Garfield,

Garfield, E.
E. "Science
“Science Citation
Citation Index."
Index.” Science
Science Citation
Citation Index
Index 1961, 1, p.v-xvi,
1961,1, p.v-xvi, 1963.
1963.

Science Citation Index

Ten years ago, the “First Symposium on Machine
Methods in Scientific Documentation” was
sponsored by the Johns Hopkins University
Machine Indexing Project and the American
Documentation Institute a t the Welch Medical
Library in Baltimore (1). Brief papers on this
pioneering project have been published (2,3).
Unfortunately, its voluminous final reports j 4 )
were not widely disseminated. However, as a
result of national publicity given to the above-
mentioned symposium, W.C. Adair, a former
Vice-president of Shepard‘s Citations, learned
of our work, Though in retirement, he wrote and
suggested that the so-called citator system
used by attorneys be tested for scientific
literature (5). .Mr. Adair‘s letter struck a highly
responsive chord as I had been working out a
method of utilizing the sentences and citations
appearing in review articles a s entries in large-
scale interdisciplinary science indexes (6).
Though the Johns Hopkins Project w a s dis-
continued in 1953, I continued to correspond
with Mr. Adair while at Columbia University
during the academic year 1953-54, and I con-
ducted some experiments on “oitation indexing”.
A s the Associate Editor of American Docwmen-
tation, I then suggested that Mr. Adair write an
article on “Citation Indexes for Scientific
Literature?” (7) which w as to be followed by
my own “Citation Indexes for Science”, a paper
sent to Professor H. Bentley Glass for review.
H i s generous editorial efforts made the article
acceptable for publication in Science (8). The
article aroused the early interest of many in-
dividual scientists especially Professor Joshua
Lederberg and Dr. Gordon Allen. In later years
they both were extremely helpful in a large-scale
experimental study of citation indexing (9,lO).
In January, 1961, with the partial support of
National Institutes of Health grant #RG-8050,
the Institute for Scientific Information expanded
the personal research I had conducted from 1955
to 1961 (11,12,13), during which t i m e the con-
cepts were greatly refined. In May, 1961, con-
tract #C-201 was also negotiated with the
National Science Foundation for the purpose of
studying the methodology of preparing citation
indexes based on comprehensive processing of
scientific journals. In addition to the prepara-
tion of experimental subject-specialized citation
indexes (14), the research program ultimateIy
led to the preparation of the initial volumes of
this comprehensive Science Citation Index,
probably the largest single international-
interdisciplinary index to the literature of
science ever compiled.
An introduction to the Science Citation Index,
therefore, would not be complete without briefly
discussing previous attempts to index the
scientific literature on an international-
interdisciplinary scale. Kathrine Murra has
discussed in detail such efforts betweel, 1850
and 1950 (15). She describes seven significant
large-scale projects, including those of the
Royal Society of London and the Brussels
Institute, the predecessor of the International
Federation of Documentation (FID). The latter
group is best known for its Universal Decimal
Classification (UDC) system while the Royal
Society produced the famous International
Catalogue of Scientific Literature, 1901-1914.
While Murra does not cover many other indexes
which surely can be characterized as either
international and/or interdisciplinary, the
bibliographic scene, especially in the past fifty
years, h a s been dominated by selective,
discipline-oriented efforts in chemistry, physics,
engineering, medicine, and biology.
The failure of various groups to compile a
so-called unified index to s c i e n c e (13) h a s been
variously attributed to probf e m s in financing,
inadequate technology, inflexibility, and lack of
imagination, It is probably true that some pro-
ponents of UDC were extreme in their belief
that i t was the solution to the information
problem, Early promoters of microfilm, com-
puters, and other hardware were equally certain
that technology was the major problem.
T h e potentialities of citation indexing are so
broad, and preliminary studies so encouraging,
that it is easy t o appear overzealous. I t is not
claimed that the Science Citation Index is a
panacea. During i t s formative years, it will
often be used in conjunction with, and supple-
mentary to, conventional searching tools. How-
ever, as the scope is increased and the techni-
cal qualities a r e improved, the Science Citation
Index promises to m e e t most of the major
desiderata of an international, interdisciplinary
s c i e n c e index.
If regular u s e of citation indexes points up t h e
need and value of selective or categorical
citation indexes as, for instance, in physics,
chemistry, or biology, techniques a r e already
available which permit u s to produce such
smaller indexes. In this respect computers have
overcome the previous inertial problems of the
older systems, which always involved enormous
manual effort for even a simple change i n
approach.
In response to user demands, the Institute for
Scientific Information plans to prepare categori-
c a l citation indexes but only after a massive
multi-disciplinary reservoir of citations h a s
been created. The categotical citation indexes
extracted from such a “unified” reservoir of
vi
citations will b e far more useful than those
prepared from any arbitrarily restricted subject
