Professional Documents
Culture Documents
by
Beverly M. Reed
December 2007
© Copyright by Beverly M. Reed 2007
All Rights Reserved
ii
A dissertation written by
Beverly M. Reed
Approved by
___________________________________
Anne Reynolds
Accepted by
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
functions during a 5-week unit in the junior-senior level History of Mathematics course.
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
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
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
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
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,
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
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
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
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
and my students,
vi
TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS ................................................................................................. iv
DEDICATION................................................................................................................... vi
CHAPTER
Instructional Implications........................................................................................ 11
Conclusion............................................................................................................... 13
vii
Summary of Chapter 1 ............................................................................................ 17
Anna Sierpinska............................................................................................. 22
Summary........................................................................................................ 29
Origins ........................................................................................................... 56
Leonhard Euler..................................................................................... 62
Conclusion ........................................................................................... 70
Summary ................................................................................................................. 70
Introduction ............................................................................................................. 72
Design...................................................................................................................... 73
Sample ..................................................................................................................... 75
Procedure................................................................................................................. 76
Triangulation ........................................................................................................... 77
Introduction ............................................................................................................. 89
ix
Initial Conceptions of Functions ............................................................................. 90
Participant 1: DB ........................................................................................... 90
xi
Participant 6: CS .......................................................................................... 229
Findings................................................................................................................. 250
FUNCTIONS................................................................................................... 310
WORKSHEET................................................................................................. 314
OF FUNCTIONS............................................................................................. 329
xiii
LIST OF FIGURES
Figure Page
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
11. Graphing Task 3 showing speed vs. time for two cars ......................................110
12. Graph for Task 2 showing an arbitrary graph with no labels ............................118
15. Graphing Task 3 showing speed vs. time for two cars ......................................125
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
19. Graphing Task 1 showing position vs. time for two objects..............................135
21. Graphing Task 3 showing speed vs. time for two cars ......................................137
23. Graph for Task 13 showing an arbitrary graph with no labels ..........................142
25. Graphing Task 3 showing speed vs. time for two cars ......................................152
26. Graphing Task 5 showing speed vs. distance along a race track .......................153
28. Graphing Task 3 showing speed vs. time for two cars ......................................158
29. Graphing Task 1 showing position vs. time for two objects..............................160
32. Graphing Task 3 showing speed vs. time for two cars ......................................164
34. DB’s sketch for graphing Task 4 after unit on history of functions ..................171
35. Graph for Task 13 showing an arbitrary graph with no labels ..........................174
37. CW’s sketch for graphing Task 4 after unit on history of functions .................186
xv
LIST OF TABLES
Table Page
3. APOS Levels of Participants Before Studying the History of Functions ........... 155
Functions............................................................................................................. 157
Functions............................................................................................................. 166
12. Participants Holding the Notion of “Function as Formula” Before and After
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
xvi
16. APOS Conception After History of Functions Unit ........................................... 257
xvii
CHAPTER I
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.
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
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
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
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
The most common reasons given in the literature are the following.
• History can help increase student motivation and develop a positive attitude
certain topic is difficult for students (Avital, 1994; Barbin, 2000a; Liu, 2003),
(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’
and problem posing while teaching (Avital, 1994). Learning the historical
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
that is, it helps students understand the reasons behind the development of
• Shows that mathematical ideas evolve over a period of time, are struggled
with and subject to change (Avital, 1994; Swetz, 1982). It shows that
math, it would influence the way she or he teaches it and thus affects the way
Backing all these claims, however, are few research studies documenting the
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
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 &
effectiveness are scarce. Furinghetti (1997) claimed that not only are the training and
experiences of such teachers are quite scattered, with no “organized network of classes
and teachers carrying out analogous experiments” (p. 55). Comparing results and
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
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
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?
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
• Does a student’s studying the history of the concept of function facilitate his
understanding?
to the learning and understanding of mathematics at the post-secondary level. The section
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
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
The acronym APOS stands for Action, Process, Object, and Schema—mental
to processes, which must come before seeing a concept as an object. The origins of this
7
concepts in von Glasersfeld’s radical constructivism. The developers of APOS see their
p. 41). To them, the following best describes what it means to learn and know something
in mathematics:
actions, processes and objects and organizing these in schemas to use in dealing
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,
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 construction is similar to Piaget’s action schemes. A student who has
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
inverses of functions and sets of functions, for example, is that the learner is not able to
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
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
theorists say the student has interiorized an action to form a process. This process relates
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
1994) and research has shown that many students do not achieve this level without
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
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—
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
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
Encapsulating processes into objects corresponds to Steffe’s and Cobb’s first level
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
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
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
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
Instructional Implications
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
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
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
utilize the ACE Teaching Cycle which has the following components:
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.
learning.
• Classroom tasks and discussions then pull together the main ideas developed
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
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
mathematical education of our teachers and are continuing to search for effective means
Board of Mathematical Sciences, 2001). National reports call for better preparation of our
Education, 2000). For the past several years, The Mathematical Association of America
The concept of function also takes center stage when it comes to mathematics
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)
algebra recommendations for grades 9-12. College textbook authors frequently take a
15
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.
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
Definition of Terms
combines, and registers in memory a collection of mental items or acts that appear in the
individual as being at least somewhat external (Asiala, Brown, et al., 1997, p. 42).
quantities.
transforms it into a unique element f(a) of a set B. The set A is called the domain and the
varying quantities.
present a sensory item without relevant sensory signals being available (Battista, 1999b,
p. 3).
the isolation of structure (form), pattern (coordination). and operations (actions) from
actions and processes may be applied (Selden & Selden, 1992, p. 19).
elements of the co-domain with the same element in the domain, that is,
understanding of the concept; a meaning which was not at all useful while performing
transformation as an action, but it is internal and hence under the control of the
individual.
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
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
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
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
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
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
The authors noted that though inter-object analysis and trans-object analysis
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:
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
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
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
According to Piaget and Garcia (1989), Galois and the development of group theory
Luis Radford
mathematical knowledge, though he claimed that several problems existed in doing such
looking for answers” to specific problems rather than general methods (Radford, p. 11).
considers the socio-cultural era during which he lived. Another difficulty is that of being
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
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
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
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
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.
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.
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
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
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
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
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
relationships between variables provided the impetus for extending the notion of function
beyond those expressible analytically. In particular, the solution to the partial differential
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
modeling situations in either science or real life), developmental contexts (using and
sufficient at the introductory level). She did not, however, provide empirical evidence for
Anna Sfard
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
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
“structural conceptions develop usually out of operational” (p. 61); that is, that abstract
objects emerge from computational processes, whether one is concerned with the
Sfard exemplified her theory with the concept of function. She claimed that
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
notion of variable. Later definitions (d’Alembert’s and Euler’s) excluded this idea and
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:
there was no reasonable link between the new idea and the existing systems of
She noted that the mathematician Frege de-legitimized the use of the term “variable,”
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
This simple description presented function as a set of ordered pairs and made no
surprisingly, this new definition, which had very little in common with its
intuitive operational origin, evoked much criticism when first proposed. But when
repeat itself once more: on the new objects new operations could now be
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
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
function with its representation. Her findings were not unlike those found by other
only superficial changes in notation are introduced; and their general neglect of domain
and range. She concluded, “pseudostructural conceptions can hardly be avoided within
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,
• Students were not able to deal formally and generally with operations on
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,
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
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
(1992) found parallels between students’ difficulties learning a concept and the
difficulties in its historical development. Radford (1997) concluded that despite the
great potential for understanding mathematics. All four valued the endeavor.
