric function, but insisted upon two
graphs rather than one graph with t as parameter.
I: Would you make a sketch of that for me?

221
MS: You have one [sketching two sets of coordinate axes] . . . so then you’d
have to refer the two. Or you could make a 3 dimensional one, but that’s
hard to draw.
He continued to insist on the need for a 3D graph throughout a lengthy discussion of this
task. He had a similar difficulty on the initial questionnaire, insisting the parametric
function in Task 14 exhibited two functions rather than one.
Despite these weaknesses and apparent lack of growth, his function conception
did deepen somewhat. His original definition of function referred to the need for an
operation, indicative of his tendency to search for some type of regular relationship, or
formula: “A function is that takes input terms, applies some operation to them, and
outputs them into a clear result.” His definition on the second questionnaire is more
general: “A function is some process that accepts an input and provides an output.” In his
reflection, MS wrote, “In general, my concept of functions has been expanded. Prior to
these worksheets I thought of functions in terms of graphs and situations, instead of
relations between processes.”
Note also his flexibility when revisiting Task 5. Recall that in the initial interview,
he was unable to find a function in this situation. During the second interview, he was
more successful.
I: Okay, what about this number 5? “Kent State Men’s basketball team makes
it to the Sweet Sixteen.” Can you find a function there?
MS: Yeah, I don’t like these.
I: These verbal ones?

222
MS: They’re just arbitrary statements. [Pause] But uh . . . I still can’t really find
anything in this one . . . other than some logical . . . if – then.
I: I didn’t know what you meant by that.
MS: Well if . . . they . . . I guess you could say if they win any in the Sweet
Sixteen, which would be the assumption.
I: Ah . . . Is that a function?
MS: You could, but it wouldn’t be very exciting. [Laughs]
I: But what do you mean it’s a function. I’m not sure what the function would
be . . .
MS: I mean you could say . . . Say if you input win, the output would be Sweet
Sixteen. But if the input is lose then, go home. But it’s not really a map
function.
I: That’s interesting.
MS: It’s just . . . like something from Discrete Math.
I: So you did see a function there.
MS: I guess, but it’s not a map. It’s not really. You can’t really make an equation
out of the thing.
I: So if you can’t make an equation out of it, it’s not a function?
MS: Well not . . . I don’t like it otherwise. Cuz it’s just a statement then . . . you
don’t need to bring functions into . . .
I: Okay. Now I’m going to have to . . . I’m going to probe you on this one a
little bit. So you said something about an equation. Now and you said you
223
like those better. Does it . . . does a function have to have . . . does it have to
have an equation to be a function?
MS: No, you could just say it . . . like a cake recipe is a function.
I: [Chuckles a little] How would that be a function?
MS: You’d input ingredients and get a cake.
I: [Laughs]
MS: And you follow the process.
I: And the process would be . . .
MS: Cooking it. The recipe . . .
I: Oh . . . But you said you like the ones with algebra formulas better . . .
MS: Right . . . That’s usually what you’re talking about.
Though in the initial questionnaire, MS found no function in Task 5, in the second
interview he admitted to one being there. When revisiting Task 19, he adjusted his
original tendency to connect the dots.
Figure 39. Graph for Task 19 showing discrete points

224
I: What about 19? I notice on 19, you had connected the dots. Um . . . The
original . . . I’m going to redraw the original in blue pen, here and just put
the dots there. Now suppose you didn’t have the line that you drew . . .
would you see a function in that situation?
MS: Uh sure. It would just have to be restricted . . . so it would only give you
points instead of a continuous . . . curve I guess.
While discussing Q2 Task 5, he revealed his increased flexibility in thinking
about functions. He reversed the inputs and outputs, and did not confuse the uniqueness
criterion with the notion of one-to-one functions.
MS: Uh . . . here I said it wouldn’t be if you had multiple test scores per person
cuz that’s not in the modern definition.
I: Uh huh.
MS: You could reverse the dependency I guess, then it would work.
I: Reverse the dependency? What do you mean?
MS: So you’d have . . . test scores would be matched up to the . . . would be the
input. Then that would give you the student’s name.
I: Oh . . . But suppose I had 5 test scores for you.
MS: As long as . . .
225
I: Would that be a function then? If my test scores were the input?
MS: They’d all be going to one output, then. So that should be fine.
I: Uh . . . huh. Okay. But suppose Matt had the same test score as you.
MS: Then you’d have a problem.
I: Then we’d have a problem.
MS: So, if everybody had a different test score . . .
I: Oh everyone would have to have different test scores. Okay.
MS: [Laughs a little]
I: Okay. But I think this wording says . . . each student’s name with a
student’s score on a test. Like on a particular test. Suppose like I took all
your . . .
MS: Yeah, I figured it probably was just for one test.
I: And if it was for one test would it change . . .
MS: There it would work.
One characteristic of MS’s function conception was noteworthy. His process
understanding appeared to be closely linked to his experience with programming, since
he repeatedly referred to computer programs when discussing functions in general. When
asked if he had thought of sequences as functions before, he responded, “Yeah . . . I’ve
seen . . . I think we did some in some early programming class.” The excerpt below also
illustrates this propensity.
MS: My definition of function is halfway between pure math and what I
remember from programming. [Laughs]

226
I: Yeah. It is an interesting point of view. You do seem to be . . . calling on
your programming knowledge.
MS: Yeah cuz in a program when you’re calling on your function, it’s just a little
piece of code that takes something and does something.
I: Uh huh.
MS: So then, just about anything’s a function.
I: Do you think that enriches your idea of a mathematical function?
MS: It just confuses me most of the time.
I: Are they the same or are they different?
MS: They’re pretty much the same.
I: In what way are they the same?
MS: Everything’s going to take an input and give you an output.
I: Okay.
MS: As long as you keep it general, they’re the same.
I: Okay . . . [pause] So is the function the actual program?
MS: Pretty much. Well, it’s . . . what’s ever inside of it, in that case.
I: What do you mean, what’s inside of it?
MS: Whatever you do, it is the function . . . the process.
I: Oh . . . Okay. So if I have an algebraic formula, is that the function?
MS: Cuz the process tells you what’s the relationship between the two . . . the
input and the output.

227
I: So what’s the difference between a function, a graph, and then the formula?
Are they the same function? Could they be the same function?
MS: One can describe the other, but they don’t have to be.
I: Okay but suppose that I took this formula, y = e x + 4 and then I plugged in
x’s and came out with y’s and then I plotted it, is that the same function?
Are they the same function?
MS: Yeah . . . why not? [Laughs a little]
I: Okay.
MS: There are different ways of looking at the same result . . . but it depends on
how you want to look at it.
Clearly, his programming work influenced his thinking. The excerpt above, though,
exhibits a clarity of thought that was absent from his comments in the initial interview.
Though he referenced computer programs there, in the above discussion he was more
focused, confident, and articulate. His study (albeit weak) of the history of functions
apparently provided him with an opportunity to reflect upon his function conception.
In conclusion, MS’s conception of function appeared to have moved from an
emerging process one to a stronger process conception. Some weaknesses still existed,
however. Significant though is the fact that in addition to not completing all required
work in a timely fashion, he had little if any interaction with his peers during the
functions project.
228
MS’s understanding of graphical representations of functions was strong before
the history of functions unit began, as evidenced by his insightful responses to the
graphing questions both on the written questionnaire and during the interview. None of
the common misconceptions had been evident. His responses on the second questionnaire
and during the second interview confirmed the strength of his understanding of graphical
representations.
Table 10
Summary of MS’s Growth in Understanding Functions
Expanded view of functions, saw function Unknown Second questionnaire, written

as process rather than function as operation reflection
More able to recognize functions in Unknown Second questionnaire; second

situations, including a graph of discrete interview
points
More able to discuss his function Unknown Second interview

understanding
229
Participant 6: CS
This section begins with an analysis of CS’s growth in understanding of functions
in general.
CS appeared to be moving toward a process conception of function, albeit a weak
one. On the second questionnaire and during the second interview, he was able to find
functions in many situations for which he was previously unable to do so. When asked to
define a function, he wrote, “A function is a relationship that provides us with a unique
output for a given input.” Note that there was no mention of equation here. During the
interview, he described a function as “any type of relationship that has an input and an
output and it’s just not graphs, equations, like . . . specific things . . . It could be . . .
something you never really thought about would be a function.”
CS readily considered alternate representation of function, including lists. Recall
that he was unable to find a function in Task 12 (the club members’ dues status) on the
initial questionnaire. On the second questionnaire and interview, he easily recognized a
function in a similar scenario:
CS wrote, “Yes because the input would be the student’s name and each name is
related to an output.” In the interview, he reiterated this new understanding and then
applied it to the club member’s dues status task on the initial questionnaire.
230
CS: Um, I said yes because again back to my definition that it’s a relationship,
and . . . in that case the student’s name would the input and the score would
be the output. So . . . [pause]
I: Okay . . .
CS: You’d just . . . you’d have a unique output for every input. That’s basically
my definition of a function.
I: You’re saying that it is a function. So here’s maybe another representation
of a function, is that what you’re saying?
CS: Uh huh . . . yeah.
I: Going back to #12 on the original questionnaire. I notice that you left that
blank. Would you change your answer now?
CS: [Immediately and confidently] Oh yeah. I was not sure at all what to do
with that, so . . . I just left it blank.
I: And now what would you say?
CS: I would definitely say that’s a function.
I: With the input of . . .
CS: Input of the name and the output of the money owed.
Recall that CS left Task 13 blank on the initial questionnaire. During the second
interview, he compared it to Q2 Task 6 and found functions in both.
I: And you would still answer the same thing? You’re still thinking that #13
on the first questionnaire is not?

231
CS: Oh no, no no . . . It is, it is a function, because . . . I thought this was a
function because . . .
I: Okay, wait . . . we’re talking about two different things. Just so when I look
at this, I’ll know. We’re talking about #6 the caterpillar problem . . . sort of
. . . And then we’re also referring to 13. You seem to want to talk about #6.
Let’s talk about #6 in detail, and then we can get back to the other one.
CS: Just the way I see it is . . . I kind of see a connection between #6 and #13.
Because you could also look at that as being the same type of thing. You
could say that that’s a caterpillar traveling on a piece of paper.
I: I see, I see.
CS: So that’s the way I kind of look at it now. I would, I would kind of compare
those and say yeah it is a function, but it’s just a lot easier to see this
relationship in #6 because of the information.
I: Input would be . . .
CS: [Long pause] . . . I think it would be like, time of when the caterpillar’s in a
certain spot. And then the output would be where it actually is.
I: If we were to write it . . . like . . . time is 1, what would the output be?
CS: Um . . . that location right there.
I: Could you describe that location in any way?
CS: Um . . . it’s in the . . . negative part of the axis.
I: Okay. So what should I write?
CS: I can’t remember what quadrant that is.

232
I: Well, it’s quadrant, [counting] one, two . . . this one, quadrant 3.
CS: Okay quadrant 3. I would take that as the beginning, just . . . I don’t know, I
guess that could be the beginning too, there . . .
I: So then you’re saying . . . now back to 13 you’re sort of saying if this was a
caterpillar then you would have the same type of mapping.
CS: Yeah.
I: You’d say the input would be . . .
CS: time. And then like, this line would be its location at different times.
Note in each of the above scenarios, there is no mention of the need for an equation. He
readily considered alternate representations of functions and discussed them in a general
way. Another scenario in which he found a function without using a formula is Q2 Task
7, the “quick fox” problem:
was written, etc.) by various characteristics (word length, frequency, etc.) For
example, one might use the number of occurrences of each vowel, a,e,i,o,u in a
text. Suppose that you were trying to analyze the following text by studying the
number of occurrences of each vowel.
whatever you think is necessary about your function to specify it. (Breidenbach et
al., 1992, pp. 281-282)

233
CS: Okay, this problem definitely confused me, cuz I really didn’t exactly sure
what they wanted.
I: Okay.
CS: So, I just kind of made up my own idea which was I just decided that my
function would be um . . . vowels to words, like for every . . . the words
would be the input . . .
I: Okay, so the words would be the input.
CS: And then every vowel in that word would be the output.
I: Oh.
CS: If you had the word “the,” then your output would be e.
I: Oh.
CS: [Laughs]
I: So if we called this function v, v . . . so your output would be . . . [writes “e”
on paper].
CS: Uh huh.
I: So now what happens then if we did your function on the word jumped?
CS: Then it would be u and e.
I: Does that satisfy your definition of function?
CS: [Sighs/laughs] The way I just looked at it is the relationship between words
and vowels in a word and . . . I would put it as my input could be each word
and then my output could be the vowels in the word.
I: all the vowels . . .