coverage. As a large-scale centralized activity
the Science Citation Index solves the dilemma
of Bradford’s Law (16) which states that it is
necessary to cover an exponentially larger
number of journals in order to cover the litera-
ture of any particular discipline. T h i s is another
version of Pareto’s Law used so heavily by
Price (17).
It is recognized that in the major universities
the Science Citation Index may reinforce the
trend towards larger centralized libraries of
science. T h e publication of subjectcategorized
citation indexes might temporarily alleviate the
probIems of departmental libraries. T h e increas-
ing interdisciplinary trend of s c i e n c e indicates
that categorical citation indexes would indeed
only b e a temporary or partial solution to the
problem. As Professor Lederberg h a s said: The
citation index is “a large undertaking but of
course no larger than the problem to which it is
addressed” (18).
Scientific research is in a transition state. I
believe that the Science Citation Index is a
major indicator of even greater changes to come.
T h e so-called Weinberg report to the President’s
Science Advisory Committee (19) recently
recommended: “Transfer of information is a n
inseparable part of research and development.
.., T h e technical community generally must
devote a larger share than heretofore of its time
and resources to the discriminating management
of the ever-increasing technical record. Doing
l e s s will lead to fragmented and ineffective
science and technology.” Your support and con-
structive criticism of our initial s t e p s c a n help
foster this kind of discriminating management.
The Science Citation Index can. be the expres-
sion of a useful partnership between the re-
searcher and the information scientist, between
government and industry, between a l l sectors
of our society concerned with scientific infor-
mation including all educational media, in-
stitutions, and the press (20).
Eugene Garfield, Director
Institute for Scientific Information
Philadelphia, Pennsylvania
November 30, 1963
 REFERENCES
(1) Hyslop, M., “Documentalists Consider Ma-
chine Techniques”, Special Libraries 44,
196-198 (1953).
(2) Larkey, S.V., ”The Army Medical Library
Research Project at the Welch Medical
Library”, Bulletin of the Medical Library
Association 37, 121-124 (1949).
(3) Larkey, S.V., “The Welch Medical Library
Indexing Project”, Bulletin of the Medical
Library Association 41, 32-40 (1953).
(4) Himwich, W.A., H.G. Field, E. Garfield,
J.M. Whittock, and S.V. Larkey, “Survey
of World Medical Serials and Coverage by
Indexing and Abstracting Services”, johns
Hopkins University, Baltimore, 1954; “Fi-
nal Report on Machine Methods for Infor-
mation Searching”, Johns Hopkins Univer-
sity, Baltimore, 1955.
( 5 ) Adair, W.C., Private Communication, March
1953.
(6) Garfield, E., “Review Literature as a
Source of Critical Entries for Scientific
Indexes”, unpublished paper, December
1952.
(7) Adair, W.C., “Citation Indexes for Scien-
tific Literature?” American Documenta-
tion 6, 31-32 (1955).
(8) Garfield, E., “Citation Indexes for Sci-
ence”, Science 122, 108-111 (1955).
(9) Garfield, E., and I.H. Sher, “New Factors
in t h e Evaluation of Scientific Literature
thmugh Citation Indexing”, Am eri c an
Documentation 14, 191201 (1963).
(10) Garfield, E., “Citation Indexes in Socio-
logical and Historical Research”, Ameri-
can Docurnentation 14, 289-291 (1963).
(11) Garfield, E., “Citation Indexes New --
Paths to Scientific Knowledge”, Chemical
Bulletin 43(4), 11-12 (1956).
(12) Garfield, E., “Breaking the Subject Index
Barrier -- A Citation Index for Chemical
Patents”, Journal of the Patent Office
Society 39, 583-595 (1957).
(13) Garfield, E., “A Unified Index to Science”,
Proceedings of the International Confer-
ence on Scientific Information, November
vi i
1958, Vol. I, p. 461-474, National Academy
of Sciences-National Research Council,
Washington, 1959.
Garfield, E., and I.H. Sher, Genetics Cita-
tion Index, 864 p., Institute for Scientific
information, Philadelphia, 1963.
Murta, K.O., “History of Some Attempts to
Organize Bibliography Internationally”, in
J.H. Shera and M.E. Egan, Bibliographic
Organization, p. 24-53, University of
Chicago Press, 1951.
Bradford, S.C., Documentation, 2d Ed.,
p. 153, Crosby Lockwood and Son Ltd.,
London, 1953.
Price, D.J. de Solla, Little Science, Big
Science, p. 49, Columbia University Press,
New York, 1963.
Lederberg, J., Preface to Genetics Cita-
tion Index (see reference 14).
Weinberg, A.M., et al., President’s Science
Advisory Committee, Science, Government,
and Infomation (The Responsibilities of
the Technical Community and the Gouem-
ment in the Transfer of Information), Part
3, ‘Titation Indexing Should be Useful”,
p. 35 and “Summary and Major Recommen-
dations”, p. 1 , Government Printing Office.
Washington, 1963 (see also review by
C.D. Leake, Science 139, 1088 (1963)).
Garfield, E., “Citations in Popular and
Interpretive Science Writing”, Science
141, 392 (1963).
ADDITIONAL PERTINENT REFERENCES
Salton, G., “Associative Document Retrieval
Techniques Using Bibliographic Information”,
Journal of the Association for Computing Ma-
chinery 10, 410-457 (1963).
Seidel, A.H., “Citation System for Patent
Office”, Journal of the Patent Office Society
31, 554 (1949).
Tukey, J.W., “Keeping Research in Contact
with the Literature: Citation Indices and Be-
yond”, Journal of Chemical Documentation 2,
34-37 (1962).
Westbrook, J .H., “Identifying Significant Re-
search”, Science 132, 1229-1234 (1960).
 GLOSSARY