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
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
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
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
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
working with the history of mathematics, she has allowed her students the time they need
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
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
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
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 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
Po-Hung Liu (2002) conducted a study whose purpose was to investigate the
college students’ views of mathematical thinking. At their first class meeting of the
questionnaire designed to examine their views about mathematical thinking. They also
During the course of the semester, the teaching of calculus followed the historical
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
individual interviews.
34
While responding to the question at the beginning of the semester, “In your
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
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
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’
(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
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
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
Their study used a theoretical model which focuses on students’ “knowledge” and
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’
‘voices’ from the history of mathematics and science may mediate some important
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
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:
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.
methodology. What henceforth will be called the ‘voices and echoes’ game is a
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
level. They believed that the game should begin with historical voices that give a
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
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:
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.
teachers concerning the topics of negative and irrational numbers and linear and quadratic
“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
where students provide the modern forms [or names] for those that appear in
other sources
were/would be teaching;
• to create awareness of, and positive attitude towards, the history of topics in
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
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
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
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
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
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
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
• early in their studies, that is, in a mathematics content course that was part of
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
Students typically take this course in their junior year. The course focuses on
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 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
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
human interest from the point of view of learners of mathematics (technical knowing,
interest from the point of view of future teachers. They performed a qualitative analysis
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
(Fleener et al., 2002, p. 77), suggesting a preference for responses in the technical
These results suggest critical and historical approaches, even sustained over
students’ past experiences with mathematics and the emphasis, becoming even
mathematics in the pre-service elementary school teachers. They expected students in the
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
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
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
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
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
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
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
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. Anna Sfard (1992) noted other misconceptions and student difficulties with the
concept:
symbolic, tabular);
• 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
Vinner (1992) studied 271 college freshmen starting calculus courses in different
science departments and noted that 56% of his subjects showed some degree of
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
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
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
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.
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
pointwise question asks for values of a function for a specific input value. An across-time
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
graph as a literal picture of the problem situation (Bell & Janvier, 1981; Clement, 1989;
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
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
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
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’
et al.’s classifications.
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
Breidenbach et al. (1992) worked with sophomore and junior mathematics majors
They compared student responses to the question “what is a function?” before and after
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
(table continues)
52
Table 1 (continued)
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
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
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
Confrey, Piliero, Rizzuti, and Smith (1994) developed software called Function
allowed students to experiment, analyze, and make conjectures about functions using
numeric procedures, fill tables to organize data, and perform transformations of function
Apparently, the exactness required to write a simple code enables the construction
order to encourage student reflection about their work. The researchers concluded that the
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
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
3. Ability to construct a process when none is explicit in the situation, and the
The authors analyzed in detail the significance of each of the function situations
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,
3. Three graphs: a single valued continuous curve (with respect to both axes;
6. Five equations: one with a single variable (#15), three with two variables (#3,
7. Three statements: two describing physical situations (#7 and #9) and one
Dubinsky and Harel’s analysis of these situations is described in Chapter 3 under the
APOS analysis. The current study attempts to discern if learning the history of the
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
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,
and n3 + n2 in the other (Kennedy & Ragan, 1989). According to Sierpinska (1992), the
kitchen door long before it was put on a pedestal by Felix Klein, who, in 1908 . . .
57
In ancient Greece, the art of constructing tables, however sophisticated, did not
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.)
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:
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
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
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
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
the University of Paris, carried this idea to its logical conclusion by introducing a two-
geometrically that the arithmetic mean of the initial and final velocities of motion with
E G
F
A B
AF = ½ AC
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
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
• lack of motivation. Why define an abstract notion of function unless one had
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
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
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
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,
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
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
quantity composed in any manner whatsoever from this variable magnitude and from
Leonhard Euler
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
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
continuous function was one that had the same analytic expression over the entire
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
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
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
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
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
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
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
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)
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
equal number of ordinates f(x). All have actual numerical values, either positive or
law; they succeed each other in any manner whatever, and each of them is given
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).
expression (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:
(algebra) is not. This may explain the freeing from geometric perception of a
He also gave profound insight into the problem of accepting a new concept, both
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
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
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
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
of the variable x in this interval there corresponds a definite value of the variable
1989, p. 291)
68
Noteworthy in this definition was the explicit restriction of the domain to an interval . . .
also a first.
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
Fourier series. Both Riemann and Weierstrass delighted in studying irregularities and
The main difference between methods of studying functions within the framework
deduces properties of any function starting from the properties of those analytical
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:
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
Bourbaki (1939):
Let E and F be two sets, which may or may not be distinct. A relation between a
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
Conclusion
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
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
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 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
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
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
group, caused consternation among researchers since controlling enough variables to get
researchers rarely consider two nearly identical classes and subject them to a single small
72
73
Design
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
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
which they attempt to describe the social and cultural factors which affect students’
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
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
• What is the function conception of junior and senior level pre-service high
school teachers?
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
• Does a student’s studying the history of the concept of function facilitate his
understanding?
research are (a) preliminary interviews that allow for the researcher to make initial
analysis to determine student’s ways of thinking, and (d) Scientific model building which
processes at particular points in time as well as changes in those concepts and processes
about current student thinking, historical studies provide the potential perturbations that
move students along toward adult competence, and APOS theory provides the framework
Sample
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
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
action—or perhaps even a pre function—conception, and three who evidenced some
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
Procedure
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
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
All students in the class completed questionnaires before and after the
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 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 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:
• 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
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
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
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
unparalleled. He and a group of researchers had developed the theory over several years
Table 2
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.
(table continues)
79
Table 2 (continued)
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
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.
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
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
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.
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
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
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
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.
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
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.
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
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
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
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
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
question, ‘What is a function?’ on the initial questionnaire, student A made the common
function. Her answer to the same question on the second questionnaire, however, was
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
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
Though the pilot study involved no interviews with students, there were enough
student understanding about the concept of function. A more detailed study was in order.
Summary of Chapter 3
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—
researcher uncovered what these six students knew about functions initially and how their
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
junior and senior level pre-service high school teachers? The second characterizes the
Dr. Dubinsky’s approval, the researcher extended APOS analysis to include responses to
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
understanding of the concept in any way and if so, in which way In particular,
does studying the history facilitate his or her move from an action level
89
90
• Does a student’s studying the history of the concept of function facilitate his
understanding?