234
CS: yeah. I . . . I’m not sure if that’s what the question wanted, but . . .
Even though his representation may not have been the one the researcher had in mind, if
one considers the output as a subset of the set of all vowels, then CS successfully created
an arbitrary mapping between two sets.
Recall also that CS was unable to consider a string as a function on the initial
questionnaire. He left Task 6 blank. In Q2 Task 9 and in the interview, he considered the
possibility of a function existing in a string, though he struggled with the idea. Again, the
representation was not what the researcher had in mind. The fact that he proposed
something indicated a broadening of his function conception, however.
functions. If not, explain why not. The string: “ZYXWVUT”
I: What about #9 . . . You have your input is . . . now do you mean your input
is that entire string or is your input . . . you know these are each individual
...
CS: Those are each individual inputs.
CS: Uh . . . I guess if you’d graph it out, and you had like a string . . . like a line
on your desk and that representing your string, then those uh, letters would
represent your locations on that graph.
I: Oh . . . so are you talking about . . . go ahead, draw the graph for me. I’m
going to give you another piece of paper so you have enough room, if you
like.
235
CS: I just looked at it as . . . um . . . if I had . . . I drew my graph and if I had a
string something like that . . .
I: Uh huh . . . He’s drawing a graph.
CS: Then the input would be the letters. And they would just be different points
on the string. So . . .
I: Uh huh.
CS: Z,Y,W . . . and your output would just be. If you input Z, your . . . output
would be where Z is on the string.
I: How would you describe that?
CS: Um . . . I would do Z as the beginning of the . . . of the string.
I: Uh huh.
CS: And I . . . I like to move from left to right.
CS appeared to be thinking of a graph as a location, similar to his discussion of
the caterpillar task.
I: Okay.
CS: [Laughs] So that’s kind of how I would start it out and then just kind of
going over here with the letters that you gave for the string.
I: Yeah, but if . . . but how would I know where y is? How do I know if y is
really really close to z or farther along? How would you describe that
output?
236
CS: Um . . . I guess again you’d just . . . You’d have these as your inputs and
you would put them . . . into an equation, I guess again, to get . . . where
you would put the location of that number. So . . . that’s how I look at it.
I: Okay. So do you need the coordinate system, here then?
CS: Uh . . . I don’t think so. [Pause] No I wouldn’t really need it.
I: Okay. That’s sort of like . . . I’m looking at the first questionnaire now. On
#6 you left, on the original questionnaire it’s also a string but it’s
ABCDEFG. Um . . . you left it blank. Would there be a function there, now,
do you think?
CS: Oh yeah. It would be the same thing as #9 on the second one.
His description was sketchy and he appeared uncertain, but he was at least willing to
consider the possibility of a function, something he did not do on the initial questionnaire.
Recall that CS was unable to find a function in Task 7 on the initial questionnaire,
but when we re-visited this task in second interview, he readily discerned a function.
I: A swimmer starts from shore and swims to the other side of the lake.
Number 10. [Q2 TASK 10, which was TASK 7 on initial questionnaire.]
Um . . . and you have something written there. Would you explain your
work, please?
CS: Um . . . I just had the . . . if you take time times average speed and that
would equal your distance, that could be a function that you would use for
that type of problem. Um . . . I guess there’s a couple different functions

237
that you could make out of that, just depending upon what relationships you
wanted to find.
I: In your original questionnaire, this was #7; um . . . you said you can’t
because you don’t know if the swimmer takes a straight line to the other
side. Or if he takes another route. Um . . . any comments about that?
CS: Uh . . . I guess . . . like what I was thinking there . . . is that you, you
couldn’t make a function out of it because you have to know like, exactly
what type of direction he’s gonna go with it. Um . . . like, I was just
thinking that you, you couldn’t have an input and output because you don’t
know . . . what your input and output would be.
I: Uh huh.
CS: So . . . I don’ know.
I: But in #10, now you’re thinking um . . .
CS: Yeah . . .
I: You could find a function?
CS: Yeah . . . I think . . . I think so. Because, if you . . . If you just took his time
and average speed that he was in the water no matter what direction he went
. . . you could still find the distance that he traveled.
I: Uh huh . . . So what would the input be?
CS: Um . . . input would be time, I guess.

238
CS: the speed . . . or no no no . . . the input would be the time and . . . kind of
. . . like related to the average speed. And then the distance would be your
final input. [meant output?]
Similarly, he was able to consider a function in Task 5 on initial questionnaire during the
second interview.
If not, explain why not: “Kent State Men’s basketball team makes it to the Sweet
Sixteen.”
I: I know we talked about this in our previous interview. Would you change
anything about that or do you have any insight about that?
CS: Again, that seems like such an abstract idea for a function that . . . I couldn’t
come up with an equation that seems like it would make any sense for that.
Like . . . it’s really not giving any relationship. It’s saying that . . . makes it
to the Sweet Sixteen . . . I guess if you put like words in it saying, like
“Kent played this team, this team, this team” . . . then you could make some
kind of function out of it, but . . .
I: Like what kind of function could you make?
CS: Uh . . . maybe like a list . . . function . . . where you had the team as the
inputs and then like maybe the final score as the output. Or win/loss as the
output.
So, though CS initially expressed the need for an equation, through conversation he
found a function without it.

239
When revisiting Task 8 in the second interview, growth in his thinking was
evident. Note his reference to an equation on the initial questionnaire, but not on the
second.
I: What about #8?
CS: I’m not 100% sure what I did there [on the initial questionnaire]. Uh . . . I
think I tried to make an equation out of it.
I: Uh huh.
CS: That fit what that was saying. And . . . I don’t think that makes any sense
looking at it again.
I: Uh huh. Would you do anything different now?
CS: Um . . . I would probably just say it is a function because if you plug in uh
. . . a number for n . . . then you will have an output of whatever number
comes out of that equation.
I: Which equation is that?
CS: the 3n + n 2 .
I: Uh huh. So you think that is a function? I’m going to mark it in blue.
CS: Cuz if you put in any number 1 to 100 . . .
I: Uh huh.
CS: you’ll get an output.

240
Another example of CS’s new found flexibility is his response to Task 10 during the
second interview.
functions. If not explain why not: A record of all MAC women’s basketball teams
giving, for the 2003-2004 season each team’s field goal shooting percentage at
home and its field goal shooting percentage away.
I: What about #10?
C: Um . . . here I just kind of came up with two different functions . . . Uh . . .
the H would be home and I just took shots made over shots missed.
I: Uh huh
CS: So, and Away, shots made over shots missed.
I: Uh huh. Would you do anything differently now? Or are you sort of happy
with that?
CS: Um . . . it might again be easier to make like a list . . . just to show the
relationship. It’s kind of hard to see the relationship if I just put it in
fractional form.
I: And what would be in that list? What would be your input?
CS: Um . . . you could just do . . . like . . . each home game. You could just
maybe who they played or what date it was . . . any . . . anything that would
show that it was a home game. And then you could put shots made or shots
missed as the output.

241
Weakness of conception. Though CS was able to discern functions in more
situations including those not requiring a formula, his conception appeared not to be a
strong process one. He was unable to consider the points in Task 19 as representing a
function, even in the second interview.
I: Would this representation be a function in any way? [Pointing to Task 19].
CS: I’m not sure . . . Um . . . just by looking at it I wouldn’t think so, but I guess
. . . are these supposed to be open or are they [pointing to the points].
I: They’re dots. They’re like dots, filled in dots. I’ll try to fill them in so that
it’s not confusing for you.
CS: I guess it could be a function . . . if you’re just plotting points. Then you
would draw like a curve or something. So that could be . . .
I: But would I have to have a curve in order to have a function?
CS: I would think so.
I: What would the input be?
CS: Um . . . I guess those points that you plotted there.
I: And what would the output be?
CS: Um . . . I’m not sure. [Pause]
I: A lot of times there’s a lot of different ways of looking at things, so . . .
CS: Yeah.
I: Please don’t feel that you’re looking for the one right answer. Just tell me
what you’re thinking.

242
CS: Now that I think about it I guess that would . . . it seems like it would be the
output if you’re actually plotting that. And the input would be the
information that you’re taking and putting into the equation to get those
points.
I: I guess you’re saying equation again. So you’re relating it to an equation to
get those points?
CS: I guess so, yeah . . .
I: Okay. As it stands, just by looking at it, you’re saying . . . you’re sort of
saying you need an equation to get that?
CS: I think you need information to show the relationship. It doesn’t make any
sense to me that a function could just be dots on the graph.
In the above excerpt, his tendency to consider an equation was still strong. He was unable
to consider the general process of taking discrete inputs and pairing them with individual
outputs. His tendency is resistant to change, also evidenced in the following.
I: Can you devise a specific situation for which a function might provide a
mathematical description?
CS: Okay, um . . . I just came up with the example . . . driving a car at different
speeds for an amount of time. And then, like if you were doing a graph
you’d have time at the bottom and you have velocity. And you’d just graph
it.
I: Okay. So that’s a function.
CS: Yeah.
243
I: On the first questionnaire I said, “Give several examples of functions” and I
notice that you have all equations. What else could you do . . . what else
could you do now?
CS: Um . . . just like you could use story problems and like, say if a person was
walking at a certain rate for a certain amount of time. Like that equation for
that you’d come up with is a function. But they’re not necessarily giving
you the equation. You have to figure out how to take what they’re saying
and figure out the equation.
I: But will the end result always be an equation?
CS: Ah . . . I’m not totally sure. It seems like it because . . . In order to graph it
you’d have to have an equation, I think. So . . . Um . . .
His comment about needing an equation before graphing is interesting. Perhaps his
conception of graphical representations was weak, rather than his function conception in
general.
Another weakness was his inability to consider Boolean functions, both on the
initial and second questionnaires and in the second interview, even when prompted. Note
also his reference to an equation in the following excerpt.
I: Okay. Number 11 . . . on the new one [second questionnaire]:
{2 n
}
> n 2 + 3n : n in [1..100] and you said there’s not a function because
there’s no output, only a statement? Um . . . and on the original one, you left
it blank. Um . . . what were you thinking when you wrote about #11?
244
CS: Um . . . just after doing the work with the functions project, I . . . I took it
the statement, this greater than statement as just been a math statement. It’s
not really um . . . having an input and output, because . . . I don’t think you
can have an output if you don’t have an equal sign in the equation.
I: Do you want to try it? What would some of your inputs be?
CS: Um . . . I guess these numbers here . . . 1 to a 100, so if we start at one.
I: Start at 1 . . .
CS: You’ll just have uh, 2 greater than 1 + 3 which would be 4, which, that’s not
even true.
I: If you put in 2?
CS: Uh, then you would have 4 greater than 18. So . . . [laughs] . . . so far none
of these are true. Um . . .
I: Is there an output?
CS: I don’t think so. I mean you’re . . . you’re still getting . . . just . . . a
statement here, saying that something is greater than something else, which,
I don’t think that’s an output . . .
I: Uh huh.
CS: Cuz also you’re having 2 here. So . . . I won’t take those as an output just
because it’s not saying anything definite. Cuz normally an output in a
function has it equal to this after you put in the equation. And here you’re
not getting that it is this, it’s just saying that something is greater than
something else.
245
I: Uh huh . . . Um . . . you mentioned that this isn’t true and this isn’t true . . .
I’m going to refer you to your definition of function. [Reading from paper]:
“It’s a relationship that provides us with a unique output for a given input.”
So if you looked at this input as 1,
CS: Okay . . . I kind of see what you’re saying. That my output could be that it’s
2 > 4? And then my output could be that it’s 4>18?
I: I don’t know, I’m just asking . . . Throwing some ideas around.
CS: I think that that just depends on how you look at it. I would still say it’s not
a function because it’s still a statement.
I: Okay. You seem pretty sure about that.
In conclusion, CS appeared to be moving toward a process conception of function
as evidenced by his ability to find functions in a greater variety of situations and to
discuss these functions as general processes of inputs and outputs. However, his repeated
reference to the need for an equation, his inability to consider a graph of discrete points
as a representation of inputs and outputs, and his inability to recognize a Boolean
function suggests that his process conception of function was weak.
CS’s understanding of graphical representations appeared to have improved as
well. Consider closely the changes he made when discussing his responses on the initial
questionnaire during the second interview.
GRAPHING TASK 3d: The graph below represents speed vs. time for two cars.
246
direction.) What is the relative position of the two cars during the time interval
between t = .75 hr. and t = 1 hr. (i.e., is one car pulling away from the other?)
Explain.
CS: Relative position . . . [long pause]. I think I answered that question wrong.
Cuz like I was saying before . . . Car A would have covered more distance
because it’s going at a faster rate than car B up until the 1 hour . . . then for
that long of a time, so I definitely would change my answer there.
I: And you would say . . .
CS: That car A would be farther ahead of car B, but car B after an hour would
eventually be catching him and passing him. That’s kind of what I was
thinking here. That car B would eventually catch him because he’s
accelerating faster.
Similarly, he corrected his error on Graphing Task 5.
I: Now since we’re talking about velocity versus time, I’m going to refer to
one of your problems on the first one, which happens to be the racing car
one. The number of bends. We have the speed, which is like velocity,
versus this is distance along the track . . . Um . . . and you said 9 bends.
Would you change that answer at all or you still think there’s 9 bends there?
C: Um . . . okay . . . [Pause] Bends, you mean . . .

247
I: Bends in the track. I’m sorry, like you know how a race car goes around a
turn.
C: Uh, uh . . . let’s see . . . One, two . . . Yeah, I would definitely change my
answer.
I: How many do you think there are?
C: I think I would say 3.
I: And where would they be?
C: Um . . . where it’s making a big decrease in speed.
I: Uh huh . . .
C: Cuz obviously you’re gonna have to slow down if you’re going around a
curve. So I would say three now.
He also easily sketched the relationship of speed versus time of a bicyclist
approaching, then going over, the top of a hill. He had no trouble sketching water height
in a symmetric bottle as a function of volume.
Some gaps still exists in CS’s graphical understanding, however. He was not able
to consider area under the curve for Task 2c, the speed versus time of a cyclist over a 10-
minute period nor was he able to consider changes in the rate of change of the height of
water vs. volume in Graphing Task 4. He made no changes to his original answers for
these tasks during the second interview.

248
Table 11
Summary of CS’s Growth in Understanding Functions
Broadened view of function. Unknown Second interview

Though “function as equation”
conception still evident, he is able
to consider arbitrary
correspondences as functions
Able to consider data lists as Unknown Second questionnaire and second

functions interview
Able to consider strings as Unknown Second questionnaire and second

functions interview
Some growth in ability to Worksheet on Oresme, small Second interview

interpret graphs of functions group discussion
Level of confidence increased Unknown Second interview
At the beginning of this study, several participants relied on the “function as
equation” notion. Only two considered an arbitrary mapping as a function. Using APOS
analysis, two held an action conception of function and three appeared to hold an
emerging process conception. Graphical understanding was particularly weak in three
participants and the ability to find functions in non-traditional situations appeared to be
undeveloped in all but one.
After the unit on the history of functions, four participants significantly
strengthened their function conceptions. None insisted that a formula was required for a
function. Two moved an entire APOS level. Other than BG, each participant exhibited an
249
increased ability to recognize a function in the given scenarios. Recall that BG’s ability to
do so was strong initially. All six participants recognized a function in an arbitrary
pairing. Growth was most profound in the area of graphical representations of functions
as two participants no longer considered a graph as a picture. Two participants came to
understand the importance of area under a curve, and of five graphing tasks, one
participant exhibited a deeper understanding on all five, one participant on three of the
five, and one participant on two.
In summary, a significant improvement in understanding functions occurred
during the course of studying its history for 4 of these 6 participants. More detail and
analysis of these findings is available in Chapter 5.