Citation Index A directory of cited references, e a c h of which

is accompanied by a list of c i t i n g s o u r c e
documents

Science Citation Index A comprehensive citation index containing all

references appearing i n t h e bibliographies o r
footnotes of s o u r c e a r t i c l e s published in a
large number of multidisciplinary a n d s p e c i a l t y
journals. E a c h reference is accompanied by its
list of c i t i n g a r t i c l e s

Author Citation Index A citation index which is alphabetized by t h e

f i r s t reference author. The Science Citation
Index is a n author citation index

Source The original c i t i n g publication from which

references a r e obtained

Reference Any item c i t e d i n t h e bibliography o r t e x t of a

source document or publication

Target Reference Any reference u s e d a s a starting point in a

search

Source Article Index An index, alphabetically arranged by first

c i t i n g (source) author, of t h o s e s o u r c e a r t i c l e s
u s e d to produce the citation index. T h e Science
Citation Index contains a s o u r c e article index

viii
 WHAT IS A CITATION INDEX?

REFERENCE STATISTICS
A citation index is a directory of cited refer-
ences where each reference i s accompanied by Total Citations 1,370,000.
a list of source documents which cite it. The Total Unique References Cited 890,000
most characteristic feature of the citation index Total Unique Reference Authors Cited 258.000
is that the user begins a search with a specific
known paper (target reference). From this *Includes 222.000 non-journal
starting point one is brought forward in time to and 20.200 anonymous references
subsequent papers related to the earlier paper.
In using the Science Citation Index, the reader
must differentiate between comprehensiveness
of sources and of references and should under- Chronological Distribution of References
stand the distinction between source year and
reference year. Cumulative Cumulative
Percentage Percentage
“Comprehensiveness of sources” applies to the Reference of Total of Unique
breadth of source journal coverage, that is, the Year Citations References
number of journals and disciplines covered, 4.14
1961 3.51
countries represented, etc., a s well a s chrono- 1960 15.09 15.04
logical depth. Source data in these initial five 1959 27.29 26.07
1958 37.30 35.21
volumes of the Science Citation Index was 1957 45.46 42.72
taken from 613 key source journals published 1956 52.34 49.15
in 1961 in 28 countries. All major fields of 1955 58.16 54.72
1954 63.01 59.46
scientific research are covered. Every 1961 67.26 63.58
1953
issue of these journals was included. 1952 70.82 67.10
##
1951 73.94 70.16
Comprehensiveness of references” refers to 1950 76.60 72.82
the scope of reference coverage. In this Science 1949 78.75 75.03
1948 80.51 76.79
Citation Index every reference citation in every 1947 82.55 78.19
journal issue processed was included, no matter 1946 82.86 79.28
where or when the reference originally appeared. 1945 83.68 80.12
1944 84.43 80.93
Over 20,000 reference journals are cited. In 1943 8 5.19 8 1.75
these first five volumes there are 1.4 miIIion 1942 85.97 82.61
citations describing articles published in any 1941 86.84 83.55
1940 87.71 84.52
period of recorded history up until 1961 on any 1939 88.57 85.50
subject ever cited by the sources used. The 1938 89.42 86.47
validity of this statement i s indicated by the 1937 90.17 87.38
1936 90.87 88.22
following summary statistics on t h e references 1935 91.50 88.99
in the Science Citation Index. The user should 1934 92.08 89.69
keep in mind that the relatively low number of 1933 92.63 90.35
1932 93.17 90.99
citations to earlier years is partly a function of 1931 93.63 91.56
the lower volume of scientiiic publication in 1930 94.07 92.12
those years. 1929 94.47 92.61

ix
For every reference appearing in every a r t i c l e U generally means “unpublished” but in t h e
in e a c h source journal a s e p a r a t e IBM punched- case of Svedberg’s book i t signifies “Ultra-
card is prepared containing both the source d a t a zentrafuge”. To illustrate the acronyms, fuller
and t h e reference data. T h e punched-cards a r e identification of some well known works is
subsequently converted to magnetic tapes. T h e given below.
t a p e s a r e then sorted and otherwise processed
on IBM 1401, 1410, and 7074 computers. T h e
sorted data a r e printed on a n IBM 1403. T h e FREQUENTLY CITED NON-JOURNALS
master copy is then photographed and reproduced
by offset printing. AEC
AMOM
A T L A S O F E X F O C I A T I V E CYTOLOGY
ANGULAR MOMENTUM I N QUANTUM
PAPANICOLAOU GN
EOMONOS AR
MECHANICS
BMOB BERGEYS MANUAL O F O E T E R M I N A T I V E ANON
BACTERIOLOGY
85 B I O C U E M I S T R V OF SEMEN MANN T
UTCW EL000 T R A N S F U S I O N I N C L I N I C b L MOLL ISON PL
MED I C I N E
R E F E R E N C E PUBLICATION POLICIES CBA CONSTITUTION OF BINARY ALLOYS bIANS.CN M
CBU C H E M I C A L B A S I S OF U E R E D I T Y VOGEL U J
COBM C U E M I C A L O Y N A M I C S OF BONE M I N E R A L NEUMAN I F
Both journal and non-journal citations a r e proc- CMBNA C U E M I C A L A N 0 MOLECULAR B A S I S OF
NERVOUS A C T I V I T Y
NACHMANSOHN D

essed. Non-journal references a r e readily iden- 00s

OHBG
O I E L E C T R I C B E H A V I O U R A N 0 STRUCTURE
DISTRIBUTION OF TUE UUMAN m o o 0
SMYTU C P
MOURANT AE
tified in t h e index by a lozenge symbol (a)at onc
GROUPS
D I S E A S E S O F T U E H E A R T AND WOO0 P
the end of the publication column. Non-journal OTEDCA
CIRCULATION
D I A G N O S I S A N 0 TREATMENT O F WILKINS L
references include references to books, meetings, E N O O C R I N E 01SORDERS I N C U l L O n O O D
A N 0 ADOLESCENCE
dissertations, reports, contracts, patents, cir- E
EA
ENZYMES
EXPERIMENTAL bTUER0SCLEROSI.S
OlXON M
K A T Z LN
culars, etc. In t h e s e f i r s t five volumes titles of EDPR E X P E R I M E N T A L DESIGN I N
PSYCUOLOGICAL RESEARCU
EDWARDS A L

books o r meetings a r e presented a s acronyms. EFAHB E P I L E P S Y AND T n E F U N C T I O N A L