Participant 1: DB
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.
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
I: I see you have something crossed off there. Do you want to expand upon
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
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
Task 18 involves Boolean values for outputs (true or false) and DB’s responses
are enlightening.
TASK 18: If possible, describe the following situations using one or more
plug in values for n from this set and then you would get a result.
DB: Um . . . [Pause] I think it would be either true or false. You would find a
DB: Um . . . an inequality . . .?
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
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
TASK 13: If possible, describe the following situations using one or more
I: Can you explain your answer please and tell me what you meant by that and
DB: Well, I mean we’re given a graph that I don’t necessarily recognize as being
and all I thought was when you would plug in the results for x and then you
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
Though DB left Task 19 blank on the written questionnaire, during the interview
TASK 19: If possible, describe the following situations using one or more
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.
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
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
I: Now if you felt uncomfortable and you couldn’t find a formula for that,
DB: [Pauses to think] Yes. You mean even if I couldn’t find the specific
function?
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?
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
In another example, she was unable to find a function in a table, where a formula
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,
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?
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.
numbers, according to DB. She exhibited a rather elementary view of functions. In Task
TASK 16: If possible, describe the following situation using one or more
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
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.
TASK 3: If possible, describe the following situation using one or more functions.
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
DB: Well the first one I think I solved for x because it’s to me easier to find the
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
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
DB: I would think one is dependent on the other. I would say y really should be
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
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
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,
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
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
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
Figure 6. Graphing Task 5 showing speed vs. distance along a race track
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.
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)
MAA Notes, 25, 175. Copyright 1992 by the Mathematical Association of America.
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
blank.”
amount of water in a bulb-shaped bottle, she drew a straight line. She also incorrectly
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
Speed4
2 4 6 8 10
Time (min)
DB stated that the cyclist “travels further in the last 5 minutes because he is
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.
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
equation.
Dubinsky and Harel (1992) claimed that this emphasis on equations, numbers, and
the finite sequence questions, CW used function notation, though note again her
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
functions. If not, explain why not: “Kent State’s Men’s basketball team makes it
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
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
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
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?”
TASK 19: If possible, describe the following situations using one or more
CW: Yeah. I would say just plotting points. Again, like I said there is no
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.
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
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
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.
TASK 12: If possible, describe the following situations using one or more
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
CW: Yeah.
TASK 10: If possible, describe the following situations using one or more
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
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
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
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
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
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)
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
GRAPHING TASK 3: 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.)
Speed
t=0 t=1
Time in hours
Figure 11. Graphing Task 3 showing speed vs. time for two cars
CW: Um . . . I looked . . . The first time I looked at this I was thinking . . . the
on this part. I know they’re not in the same position there. So . . . I think . . .
[pause].
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
I: Uh Huh . . .
CW: But if B has a . . . then vice versa. But here I was just getting myself
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: Then we know that B has a greater slope than A from that picture so that B
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
She left the bottle problem blank on the initial questionnaire, but had no difficulty
Participant 3: MJ
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
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
MJ: If I gave you a number and then the uh . . . the thing would be . . . that’s the
I: So what happens if I give you this string? So, if I gave you, if this was a
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.
troubleshooting problems with circuit boards at the component level. His previous work
with computers may have enriched his notion of function. Research supports this
illustrated this ability in his response to Task 11, albeit after some struggling.
TASK 11: If possible, describe the following situations using one or more
I: What about #11? I noticed you tried to start doing something there. I’m not
MJ: What I was . . . my thought process on it was, I could solve for y and uh . . .
. . . [pause]. I mean I know that you have to solve that by maybe doing the
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
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 . . .
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.
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
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?
When questioned further, he concluded that functions are different than mappings.
116
you’re thinking?
that.
I: But yet you call that a mapping opposed to a formula is there any . . .
from set A, from an element in set A, and put it into a unique element into B.
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
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
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
something that tells you the relationship between two sets of items.
MJ: Yeah, that’s like saying here, mapping is more like a procedural thing, it
people and I say “is the brother of,” is that the type of verbal thing you
mean?
MJ: Yes.
MJ: No I’m saying like that, that could be a function. If there’s a unique
relationship?
MJ: Yes.
MJ’s conception of function, then, does not require a formula, only a regular
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
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
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
well.
TASK 10: If possible, describe the following situations using one or more
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
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]
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).
TASK 12: If possible, describe the following situation using one or more
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
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
sequences on compute—as inputs and outputs, but nothing in his computer experience
MJ was also unable to consider a Boolean function. Note his response to Task 15.
TASK 15: If possible, describe the following situation using one or more
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.
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 . . .
MJ: (Long Pause) No, I don’t think . . . I think its just one solution to it.
Notice in the following excerpt, MJ was unable to consider Boolean outputs, even
TASK 18: If possible, describe the following situation using one or more
MJ: Um . . . [pause]. No, not really. [Long pause] Then you would be having a
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: So when that is true, then there is a function and the function is that
MJ: Yes. That is what I would think the function would do.
MJ: No.
TASK 16: If possible, describe the following situation using one or more
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
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
Recall that Dubinsky and Harel (1992) claimed that one’s ability to find a
as evidenced below.
TASK 19: If possible, describe the following situation using one or more
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
In conclusion, though MJ was able to find functions in strings and discuss the
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
one.
124
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
given a speed versus time graph, the area under the curve represents the distance traveled.
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
Speed4
2 4 6 8 10
Time (min)
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
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
direction.) State the relationship between the position of car A and car B at t = 1
hr. Explain.
Speed
A
B
t=0 t=1
Time in hours
Figure 15. Graphing Task 3 showing speed vs. time for two cars
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
the changes in concavity he sketched while making a graph of height as a function of the
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
of the area curve as representing distance, he did not apply that knowledge in this
height of a line, that is, its output value, rather than its slope, as determining speed in a
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
MAA Notes, 25, 175. Copyright 1992 by the Mathematical Association of America.
He sketched the points (2, y1 ) on line B and (2, y2 ) on line A, then claimed “A is greater
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
began the unit on the history of the concept of function. His understanding of graphs was
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
just never had to . . . really think of what a function is, just . . . going back to 11th
things. The equation thing I realized later even when I was taking the, . . . the uh
128
. . . 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
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
discussing the graph in Task 2, he wrote f (x ) = e x , but in the interview, he reiterated that
TASK 2: If possible, describe the following situation using one or more functions.
Figure 17. Graph for Task 2 showing an arbitrary graph with no labels
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
um . . . oh gosh . . . I would have described it in terms of, well for the values
I: So it is just sort of, well it’s nice if there’s a formula but it’s not necessary.
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
representation does not suggest a function, so the construction came from BG himself.
TASK 6: If possible, describe the following situation using one or more functions.
BG: Oh . . . the string ABCDE . . . uh . . . I have something here, but I’m not
I: That’s okay. That’s really why I want to talk to you about it so that I
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
same. Now if you have like, p, q, p . . . oh no you can have it. I think it was
BG: Yes you could do the same thing. [Pause] Now if you had uh like . . . “p or
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.