CHAPTER V
DISCUSSION
Introduction
This study asked the questions:
school teachers?
understanding to a process level understanding as described by APOS theory?
understanding?
This chapter discusses the findings for each question, the implications for teaching and
intervention, cautions in interpretations, and the implications for future research.
Findings
Although each participant was considered separately in the narratives in Chapter
4, this section analyzes them as a group to see what patterns emerged regarding the
250
251
growth in their understanding of functions as they studied the history of the concept. This
section has four subsections. The first part is a glimpse at the participants’ growth
concerning their definition of function, the second is a look at their increased abilities to
recognize functions in situations, and the third summarizes changes in APOS levels. The
section ends with comments concerning changes in understanding graphs of functions.
Definitions of Function
Recall that Dubinsky and Harel (1992) claimed that an emphasis on equations,
numbers, and evaluating expressions is evidence of an action conception of function. On
the initial questionnaire, several participants exhibited this tendency. The notion that a
function was an equation was strongest in CW and CS. DB and MJ exhibited the
tendency to a lesser degree and DB thought that a graph or a formula is the function,
rather than just a representation of a function. Interestingly, DB and MJ at times
discussed a function as a general process, but reverted to a “function as equation” as they
attempted to recognize functions in situations. This tendency suggests that DB and MJ
had an emerging, but not strong, process conception initially. MS at times referred to the
need for “some operation” but he, like BG, was able to recognize that a function may
have more than one representation. BG did not exhibit the “function as equation”
understanding.
On the second questionnaire and interview, DB made no mention of function as
equation. In her words, “now I feel that a function simply stated is something that takes
an input and produces an output.” She came to understand the uniqueness criterion,
something she overlooked on the initial instruments. Similarly, CW’s new definition of
252
function is “a relationship between variables where one input produces one output and
the relationship can be shown in a number of ways, not just an analytic expression.” MJ
and CS also let go of their tendency to look for an equation, with MJ noting, “the
mapping can be arbitrary” and CS claiming that a “function is a relationship that provides
us with a unique output for a given input.” Note that there is no mention of equation in
these definitions.
These results suggest that studying the history of functions broadened the
participants’ definition of function. After working on the worksheet entitled “Definitions
of Functions” and the readings associated with it, the class had an extensive discussion
about the change in the definition of function over the years and what prompted such
change. It is reasonable to conclude that this worksheet and discussion facilitated a move
away from the “function as equation” notion in the participants. Table 12 summarizes
these results.
Functions in Situations
The current study provided other evidence that studying the history of function
facilitate a move away from a narrow view of function. Not only did the participants’
definitions change, but their ability to find functions in situations improved as well. The
following discussion focuses on four tasks on the initial questionnaire and the
corresponding tasks on the second questionnaire that may be considered bellwether
indicators of a process conception (Dubinsky & Harel, 1992).

253
Table 12
Participants Holding the Notion of “Function as Formula” Before and After History of
Functions Unit
Before After
DB *
CW *
MJ *
BG
MS
CS *
Table 13 indicates which participants recognized a function in the given scenario.
An asterisk (*) means that participant clearly articulated an appropriate function. The
phrase “considered the possibility” indicates that the participant had a vague notion of
function in the scenario, but was either unable to articulate it or gave an inappropriate
formulation.
Only two participants considered an arbitrary mapping as a function. The
“function as equation” notion was evident in DB (“you apply the function of
subtraction”), CW (“if you’d have the previous one plus five, you would have a
function”), and MJ (“starting to graph it a little bit and see if there was any kind of
relationship”). CS left the task completely blank. On the discrete set of points graph, MS
did successfully find a function, but insisted on connecting the dots and trying to find a
254
Table 13
Appropriate Answers to Specific Tasks Before History of Functions Unit
Arbitrary pairing Recognizes a Discrete set of

as function Boolean (true-false points on graph as
String as function (i.e., list) output) function function
DB considered the considered the

possibility possibility
CW
MJ *
BG * * * *
MS considered the * considered the

CS
formula so that he could “find a pattern of how to change it” [the inputs]. Note also that
other than BG and MS, the participants had little success with these tasks as a whole.
Table 14 indicates the participants who successfully recognized a function in
these scenarios on the second questionnaire or while revisiting the initial questions during
the second interview.
Other than BG, each participant exhibited an increased ability to recognize a
function in the given scenarios. The data above shows remarkable progress in the abilities
of DB, CW, MJ, and CS. It is worth noting that BG, who recognized functions in each of
the initial scenarios, exhibited evidence of an object conception of function in the second
interview. Also noteworthy is the fact that MS, who showed the least growth, had put
255
Table 14
Correct Answers to Specific Tasks after the Unit on History of Functions
Arbitrary pairing Recognizes a Discrete set of

as function Boolean (true-false points on graph as
String as function (i.e., list) output) function function
DB * * * not discussed
CW * * * *
MJ not discussed * * *
BG * * * *
MS considered the * *
possibility
*
CS considered the considered the
forth the least effort during the functions project, insisting on working alone, and turning
in work of mediocre quality. The others, DB and MJ in particular, worked incredibly hard
on the project, seeking help when necessary, struggling with the ideas presented. Not
surprisingly, they exhibited impressive growth.
Of interest to the researcher is the fact that the class never discussed Boolean
functions, nor was topic covered on any of the worksheets. It is possible that participants
discussed this scenario in small groups, since they worked together on the function
project. It is also likely that the worksheet concerning “Definitions of Functions” and the
subsequent class discussion facilitated an ability to recognize this type of function.

256
APOS Level Changes
The initial APOS conception of each participant is summarized in Table 15. The
reader may refer to the first part of the Chapter 4 for a detailed analysis of each
participant’s responses characterizing their function conception.
Table 15
APOS Conception Before History of Functions Unit
DB *
CW *
MJ *
BG *
MS *
CS *
Table 16 lists each participant’s conception after the unit on the history of
functions. The reader may refer to the detailed analyses that led to these
characterizations. A double asterisk (**) indicates a change in level from the initial
conception. Note that four participants notably strengthened their function conceptions.
Two participants moved an entire level: CS from an action conception to an emerging
process one and BG from a process conception to an object. Admittedly, however, the
evidence from this study is insufficient to claim that the unit on history of functions
257
Table 16
APOS Conception After History of Functions Unit
DB **
CW *
MJ **
BG **
MS *
CS **
enabled BG’s move to an object level, since the initial questionnaires did not test for this
understanding. All one can claim is that after the unit on history of functions, BG
appeared to hold an object conception. DB and MJ appear to have strengthened their
process conception. Recall that though CW did not advance a level in her APOS
conception, she appeared to be moving toward a process conception. Recall also that MS
was the weakest student among the participants and not surprisingly, showed little
growth.
Changes in Understanding Graphical Representations
This section considers the participants’ responses to individual graphing tasks and
characterizes the APOS level for that task. A summary of the participants’ understanding
of graphs is in Table 17. Notable is the low level of graphical understanding for two or
more questions in three of the participants, DB, CW, and MJ. Compare these levels
258
Table 17
Interpretation of Graphs Before History of Functions Unit
Task
1 2c 3a 3d 5
Participant
DB (no answer) Pre-Function Pre-Function Pre-Function Pre-Function
CW Process Process Pre-Function Pre-Function Process
MJ Action Object Object Pre-Function Pre-Function
BG Process Object Object Process Process
MS Process Object Process Process Process
CS Process Pre-Function Object Pre-Function Pre-Function
with those following the study of the history of functions, in particular, after completion
of the worksheet concerning Oresme’s techniques. Table 18 lists only those conceptions
which changed during the course of study.
What Caused the Improvement
There appears to be little question that a marked improvement in understanding
functions occurred during the course of studying its history for four of these six
participants. Less clear is the reason this improvement occurred. This section attempts to
answer the question,
In other words, what specific uses of historical material facilitated the change?
259
Table 18
Interpretation of Graphs After History of Functions Unit
Task 1 2c 3a 3d 5
Participant
DB Process Object Object Process Process
CW Action
MJ Process Process Process
BG
MS
CS Process Process
Growth was most profound in the area of graphical representations of functions.
Interestingly, of all the worksheets, the Oresme worksheet on the history of graphs was
the one most dependent upon original sources and provided the most in-depth
information about the germination of the concept. It offered a combination of historical
reading pinpointing the rationale for a new technique, activities comparing Oresme’s
techniques to modern ones, and Oresme’s proofs with the details omitted. It appears to
have cured the “graph as picture” tendency in DB and CW and enabled the understanding
of area under the curve for MJ and DB. According to MJ, it helped him understand the
“connection between/motivation for integration and area.” DB commented that this
particular set of exercises clarified her thinking about graphs. One can reasonably
conclude that this use of primary sources revealing the germination of an idea along with
activities relating original methods to modern-day ones enabled conceptual growth. With
260
the other worksheets it is difficult to say whether the history, the class discussion, or
group interaction caused growth, but here, it is evident that the use of these materials
facilitated growth. This finding validates the work described by Jahnke (2000)
concerning the benefits of using original sources.
The Leibniz worksheet concerned the first use of the word “function.” Perhaps
this exercise suggested the idea that a function need not be definable by equations alone.
The Fibonacci worksheet was not historical in the sense that it did not delve into the
beginnings of an idea. It was just a problem from history. The learning that occurred as a
result of this worksheet supports the claim that history is a good source of problems,
however. Both MJ and CW had not considered sequences to be functions until after their
work on this assignment. Perhaps it facilitated their new-found ability evidenced on the
second questionnaire and in the second interview to see strings as functions as well.
The Definitions of Function worksheet showed participants that the definitions of
function changed, but did not really go into depth about why they changed. One may
therefore wonder if the change in participants’ understanding was superficial, though
their comments in the second interview suggested otherwise.
The above evidence suggests that a wide variety of materials may pull students
along in their understanding of a concept: primary source readings about the germination
of a concept, problems from history, or simply reading about the changes of a concept
over time. The emphasis here, however, is clearly on a thematic approach to history—the
germination and development a single mathematical concept. To those who view the
261
history of mathematics as a disjoint collection of anecdotes, facts, and dates, this study
offers nothing.
Comparisons to Other Studies
The reader may wonder how the results of this study compare with similar
studies, in particular those concerned with student learning as a result of studying the
history of mathematics or those concerned with student understanding the concept of
function. As mentioned in Chapter 2, finding studies that attempted to show the
mathematical content learning that occurred as a result of studying the history of
mathematics was difficult. Only the van Gulik (2005) and Arcavi studies (1985) claim to
have done so to some degree.
Recall that Barbin’s (2000b) overview of nine case studies reported results from
teacher introspection, rather than objective analysis, and most dealt with changes in views
of mathematics, rather than mathematics content learning. Furinghetti (1997, 2002)
reported four case studies on different uses of the history of mathematics in teaching, yet
did not reveal empirical evidence about students’ growth in understanding. Po-Hung Liu
(2002) conducted a study whose purpose was to investigate the relationships between a
problem-based, historical approach to teaching calculus and college students’ views of
mathematical thinking. Again, the study focused on general views of mathematics rather
than specific content. Boero et al. (1997) conjectured that the introduction of ‘voices’
from the history of mathematics and science may mediate some important elements of
theoretical scientific knowledge. The researchers noted that the resultant learning was
“better and more extensive” than those usually achieved when eighth graders approach
262
theoretical knowledge (p. 2-85), though they did not indicate the instrument used to draw
these conclusions, nor did they specify what they meant by “better and more extensive.”
Van Gulik’s study (2005) was similar in design and purpose to the current study.
She collected data through questionnaires preceding and following the series of lessons,
through lesson observations, and through discussions with students and teachers. She
concluded that her subjects gained a deeper insight into the origin of geometry, the
teachers found the subject challenging and inspiring, and that “the ‘reinvention’ of the
basic assumptions of geometry results in a more lively learning process and better
motivation” (personal communication, February 23, 2006). Note, however, the emphasis
on the process of learning and motivation, rather than learning outcomes.
Arcavi’s (1985) study also was similar in design and purpose to the current study.
He showed that the mathematical reading ability of pre-service and in-service teachers
improved as a result of their using his historical materials and that they were better able
to identify irrational numbers in a given list of numbers. The current study focused on a
more advanced mathematical topic and attempted to analyze student thinking in more
detail.
One might compare the results of the current study to others that focused on the
concept of function. The design and purpose of the Breidenbach et al. (1992) study was
similar to the current one. Recall that in this study participants experienced a 4-week
computer-intensive environment and wrote programs for functions in ISETL.
The initial conceptions of functions of the participants in the two studies was
similar. The researchers noted that “of 59 subjects, 7 [11.9%] appeared to start course
263
with strong process conception” (p. 274). In the current study, one of seven appeared to
start the course with a strong process conception [14.2%], though one needs to be
cautious while making statistical comparisons with such a small sample size. Interesting
is the fact that though the Breidenbach participants were at least one year younger and
presumably one year less mathematically mature, the initial function conceptions of
students were similar. Both sets of participants exhibited many of the common
misconceptions mentioned previously in the literature (see Chapter 2).
Notable also is that the current study obtained results similar to this study.
Breidenbach et al. (1992) claimed that “24 [40%] showed clear progress throughout the
semester that seemed to be more than one might expect from ordinary instruction” (p.
274). In the current study, three of six (50%) significantly strengthened their conception
of function; two (33%) moved up an APOS level. Recall that BG appeared to move up to
an object conception, but the evidence that this growth was caused by studying the
history was insufficient. It appears that these particular students in this particular study
showed comparable growth to those in the Breidenbach et al. study.
Interesting to note is that the instructional activities in Breidenbach et al. (1992)
focused on constructions derived directly from APOS theory. Subjects wrote ISETL code
for various functions, for testing equality of functions, for point-wise arithmetic of
functions, composition, and so forth. The instructional activities in the current study
focused on the historical development of functions, which is remarkably similar to the
development of understanding in some students. The early conception of function (up to
and including that found in much of Euler’s early work) was that of a formula and
264
became more general only as the need arose. Once the notion of function as a single
formula was found to be insufficient to describe the motion of a vibrating string, the
concept expanded. Functions were considered objects in their own right only much later.
In this particular instance, a similarity exists between the historical development and the
development of student understanding according to APOS Theory. Perhaps reading about
the historical perturbation caused participants to reflect upon their own understanding.
Interesting to note also was that two participants in the current study referred to
their experiences with computers while discussing the notion of function. MS repeatedly
referred to computer programs when discussing functions in general. MJ created a mental
image of a computer program as he found functions in the string problems on the initial
questionnaire. Clearly their previous work with computers enriched their notions of
function. Their comments support the findings of the Breidenbach et al. (1992) study.
Contributions to the Field
The professional mathematical societies are deeply concerned about the
mathematical education of our teachers and are continuing to search for effective means
to deepen students’ understanding of fundamental mathematical concepts (Conference
Board of Mathematical Sciences, 2001). For those mathematicians who question the
worth of requiring a History of Mathematics course of future teachers, the current study
provides evidence that taking such a course can strengthen pre-service teachers’
understating of fundamental concepts, clarify misconceptions, and bring to the surface
long-forgotten notions.
265
It offers support to those who claim that studying the history of a mathematical
concept deepens mathematical understanding and it offers a foundation for future
research in this area. Recall Furinghetti’s comment, “Though teachers who use the
history of mathematics are enthusiastic about its effect, these opinions are subjective, not
a result of regular and systematic studies” (1997, p. 55). The current study may spawn
interest in similar studies concerning other topics, or perhaps concerning a more
advanced study of functions. If studying the early development of functions encouraged a
move toward a process conception of function, might not a more in-depth study of its
later development facilitate a move toward an object conception? BG’s comments
suggest that this may indeed be the case, though more research is needed in this area.
Also noteworthy is that the Oresme worksheet, which relied heavily on original
sources rather than adaptations of historical problems, was the one that spawned the most
growth. This finding supports the claims mentioned in Jahnke (2000) about the value of
using original sources. Note also that the entire unit focused on a single mathematical
concept rather than on isolated historical anecdotes, famous mathematicians, or a
chronological and encyclopedic compilation of thousands of years of mathematics. This
suggests that the study of the development of an idea, rather than a chronological
narrative of disjoint historical facts encourage growth in mathematical understanding.
Arcavi’s work (1985) further supports this claim.
Another contribution of this study was the adaptation of APOS theory to the
knowing and learning of graphical representations of functions. Though Baker, Cooley,
and Trigueros (2000) analyzed students’ understanding of graphs of functions in the