ANATOMY O F T U E HUMAN B R A I N
PENFIELD I

In subsequent volumes acronyms will be re- En E L E K T R O P H V S I O L O G I E OER

U€RZMUSKELFASER
YEIOMANN S

placed by full t i t l e s or easily understood abbre- E l I P ELECTRONIC AN0 I O N I C IMPACT

PUENOMENA
MASSEY US

viations. Reports and circulars may be identified ES

ETAM
E L E C T R O L Y T I C S D L U T I ONS
E L E M E N T A R V TUEORV O F A N W L A R
ROBINSON RA
ROSE ME
either by numbers, or, if t h e reference to t h e re- FRS
MOMENTUM
FREE RADICALS I N SOLUTION WALLING C
port included a title, by acronym. Dissertations GU
nE
GENETIC UOMEOSTASIS
HYDROGEN BOND
L E R N E R IM
P I M E N T E L GC
and t h e s e are recorded a s “D~SS.’’ The fol- UBP
UD
U IG U BLOOD PMSSURE
UEMORRUAGIC D I S E A S E S
P I C K E R I N G GW
Q U I C K AJ
lowing is a l i s t of common abbreviations i n noic U E A R T D I S E A S E I N I N F A N C Y AND
cn I LDUOOO
K E I TU J O

non-journal citations: URNMR U l G U RESOLUTION NUCLEAR MAGNETIC

RESONANCE
POPLE JA
URP nEMOCYS1.S AND R E L A T E D PUENOMENA PONDER E
IP In P r e s s UTPU UISTOPATUOLOGIC TECUNIC AN0 L I L L I E RD
PRACTICAL UISTOCUEMISTRY
PC Private Communication I
IE
cc INTERNAL CONVERSION C O E F F I C I E N T S
I O E N T I F I C A T I O N O F TME
ROSE ME
CDwAROS PR
TBP To Be Published LLLT
ENTEROUACTER I ACEAE
L Y M P U A T I C S LYMPH A N 0 L Y M P P H O I D YOFFEV JM
TISSUE
U Unpublished MA I MEASUREMENT OF A D U L T I N T E L L I G E N C E WECUSLER 0
MBA METHODS O F B I O C H E M I C A L A N A L Y S I S W I N Z L E R RJ
UD Unpublished Data MBA
MBA
METHODS OF B I O C H E M I C A L A N A L Y S I S
METUODS OF BIOCHEMICAL ANALYSIS
KUNKEL n G
FRAENKELCONRAT U
MC MAST C E L L S R I L E Y JF
ME METHODS I N ENZYMOLOGY SCHbCHMAN HK
T h e original rationale for using acronyms, how- ME METHODS I N ENZYMOLOGY LAYNE E
ME METHODS I N ENZYMOLOGV KORNBERG A
ever, is better appreciated if the user k e e p s in ME METHODS I N ENZYMOLOGY KORNBERG A
MM M E I S U R E M E N T OF M E A N I N G OSGOOO CE
mind t h a t in the Science Citatwn Index t h e MMS M A T H E M A T I C A L METUOOS O F S T A T I S T I C S CRAMER U
MPPS MICROSOMAL P A R T I C L E S A N 0 P R O T E I N H A L L UO
target reference i s usually known. P l e a s e note SYNTUES I S
MS MICROWAVE SPECTROSCOPY TOWNES cn
t h a t acronyms may not b e unique. For example, MT MANOMETRIC TECUNIOUES UMBREIT WW