TASK 11: If possible, describe the following situation using one or more
BG: [Laughs] I . . . I tried for a little while to solve for one of the variables and I
pause] I’m not sure if I . . . well, hold on. I could’ve, I could have done it.
131
BG: [Long pause] Wait a minute, never mind, I don’t think I could. [Inaudible]
BG: Not necessarily. There could be. I mean if you plotted that out on a graph,
I: Which one would be, like you were talking that the input would be n here.
BG similarly had no difficulty identifying a function in Task 12, the club members’ dues
status.
TASK 12: If possible, describe the following situation using one or more
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
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
Note also the flexibility in his thinking when discovering a function in Task 13.
TASK 13: If possible, describe the following situation using one or more
Figure 18. Graph for Task 13 showing an arbitrary graph with no labels
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
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.
TASK 18: If possible, describe the following situation using one or more
I: What about 18? Do you remember what you are thinking as you were
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
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
Finding a function in Task 19 was similarly easy for him. Recall that Dubinsky and Harel
function.
TASK 19: If possible, describe the following situation using one or more
functions. If not, explain why not. [The graph of discrete points is given.]
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
TASK 15: If possible, describe the following situation using one or more
BG: Well . . . it’s . . . it’s an equation. There is only one variable. You’ve got
reason I was thinking, it’s just a sentence, it’s not . . . you’re not really
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
given in Part 2 of the initial questionnaire, including those that Dubinsky and Harel
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
position (cm) B
2 4 6 8 time (sec)
Figure 19. Graphing Task 1 showing position vs. time for two objects
MAA Notes, 25, 175. Copyright 1992 by the Mathematical Association of America.
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,
Graphing Task 2.
GRAPHING TASK 2: This graph shows the speed in meters per second of a
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
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
cars. (Assume the cars start from the same position and are traveling in the same
direction.)
Speed
A
B
t=0 t=1
Time in hours
Figure 21. Graphing Task 3 showing speed vs. time for two cars
I: What about the relative position between the points t = .75 and t = 1 hour?
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
He drew an accurate sketch in Graphing Task 4 and correctly identified the three bends in
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
Participant 5: MS
Evidence suggests that MS began the unit with an emerging process conception of
following analysis is based on his responses to the initial questionnaire and his comments
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].
Usually you put some data in then you get something out. You get an
answer.
MS: You’d get the same result. If you used the same input.
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
MS: Hmm . . . I was in topology and he just showed us that so I thought I’d use
function?
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
MS: Yeah. If it’s got two variables. You put in one so that you can see the
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?
solving for x or y.
TASK 11: If possible, describe the following situation using one or more
I: Number 11. I notice that you started writing something and then you left it
MS: Well, I was going to try and isolate y again, but then it just looked like a
MS: Uh . . . It could be, but if you wanted to do it with one variable, you would
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
TASK 12: If possible, describe the following situation using one or more
functions. If not, explain why not: The club members’ dues status. [A chart with
I: Okay 12. The club members’ dues status. What were you thinking here and
MS: Well, if you had a database, say, with all these names, they would have a
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
TASK 13: If possible, describe the following situation using one or more
Figure 23. Graph for Task 13 showing an arbitrary graph with no labels
MS: That just looks like a cosine function. But it’s turned on its side. So if you
I: So if I turn the paper this way, then you would say there's a function. What
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.
TASK 17: If possible, describe the following situation using one or more
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
MS: Sure. When you're inputting 1 through 100, and they would get a point.
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
The following two excerpts show his inability to consider Boolean outputs.
TASK 15: If possible, describe the following situation using one or more
MS: It was a quadratic and I factored it, and then I found the two solutions.
MS: Yes, but it only would be defined at those two numbers, −1 and . . .
I: Oh.
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.
TASK 18: If possible, describe the following situation using one or more
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].
MS: Sure. Cuz then it would work. But it doesn't make sense. Because right now
MS: Right.
While discussing Task 5, MS came close to finding a function there, but was
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: But how does “If x then y” [reading paper] relate to that statement?
I: Oh, so x would be . . .
I: So like x would be “if the team wins” then they’d go to the Sweet 16, is that
MS: Right.
MS: Uh . . . this wouldn’t really be a function but it would just be . . . [pause] not
MS: Right.
146
MS: [Pause] I guess it could be a function, but I’d come up with some
MS: No idea.
function.
. . . [pause].
I: This is interesting.
MS: Then, whatever you want, say, then go to the next bracket [inaudible]. It
When discussing the Task 19, MS’s first tendency was to connect the dots and find a
formula.
147
TASK 19: If possible, describe the following situation using one or more
I: I notice that you've connected the dots in No. 19. What were you thinking
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–
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
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
MS: Well, it could be predictable, because when you have a function, you know
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
function before the beginning of the unit on the history of functions. A few minor
weaknesses exist.
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
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
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.
TASK 7: If possible, describe the following situation using one or more functions.
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
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
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
functions. If not, explain why not: “Kent State’s Men’s basketball team makes it
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.
His discussion about Task 8, which he left blank on the initial questionnaire, is
strikingly similar.
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
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.
interpreted some graphs correctly (Graphing Tasks 1, 2a, 2b, 3a, 3b, 3c), he showed
A would be ahead of Car B because he has covered more area,” yet he is unable to apply
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
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.
(Assume the cars start from the same position and are traveling in the same
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
Figure 25. Graphing Task 3 showing speed vs. time for two cars
CS thought that “Car B is pulling ahead because it is exponentially growing while the
While answering Graphing Task 5, he made the classic error of mistaking a graph
GRAPHING TASK 5: Can you indicate, from the graph below, how many bends
there are along the track on which the car was driven?
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
Figure 26. Graphing Task 5 showing speed vs. distance along a race track
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.
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
Thus CS exhibited a relatively weak conception of the notion of area under a curve and of
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
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
Table 3
DB *
CW *
MJ *
BG *
MS *
CS *
of functions using APOS terminology. The researcher proposes the descriptions and
analysis of student answers that characterize each of the APOS levels. This
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
interferes with attention to the meanings of the abstract features of the graphs.
(p. 37)
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
builds upon the work of these researchers. The researcher proposes following
Table 4
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.
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.
as a relationship between two varying quantities and will frequently mistake a graph as a
GRAPHING TASK 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
158
direction.) State the relationship between the position of car A and car B at t = 1
hr. Explain.
Speed
A
B
t=0 t=1
Time in hours
Figure 28. Graphing Task 3 showing speed vs. time for two cars
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
GRAPHING TASK 5: Can you indicate, from the graph below, how many bends
there are along the track on which the car was driven? [The graph of speed vs.
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.
graphical representation.
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
corresponding inputs on the graph. She does not refer to the global features of a graph.