266
context of APOS theory, they based their interpretation on the triad of Piaget and Garcia
(1989). Asiala, Cottrill, et al. (1997) discussed students’ graphical understanding of the
derivative within the context of APOS theory but used APOS terminology only as it
relates to functions. The problems that students had with graphs in the current study have
been well documented in the literature (Bell & Janvier, 1981; Carlson, 1998; Clement,
1989; Goldenberg et al., 1992; Monk, 1992), but applying APOS terminology to this
analysis was an innovation, to this researcher’s knowledge.
Also significant is the current study’s emphasis on functions, a fundamental
mathematical concept. This concept takes center stage when it comes to mathematics
education. Guershon Harel and Ed Dubinsky argued (1992) that
The concept of function is the single most important concept from kindergarten to
graduate school and is critical throughout the full range of education. Arithmetic
in early grades, algebra in middle and high school, and transformational geometry
in high school are all coming to be based on the idea of function. (p. vii)
Though some researchers have obtained positive results for student construction of a
process conception of function (Breidenbach et al., 1992), others have not (Sfard, 1992)
and still others note the continued difficulty students have with the concept (Breidenbach
et al., 1992; Carlson, 1998; Even, 1993; Norman, 1992; Sierpinska, 1992; Wilson, 1994).
The fact that a historical study significantly deepened participants’ understanding will be
of interest to those concerned about student learning of this topic. This study suggests
that studying its history is one vehicle toward understanding, not the—or the best—
vehicle, but a viable one.

267
This study also reinforced that at least one of Sierpinska’s epistemological
obstacles common in history is evident in today’s students as well. The belief that “only
relationships describable by analytic formulae are worthy of being given the name of
function” (Sierpinska, 1992, p. 46) evident in the 17th and 18th centuries was similarly
found in participants in this study. This finding is not new, however, as Siepinska herself
noted the same and other researchers have documented this student belief about functions
(Breidenbach et al., 1992; Carlson, 1998).
Finally, this study confirmed what other studies have shown in terms of
motivation and attitude toward mathematics. Confidence level of the participants was
notably increased after studying the history of the topic.
Future Studies
Suggestions for future research are numerous. First and foremost, the materials in
this study need to be tested in other venues. The instructor-as-researcher design of this
study necessitates similar studies to validate results. Secondly, researchers might consider
conducting similar studies using the history of other mathematical concepts. How does
studying the history of algebra, for example, change students’ thinking about it? In
particular, by studying the early motivation for and development of abstract algebra do
they deepen their understanding and appreciation of the need for structure? Does a
focused study on 17th century calculus develop a deeper conception of the fundamental
ideas behind this most impressive mathematical achievement? The possibilities are
endless.
268
Another reasonable question arises from this study. If an in-depth look at the
beginnings of the concept of function strengthened an emerging process conception in
some participants, might a more advanced study enable a move to an object conception?
This study has only begun to answer that question.
Researchers may also pursue the notion of historical development of a concept
paralleling the conceptual development of the concept in an individual. Recall that at the
beginning of this study, several participants relied on the “function as equation” idea.
Interesting to note is the fact that the concept of function originally was definable as an
equation relating two changing quantities. Not until the vibrating string problem did this
idea expand. This suggests a need to recreate in our teaching a need for students to
change their ideas, for their concept to expand. We need to show them that their current
understanding is not sufficient to explain more complex ideas. One way may be to
recreate for our students the environment that caused a perturbation in the history of a
concept by pinpointing those specific events, questions, and situations that preceded its
development. Anna Sierpinska (1992) has already laid the groundwork. What needs to be
done is to create, then study the effectiveness of, learning materials that focus on her
epistemological obstacles.
Finally, researchers may explore further the question of the relative effectiveness
of certain historical materials. What type of materials encourages the most learning—
activities, solving problems, reading original sources? Why is this the case? What
difficulties do students encounter when doing so? Jahnke (2000) echoed this sentiment as
269
he calls for research on the benefits of, the strategies used, and difficulties encountered as
students work with original source materials.
Cautions With Interpretations
One limitation of the study is that the researcher is the instructor for the course, a
difficult limitation to overcome since only one section of the History of Mathematics
course is offered per semester and few other courses offer sufficient time for the
instructional activities suggested in this study. The inclination may be to be overly
optimistic about alleged results. However, extensive inter-rater reliability counteracted
this inclination.
Another limitation is that third and fourth year students may not be aware of their
knowledge about functions and a poor showing on the initial questionnaire may simply
be a result of simply not thinking about these issues for some time. Critics might also
argue that ANY in-depth study of the concept would increase, or help students to become
aware of, their understanding. These arguments are not really limitations, however, since
the current study is investigating whether or not history is one vehicle toward
understanding, not the—or even the best—vehicle.
This study has shown that it is possible to create a positive and productive
learning environment using historical materials. One needs to be cautious about
generalizing the results, though all reasonable attempts have been made to be transparent
in this research and to offer specifics to allow for replication.

270
Conclusion and Summary
In conclusion, this study confirmed what countless mathematicians have
conjectured about the value of studying the history of mathematics. History can be a
vehicle to refresh and strengthen a student’s understanding of a mathematical concept.
An in-depth study of the birth of an idea and its early development provide the impetus.
A thematic approach, studying the development of an idea, reading original sources, and
working on problems of former mathematicians did lend insight into the conceptual base.
Three students showed significant growth. Five of the six showed improvement in
finding functions in situations. One claimed that the unit helped refresh ideas long
forgotten. These results support the requirement that pre-service teachers study history to
deepen understanding. A History of Mathematics course is a reasonable substitute for
another mathematics course at the junior/senior level.
This study, then, provided evidence that studying the history of mathematics
enables a deep reflection of ideas, or as Von Glasersfeld (1995) suggested, a “re-
presentation” or reconstruction of ideas. For those that study it for pure enjoyment, this
research suggests another good reason to study it, to teach it, to preach it.
APPENDICES
APPENDIX A
QUESTIONS ABOUT FUNCTIONS: INITIAL QUESTIONNAIRE

QUESTIONS ABOUT FUNCTIONS
Initial Questionnaire
Part 1.
Please write your response to the following questions in the spaces below. Be detailed,
but concise.
1. What is a function?
2. Describe the ways a function can be represented.
3. Give several examples of functions.
Part 2. Functions in situations1
Directions. If possible, describe the following situations using one or more functions. If
not, explain why not.
1. {(x, 3x + 2): x in the set of all integers}
2.
Figure A1. Graph for Task 2
3. y4 = x2
4. y4 = x3
5. “Kent State’s Men’s basketball team makes it to the Sweet Sixteen.”

1
From “The Nature of the Process Conception of Function by E. Dubinsky & G. Harel
(1992). The Concept of Function. Aspects of Epistemology and Pedagogy, MAA Notes,
25, 88- 90. Copyright 1992 by the Mathematical Association of America. Adapted with
permission.
273
274
6. The string “ABCDEFG.”
7. A swimmer starts from shore and swims to the other side of Walden Pond.
8. {3n + n 2 : n in [1...100]}
9. A square in the plane centered at the origin is rotated clockwise by 90o.
10. A record of all MAC women’s basketball teams giving, for the 2003-2004 season
each team’s field goal shooting percentage at home and its field goal shooting
percentage away.
11. 2 x 3 y − x log y = 2
12. The club member’s dues status.
Name Owed
John $15
Julie 10
Jen 0
Dave 7.50
Brittany 10
Alex 5
Mary Ann 25
Sam 20
Sally 17
Figure A2. Table showing dues owed

275
13.
Figure A3. Graph for Task 13
x = t3 + t ⎫⎪
14. ⎬ t , a real number
y = 1 − 3t + 2t 4 ⎪⎭
15. x 2 + 3 x+ 2 = 0
16. {(x,y): x, y in the set of all rational numbers}
17. {(1, 2x): x in [1 . . . 100]}
18. {2n > n2 + 3n: n in [1 . . . 100]}
19.
Figure A4. Graph for Task 19 showing discrete points

276
Part 3. Graphing Tasks
1. Consider the graph below. At the instant t = 2 seconds, is the speed of object A greater
than, less than, or equal to the speed of object B? Explain
position (cm) B
2 4 6 8 time (sec)
Figure A5. Graph showing position vs. time

by S. Monk (1992). The Concept of Function. Aspects of Epistemology and Pedagogy,
2. This graph shows the speed in meters per second of a cyclist over a 10-minute period.
speed
2 4 6 8 10
time
Figure A6. Graph showing speed vs. time
a) When is the speed the greatest?

b) When is the speed changing most rapidly?
277
c) Does the cyclist travel further during the first five minutes or during the last 5
minutes? Please explain your answer.
3. The graph below represents speed vs. time for two cars. (Assume the cars start from
the same position and are traveling in the same direction.)
speed
A
B
t = 0 hrs. t = 1 hr.
Time in hours
Figure A7. Graph showing speed vs. time

a) State the relationship between the position of car A and car B at t = 1 hr. Explain.
b) State the relationship between the speed of car A and car B at t = 1 hr. Explain.
c) State the relationship between the acceleration of car A and car B at t =1 hr.
Explain.
d) What is the relative position of the two cars during the time interval between t =
.75 hr. and t = 1 hr. (i.e., is one car pulling away from the other?) Explain.
278
4. Imagine this bottle filling with water. Sketch a graph of the height as a function of the
amount of water that’s in the bottle.
height
volume
Figure A8. Sketch of bottle and grid for height vs. volume
5. Can you indicate, from the graph below, how many bends there are along the track on
which the car was driven?
Speed (km/hr)
200
180
160
140
120
100
80
60
40
20
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3

(km)
Figure A9. Speed of a race car along a 3 km. track

Note. From “Use of Situations in Mathematics Education” by C. Janvier (1981).
Educational Studies in Mathematics, 12(1), 115, graph 1.1. Copyright 1981 by D. Reidel.
APPENDIX B
QUESTIONS ABOUT FUNCTIONS II: SECOND QUESTIONNAIRE (Q2)

QUESTIONS ABOUT FUNCTIONS II
Second Questionnaire (Q2)
Please write your response to the following questions in the spaces below. Be detailed,
but concise.
1. Describe any insights you gained concerning the concept of function over the past
couple weeks as you worked on the functions project or term paper.
2. What is a function?
3. Describe the ways a function can be represented.
4. Can you devise a specific situation for which a function might provide a
mathematical description?
5. Suppose you were to make a list pairing each student’s first name with that
student’s score on a test. Could this pairing represent a function? Explain why or
why not.
6. A caterpillar is crawling around on a piece of [coordinatized] graph paper, as

represented below. If we were to determine the creature’s location on the paper
with respect to time, would this location be a function of time? Why or why not?
Can time be described as a function of its location? Explain.
a) b)
Figure B1. Path of a caterpillar
280
281
7. One can sometimes tell a great deal about a text (its author, when it was written,
etc.) by various characteristics (e.g., word length, frequency of occurrence of
certain words, etc.). For example, one might use the number of occurrences of
each vowel, a,e,i,o,u, in a text. Suppose that you were trying to analyze the
following text (of course, in actual practice, one uses a much larger amount of
material) by studying the number of occurrences of each vowel.
whatever you think is necessary about your function to specify it. Evaluate your
function once (Breidenbach et al., 1992, pp. 281-282).
8. Are the following descriptions of functions? Explain.
⎧⎪ x 2 + 1, if x ≤ 0
f ( x) = ⎨
⎪⎩ − 2 x + 4 if x > 0
⎧ sin x
⎪ if x ≠ 0
g ( x) = ⎨ x
⎪⎩1 if x = 0
Directions. If possible, describe the following situations using one or more functions. If
not, explain why not.
⎧1 1 1 1 ⎫
9. ⎨ , , , ,...⎬
⎩ 2 4 8 16 ⎭
10. The string “ZYXWVUT”
11. A swimmer starts from shore and swims to the other side of the lake.
12. {2 n > n 2 + 3n : n in [1..100]}
13. Make a sketch of the relationship between velocity and time of a bicyclist as that
bicyclist is approaching the top of a hill, then goes over the top and down.
282
14. Suppose this is the graph of the height as a function of volume as a symmetric
bottle is filling with water. Sketch the shape of the bottle.
height
volume
Figure B2. Graph showing height vs. volume

APPENDIX C
WORKSHEET #1
GRAPHS OF FUNCTIONS
WORKSHEET #1: GRAPHS OF FUNCTIONS
Victor Katz, mathematics historian, claims that the following statement made by
Heytesbury in 1335 provides “in nascent form at least,” the notion of velocity as a
function of time, i.e., velocity changing with time (Katz, 1998, p. 319):
Any motion whatever is uniformly accelerated if, in each of any equal parts of the
time whatsoever, it acquires an equal increment of velocity . . . But a motion is
nonuniformly accelerated . . . when it acquires . . . a greater increment of velocity
in one part of time than in another equal part . . . And since any degree of velocity
whatsoever differs by a finite amount from zero velocity . . ., therefore any mobile
body may be uniformly accelerated from rest to any assigned degree of velocity.
(Heytesbury, in Katz, 1998, p. 319)
Activity 1 (Please do all your work neatly on separate sheets of paper.)
1. Make a detailed sketch for each of the two situations above, i.e., uniform and non-
uniform acceleration. Let time be the independent variable, i.e., that on the horizontal
axis, and velocity be the dependent variable. Be sure to label important aspects of
your graph, explaining each part.
2. Make a detailed sketch explaining the notion of instantaneous velocity as discussed
by Heytsbury in the following reading. Again use time as the independent variable,
but this time use distance traveled on the vertical axis as the dependent variable.
Which modern day calculus concept is he describing?
284
285
In non-uniform motion . . . the velocity at any given instant will be measured by
the path which would be described by the . . . point if, in a period of time, it were
moved uniformly at the same degree of velocity with which it is moved in that
given instant, whatever [instant] be assigned. (in Katz, 1998, p. 318)
Historian A. P. Youschkevitch (1976) claims, “The notion of function first
occurred in a . . . general form the schools of natural philosophy at Oxford and Paris . . .”
(p. 45). The notion that velocity was measurable, in particular instantaneous velocity,
appeared in the beginning of the 14th century (Katz, 1998). The ancient Greeks never
considered velocity or acceleration as independent, measurable quantities. As
mathematicians attempted to quantify Aristotle’s ideas on motion, they began to explore
the idea of representing velocity, as well as other varying quantities, by line segments.
Notions of time, distance, and length of line segments were still considered to be
continuous magnitudes, not discrete numbers. Thus, representing the abstract notion of
velocity, which was clearly continuous, as a line segment seemed reasonable. Velocities
of varying magnitudes would be represented by line segments of different lengths.
Nicole Oresme (1320-1382), a French cleric and mathematician associated with
the University of Paris, carried this idea to its logical conclusion by introducing a two-
dimensional representation of velocity changing with respect to time. His base line, or
horizontal axis, represented instants in time and the vertical or perpendicular represented
velocity or degree of intensity of some other characteristic.