X
MTGL MOLECULAR THEORY O F G A S E S A N 0
LIOUIDS
HIRSCHFELOER J O
In t h e s e first f i v e volumes, references a r e con-
MTS ___
MODFRN THEORY O F SOLIDS
MJCLEIC ACIDS
SEITZ F
HOAGLAND MB
sidered identical if they match on minimal dis-
NA
CHARGAFF E
NA
re€
W C L E I C ACIDS
NEW B A S E S OF E L E C T R O C A R D I O G R A P H Y SODIPULARES 0
tinctive citation data. T h e computer then t a l l i e s
NATURE O F T H E C H E M I C A L B O W
NCU
Nce NATURE OF THE C n E M I C A L B O W
PAULING C
PAULING L
variations in t h e presentation of authors a n d
NCPG NEURAL CONTRQ
GLAND
O F TME P I T U I T A R Y H A R R I S GY
journals within e a c h c l u s t e r of “identical”
NSR
oe
NUCLEAR S T R I P P I N G R E A C T I O N
O R-G A N.-I Z A T I O N OF B E H A V I O R
B U T L E R ST
HEBU 0
references, selects the most frequently c i t e d
OMA
PA
O F F I C I A L METHODS OF A N A L Y S I S
P L A N T ANATOMY
6NON
ESAU K
or, i n case of t i e s , the longest version of t h e s e
PRINCIPLES OF BEHAVIOR nuLL CL
PB
PUS PnYSIOLOGY AN0 BIOCHEMISTRY OF THE ROTHMAN S
elements a n d s t a n d a r d i z e s the reference. A s a
PBT
SKIN
PnARMACOLOGlCAL B A S I S OF GOODMAN L S
result, i n most cases, t h e common reference
THERAPEUTICS
PFIG PnysIcs O F F U L L Y I O N I Z E D G A S E S SPITZER L
will be l i s t e d only once, and all t h e descrip-
P U A C T I C A L HAEMATOLOGY
PH
PNC P n Y S I O L O G Y OF N E R V E C E L L S
DACIE JV
ECCLES JC
tions of a r t i c l e s c i t i n g i t will follow in o n e
P R I N C I P L E S OF O P T I C S BORN M
PO
POC P H Y S I C A L ORGANIC CHEMISTRY HAMMETT LP group. In s u b s e q u e n t volumes even greater
POCF P H Y S I O L O G Y O F T H E OCULAR A N 0 DAVSON n
CEREBROSPINAL F L U l U S standardization will be achieved by u s e of new
PPC P R I N C I P L E S O F POLYMER C H E M I S T R Y FLORY P J
PRP P R I N C I P L E S OF RENAL PHYSIOLOGY s M i T n nu s y s t e m s of processing and new computer pro-
PT PATHOCOGY O F TUMOURS W I L L I S RA
RACS R E F L E X O G E N I C AREAS OF T H E HEYMANS C grams. Nevertheless, i n s o m e cases, c i t a t i o n s
CARDIOVASCULAR SYSTEM
SCB STRENGTHS O F C H E M I C A L BONDS COTTRELL TL to the s a m e reference may be separated. T h i s
SO* SPECTRA O F D I A T O M I C M O L E C U E S HERZUERG G
SI S T A T I S T I C A L INFERENCE WALKER tin o c c u r s whenever the reference is presented in a
SM S T A T I S T I C A L METHOOS SNEDECOR G I
SMBA S T A T I S T I C A L METHOD I N B I O L O G I C A L FINNEV O J variation t h a t defied the unification procedure.
ASSAY
SMOC STRUCTURE A N 0 M E C H A N I S M I N O R G A N I C I N G O L O CK F o r example, though a reference author’s name
CHEMISTRY
SP STOCHASTIC PROCESSES OOOtl JL w i l l usually appear i n the literature with two
SP SUUCELLULAR P A R T I C L E S
SSR S P E C I F I C I T Y OF SEROLOGICAL
OEOUVE C
L A N D S T E I N E R 1.
initials (if indeed h e h a s two initials), it may
REACT I O N S
Tn T n Y R O I o noRMouEs PITTRIVERS R
occasionally be c i t e d with only o n e initial or
TLP THEORETICAL NUCLEAR P H Y S I C S ULATT JM
TRP THEORY OF R A T E P R O C E S S E S GLASSTONE 5
e v e n none. T h e s e a l t e r n a t i v e s should b e checked.
TTCBJ TUMORS A N 0 TUnOROUS C O N D I T I O N S OF JAFFE nL
T n E BONES AND JOINTS Differences i n s p e l l i n g o r transliteration of a n
U U L T R A Z E N T R IFUGE SVEDBERG T
author’s name from country to country may also
defy t h e unification procedure and c a u s e some
duplication. Cases of references which a r e n o t
“Cited Indirectly” appearing in t h e reference unified d u e to incomplete or incorrect reference
publication column means that the c i t i n g author d a t a , different page numbers and variant titles
learned of t h e reference work through reading a will also occur. However, t h e so-called “unifi-
third work, which may or may n o t h a v e given a cation’’ procedure actually corrects many errors
complete citation. T h e user c a n retrieve t h e i n reference journal titles and in reference
citing paper to identify t h i s third work, which authors’ names. Eventually our processing s y s -
also appears in the index a s a reference. The tem will make p o s s i b l e t h e correction of litera-
latter may be u s e d as a n additional target refer- ture errors i n d a t e s , volume numbers, etc,
e n c e i n s e a r c h i n g with t h e citation index.
FORMAT
The Science Citation Index is arranged alpha-
betically by first reference author, year, publi-
UNIFICATION AND STANDARDIZATION cation (usually a journal) , volume, a n d page.
The s o u r c e a r t i c l e s c i t i n g a particular reference
Frequently, a given reference is c i t e d i n dif- article a r e arranged alphabetically by t h e first
ferent ways in t h e literature. F o r example, dif- s o u r c e author - - immediately under each refer-
ferent abbreviations for t h e s a m e journal may e n c e line. Citations to anonymous reference
occur. In compiling a citation index, we must a r t i c l e s follow the author s e c t i o n of t h e citation
pull together all c i t a t i o n s to t h e s a m e reference, index. Descriptions of anonymous source articles
i n s p i t e of these variations. Unification and appear immediately under t h e corresponding
partial standardization of reference citations i s reference line, preceding t h e other s o u r c e s
accomplished by s p e c i a l computer programs. c i t i n g t h e s a m e reference.