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.
the speed of object A greater than, less than, or equal to the speed of object B?
position (cm) B
2 4 6 8 time (sec)
Figure 29. Graphing Task 1 showing position vs. time for two objects
MAA Notes, 25, 175. Copyright 1992 by the Mathematical Association of America.
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
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.
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 five
Speed4
2 4 6 8 10
Time (min)
CW: I said because they were going the fastest between those times so I said they
and 4 seconds and they’re going a lot slower even between 4 and 10
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
Similarly, in the brief excerpts below BG, MS, and CS considered the qualitative
the speed of object A greater than, less than, or equal to the speed of object B?
[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,
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
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
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
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
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
direction.) State the relationship between the position of car A and car B at t = 1
hr. Explain.
164
Speed
A
B
t=0 t=1
Time in hours
Figure 32. Graphing Task 3 showing speed vs. time for two cars
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
CS was initially not quite so certain, but appeared to understand the connection between
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
same amount of time car B has went at a slower speed. So yeah, I guess I
I: Now when you say covering more area, what do you mean?
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.
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
Task
1 2c 3a 3d 5
Student
NA
DB (no answer) Pre-Function Pre-Function Pre-Function Pre-Function
Participant 1: DB
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
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
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
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
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
meeting signifies how I looked at my worksheet. Now I don’t think they are
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
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.
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
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
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
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
After correcting her work on the bottle problem (Graphing Task 4 on the initial
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
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.
and her increased ability to identify functions in situations. Her definition of function had
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
[Pause] Um, and I wasn’t clear on that the output is going to be a unique
output . . . [pause].
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
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
evaluation, which as a list with the students’ name and amount owed. You
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
I: So there is a difference.
DB: I think I was originally thinking that a function involved, kind of, . . .
One may wonder what particular aspect of the function project spurred this growth. Her
different formulas. If the output is the same in each representation, we have the
same function.
the uniqueness criterion and her increased ability to think creatively to find a function.
TASK 13: If possible, describe the following situations using one or more
Figure 35. Graph for Task 13 showing an arbitrary graph with no labels
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 . . .
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.
Q2 TASK 7: One can sometimes tell a great deal about a text (its author, when it
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
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
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.
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
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
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
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
to a function. Judging from the following comments, she did not have this understanding
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
how the graphs would become more like the original function as n got large.
function with Boolean outputs in Q2 Task 11 and, with probing, was able to find a
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
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
Table 6
Ability to consider arbitrary Class and small group discussion; occurred Written reflection
correspondences as functions between work on the Descartes and Euler
worksheets
Participant 2: CW
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
variables in which one input produces exactly one output, and the relationship can
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
the paper with respect to time, would this location be a function of time? Why or
a) b)
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 . . .
CW: So this is time. And then this would be location. It’s at like . . . two different
I: And what would the, the actual description of the vertical axis be?
CW: 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
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]
it would be a function. But if you just . . . I was just looking at these graphs
as . . . the function, so . . .
I: uh hum . . .
Her comments during the second interview while discussing Task 13 on the initial
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.
right? If it’s inverse and it’s . . . so it’s not . . . it does not have one-to-one
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
Similarly, her comments about the “quick fox” problem exhibited her newly
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
I: Uh huh.
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
And like . . . for a function, if you wanted to show it as a function, then the
Recall that strings, particularly those whose representations do not suggest functions,
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
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
Note also her ability to find a Boolean function in Q2 Task 11. She was a bit
Q2 TASK 11: If possible, describe the following situations using one or more
CW: And the output is the inequality, regardless of its truth value.
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
I: Uh huh . . .
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 . . .
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 . . .
CW: The input would be our . . . our scores for each of our names and then the
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
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
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
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
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
GRAPHING TASK 4: Imagine this bottle filling with water. Sketch a graph of the
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
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
Table 7
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
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
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
MJ: Um . . . Over the last few weeks, I think that the main thing that kinda
the definition.
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
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.
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
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
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?
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
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
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
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
arbitrary pairing:
MJ: Um . . . the only thing that I really don’t understand about those types 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
I: Uh huh . . . uh huh . . .
I: You’re not sure if it’s one to one? Is that what you’re . . . You’re not seeing
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
I: What could be the domain here? What could we identify as the domain?
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
I: Uh huh . . .
MJ: Junior?
MJ: Yeah. You’d actually have to have individual names. But that would be,
I: Uh huh . . .
MJ: You couldn’t . . . you just couldn’t . . . you don’t have like, real numbers.
MJ: Yeah.
I: Right?
I: Is it?
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
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
I: Uh . . . huh.
193
MJ: Um . . . now . . . since you have two values of 10, that means that 10 is
MJ: I guess . . . I mean I guess that is true. I guess, yeah, um . . . You just don’t
I: Uh huh . . . But are the . . . the fundamental ideas . . . are able to be defined.
MJ: Yeah.
pairings.
the paper with respect to time, would this location be a function of time? Why or
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
I: Uh huh . . .
MJ: time.
MJ: And now on part (b), you could follow the same notion, but since there’s an
different ts . . . well that could be a function that way . . . You couldn’t have
it come back.
I: Uh huh . . . uh . . . huh . . .
I: Uh huh . . .
MJ: . . . because you have one point pointing at two different values of time.
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].
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.
Recall that Dubinsky and Harel (1992) believed that this problem is likely to provide
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
MJ: I don’t know what to do with that one . . . [pause]. Each x corresponds to a
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
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
I: I guess there’s probably more than one way of doing . . . you know. It’s
MJ: I mean I kinda see it could . . . it’s a function but . . . I don’t know really . . .
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
Q2 TASK 11: If possible, describe the following situation using one or more
MJ: W ell what I thought on that one what you have is an inequality and it’s either
I: Uh huh . . .
MJ: So every n is gonna have one result . . . true or false, so therefore it’s a
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
I: Is it one-to-one?
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]
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
MJ’s reflection written after the Definition of Functions worksheet revealed his
My understanding of what a function is, often gets confused with the properties of
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
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
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
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
algebraic expressions. By his own admission and by the evidence, the study of history has
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
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
mistake on Graph Task 5 on the initial questionnaire. When asked in the second interview
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.
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
I: Oh . . .
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
I: Uh huh . . .
MJ: Yeah.
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
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
Clearly, working on the Oresme worksheet helped him make a connection between area
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
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.
realized [laughs] as I came close to the top [laughs], I’m slowing down.
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
I: Uh huh . . .
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
Table 8
Participant 4: BG
second questionnaire he exhibited the same. During the second interview, however, he
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
where there is a unique output for any given input.” Um . . . On your first
see that there’s a change from your first one to your second one?
I: Uh huh.
BG: . . . written out. I guess you could say more, more like, Euler’s first
definition..
I: Uh huh.
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
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
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.
determine the creature’s location on the paper with respect to time, would this
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 . . .
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
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
I: [Laughs]
207
BG: [Laughs too] Well I mean, . . . When I just looked at the graphs by
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
BG: Right.