286
The first reading below is from Oresme, N. (c.1350) in Clagett (1968). In it
Oresme explains why he thinks that line segments are appropriate representations for
measures of intensity of a characteristic or quality.
Activity 2
Again, intensity is that according to which something is said to be “more such and
such,” as “more white” or “more swift.” Since intensity, or rather the intensity of
a point, is infinitely divisible in the manner of a continuum in only one way,
therefore there is no more fitting way for it to be imagined than by that species of
a continuum which is initially divisible and only in one way, namely by a line.
And since the quantity or ratio of lines is better known and is more readily
conceived by us—nay the line is in the first species of continua, therefore such
intensity ought to be imagined by lines and most fittingly by those lines which are
erected perpendicular to the subject. The consideration of these lines naturally
helps and leads to the knowledge of any intensity . . . Therefore, equal intensities
are designated by equal lines, a double intensity by a double line, and always in
the same way if one proceeds proportionally.
3. In your own words, explain why Oresme thought the lines are appropriate
representations for measures of intensity of a subject.
From these straight lines, Oresme constructed what he called configurations
(Katz, 1998), a geometrical figure consisting of all perpendicular lines drawn over the
base line. If the dependent variable was velocity, the base line represented time and the
perpendiculars represented the velocities at each instant. In our modern rectangular

287
coordinate system, our ordinates (y-coordinates) would be the endpoints of Oresme’s
perpendicular line segments. A configuration might look something like this:
time
Figure C1. A possible configuration of Oresme
4. Suppose we had two variables: the number of years a couple is married and the level
of marital satisfaction. In the table below is data collected from 10 couples (Hinkle et
al., 1998, p. 128).
Years of 2 3 4 6 7 8 9 10 12 14
marriage
Level of 55 52 48 45 43 30 27 30 33 36
marital
satisfaction
=Suppose further that you were tutoring a middle school student in the creation of graphs.
How might you use Oresme’s ideas of lines representing intensity to explain to this
student how to graph this data? Create two graphs:
i. one showing how Oresme would represent this data.
ii. a modern-day scatterplot of this data.

288
5. Create three other scenarios that you might use with middle school students in an
exercise like #4 above. Repeat exercise #4 for one of these scenarios.
In the reading below, Oresme uses the term “difform” to mean rate of change, i.e.,
if velocity was “uniformly difform,” the velocity was changing at a constant rate.
Activity 3
Reading from Oresme (1350):
And so every uniform quality is imagined by a rectangle and every quality
uniformly difform terminated at no degree is imaginable by a right triangle.
Further, every quality uniformly difform terminated in both extremes at some
degree is to be imagined by a quadrangle having right angles on its base and the
other two angles unequal. Now every other linear quality is said to be “difformly
difform” and is imaginable by means of figures otherwise disposed according to
manifold variation. (p. 193)
6. Make a sketch of a configuration for each of the following:
(a) Uniform quality
(b) Uniformly difform quality (two possibilities – one whose initial value of the
output is 0, and one whose initial (and final values are greater than 0)
(c) Difformly difform quality

289
7. Given the same two variables as in Activity 4 (i.e., the marital satisfaction vs. number
of years of marriage). Make a new table of data for each of the configurations you
sketched in problem #6 above.
8. Suppose again that you were tutoring a middle school student in math. Create another
real-world scenario for each configuration and explain how you would clarify the
difference between each type of rate of change.
In the reading below, Oresme discusses in an indirect way our notion of slope.
Activity 4
Reading from Oresme in Clagett, p. 193:
A uniform quality is one which is equally intense in all parts of the subject. while
a quality uniformly difform is one in which if any three points [of the subject line]
are taken, the ratio of the distance between the first and the second to the distance
between the second and the third is as the ratio of the excess in intensity of the
first point over that of the second point to the excess of that of the second point
over that of the third point, calling the first of those three points the one of
greatest intensity.
9. (a) How is Oresme’s description of a constant rate of change different than our
modern-day version of slope? You might find it helpful to refer to the sketch in
(b) below.
(b) Using the sketch below, verify Oresme’s claim stated in the reading, i.e., Given
Δ ABC. Let AC, DG, and FE be perpendicular to AB. Let HE be drawn parallel to
290
line DF and also KG parallel to AD. You need to verify that “the ratio of the
distance between the first and the second [AD] to the distance between the second
and the third [DF] is as the ratio of the excess in intensity of the first point over
that of the second point [CK] to the excess of that of the second point over that of
the third point [GH].” (p. 193)

C
G
K
E
H
D B
A F
Figure C2. Oresme’s triangle

Note. From Nicole Oresme and the Geometry of Qualities and Motions (p. 193),
by M. Clagett (1968). Madison, WI: University of Wisconsin Press. Copyright 1968 by
the Regents of the University of Wisconsin. Reprinted with permission.
10. If the graph above represented the marital satisfaction data, how would you interpret
the ratio GH/HE?
Activity 5
Suppose you wanted to determine your average speed on a bicycle if you were
accelerating uniformly over an (very brief) interval of time. Using Oresme’s graphing
technique would result in a graph like the following:

291
c
20 M/SEC
e
Va d
a b
t
4 SEC
Figure C3. The mean speed rule

Note. From “The Concept of Variation and Misconception in Cartesian Graphing”
by J. Clement (1989). Focus on Learning Problems in Mathematics, 11(2), 84. Copyright
1989 by Center for Teaching/Learning of Mathematics. Reprinted with permission.
11. a) What do the vertical lines represent?
b) What do the areas between the vertical lines represent?
c) What does the horizontal dotted line represent?
12. a) What would be the distance traveled over the entire interval? Explain.
b) How does this distance compare to that traveled over the same amount of time at
the uniform velocity represented by the horizontal line interesting the curve at t =
2?
13. Make a conjecture for the general case (i.e., time interval of b seconds, max speed of
y meters per second) concerning the total distance traveled after uniform acceleration
vs. distance traveled at uniform velocity, in particular, at the velocity at the midpoint
of the time interval. Prove your conjecture using a sketch similar to that above.
292
14. a) Show that under your assumption above, if you divide the time interval into four
equal subintervals, the distances covered in each interval will be in the ratio
1:3:5:7.
b) Generalize this statement of a division of the time interval into n equal
subintervals. (Katz, 1998, p. 323)
Summary and Reflection
Please type up a page (or so), answering the following questions:
1. Do you think Oresme’s approach is more or less intuitive than our modern-day
technique of graphing? Consider how a very young student might represent such
data if given absolutely no prompts.
2. a) Has Oresme’s approach enriched your understanding of graphs in any way?
Explain.
b) Have the above exercises enriched your understanding of the nature of graphs in
any way? Explain.
3. Think about the graphs that were on the initial questionnaire you wrote. They are
repeated below. Would you change any of your original answers now? Explain.
(The sketches are reproduced below for your convenience. You need only
comment on the ones for which your answers would be different from your
original ones.)
APPENDIX D
WORKSHEET #2
THE ROLE OF FERMAT’S ANALYTIC GEOMETRY

WORKSHEET #2 THE ROLE OF FERMAT’S ANALYTIC
GEOMETRY
Background Information
The following to two readings by noted historians Israel Kleiner (1989) and A. P.
Youschkevitch (1976) provide us with interesting background information.
From Kleiner: p. 283:
The 17th century witnessed the emergence of modern mathematized science and
the invention of analytic geometry. Both of these developments suggested a dynamic,
continuous view of the functional relationship as against the static, discrete view held by
the ancients. In the blending of algebra and geometry, the key elements were the
introduction of variables and the expression of the relationship between variables by
means of equations. The latter provided a large number of examples of curves (potential
functions) for study and set the final stage for the introduction of the function concept.
What was lacking was the identification of the independent and dependent variables in an
equation. Read carefully the following development according to noted historian, A.P.
Youschkevitch.
Note. From “The concept of function up to the middle of the 19th century” by
A.P. Youschkevitch. In C. Truesdell (Ed.), Archive for history of exact sciences, 16(1),
37-85. New York: Springer-Verlag. pp. 50-52. Reprinted with kind permission from
Springer Science and Business Media.
294
295
A new interpretation of functionality came to the fore in the 17th century decisive
significance for the further development of the doctrine of functions was played,
on the one hand, by the impetuous growth of computational mathematics and, on
the other, by the creation of literal, symbolic algebra along with the corresponding
extension of the concept of number, so as, by the end of the 16th century, to
embrace not only the whole field of real numbers but also imaginary and complex
numbers. These were, so to say, preliminaries in mathematics itself to the
introduction to the concept of function as a relation between sets of numbers
rather than “quantities” and for analytical representation of functions by formulae.
. . . What should be especially emphasized, though, is the introduction of
numerous signs for mathematical operations and relations (in the first place, those
of addition, subtraction, of powers and of equality) and, above all, of signs for
unknown quantities and parameters, which Viète in 1591 denoted by vowels A, E,
I, . . . and consonants B,G,D, . . . of the Latin alphabet, respectively. The
importance of this notation, which, for the first time ever, made it possible to put
on paper in symbolic form algebraic equations and expressions containing
unknown quantities and arbitrary coefficients (a word also originating with Viète)
could hardly be overestimated. However, the creator of the new algebra did not
use his remarkable discovery to further the concept of function; “functional
thought” was not characteristic in his mind.
Viète’s symbolism suffered from serious shortcomings and soon was amended by
number of scholars, then extended beyond the realms of algebra and use in the
296
infinitesimal calculus. Descartes, Newton, Leibniz (who attached utmost
importance to the appropriate selection of signs), Euler and other scholars of the
highest caliber participated in this process of perfecting mathematical symbolism
and this process continues in our time in all branches of mathematics.
On the other hand, in the exact sciences of former times, especially from the
beginning of the 17th century, the new conception of quantitative laws of nature as
establishing functional relations between numerical values of physical quantities
had been gathering strength in ever-increasing measure and becoming more and
more distinctive. In this process the creation of a broader and broader field of
physical metrology with the introduction of quantitative measures of experiments
and observations, brought about the invention of various scientific instruments.
Among the sciences, mechanics, overtaking astronomy came to the fore and, with
it, its new branch, dynamics, soon to be joined by celestial mechanics. To study
the relation between curvilinear motion and the forces affecting motion had
become the chief problem of science. This problem gave rise to a series of
problems in infinitesimal analysis, the solution of which had to be carried through
to numerical answers.
As a consequence of all this, a new method of introducing functions was brought
into being, to become for a long time the principal method in mathematics and
especially, in its applications. As before, functions not infrequently were
introduced verbally; by a graph; kinematically; and, as before, tables of functions
continued to be used most extensively. However, in theoretical research, the

297
analytical method of introducing functions by means of formulae and equations
came to the foreground.
We are able to tell almost exactly when this reversal of ideas took place. Even by
the turn of the 16th century functions were being introduced only by means of old
methods. . . .
But then, only 15 to 20 years after this, independently of each other, both Fermat
and Descartes in applying the new algebra to geometry presented the analytical
method of introducing functions, thus opening a new era in mathematics.
What follows are samples of the work, first of Fermat, then Descartes, to give you a feel
for these developments.

298
Worksheet Activities
Please do all work neatly on your own paper.
FERMAT in Introduction to plane and solid loci: (written before 1637)
“As soon as two unknown quantities appear in a final equation, there is a locus,
and the end point of one of the two quantities describes a straight or curved line.”
(Youschkevitch, 1976, p. 52)
In the following brief reading, assume B and D are known, but arbitrary
D
quantities. Consider the constant ratio as the point Z moves to the right along the line
B
NZM.
Let NZM be a straight line given in position, with the point N fixed. Let NZ be
equal to the unknown quantity A, and ZI, the lines drawn to form the angle NZI,
the other unknown quantity E. If D times A equals B times E, the point I will
describe a straight line given in position. (Katz, 1998, p. 435)
He assumes also that ∠INZ is fixed.
T
A
N Z W M
Figure D1. Fermat’s analysis of the equation D times A equals B times E

Note. From A History of Mathematics, An Introduction. (p. 435), by V. Katz
(1998). Reading, MA: Addison Wesley Longman, Inc. Copyright 1998 by Addison
Wesley Longman, Inc. Reprinted with permission.
299
1. If Fermat used Viète’s convention of vowels representing unknowns and consonants
representing known quantities, translate the last statement above using modern
notation.
D
Note that Fermat uses a single axis, a constant ratio , and an equation, which he
B
wants to show represents a straight line. He generates his curve by the motion of the
“endpoint I of the variable line segment ZI as Z moves along the given axis” (Katz, 1998,
p. 435). His method is unlike the modern method of plotting points with respect to two
axes.
2. Explain why that for any such points Z and W along the segment NM and the
corresponding points I and T on NI, ΔNIZ ~ ΔNTW .
3. Create a straight line using Fermat’s method with D =3 and B =2. Show at least 3
different positions of Z.
N M
Fermat also wanted to show that, in modern notation: x 2 = Dy determines a
parabola. “If Aq = DE, point I lies on a parabola” (Katz, 1998, p. 435). He began with the
two line segments NZ and ZI as he did in the linear case, but this time let NZ ⊥ ZI . He
draws NP parallel to ZI and then claims that the parabola with vertex at N, axis NP, and
latus rectum (segment through the focus with endpoints on the parabola) D is the
parabola determined by the given equation.