xi
In the source lines, t h e y e a r follows the publi-
cation, instead of preceding it, a s in the refer-
e n c e lines. While reference years may be any
year in recorded history, t h e source y e a r in
t h e s e first five volumes will always b e 1961. In
subsequent volumes the s o u r c e y e a r will be the
current year.
A maximum of eighteen characters w a s allotted
for the reference author and eleven characters
for t h e source author. In both cases, a maximum
of two initials w a s allowed. Twenty characters
are allotted for the reference publication, a n d
eleven characters for the source publication.
Thus, t h e source authors and journals may b e
shorter than corrcsponding reference entries. A
list of the full titles of all source journals pre-
c e d e s the Source Article Index.
Asterisks flanking t h e reference year identify
t h e first reference line, and usually t h e first
year cited, for a given author. Additional dashed
l i n e s indented under the reference author iden-
tify other c i t e d a r t i c l e s or books published by
t h e s a m e author. Additional uncited works c a n
sometimes be found in the source article index.
An illustrative page of the Science Citation
Index will .be found inside t h e covers of e a c h
volume.
Those of u s who helped prepare the index are
well aware of imperfections such as truncated
xii
source authors or journals. Source authors’ last
names may sometimes be m n c a t e d to e i g h t
characters. T h e source journal abbreviations are
limited to e l e v e n characters. T h e s i z e of print
is much smaller than the average index, but
t h i s h a s s a v e d shelf space. And surely it would
h a v e been most desirable, though not economi-
cally feasible, to include t h e full titles of
source articles in t h e s e f i r s t volumes. However,
improvements will be incorporated in subsequent
volumes of t h e Science Citation Index. Thus,
t h e s o u r c e article index in future i s s u e s of the
Science Citation Index will include s o u r c e
article titles. T h i s decision is important to t h e
s c i e n t i s t a n d librarian as t h i s feature w i l l con-
stitute the first compkehensive, interdisciplinary
inventory of current scientific publications
available. Furthermore, future volumes of t h e
Science Citation Index w i l l include all s o u r c e
co-authors’ names. T h e cumulation of t h i s im-
proved source a r t i c l e index will obviate the
search through numerous fragmentary, selective
indexes to identify a scientist’s publications --
whether they have been c i t e d or not. T h e reader
should note, however, t h e function of t h e citation
index i s primarily to d i s c l o s e by whom and where
individual papers h a v e been cited. Thus, while
Lineweaver and Burk are t h e co-authors of “ T h e
Determination of Enzyme Dissociation Con-
s t a n t s ” J . Am. Chem. SOC. %, 658 (1934), it is
only necessary to refer to Lineweaver, the first
author, to find who h a s cited this paper ( s e e
page 1514).
In the source lines, t h e y e a r follows the publi-
cation, instead of preceding it, a s in the refer-
e n c e lines. While reference years may be any
year in recorded history, t h e source y e a r in
t h e s e first five volumes will always b e 1961. In
subsequent volumes the s o u r c e y e a r will be the
current year.
A maximum of eighteen characters w a s allotted
for the reference author and eleven characters
for t h e source author. In both cases, a maximum
of two initials w a s allowed. Twenty characters
are allotted for the reference publication, a n d
eleven characters for the source publication.
Thus, t h e source authors and journals may b e
shorter than corrcsponding reference entries. A
list of the full titles of all source journals pre-
c e d e s the Source Article Index.
Asterisks flanking t h e reference year identify
t h e first reference line, and usually t h e first
year cited, for a given author. Additional dashed
l i n e s indented under the reference author iden-
tify other c i t e d a r t i c l e s or books published by
t h e s a m e author. Additional uncited works c a n
sometimes be found in the source article index.
An illustrative page of the Science Citation
Index will .be found inside t h e covers of e a c h
volume.
Those of u s who helped prepare the index are
well aware of imperfections such as truncated
xii
source authors or journals. Source authors’ last
names may sometimes be m n c a t e d to e i g h t
characters. T h e source journal abbreviations are
limited to e l e v e n characters. T h e s i z e of print
is much smaller than the average index, but
t h i s h a s s a v e d shelf space. And surely it would
h a v e been most desirable, though not economi-
cally feasible, to include t h e full titles of
source articles in t h e s e f i r s t volumes. However,
improvements will be incorporated in subsequent
volumes of t h e Science Citation Index. Thus,
t h e s o u r c e article index in future i s s u e s of the
Science Citation Index will include s o u r c e
article titles. T h i s decision is important to t h e
s c i e n t i s t a n d librarian as t h i s feature w i l l con-
stitute the first compkehensive, interdisciplinary
inventory of current scientific publications
available. Furthermore, future volumes of t h e
Science Citation Index w i l l include all s o u r c e
co-authors’ names. T h e cumulation of t h i s im-
proved source a r t i c l e index will obviate the
search through numerous fragmentary, selective
indexes to identify a scientist’s publications --
whether they have been c i t e d or not. T h e reader
should note, however, t h e function of t h e citation
index i s primarily to d i s c l o s e by whom and where
individual papers h a v e been cited. Thus, while
Lineweaver and Burk are t h e co-authors of “ T h e
Determination of Enzyme Dissociation Con-
s t a n t s ” J . Am. Chem. SOC. %, 658 (1934), it is
only necessary to refer to Lineweaver, the first
author, to find who h a s cited this paper ( s e e
page 1514).
A s with any conventional index, experience in
using a citation index increases one’s ability
to find information quickly. However, unlike
language-oriented indexes, training in medical,
chemical, or other nomenclature i s not required
t o u s e a citation index effectively.
[n the citation index t h e “subject” i s the infor-
mation or concept contained in t h e starting
point - - the target reference. T h u s t h e s e a r c h i s
independent of s u b j e c t nomenclature and quickly