Note also the ease with which he found a function in the scenario in Q2 TASK 7,
Q2 TASK 7: One can sometimes tell a great deal about a text (its author, when it
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
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
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
I: Okay . . .
BG: Uh . . . [pause]. I really struggled with how to write this out . . . [laughs a
bit].
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
final test . . . what is . . . when I say “for n within N.” Like this [pointing to
I: Um . . . Is an element of . . .?
I: Yeah. Um . . . that’s fine. So, was that . . . Have you been in the habit of
Q: Uh huh.
I: Uh huh.
Q: Uh huh.
I: . . . with the domain being the natural numbers as you have indicated here
BG: Right.
Fibonacci worksheets, since sequences were the main topic of study on these sheets and
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
BG: [Laughs] I was trying to be more creative than I was in the last one. There’s
I: Okay.
211
BG: Yeah . . . Um . . . and so I said, well all right, this one . . . I didn’t write this
BG: Yeah.
BG: Uh . . . these capital letters. Yeah . . . That’s fine. That’s what I was trying
to write.
I: How would you have written it if um . . . these weren’t in that order? Would
it have mattered?
BG: I’d probably have done it the same way as I did over here.
BG: Yeah.
Q2 TASK 12: If possible, describe the following situation using one or more
I: Now when you first did the first questionnaire, you didn’t have one [find a
BG: Right.
I: Remember in the conversation? I remember you doing that. So, and I notice
BG: Right.
I: Were you thinking anything different? Or you just saw that and said oh, you
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,
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 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.”
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
something. What other . . . where else in calculus are you familiar with
calculus?
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
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?
I: Uh huh.
214
BG: Either of one power lower. You know, the transcendentals have different
I: And what about when you use the chain rule . . . or taking derivatives in
general?
I: Yeah. What’s your interpretation of the chain rule? Does that . . . I don’t
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).
BG: Right. Yeah. There we go. [Laughs] Yeah that’s right. f '( g ( x)) g '( x)
composition of functions.
BG: Right.
I: You work with one function then you do the other function . . . and your
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
BG: Yeah.
215
Though the above dialogue skimmed the surface of BG’s thinking about functions
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
conception before the history of functions unit was unknown since the researcher did not
Table 9
Participant 5: MS
on time and had low grades on most of them. He did not interact with others, despite
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.
Q2 TASK 11: If possible, describe the following situations using one or more
I: n is one.
I: . . . if n is one. Okay so . . .
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.
MS: No. You would have to change the range . . . or whatever that is . . . domain.
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 . . .
I: [Laughs] You don’t have to like it, it’s okay. It doesn’t work? Because it’s
not always . . .
MS: It just gives you two numbers that are either greater than or less than.
MS: I guess you could have one number . . . like one number on the left could be
I: Uh huh. Okay.
MS: You could do it the other way around. The right side is the input. You could
switch them.
MS: 2n
I: Okay.
MS: So I guess you could say that one side’s one x and this one side’s the other.
Similarly, when revisiting Task 15 from the initial questionnaire, he was unable to
TASK 15: If possible, describe the following situations using one or more
MS: No. . . . but it looks like I just factored it. I found the solutions.
MS: Yeah, but it would only give . . . you’d only have two inputs . . . that are
possible.
MS: Always 0?
I: Okay. Anything else about that? I mean could there be a different type of
function there?
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
the paper with respect to time, would this location be a function of time? Why or
a) b)
MS: Uh . . . well you could, but I don’t think I would want to use a graph that
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
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
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
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
MS: They’re just arbitrary statements. [Pause] But uh . . . I still can’t really find
MS: Well if . . . they . . . I guess you could say if they win any in the Sweet
I: Ah . . . Is that a function?
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: I guess, but it’s not a map. It’s not really. You can’t really make an equation
MS: Well not . . . I don’t like it otherwise. Cuz it’s just a statement then . . . you
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
MS: No, you could just say it . . . like a cake recipe is a function.
I: [Laughs]
I: Oh . . . But you said you like the ones with algebra formulas better . . .
interview he admitted to one being there. When revisiting Task 19, he adjusted his
TASK 19: If possible, describe the following situations using one or more
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 . . .
MS: Uh sure. It would just have to be restricted . . . so it would only give you
about functions. He reversed the inputs and outputs, and did not confuse the uniqueness
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?
MS: Uh . . . here I said it wouldn’t be if you had multiple test scores per person
I: Uh huh.
MS: You could reverse the dependency I guess, then it would work.
MS: So you’d have . . . test scores would be matched up to the . . . would be the
MS: As long as . . .
225
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.
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 . . .
seen . . . I think we did some in some early programming class.” The excerpt below also
MS: Yeah cuz in a program when you’re calling on your function, it’s just a little
I: Uh huh.
I: Okay.
MS: Pretty much. Well, it’s . . . what’s ever inside of it, in that case.
MS: Cuz the process tells you what’s the relationship between the two . . . the
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?
I: Okay.
MS: There are different ways of looking at the same result . . . but it depends on
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.
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
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
Participant 6: CS
in general.
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
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 . . .
that he was unable to find a function in Task 12 (the club members’ dues status) on the
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?
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
I: Okay . . .
CS: You’d just . . . you’d have a unique output for every input. That’s basically
my definition of a function.
I: Going back to #12 on the original questionnaire. I notice that you left that
CS: [Immediately and confidently] Oh yeah. I was not sure at all what to do
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
I: And you would still answer the same thing? You’re still thinking that #13
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
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
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.
CS: Okay quadrant 3. I would take that as the beginning, just . . . I don’t know, I
I: So then you’re saying . . . now back to 13 you’re sort of saying if this was a
CS: Yeah.
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
way. Another scenario in which he found a function without using a formula is Q2 Task
Q2 TASK 7: One can sometimes tell a great deal about a text (its author, when it
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
Explain how you might use a function to organize this information. Write
whatever you think is necessary about your function to specify it. (Breidenbach et
CS: Okay, this problem definitely confused me, cuz I really didn’t exactly sure
I: Okay.
CS: So, I just kind of made up my own idea which was I just decided that my
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]
on paper].
CS: Uh huh.
I: So now what happens then if we did your function on the word jumped?
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
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
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
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: 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
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: 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
I: Uh huh.
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. 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
do you think?
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 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
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
I: Uh huh.
CS: Yeah . . .
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
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
Similarly, he was able to consider a function in Task 5 on initial questionnaire during the
second interview.
TASK 5: If possible, describe the following situation using one or more functions.
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
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
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
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.
CS: I’m not 100% sure what I did there [on the initial questionnaire]. Uh . . . I
I: Uh huh.
CS: That fit what that was saying. And . . . I don’t think that makes any sense
looking at it again.
CS: the 3n + n 2 .
I: Uh huh.
Another example of CS’s new found flexibility is his response to Task 10 during the
second interview.
TASK 10: If possible, describe the following situations using one or more
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
the H would be home and I just took shots made over shots missed.