300
P I
A
N Z
Figure D2. Fermat’s analysis of the equation Aq equals D times E

Note. From A History of Mathematics, An Introduction. (p. 435), by V. Kat
3. Let D = 4 and pick a point on the parabola, say (2,1). Show a geometric interpretation
to the equation, i.e., show that the area on left hand side of the equation, i.e., the
square, is equal to the area on the right hand side, i.e., the rectangle. Sketch in the
actual square and rectangle, with the length of each lying on NZ. Repeat for another
set of points.
4. Fermat analyzed the two-point case of a theorem from Apollonius: If from any
number of given points, straight lines are drawn to a point, and if the sum of the
squares of the lines is equal to a given area, the point lies on a circumference [circle]
given in position (Katz, 1998, p. 432).
In other words, given two points, A and B. Draw straight lines from these to a point,
P. If the sum AP 2 + BP 2 is equal to a given area, then all the points P that satisfy
this condition lie on a circle.
Fermat bisects the segment AB, calling the midpoint E. He finds the point I such that
( )
2 AE 2 + IE 2 = a given area, M. With IE as radius and E as center, he describes a circle.
301
He then shows that any point P on this circle satisfies the condition, i.e., AP 2 + BP 2 =
M.
His analysis is below (Katz, 1998, p. 432). PROVIDE REASONS FOR EACH OF HIS
STEPS.
(Note that AZ ⊥ PZ and BV ⊥ PZ by construction).
V
E
I A B
Z
(a) AP 2 + BP 2 = PZ 2 + AZ 2 + PV 2 + BV 2
= (PE + EZ )2 + AE 2 − EZ 2 + (PE − EV ) + BE 2 − EV 2
2
(b)
(c) = PE 2 + 2 PE ⋅ EZ + EZ 2 + AE 2 − EZ 2 + PE 2 − 2 PE ⋅ EV + EV 2 + BE 2 − EV 2
(d) = 2 PE 2 + 2 AE 2
(e) = 2( AE 2 + IE 2 )
Figure D2. Fermat’s analysis of a special case of a theorem of Apollonius

302
The significant aspects of the above proof are that the circle is determined by the sums of
squares of two changing quantities, AP and BP, and that the point I is determined in terms
of its distance from the “origin” E (Katz, 1998).
5. Using the above analysis, determine the equation of the circle that solves the problem.
(Hint: Let the coordinates A and B in the figure be (-a,0) and (a,0), respectively and
let the coordinates of P be (xo,yo). Begin with the equation AP 2 + BP 2 = M )

(problem from Katz, 1998, p. 462).
P (xo,yo)
V
E
I A B
Z
Figure D3. Fermat’s analysis of a special case of a theorem of Apollonius

6. Comment about the following statement, as related to the work of Fermat.
Variables are not functions. The concept of function implies that a unidirectional
relation between an ‘independent’ and ‘dependent’ variable. But in the case of
variables as they occur in mathematical or physical problems, there need not be
such a division of roles. And as long as no special independent role is given to

303
one of the variables involved, the variables are not functions but simply variables.
(Bos, in Kleiner, 1989, p. 283)
Write a brief (one page) summary:
1. Reflect on your understanding of the concept of function and the ideas
presented in this worksheet. Comment on any connection you may find.
2. Did this worksheet led to new insights concerning the concept of function?
APPENDIX E
WORKSHEET #3
DESCARTES AND THE LOCI WITH RESPECT TO THREE OR FOUR LINES
Note. From “Functions,” by D. Anderson, R. Berg, A. Sebrell, & D. Smith, 2005. In
Historical modules for the teaching and learning of mathematics (compact disc) edited
by V. J. Katz and K. D. Michalowicz (pp. 90, 91, 94, 97). Washington, DC: The
Mathematical Association of America. Copyright 2005 by the Mathematical Association
of America. Reprinted with permission.

APPENDIX F
WORKSHEET #4
LEIBNIZ AND THE HISTORY OF FUNCTIONS

WORKSHEET #4
LEIBNIZ AND THE HISTORY OF FUNCTIONS
Gottfried Leibniz was first to use the term “function” in 1694 to describe six
specific line segments associated with a variable point on a given plane curve (Usiskin,
Peressini, Marchhisotto, & Stanley, 2003). Though he was not responsible for modern
function notation, “it is to him that the word ‘function’ in much the same sense as it is
used today, is due” (Boyer, 1968, p. 444).
The Leibniz segments (Usiskin, Peressini, Marchhisotto, & Stanley, 2003, p. 130)
Given a curve C in a coordinate plane and a point P on C, let T be the point of

intersection of the tangent line to C at P with the horizontal axis, let N be the point of
intersection of the normal line to C at P with the horizontal axis, and let Q be the foot of
the perpendicular from the horizontal axis through P. See figure below.
T Q N
Figure F1. Leibniz’s six functions of a point.

Note. From Mathematics for High School Teachers, An Advanced Perspective.
(p. 130) by Usiskin et al., 2003. Reprinted with permission of Pearson Education, Inc.,
Upper Saddle River, New Jersey.
311
312
Leibniz introduced the following six “functions” of the point P on C.
i. The abscissa at P is the line segment OQ .
ii. The ordinate of P is the line segment PQ .
iii. The tangent at P is the line segment PT .
iv. The normal at P is the line segment PN
v. The subtangent at P is the line segment TQ .
vi. The subnormal at P is the line segment NQ .
1. (a) Leibniz was interested in the notion that the slope of PT is equal to three
different ratios of these six functions. Identify these ratios.
(b) Prove that the length PQ of the ordinate at P is the geometric mean of the length
QT of the subtangent and the length QN of the subnormal at P.
2. Find the lengths of all six of the line segments defined above for the point P = (3,6)
on the parabola defined by the equation
2y = x2 + 3.
3. Recall that a parabola is the set of points equidistant from a given point (its focus) and
a given line (its directrix). For the parabola described by y2 = 2px, explain why the
⎛p ⎞ p
focus is at ⎜ ,0 ⎟ and the directrix is the line x = − . For a given point P = (x1, y1)
⎝2 ⎠ 2
on this parabola, find the x-coordinates of the points T, Q, and N in terms of y1.
4. Do you think there is any connection between Leibniz’s use of the term “function”
and our modern usage?

313
Write a brief (one page) summary:
1. Reflect on your understanding of the concept of function and the ideas
2. Did this worksheet led to new insights concerning the concept of function?
APPENDIX G
WORKSHEET #5
RABBITS, RABBITS EVERYWHERE!
by V.J. Katz and K. D. Michalowicz (pp. 72-73, 76). Washington, DC: The Mathematical
Association of America. Copyright 2005 by the Mathematical Association of America.
Reprinted with permission.

APPENDIX H
WORKSHEET #6
EVEN MORE FIBONACCI:
FINDING A FUNCTION TO REPRESENT THE FIBONACCI SEQUENCE
Note. From “Functions,” by D. Anderson, R. Berg, A. Sebrell, and D. Smith, 2005. In
by V.J. Katz and K. D. Michalowicz (pp. 81-83). Washington, DC: The Mathematical

APPENDIX I
WORKSHEET #7
EULER ON FUNCTIONS
by V.J. Katz and K. D. Michalowicz (pp. 59-62). Washington, DC: The Mathematical

APPENDIX J
WORKSHEET #8
FOURIER SERIES REPRESENTATIONS OF FUNCTIONS
by V.J. Katz and K. D. Michalowicz (pp. 99, 103, 104). Washington, DC: The
Mathematical Association of America. Copyright 2005 by the Mathematical Association
of America. Reprinted with permission.
“The Arithmetization of Analysis” reading from A History of Mathematics (pp. 598-599)
by C. Boyer, 1968, New York: John Wiley & Sons. Copyright 1968 by John Wiley &
Sons, Inc. Reprinted with permission.

APPENDIX K
WORKSHEET #9
DEFINITIONS OF FUNCTIONS
Worksheet #9: Definitions of Functions
Euler (1748): A function of a variable quantity is an analytic expression composed in any
way whatsoever of the variable quantity and numbers or constant quantities. Hence every
analytic expression, in which all component quantities except the variable x are
constants, will be a function of that x; thus a + 3 x; ax − 3 x 2 ; ax + b a 2 − x 2 ; c x ; etc. are
functions of x. (Katz & Michalowicz, 2005)
Euler (1755): If however, some quantities depend on others in such a way that if the
latter are changed the former undergoes changes themselves then the former quantities
are called functions of the latter quantities. This is a very comprehensive notion and
comprises in itself all the modes through which one quantity can be determined by others.
If, therefore, x denotes a variable quantity then all the quantities which depend on x in
any manner whatsoever or are determined by it are called functions. (Rüthing, 1984, pp.
72-73)
1. Compare Euler’s second definition with his first. Recall that you worked with the first
definition in an earlier set of worksheets. What idea expressed in the first definition is
absent in the second definition?
2. Recall the “function situations” in the function questionnaire you took before you
started this project. Several are listed below. Which of the following would be
functions as defined originally by Euler (1748). (We’ll call this definition “early
Euler”). Which of the following might be functions as defined above by Euler (late
Euler). Briefly explain your reasoning.
336
337
(a) {(x, 3x + 2): x in the set of all integers}

(b) The club member’s dues status.
Name Owed
John $15
Julie 10
Jen 0
Dave 7.50
Brittany 10
Alex 5
Mary Ann 25
Sam 20
Sally 17
Figure K1. Table showing club members and dues owed
x = t3 + t ⎫⎪
c) ⎬ t , a real number
y = 1 − 3t + 2t 4 ⎪⎭
d) {3n + n 2 : n in [1...100]}
e) A swimmer starts from shore and swims to the other side of Walden Pond
f) {2n > n2 + 3n: n in [1 ..100]}
3. Conjecture about what might have prompted Euler to change his definition.
4. Please read the article “What prompted Euler to change his definition of function?”
Reflect on your own understanding of function. Did you gain any insight concerning
the concept after reading the article?
***************
J. L. Lagrange (1797). One calls function of one or several quantities any expression for
calculation in which these quantities enter in any many whatsoever, mingled or not with
338
some other quantities which are regarded as being given and invariable values, whereas
the quantities of the function can take all possible values.
...
We denote, in general, by the letter f or F place before a variable any function of this
variable, that is to say any quantity depending on this variable and which varies with it
together according to a given law. (Rüthing, 1984, p. 73)
5. How does this definition of Lagrange compare with each of those given by Euler? To
which is it more familiar, early Euler or late Euler?
**************
J.B.J. Fourier (1822). In general, the function f(x) represents a succession of values or
ordinates each of which is arbitrary. An infinity of values being given to the abscissa x,
there are an equal number of ordinates f(x). All have actual numerical values, either
positive or negative or null. We do not suppose these ordinates to be subject to a common
law; they succeed each other in any manner whatever, and each of them is given as it
were a single quantity. (Rüthing, 1984, p. 73)
6. Compare and contrast this definition with that of early Euler.
7. Repeat exercise 2 above using Fourier’s definition.
8. Read the few paragraphs entitled Fourier Series and summarize the main effect of
Fourier’s work on the development of the function concept.
*****************
339
G. L. Dirichlet (1837). Let us suppose that a and b are two definite values and x is a
variable quantity which is to assume, gradually, all values located between a and b. Now,
if to each x there corresponds a unique, finite y in such a way that, as x continuously
passes through the interval from a to b, y = f(x) varies likewise gradually, then y is called
a continuous . . . function of x for this interval. It is, moreover, not at all necessary, that y
depend on x in this whole interval according to the same law; indeed, it is not necessary
to think of only relations that can be expressed by mathematical operations. (Rüthing,
1984, p. 74)
9. Create three functions that would agree with Dirichlet’s characterization of functions,
but not that of early Euler.
10. Please read the excerpt entitled Cauchy and Dirichlet.
11. Read the excerpt entitled Riemann and Weierstrass and explain in your own words
the difference between classical analysis and the theory of functions.
******************
H. Hankel (1870). . . . Moreover it does not matter at all from where and how f(x) is
determined, whether by an analytical operation of quantities r in other ways. The value of
f(x) must only be uniquely determined everywhere. (Rüthing, 1984, p. 75)
G. Peano (1911). . . . the function is a special relation, by which to each value of the
variable there corresponds a unique value. (Rüthing, 1984, p. 76-77)

340
E. Goursat (1923). The modern definition of the word function is due to Cauchy and
Riemann. One says that y is a function of x if a value of x corresponds to a value of y.
One indicates this dependency by the equation y = f(x). The majority of functions which
we shall examine are defined analytically, that is to say by the indication of the
operations which one would have to carry out in order to deduced the value of y from x,
but very frequently, this does not enter the argument. (Rüthing, 1984, p. 77)
N. Bourbaki (1939). Let E and F be two sets, which may or may not be distinct. A
relation between a variable element x of E and a variable element y of F is called a
functional relation if, for all x ∈ E , there exists unique y ∈ F , which is in the given
relation with x. (Rüthing, 1984, p. 77)
13. Please redo exercise #2 one more time, now using the above modern definitions.
They are all similar to each other, so chose the one that makes the most sense to you.
14. Write a brief (one page) summary:
a. Reflect on your understanding of the concept of function and the ideas
b. Did this worksheet led to new insights concerning the concept of function?
c. Comment on the closing words of Richard Courant expressed in the excerpt
below entitled Summary and Conclusion.
*************************
341
Summary and Conclusion
Kleiner (1989) summarizes the development of the function concept succinctly.
He claims that implicit manifestations of the function concept date as far back as 2000
B.C., though its explicit form did not emerge until the beginning of the 18th century. He
views the evolution of the function concept as “a tug of war between two elements, two
mental images: the geometric (expressed in the form of a curve) and the algebraic
(expressed as a formula – first finite and later allowing infinitely many terms . . .)” (p.
282). Later, the “logical” definition of function appears and the geometric conception is
gradually abandoned.
Siu asks a pertinent question at the conclusion of his paper (Siu, 1995, p. 188):
“What implications in teaching can we learn from this tug-of-war . . .?” He concludes
with thoughtful insight with the words of Richard Courant (p. 118):
The presentation of analysis as a closed system of truths without reference to their
origin and purpose has, it is true, an aesthetic charm and satisfies a deep
philosophical need. But the attitude of those who consider analysis solely as an
abstractly logical, introverted science is not only highly unsuitable for beginners
but endangers the future of the subject.