leads t o current papers related to the target
reference.
The citation index is concept oriented. I t is,
therefore, a relatively sophisticated searching
tool which presupposes some knowledge of a
starting point. A target reference must either be
known or first identified through a footnote, a n
encyclopedia, book, o r s u b j e c t index. T h e cita-
tion index then a n s w e r s t h e question, “What h a s
happened s i n c e ? ” We believe t h i s question i s
fundamental to research activity.
SEARCH STRATEGY
To locate sources which cite a particular target
paper, i t will usually suffice to know the name
of t h e first author, year, and page. It i s general-
ly not even n e c e s s a r y to know t h e title of t h e
journal or publication in which the paper ap-
peared. In the citation index t h e cited or refer-
e n c e author is e a s i l y located on t h e left. F o r
e a c h cited paper by that author there is a d a s h e d
l i n e which continues to the column reserved for
the year of the reference publication. T h e jour-
nal abbreviation (or non-journal acronym), volume
and page a r e found to the right of the reference
year, Indented under e a c h dashed l i n e a r e d a t a
identifying source articles which h a v e cited
that specific reference. When a given reference
...
Xlll
h a s been c i t e d by several s o u r c e s t h e l a t t e r a r e
arranged alphabetically by source authors.
On the average, the u s e r will find only a few
s o u r c e papers which cite a particular target
paper. T h e citation index is highly s p e c i f i c and
the retrieved s o u r c e s h a v e a direct relationship
to t h e starting point. By various techniques a
s e a r c h may be readily expanded i n order to build
more extensive bibliographies. For instance,
one may select relevant c i t a t i o n s appearing in
the bibliographies of the source articles dis-
c l o s e d by t h e first stage of t h e search. T h e s e
citations, as well as the s o u r c e s which con-
tained them, c a n now become new target refer-
e n c e s or entry points to t h e citation index. T h i s
66
aycling” procedure may b e repeated.
Another technique t a k e s advantage of the fact
that t h e s a m e author will frequently w r i t e more
than o n e closely related paper. Therefore when
u s i n g t h e citation index, additional relevant
a r t i c l e s by t h e author of t h e first target refer-
e n c e c a n be u s e d as target references and the
s o u r c e s citing these articles may be examined.
Additional current articles by a given author
may also be found in the s e p a r a t e source article
index even though they may be too current to
have been c i t e d yet.
Starting with a given target reference paper one
may find that i t is n o t cited a t all. T h i s may or
may not b e good news. Do not give up! Examine
t h e target paper itself and look over the cita-
tions contained in it. Identify t h e reference most
pertinent to your search. Then, i n the citation
index, locate s o u r c e s which c i t e this new refer-
ence. If not, try another relevant reference from
the s a m e target paper. Since the average paper
h a s a bibliography of fourteen items i t is rare
that a pertinent source i s not located, provided
there h a s been some work done on the subject.
When u s i n g a citation index a negative result
A s with any conventional index, experience in
using a citation index increases one’s ability
to find information quickly. However, unlike
language-oriented indexes, training in medical,
chemical, or other nomenclature i s not required
t o u s e a citation index effectively.
[n the citation index t h e “subject” i s the infor-
mation or concept contained in t h e starting
point - - the target reference. T h u s t h e s e a r c h i s
independent of s u b j e c t nomenclature and quickly
leads t o current papers related to the target
reference.
The citation index is concept oriented. I t is,
therefore, a relatively sophisticated searching
tool which presupposes some knowledge of a
starting point. A target reference must either be
known or first identified through a footnote, a n
encyclopedia, book, o r s u b j e c t index. T h e cita-
tion index then a n s w e r s t h e question, “What h a s
happened s i n c e ? ” We believe t h i s question i s
fundamental to research activity.
SEARCH STRATEGY
To locate sources which cite a particular target
paper, i t will usually suffice to know the name
of t h e first author, year, and page. It i s general-
ly not even n e c e s s a r y to know t h e title of t h e
journal or publication in which the paper ap-
peared. In the citation index t h e cited or refer-
e n c e author is e a s i l y located on t h e left. F o r
e a c h cited paper by that author there is a d a s h e d
l i n e which continues to the column reserved for
the year of the reference publication. T h e jour-
nal abbreviation (or non-journal acronym), volume
and page a r e found to the right of the reference
year, Indented under e a c h dashed l i n e a r e d a t a
identifying source articles which h a v e cited
that specific reference. When a given reference
...
Xlll
h a s been c i t e d by several s o u r c e s t h e l a t t e r a r e
arranged alphabetically by source authors.
On the average, the u s e r will find only a few
s o u r c e papers which cite a particular target
paper. T h e citation index is highly s p e c i f i c and
the retrieved s o u r c e s h a v e a direct relationship
to t h e starting point. By various techniques a
s e a r c h may be readily expanded i n order to build
more extensive bibliographies. For instance,
one may select relevant c i t a t i o n s appearing in
the bibliographies of the source articles dis-
c l o s e d by t h e first stage of t h e search. T h e s e
citations, as well as the s o u r c e s which con-
tained them, c a n now become new target refer-
e n c e s or entry points to t h e citation index. T h i s
66
aycling” procedure may b e repeated.
Another technique t a k e s advantage of the fact
that t h e s a m e author will frequently w r i t e more
than o n e closely related paper. Therefore when
u s i n g t h e citation index, additional relevant
a r t i c l e s by t h e author of t h e first target refer-
e n c e c a n be u s e d as target references and the
s o u r c e s citing these articles may be examined.
Additional current articles by a given author
may also be found in the s e p a r a t e source article
index even though they may be too current to
have been c i t e d yet.