I: Uh huh
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
fractional form.
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
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
CS: I’m not sure . . . Um . . . just by looking at it I wouldn’t think so, but I guess
I: They’re dots. They’re like dots, filled in dots. I’ll try to fill them in so that
CS: I guess it could be a function . . . if you’re just plotting points. Then you
CS: Yeah.
I: Please don’t feel that you’re looking for the one right answer. Just tell me
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.
CS: I think you need information to show the relationship. It doesn’t make any
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
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.
CS: Yeah.
243
notice that you have all equations. What else could you do . . . what else
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
CS: Ah . . . I’m not totally sure. It seems like it because . . . In order to graph it
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
{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?
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
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: Uh huh.
CS: Cuz also you’re having 2 here. So . . . I won’t take those as an output just
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.”
CS: Okay . . . I kind of see what you’re saying. That my output could be that it’s
CS: I think that that just depends on how you look at it. I would still say it’s not
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
well. Consider closely the changes he made when discussing his responses on the initial
GRAPHING TASK 3d: The graph below represents speed vs. time for two cars.
(Assume the cars start from the same position and are traveling in the same
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
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.
GRAPHING TASK 5: Can you indicate, from the graph below, how many bends
there are along the track on which the car was driven?
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?
I: Bends in the track. I’m sorry, like you know how a race car goes around a
turn.
answer.
I: Uh huh . . .
C: Cuz obviously you’re gonna have to slow down if you’re going around a
approaching, then going over, the top of a hill. He had no trouble sketching water height
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
Table 11
Summary of Chapter 4
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
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
pairing. Growth was most profound in the area of graphical representations of functions
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
during the course of studying its history for 4 of these 6 participants. More detail and
DISCUSSION
Introduction
• What is the function conception of junior and senior level pre-service high
school teachers?
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
• Does a student’s studying the history of the concept of function facilitate his
understanding?
This chapter discusses the findings for each question, the implications for teaching and
Findings
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
Definitions of Function
Recall that Dubinsky and Harel (1992) claimed that an emphasis on equations,
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,
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.
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
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
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 *
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.
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
CW
MJ *
BG * * * *
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.
these scenarios on the second questionnaire or while revisiting the initial questions during
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
DB * * * not discussed
CW * * * *
MJ not discussed * * *
BG * * * *
MS considered the * *
possibility
*
CS considered the considered the
possibility possibility
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
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
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
Table 15
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.
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
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
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.
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
Task
1 2c 3a 3d 5
Participant
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
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
In other words, what specific uses of historical material facilitated the change?
259
Table 18
Task 1 2c 3a 3d 5
Participant
CW Action
BG
MS
CS Process Process
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
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
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)
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.
function changed, but did not really go into depth about why they changed. One may
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.
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
mathematics was difficult. Only the van Gulik (2005) and Arcavi studies (1985) claim to
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
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
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
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
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
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
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
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
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
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.
mathematical education of our teachers and are continuing to search for effective means
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’
long-forgotten notions.
265
It offers support to those who claim that studying the history of a mathematical
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
move toward a process conception of function, might not a more in-depth study of its
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
suggests that the study of the development of an idea, rather than a chronological
Another contribution of this study was the adaptation of APOS theory to the
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
mathematical concept. This concept takes center stage when it comes to mathematics
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—
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
Finally, this study confirmed what other studies have shown in terms of
motivation and attitude toward mathematics. Confidence level of the participants was
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
some participants, might a more advanced study enable a move to an object conception?
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
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
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
This study has shown that it is possible to create a positive and productive
generalizing the results, though all reasonable attempts have been made to be transparent
conjectured about the value of studying the history of mathematics. History can be a
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
This study, then, provided evidence that studying the history of mathematics
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
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?
Directions. If possible, describe the following situations using one or more functions. If
not, explain why not.
2.
3. y4 = x2
4. y4 = x3
273
274
7. A swimmer starts from shore and swims to the other side of Walden Pond.
8. {3n + n 2 : n in [1...100]}
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
Name Owed
John $15
Julie 10
Jen 0
Dave 7.50
Brittany 10
Alex 5
Mary Ann 25
Sam 20
Sally 17
13.
x = t3 + t ⎫⎪
14. ⎬ t , a real number
y = 1 − 3t + 2t 4 ⎪⎭
15. x 2 + 3 x+ 2 = 0
19.
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)
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
Note. From “An Investigation of the Function Concept” by M. Carlson (1998).
Research in Collegiate Mathematics Education, III, 7, 146. Copyright 1998 by American
Mathematical Society. Adapted with permission.
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
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
Note. From “An Investigation of the Function Concept” by M. Carlson (1998).
Research in Collegiate Mathematics Education, III, 7, 148. Copyright 1998 by American
Mathematical Society. Reprinted with permission.
5. Can you indicate, from the graph below, how many bends there are along the track on
which the car was driven?
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
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?
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.
a) b)
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.
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 (Breidenbach et al., 1992, pp. 281-282).
⎧⎪ 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 ⎭
11. A swimmer starts from shore and swims to the other side of the lake.
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
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
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.
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
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.
284
285
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
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
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
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
Oresme explains why he thinks that line segments are appropriate representations for
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
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
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
3. In your own words, explain why Oresme thought the lines are appropriate
(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
time
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
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
5. Create three other scenarios that you might use with middle school students in an
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
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
(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)
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
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
In the reading below, Oresme discusses in an indirect way our notion of slope.
Activity 4
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
G
K
E
H
D B
A F
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
c
20 M/SEC
e
Va d
a b
t
4 SEC
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.
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
Explain.
b) Have the above exercises enriched your understanding of the nature of graphs in
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
GEOMETRY
Background Information
The following to two readings by noted historians Israel Kleiner (1989) and A. P.
The 17th century witnessed the emergence of modern mathematized science and
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
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
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,
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
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
importance of this notation, which, for the first time ever, made it possible to put
unknown quantities and arbitrary coefficients (a word also originating with Viète)
could hardly be overestimated. However, the creator of the new algebra did not
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
importance to the appropriate selection of signs), Euler and other scholars of the
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
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
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
to numerical answers.
into being, to become for a long time the principal method in mathematics and
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
What follows are samples of the work, first of Fermat, then Descartes, to give you a feel
Worksheet Activities
“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.”
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
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
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
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
P I
A
N Z
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]
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
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 )
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
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
V
E
I A B
Z
Variables are not functions. The concept of function implies that a unidirectional
one of the variables involved, the variables are not functions but simply variables.