APPENDIX L
READINGS FOR WORKSHEET #9

READINGS FOR WORKSHEET #9
What prompted Euler to change his definition of function?
Leonhard Euler, in his book Introduction in Analysis Infinitorum written in1748,
claimed that Analysis is the science of variables and their functions. His entire approach
was algebraic, not geometric; not a single drawing appears (Kleiner, 1989, p. 184). His
definition of function was almost identical to Bernoulli’s. He used the term “analytic
expression” and though he does not define it, he states the admissible ones: the four
algebraic operations, roots, exponentials, logarithms, trig functions, derivatives, and
integrals. He classifies functions as being either algebraic or transcendental and claims
that any function can be expanded in a power series. His work exhibited one of the
earliest treatment of trig functions as numeric ratios and was the earliest interpretation of
logarithms as exponents (Kleiner, 1989, p. 184). Interestingly, Euler’s definition of a
continuous function was one that had the same analytic expression over the entire
domain. Historian Hawkins (1975) sums up Euler’s contributions to the development of
the function concept: “Although the notion of function did not originate with Euler, it
was he who first gave it prominence by treating the calculus as a formal theory of
functions” (p. 3).
The Vibrating String Problem. The next major influence on the development of
the function concept was the controversy over the solution to the vibrating string
problem. Given an elastic string with fixed ends, which is pulled into some initial shape
then released to vibrate. The problem is to determine the function that describes the shape
of a string at time t. The controversy centered on the meaning of “function” and the type
343
344
of functions which could be allowed in analysis from the standpoint of a mathematician
(d’Alembert), a physicist (Bernoulli), and a mathematical physicist (Euler). In 1747,
d’Alembert solved the problem by showing that the motion of the string is governed by
the so-called “wave” partial differential equation
∂ 2y ∂2 y
=a 2 (a is constant, y(0,t) = y(L,t) = 0).
∂t 2 ∂x 2
In this representation y, the dependent variable, represents the displacement from
equilibrium, x represents the distance from the origin and t indicates time. The most
“general solution” is y(x,t) = f(x + t) + f(x – t), where f is an arbitrary function. The only
restrictions on it were that it be periodic, odd, and everywhere (twice) differentiable.
D’Alembert thought that the function describing the initial form of the string must be a
single analytic expression, i.e., given by the same formula over the entire length of the
string.
In 1748 Euler solved the same problem and showed that his solution gives the
shape of the string for different values of t even when the initial shape is NOT given by a
single formula. Euler argued that the initial shape can be given by different analytic
expressions in different subintervals, or even drawn free-hand. Different pieces of the
string might be described by circular arcs of varying radii, for example. The belief at the
time, however, was that if two analytic expressions agreed on an interval, they agreed
everywhere. In other words, if a single analytic expression determined the shape of the
entire curve, looking at a single interval, no matter how small, would be sufficient to
determine behavior on the entire string. If the initial shape of the string were given by
345
several analytic expressions or drawn free-hand, it could not possibly be given by a single
expression. Thus d’Alembert’s solution could not account for all possibilities.
Bernoulli entered the debate, which lasted for several more years. The major
outcome of this debate was to extend the function concept to include:
• functions defined piecewise by analytic expressions in different intervals
• functions drawn freehand and possible not given by any combination of
analytic expressions. (Kleiner, 1989, p. 288)
Another eventual result of this debate was the change in Euler’s definition of
function. Recall that his 1748 definition used the term “analytic expression” but the later
one (1755) did not.
Fourier series
Fourier studied heat conduction and his work, in addition to winning him a prize
from the Institut de France in 1812, was “revolutionary in the evolution of the function
concept” (Kleiner, 1989, p. 289). He claimed that any function f(x) defined over an
interval is representable over this interval by a series of sines and cosines. Both Lagrange
and Euler, among others, had previously recognized Fourier’s results as true for certain
functions. Fourier’s claim that it was true for all functions was revolutionary. His
definition (1822) of function is as follows:
In general, the function f(x) represents a succession of values or ordinates each of
which is arbitrary. An infinity of values being given to the abscissa x, there is an
equal number of ordinates f(x). All have actual numerical values, either positive or
negative or null. We do not suppose these ordinates to be subject to a common

346
law; they succeed each other in any manner whatever, and each of them is given
as if it were a single quantity. (Katz, 1998, p. 724)
The influence of Fourier’s work is as follows:
• It disproved the belief that if two analytic expressions agreed on an interval,
they agreed everywhere. He calculated the Fourier coefficients for small
values of n for a great variety of functions and noted the close agreement in an
interval, but not outside the interval, between the initial segments of the
Fourier series and the function values of the given function. (Kleiner, 1989, p.
290).
• There was a renewed emphasis on analytic expressions (Siu, 1995, p. 112).
• The recognition that an “arbitrary” function is representable by an analytical
expression (Siu).
• Reexamination of the function concept as a whole (Siu).
Both Siu and Kleiner claim that Fourier’s work in heat conduction set the analytic
expression of a function on (at least) equal footing with its geometric representation. Siu
compares the vibrating string problem with the heat conduction problem:
The shape of the string (geometry) is visible, while temperature distribution
(algebra) is not. This may explain the freeing from geometric perception of a
function and the emergence of a general notion of function in the nineteenth
century. (Siu, 1995, p. 111)
He also gives profound insight into the problem of accepting a new concept, both
historically and in students:

347
Why was the “Eulerian” concept of function maintained so long after the
realization that is was inadequate? What lesson do we learn from this experience?
(If only a particular form is used, students unconsciously accept that particular
form as the definition. We witness the same psychological effect in
mathematicians of the seventeenth/eighteenth centuries. A new concept receives
recognition only when it is relevant to current usage. This is as true in research as
it is in teaching). (p. 113)
Cauchy and Dirichlet
Cauchy was one of the first mathematicians to usher in a “spirit of rigor” in
mathematical analysis and defined the concepts of continuity, differentiability, and
integrability of a function in terms of limits (Kleiner, 1989, p. 291). His definition of
function in 1821 was not that different from his predecessors:
When the variable quantities are linked together in such a way that, when the
value of one of them is given, we can infer the values of all the others, we
ordinarily conceive that these various quantities are expressed by means of one of
them which then takes the name of independent variable; and the remaining
quantities, expressed by means of the independent variable, are those which one
calls functions of this variable. (p. 291)
Another main player of the time was Dirichlet. He questioned Fourier’s result and
in 1829, gave sufficient conditions for Fourier’s representability of functions, i.e., that the
function needs to have only finitely many discontinuities and finitely many maxima and
348
minima on the interval. Dirichlet was the first to take seriously the notion of function as
an arbitrary correspondence (Siu). He is known for the celebrated “Dirichlet function:
⎧c , x is rational
D(x) = ⎨
⎩d , x is irrational
This function was the first explicit example(s) of one not given by an analytic
expression, nor was it a curve drawn freehand. It was also the first example of a function
discontinuous everywhere. It also illustrated the concept of a function as an arbitrary
pairing (Kleiner, 1989, p. 292). Noteworthy in his definition was the explicit restriction
of the domain to an interval . . . also a first.
Riemann and Weierstrass
In 1854, Riemann dealt with the representations of functions in Fourier series.
The coefficients of a Fourier series are given by integrals. Cauchy had developed his
integral only for continuous functions, but Riemann extended this concept to
discontinuous functions. As a result, he enlarged the class of functions representable by
Fourier series. Both Riemann and Weierstrass delighted in studying irregularities and
discovering exceptions in analysis (Kleiner, 1989, p. 292). Weierstrass found an example
of a continuous, nowhere-differentiable function and thus began the “disengagement of
the continuous from the differentiable” in analysis (Kleiner, 1989, p. 293).
These “pathological” examples of functions characterized the change in emphasis
in analysis during the late 19th century:
The main difference between methods of studying functions within the framework
of mathematical analysis and the theory of functions is that classical analysis

349
deduces properties of any function starting from the properties of those analytical
expressions and formulae by which it is defined, while the theory of functions
determines the properties of function starting from that property which a priori
distinguishes the class of functions considered. (Luzin, as cited in Siu, 1995, pp.
114-115)
APPENDIX M
LETTERS OF PERMISSION TO REPRINT

REFERENCES
REFERENCES
Anderson, D. L., Berg, R., Sebrell, A. W., & Smith, D. (2004). Functions. In V. Katz &
K. Michalowicz (Eds.), Historical modules for the teaching and learning of
mathematics (pp. 56-62, 69-84, 89-105). Washington, DC: Mathematical
Association of America.
Arcavi, A. (1985). History of mathematics as a component of mathematics teachers
background, Part I. Unpublished doctoral dissertation, Weizmann Institute of
Science, Rehovot, Israel.
Arcavi, A., Bruckheimer, M., & Ben-Zvi, R. (1982). Maybe a mathematics teacher can
profit from the study of the history of mathematics. For the Learning of
Mathematics, 3(1), 30-37.
Artigue, M. (1992). Functions from an algebraic and graphic point of view: Cognitive
difficulties and teaching practices. In G. Harel & E. Dubinsky (Eds.), The concept
of function: Aspects of epistemology and pedagogy. MAA Notes, 25 (pp. 109-132).
Washington, DC: Mathematical Association of America.
Asiala, M., Brown, A., DeVries, D., Dubinsky, E., Mathews, D., & Brown, K. (1997). A
framework for research and curriculum development in undergraduate
mathematics education. In E. Dubinsky, D. Mathews, & B. Reynolds (Eds.),
Readings in cooperative learning for undergraduate mathematics. MAA Notes, 44
(pp. 37-53). Washington, DC: Mathematical Association of America.
366
367
Asiala, M., Cottrill, J., Dubinsky, E., & Schwingendorf, K. (1997). The development of
students’ graphical understanding of the derivative. Journal of Mathematical
Behavior, 16(3). Retrieved on January 26, 2006, from
http://www.math.kent.edu/~edd/publications.html#C.)%20Mathematics%20
Education%20-%20Refereed
Avital, S. (1994). History of mathematics can help improve instruction and learning. In F.
Swetz, J. Fauvel, O. Bekkin, B. Johansson, & V. Katz (Eds.), Learn from the
masters (pp. 3-12).Washington, DC: Mathematical Association of America.
Baker, B., Coolery, L., & Trigueros, M. (2000). A Calculus Graphing Schema. Journal
for Research in Mathematics Education, 31(5), 557-578.
Barbin, E. (2000a). Integrating history: Research perspectives. Introduction. In J. Fauvel
& J. van Maanen (Eds.), History in mathematics education, The ICMI Study (pp.
63-66). Boston: Kluwer Academic.
Barbin, E. (2000b). The historical dimension: From teacher to learner. In J. Fauvel & J.
van Maanen (Eds.), History in mathematics education, The ICMI Study (pp. 66-
70). Boston: Kluwer Academic.
Barbin, E. (2000c). Suggestions for future research. In J. Fauvel & J. van Maanen (Eds.),
History in Mathematics Education, The ICMI Study (p. 90). Boston: Kluwer
Academic.
Battista, M. T. (1999a). Notes on the constructivist view of learning and teaching
mathematics. Kent, OH: Kent State University.

368
Batttista, M. T. (1999b). The constructivist theory for how children learn mathematics.
Kent, OH: Kent State University.
Bell, A., & Janvier, C. (1981). The interpretation of graphs representing situations. For
the Learning of Mathematics, 2, 34-42.
Bell, E. T. (1945). Development of mathematics. New York: McGraw-Hill.
Boero, P., Pedemonte, B., & Robotti, E. (1997). Approaching theoretical knowledge
through voices and echoes; a Vygotskian perspective. International Conference
for the Psychology of Mathematics (pp. 2-81 to 2-88). Lahti, Finland.
Boyer, C. (1949). The history of the calculus and its conceptual development. New York:
Dover.
Boyer, C. (1968). A history of mathematics. New York: John Wiley & Sons.
Breidenbach, D., Dubinsky, E., Hawks, J., & Nichols, D. (1992). Development of the
process conception of function. Educational Studies in Mathematics, 23, 247-
285.
Brummelen, G. V. (1995). Using ancient astronomy to teach trigonometry: A case study.
Histoire et épistémologie dans l’éducation mathématique (pp. 275-281). IREM de
Montpellier.
Bueno, M., & Lins, R. (2002). The history of mathematics in the education of
mathematics teachers: An innovative approach. Second International Conference
on the Teaching of Mathematics (at the undergraduate level). University of Crete,
Hersonissos, Crete, Greece. Retrieved on September 28, 2005, from
http://www.math.uoc.gr/~ictm2/Proceedings/pap402.pdf
369
Calinger, R. (1995). Classics of mathematics. Englewood Cliffs, NJ: Prentice Hall.
Calinger, R. (Ed.). (1996). Vita mathematica: Historical research and integration with
teaching. Washington, DC: Mathematical Association of America.
Carlson, M. P. (1998). A cross-sectional investigation of the development of the function
concept. In A. H. Scheoenfeld, J. Kaput, & E. Dubinsky (Eds.), Research in
collegiate mathematics education III (pp. 114-162). Washington, DC:
Mathematical Association of America.
Clagett, M. (1968). Nicole Oresme and the medieval geometry of qualities and motions. A
treatise on the uniformity and difformity of intensities know as Tractatus de
configurationibus qualitatum. Madison: University of Wisconsin Press.
Clement, J. (1989). The concept of variation and misconceptions in Cartesian graphing.
In T. Eisenberg & T. Dreyfus (Eds.), Visualization and mathematics education
part two. On learning problems in mathematics (pp. 77-87). Framingham, MA:
Center for Teaching/Learning of Mathematics.
Conference Board of the Mathematical Sciences. (2001). The mathematical education of
teachers. Washington, DC: Mathematical Association of America.
Confrey, J., Piliero, S., Rizzuti, J., & Smith, E. (1994). Development of teacher
knowledge and implementation of a problem-based mathematics curriculum using
multirepresentational software. Apple classrooms of tomorrow. Report number
11. High school mathematics. Retrieved on July 7, 2004, from
http://images.apple.com/education/k12/leadership/acot/pdf/rpt11.pdf
370
Crombie, A. C. (1959). Medieval and early modern science, Volume 1. Science in the
middle ages: V-XII centuries. Garden City, NY: Doubleday.
Dubinsky, E., & Harel, G. (1992). The nature of the process conception of function. In G.
Harel & E. Dubinsky (Eds.), The concept of function: Aspects of epistemology
and pedagogy. MAA Notes, 25 (pp. 85-106). Washington, DC: Mathematical
Eisenberg, T. (1992). On the development of a sense for functions. In G. Harel & E.
Dubinsky (Eds.), The concept of function: Aspects of epistemology and pedagogy.
MAA Notes, 25 (pp. 153-174). Washington, DC: Mathematical Association of
America.
Engels, F. (1940). Dialectics of nature (C. Dutt, Trans.). New York: International
Publishers.
Even, R. (1993). Subject-matter knowledge and pedagogical knowledge: Prospective
secondary teachers and the function concept. Journal for Research in
Mathematics Education, 24(2), 94-116.
Fauvel, J., & van Maanen, J. (Eds.). (2000). History in mathematics education: The ICMI
study. Dordecht, The Netherlands: Kluwer Academic.
Fleener, M., Reeder, S., Young, E., & Reynolds, A. (2002). History of mathematics:
Building relationships for teaching and learning. Action in Teacher Education,
24(3), 73-84.
Freudenthal, H. (1981). Should a mathematics teacher know something about the history
of mathematics? For the Learning of Mathematics, 2(1), 30-33.