2. Did this worksheet led to new insights concerning the concept of function?
APPENDIX E
WORKSHEET #3
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
WORKSHEET #4
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)
T Q N
311
312
1. (a) Leibniz was interested in the notion that the slope of PT is equal to three
(b) Prove that the length PQ of the ordinate at P is the geometric mean of the length
2. Find the lengths of all six of the line segments defined above for the point P = (3,6)
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”
2. Did this worksheet led to new insights concerning the concept of function?
APPENDIX G
WORKSHEET #5
Historical modules for the teaching and learning of mathematics (compact disc) edited
by V.J. Katz and K. D. Michalowicz (pp. 72-73, 76). Washington, DC: The Mathematical
WORKSHEET #6
Historical modules for the teaching and learning of mathematics (compact disc) edited
by V.J. Katz and K. D. Michalowicz (pp. 81-83). Washington, DC: The Mathematical
WORKSHEET #7
EULER ON FUNCTIONS
Historical modules for the teaching and learning of mathematics (compact disc) edited
by V.J. Katz and K. D. Michalowicz (pp. 59-62). Washington, DC: The Mathematical
WORKSHEET #8
Historical modules for the teaching and learning of mathematics (compact disc) edited
by V.J. Katz and K. D. Michalowicz (pp. 99, 103, 104). Washington, DC: The
by C. Boyer, 1968, New York: John Wiley & Sons. Copyright 1968 by John Wiley &
WORKSHEET #9
DEFINITIONS OF FUNCTIONS
Worksheet #9: Definitions of Functions
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
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
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
336
337
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
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
***************
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
...
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
5. How does this definition of Lagrange compare with each of those given by Euler? To
**************
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
law; they succeed each other in any manner whatever, and each of them is given as it
8. Read the few paragraphs entitled Fourier Series and summarize the main effect of
*****************
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,
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
1984, p. 74)
9. Create three functions that would agree with Dirichlet’s characterization of functions,
11. Read the excerpt entitled Riemann and Weierstrass and explain in your own words
******************
H. Hankel (1870). . . . Moreover it does not matter at all from where and how f(x) is
G. Peano (1911). . . . the function is a special relation, by which to each value of the
E. Goursat (1923). The modern definition of the word function is due to Cauchy and
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
functional relation if, for all x ∈ E , there exists unique y ∈ F , which is in the given
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.
b. Did this worksheet led to new insights concerning the concept of function?
*************************
341
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):
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
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
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
continuous function was one that had the same analytic expression over the entire
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
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
d’Alembert solved the problem by showing that the motion of the string is governed by
∂ 2y ∂2 y
=a 2 (a is constant, y(0,t) = y(L,t) = 0).
∂t 2 ∂x 2
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
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
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
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
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
equal number of ordinates f(x). All have actual numerical values, either positive or
law; they succeed each other in any manner whatever, and each of them is given
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).
expression (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:
(algebra) is not. This may explain the freeing from geometric perception of a
He also gives profound insight into the problem of accepting a new concept, both
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
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
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
⎧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
pairing (Kleiner, 1989, p. 292). Noteworthy in his definition was the explicit restriction
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
Fourier series. Both Riemann and Weierstrass delighted in studying irregularities and
The main difference between methods of studying functions within the framework
deduces properties of any function starting from the properties of those analytical
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
Anderson, D. L., Berg, R., Sebrell, A. W., & Smith, D. (2004). Functions. In V. Katz &
Association of America.
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
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
Asiala, M., Brown, A., DeVries, D., Dubinsky, E., Mathews, D., & Brown, K. (1997). A
366
367
Asiala, M., Cottrill, J., Dubinsky, E., & Schwingendorf, K. (1997). The development of
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
Baker, B., Coolery, L., & Trigueros, M. (2000). A Calculus Graphing Schema. Journal
& J. van Maanen (Eds.), History in mathematics education, The ICMI Study (pp.
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-
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.
Batttista, M. T. (1999b). The constructivist theory for how children learn mathematics.
Bell, A., & Janvier, C. (1981). The interpretation of graphs representing situations. For
Boero, P., Pedemonte, B., & Robotti, E. (1997). Approaching theoretical knowledge
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
285.
Montpellier.
Bueno, M., & Lins, R. (2002). The history of mathematics in the education of
http://www.math.uoc.gr/~ictm2/Proceedings/pap402.pdf
369
Calinger, R. (Ed.). (1996). Vita mathematica: Historical research and integration with
Clagett, M. (1968). Nicole Oresme and the medieval geometry of qualities and motions. A
Confrey, J., Piliero, S., Rizzuti, J., & Smith, E. (1994). Development of teacher
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
Dubinsky, E., & Harel, G. (1992). The nature of the process conception of function. In G.
Association of America.
America.
Engels, F. (1940). Dialectics of nature (C. Dutt, Trans.). New York: International
Publishers.
Fauvel, J., & van Maanen, J. (Eds.). (2000). History in mathematics education: The ICMI
Fleener, M., Reeder, S., Young, E., & Reynolds, A. (2002). History of mathematics:
24(3), 73-84.
Freudenthal, H. (1981). Should a mathematics teacher know something about the history
17(1), 55-61.
http://www.math.uoc.gr/~ictm2/Proceedings/panFur.pdf
Goldenberg, P., Lewis, P., & O’Keefe, J. (1992). Dynamic representations and the
Hagelgans, N., Reynolds, B. E., Schwingendorf, K., Vidakovic, D., Dubinsky, E., Shabin,
America.
372
Harel, G., & Dubinsky, E. (Eds.). (1992). The concept of function. Aspects of
Association of America.
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
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
Addison-Wesley.
Katz, V. J., & Michalowicz, K. D. (Eds.). (2005). Historical modules for the teaching and
Association of America.
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).
Kleiner, I. (1989). Evolution of the function concept: A brief survey. The College
Liu, P-H. (2003). Do teachers need to incorporate the history of mathematics in their
from http://mathdl.maa.org/convergence/1/
Harel & E. Dubinsky (Eds.), The concept of function. Aspects of epistemology and
Association of America.
Mathematics.
374
Harel & E. Dubinsky (Eds.), The concept of function. Aspects of epistemology and
Association of America.
Piaget, J., & Garcia, R. (1989). Psychogenesis and the history of science. New York:
RAND Mathematics Study Panel. (2003). Mathematical proficiency for all students:
Reed, B. (2004). How do functions change? Let me count the ways. Unpublished
J. Fauvel, O. Bekkin, B. Johansson, & V. Katz (Eds.), Learn from the masters (pp.
Rüthing, D. (1984). Some definitions of the concept of function from John Bernoulli to
Schwingendorf, K., Hawks, J., & Beineke, J. (1992). Horizontal and vertical growth of
processes and objects as different sides of the same coin. Educational Studies in
reification – The case of function. In G. Harel & E. Dubinsky (Eds.), The concept
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-
Springer, L., Stanne, M. E., & Donovan, S. (1998). Measuring the success of small-group
http://www.wcer.wisc.edu/archive/cl1/CL/resource/scismet.htm
Swetz, F., Fauvel, J., Bekken, O., Johansson, B., & Katz, V. (Eds.). (1995). Learn from
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
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
Vinner, S., & Dreyfus, T. (1989). Images and definitions for the concept of function.
century. In C. Truesdell (Ed.), Archive for history of exact sciences, 16(1), 37-85.