371
Furinghetti, F. (1997). History of mathematics, mathematics education, school practice:
Case studies in linking different domains. For the Learning of Mathematics,
17(1), 55-61.
Furinghetti, F. (2002). On the role of the history of mathematics in mathematics
education. Panel discussion from Second International Conference on the
Teaching of Mathematics (at the undergraduate level). University of Crete,
Hersonissos, Crete, Greece. Retrieved on September 28, 2005, from
http://www.math.uoc.gr/~ictm2/Proceedings/panFur.pdf
Garuti, R. (1997). A classroom discussion and an historical dialogue: A case study.
International Conference for the Psychology of Mathematics Education, 21st, pp.
297-304. Lahti, Finland.
Goldenberg, P., Lewis, P., & O’Keefe, J. (1992). Dynamic representations and the
development of a process understanding of function. In G. Harel & E. Dubinsky
(Eds.), The concept of function: Aspects of epistemology and pedagogy, MAA
Notes, 25 (pp. 235-260). Washington, DC: Mathematical Association of America.
Grattan-Guinness, I. (1977). The history of mathematics and mathematics education. The
Australian Mathematics Teacher, 33(3/4), 117-127.
Hagelgans, N., Reynolds, B. E., Schwingendorf, K., Vidakovic, D., Dubinsky, E., Shabin,
M., et al. (1995). A practical guide to cooperative learning in collegiate
mathematics. MAA Notes, 37. Washington, DC: Mathematical Association of
America.
372
Harel, G., & Dubinsky, E. (Eds.). (1992). The concept of function. Aspects of
epistemology and pedagogy, MAA Notes, 25. Washington, DC: Mathematical
Hawkins, T. (1975). Lebesgue’s Theory of Integration: Its origins and development. New
York: Chelsea.
Hinkle, D., Wiersma, W., & Jurs, S. (1998). Applied statistics for the behavioral
sciences. Boston: Houghton Mifflin.
Jahnke, H.M. (2000). The use of original sources in the mathematics classroom. In
Fauvel, J., & van Maanen, J. (Eds.), History in Mathematics Education, The ICMI
Study (pp. 291-328). Boston: Kluwer Academic.
Jones, P. (1989). The history of mathematics as a teaching tool. In J. Baumgart, D. Deal,
B. Vogeli, & A. Hallerberg (Eds.), Historical topics for the mathematics
classroom. Reston, VA: National Council of Teachers of Mathematics.
Katz, V. J. (1998). A history of mathematics. An introduction (2nd ed.). New York:
Addison-Wesley.
Katz, V. J. (Ed.). (2000). Using history to teach mathematics. An international
perspective. Washington, DC: Mathematical Association of America.
Katz, V. J., & Michalowicz, K. D. (Eds.). (2005). Historical modules for the teaching and
learning of mathematics (compact disc). Washington, DC: Mathematical
373
Kennedy, J., & Ragan, E. (1989). Function. In J. Baumgart, D. Deal, B. Vogeli, & A.
Hallerberg (Eds.), Historical topics for the mathematics classroom (pp. 312-313).
Reston, VA: National Council of Teachers of Mathematics.
Kleiner, I. (1989). Evolution of the function concept: A brief survey. The College
Mathematics Journal, 20, 282-300.
Liu, P-H. (2002). Developing college students’ views on mathematical thinking in a
historical approach, problem-based calculus course. Second International
Conference on the Teaching of Mathematics (at the undergraduate level).
University of Crete, Hersonissos, Crete, Greece. Retrieved on September 28,
2005, from http://www.math.uoc.gr/~ictm2/Proceedings/pap191.pdf
Liu, P-H. (2003). Do teachers need to incorporate the history of mathematics in their
teaching? Mathematics Teacher, 96(6), 416-421.
Mathematical Association of America. (n.d.). Convergence. Retrieved on March 4, 2006,
from http://mathdl.maa.org/convergence/1/
Monk, S. (1992). Students understanding of a function given by a physical model. In G.
Harel & E. Dubinsky (Eds.), The concept of function. Aspects of epistemology and
pedagogy. MAA Notes, 25 (pp. 175-193). Washington, DC: Mathematical
National Council of Teachers of Mathematics. (1998). Historical topics for the
mathematics classroom. Reston, VA: National Council of Teachers of
Mathematics.
374
National Council of Teachers of Mathematics. (2000). Principles and standards for
school mathematics. Reston, VA: National Council of Teachers of Mathematics.
Norman, A. (1992). Teachers’ mathematical knowledge of the concept of function. In G.
Harel & E. Dubinsky (Eds.), The concept of function. Aspects of epistemology and
pedagogy MAA Notes, 25 (pp. 215-232). Washington, DC: Mathematical
Piaget, J., & Garcia, R. (1989). Psychogenesis and the history of science. New York:
Columbia University Press.
Radford, L. (1997, February). On psychology, historical epistemology, and the teaching
of mathematics: Toward a socio-cultural history of mathematics. For the
Learning of Mathematics, 17(I), 26-33.
RAND Mathematics Study Panel. (2003). Mathematical proficiency for all students:
Toward a strategic research and development program in mathematics education.
D. L. Ball, Chair. Santa Monica, CA: RAND Corporation.
Reed, B. (2004). How do functions change? Let me count the ways. Unpublished
manuscript, Kent State University, Kent OH.
Rickey, V. F. (1995). My favorite ways of using history in teaching calculus. In F. Swetz,
J. Fauvel, O. Bekkin, B. Johansson, & V. Katz (Eds.), Learn from the masters (pp.
123-134). Washington, DC: Mathematical Association of America.
Rüthing, D. (1984). Some definitions of the concept of function from John Bernoulli to
N. Bourbaki. The Mathematical Intelligencer, 6(4), 72-77.

375
Schoenfeld, A. (2002). Research methods in (mathematics) education. In L. English
(Ed.), Handbook of international research in mathematics education (pp. 435-
475). Mahwah, NJ: Lawrence Erlbaum.
Schwingendorf, K., Hawks, J., & Beineke, J. (1992). Horizontal and vertical growth of
the student’s conception of function. In G. Harel & E. Dubinsky (Eds.), The
concept of function. Aspects of epistemology and pedagogy, MAA Notes, 25 (pp.
133-149). Washington, DC: Mathematical Association of America.
Selden, A., &. Selden, J. (1992). Research perspectives on conceptions of function:
Summary and overview. In G. Harel & E. Dubinsky (Ed.), The concept of
function: Aspects of epistemology and pedagogy, MAA Notes, 25 (pp. 1-16).
Sfard, A. (1991). On the dual nature of mathematical conceptions: Reflections on
processes and objects as different sides of the same coin. Educational Studies in
Mathematics, 22(1), 1-36.
Sfard, A. (1992). Operational origins of mathematical objects and the quandary of
reification – The case of function. In G. Harel & E. Dubinsky (Eds.), The concept
of function: Aspects of epistemology and pedagogy, MAA Notes, 25 (pp. 59-84).
Sierpinska, A. (1992). On understanding the notion of function. In G. Harel & E.
Dubinsky (Eds.), The concept of function: Aspects of epistemology and pedagogy
(pp. 25-58). Washington, DC: Mathematical Association of America.

376
Siu, M. K. (1995). Concept of function – its history and teaching. In. F. Swetz, J. Fauvel,
O. Bekken, B. Johansson, & V. Katz (Eds.), Learn from the masters (pp. 105-
121). Washington, DC: Mathematical Association of America.
Springer, L., Stanne, M. E., & Donovan, S. (1998). Measuring the success of small-group
learning in college-level SMET teaching: A meta analysis. National Institute for
Science Education. Wisconsin Center for Educational Research: University of
Wisconsin-Madison. Retrieved on February 26, 2006, from
http://www.wcer.wisc.edu/archive/cl1/CL/resource/scismet.htm
Swetz, F. (1982, November). What ever happened to the history of mathematics?
American Mathematical Monthly, 89(9), 695-696.
Swetz, F., Fauvel, J., Bekken, O., Johansson, B., & Katz, V. (Eds.). (1995). Learn from
the masters. Washington, DC: Mathematical Association of America.
Thompson, P. W. (1994). Students, functions, and the undergraduate curriculum. In E.
Dubinsky, A. Schoenfeld, & J. Kaput (Ed.), Research in collegiate mathematics
education, I(4), 21-44. Washington, DC: Mathematical Association of America.
U.S. Department of Education. (2000). Before it’s too late: A report to the nation from
the national commission on mathematics and science teaching for the 21st century
(J. Glenn, Commissioner). Washington, DC: Author. Retrieved on March 4, 2006,
from http://www.ed.gov/inits/Math/glenn/report.pdf
Usiskin, Z., Peressini, A., Marchhisotto, E. A., & Stanley, D. (2003). Mathematics for
high school teachers, an advanced perspective. Upper Saddle River, NJ: Pearson
Education.
377
Van Gulik, I. (2005). Reinvention of geometry. A study into the value and use of history
of geometry in teaching mathematics. Unpublished doctoral dissertation,
University of Groningen, The Netherlands.
Vinner, S. (1992). The function concept as a prototype for problems in mathematics
learning. In G. Harel & E. Dubinsky (Eds.), The concept of function: Aspects of
epistemology and pedagogy. MAA Notes, 25 (pp. 195-213). Washington, DC:
Mathematical Association of America.
Vinner, S., & Dreyfus, T. (1989). Images and definitions for the concept of function.
Journal for Research in Mathematics Education, 20(4), 356-366.
von Glasersfeld, E. (1987). Learning as a constructive activity. In C. Janvier (Ed.),
Problems of representation in the teaching and learning of mathematics (pp. 3-
32). Hillsdale, NJ: Lawrence Erlbaum.
Von Glasersfeld, E. (1995). Radical constructivism: A way of knowing and learning.
Washington, DC: Falmer Press.
Wilson, M. (1994). One preservice secondary teacher’s understanding of function: The
impact of a course integrating mathematical content and pedagogy. Journal for
Research in Mathematics Education, 25(4), 346-370.
Youschkevitch, A. P. (1976). The concept of function up to the middle of the 19th
century. In C. Truesdell (Ed.), Archive for history of exact sciences, 16(1), 37-85.
New York: Springer-Verlag.

Reed Beverly M

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Reed Beverly M

Uploaded by

Copyright:

Available Formats

THE EFFECTS OF STUDYING THE HISTORY

OF THE CONCEPT OF FUNCTION

A dissertation submitted to the

B.A., John Carroll University, 1975

M.A., John Carroll University, 1979

Ph.D., Kent State University, 2007

___________________________________ Co-Director, Doctoral

___________________________________ Co-Director, Doctoral

___________________________________ Members, Doctoral Dissertation Committee

__________________________________, Interim Chairperson, Department of

__________________________________, Interim Dean, College and Graduate School

THE EFFECTS OF STUDYING THE HISTORY OF THE CONCEPT OF FUNCTION

Co-Directors of Dissertation: Ed Dubinsky, Ph.D.

on the concept of function. Students experienced an in-depth study of the history of

They completed a series of worksheets, readings, and problems.

participate in an in-depth interview to reveal their understanding of the function concept.

most profound in the area of graphical representations.

First and foremost, I would like to thank my co-director Ed Dubinsky. To have

speak. Thank you, Ed.

Secondly I couldn’t have completed a project of this magnitude without the

with a listening ear, his emotional support was invaluable.

Special thanks go to my colleagues Austin Melton and Bathi Kasturiarachi who

friendship, and encouragement throughout the entire process.

Thanks also to Michael Mikusa, my co-director for keeping me on track, to Anne

the final document.

was of me and how cool she thought my research was.

herself, knows the value of education and kept encouraging me to continue.

To my children, Joe and Jen,

may you continue your life-long journey of learning

LIST OF FIGURES ......................................................................................................... xiv

LIST OF TABLES........................................................................................................... xvi

I. INTRODUCTION TO THE STUDY............................................................................. 1

Claims About the Value of the History of Mathematics ........................................... 1

Statement of the Problem .......................................................................................... 3

Purpose of the Study ................................................................................................. 5

Theoretical Basis for the Study: APOS Theory ........................................................ 6

The Action Construct....................................................................................... 7

The Process Construct ..................................................................................... 8

The Object Construct..................................................................................... 10

The Schema Construct................................................................................... 10

Significance of the Study ........................................................................................ 14

Definition of Terms ................................................................................................. 15

II. REVIEW OF RELATED LITERATURE ................................................................... 18

Four Theoretical Frameworks: History and Learning............................................. 18

Piaget and Garcia........................................................................................... 18

Luis Radford .................................................................................................. 20

Anna Sfard ..................................................................................................... 25

Studies Using History of Mathematics as a Tool.................................................... 30

Studies About Learning the Concept of Function ................................................... 45

History of the Concept of Function......................................................................... 56

The Fourteenth Through Sixteenth Centuries................................................ 58

Seventeenth and Eighteenth Centuries .......................................................... 60

Rene Descartes ..................................................................................... 60

Newton and Leibniz ............................................................................. 61

The Vibrating String Problem.............................................................. 62

The Nineteenth Century................................................................................. 64

Fourier Series ....................................................................................... 64

Cauchy and Dirichlet ........................................................................... 66

Riemann and Weierstrass..................................................................... 68

III. METHODOLOGY ..................................................................................................... 72

Inter-Rater Reliability ............................................................................................. 78

Data Analysis .......................................................................................................... 81

Finite Sequences (#8, #18) ............................................................................ 81