Starting with a given target reference paper one
may find that i t is n o t cited a t all. T h i s may or
may not b e good news. Do not give up! Examine
t h e target paper itself and look over the cita-
tions contained in it. Identify t h e reference most
pertinent to your search. Then, i n the citation
index, locate s o u r c e s which c i t e this new refer-
ence. If not, try another relevant reference from
the s a m e target paper. Since the average paper
h a s a bibliography of fourteen items i t is rare
that a pertinent source i s not located, provided
there h a s been some work done on the subject.
When u s i n g a citation index a negative result
may have more significance than in any other
indexing system since the problem of synonymy
does not exist. There will be therefore, less
uncertainty about the meaning of a negative re-
sult. No index can locate information that does
not exist. A key problem is to obtain this nega-
tive information quickly.
Naturally a citation index w i l l be most helpful
when used in conjunction with a large library.
Since the scope of t h e s e indexes i s quite broad,
source articles may be encountered which a r e
not readily available. T h e Institute for Scientific
Information will assist in the procurement of
articles in any source publication through the
Original Article Tear Sheet service (OATS) of
t h e Institute for Scientific Information.
xiv
Citation indexing is not recommended merely as
a substitute for conventional indexes. Rather,
the citation index adds a new dimension to the
pursuit of scientific knowledge. T h e citation
index is your historical roadmap of the litera-
ture. Where you travel i s primarily your decision.
What you find depends on you and what is avail-
able. Only you can measure its relevance. What
may be valuable to o n e man is irrelevant to
another. Two otherwise unrelated scientific
observations may be correlated in the citation
index through a common reference or through a
chain of source-reference links. However, you
must select the starting point, trace t h e network,
and detect the correlation - - if it exists.
A few actual searches will quickly demonstrate
the simplicity of the citation index system,
 APPLICATIONS OF THE SCIENCE CITATION INDEX
The fundamental question o n e can answer quick-
ly through t h e Science Citation Index i s , “Where,
and by whom, h a s this paper been cited in the
literature?” T h e significance of t h e answer is
fundamental t o the method and the scholarly
tradition of s c i e n c e i n t h a t appropriate refer-
ences -- methodological, ideational, historical,
or otherwise -- are cited to support arguments
or facts. We have found this tradition is, i n
general, closely followed. Naturally there a r e
violations and a b u s e s but t h e s e a r e the excep-
tions and not common practice.
The Science Citation Index w i l l e n a b l e the u s e r
to trace new applications of theories, methods,
instruments, chemicals, etc. Corrections, errata,
amendments, refutations, letters, editorials,
discussions, translations, reviews, etc. c a n also
b e located though, in the p a s t , they may h a v e
been buried in the literature.
T h e Science Citation lndex follows the thread
of a theme through t h e years and a c r o s s artifi-
cial boundaries imposed by classification
schemes and journal titles. For example, con-
s i d e r Einstein’s 1905 paper analyzing Brownian
movement a n d diffusion (Ann. Fhysih. 17, 549).
T h i s paper w a s cited in 1961 not only in chem-
istry, mathematics a n d physics journals but also
in a biological study on t h e immunological ap-
plication of double diffusion (C.J. Van Oss,
Zeit. Immunitatsfotsch. 122, 44 (1961)).
T h e disclosure by the Science Citation lndex of
current improvements in methodology may b e
illustrated by the literature on t h e determination
of thiamine and i t s phosphoric acid esters in
animal tissues. Using any one of several target
papers describing various earlier methods, one
would b e led to the article by G. Rindi, Biochem.
J. 78, 602 (1961) in which h e reports a newchro-
matographic method which i s claimed to b e an
improvement over the previous quantitative and
qualitative methods which h e cites.*
xv
Furthermore, continuing another c y c l e and u s i n g
Rindi’s article as a new target reference, t h e
Science Citation Index reveals four more recent
papers which cite this work. In three, t h e new
method h a s been applied and the fourth describes
a method for determining the related compound,
pyrithiamine.
A secondary application of the Science Citation
Index is its u s e a s a conventional author index
to help identify an author’s publications. Note,
however, that i n t h e s e first five volumes only
papers in which h e appears as first author c a n
be found under a specific author’s name. Fur-
ther, only if t h e work was cited will it b e found
in t h e main body of t h e citation index.
T h e citation index is invaluable in preparing
historical introductions to scientific papers,
critical reviews, and books. Related to histori-
c a l s t u d i e s are the sociological applications of
citation indexes for evaluation of s c i e n t i s t s
and their work. T h e s e applications a r e legiti-
mate to the extent that o n e judiciously u s e s a
citation index a s a retrieval tool to facilitate
the location of criticisms of a man’s work.
Qualitative judgements must always temper
quantitative data. T h e Science Citation Index
c a n a l s o b e used to identify quickly, s c i e n t i s t s
currently working on s p e c i a l problems.
A great d e a l h a s been s a i d about the u s e of
citation indexes for ego gratification a n d i t is
important to briefly d i s c u s s the s u b j e c t here.
Well-meaning individuals may reject citation in-
* Bartley W., B i o c h e m . J., 56, 3 7 9 (1954)
B e r n a b e i , O . , N a t u r w i s s . , 46, 2 2 9 ( 1 9 5 9 )
Gurtner, H.P..H e l v . P h y s i o l . A c t a , 15, C 6 6 ( 1 9 5 7 )
K a z i r o , Y . , J . B i o c h e m . , T o k y o , 44, 8 2 7 ( 3 9 5 7 )
K i e s s l i n e , K.H., A r k . K e m l , 6, 2 7 1 ( 1 9 5 3 )
R o s s i - F a n e l l i . A . . Science, 116, 7 1 1 ( 1 9 5 2 )
Spadoni, M.A., Quad. N u t r . , 11, 2 6 ( 1 9 5 0 )
W e s t e n b r i n k . H . G . K . . I n l . 2. V l t a m l n f o r s c h . , 21, 4 6 1
(1949).
dexing because they a s s u m e it w a s designed for
vainglorious purposes. We do not believe t h e
Science Citation Index will b e supported purely
for ego gratification of subscribers. Ego gratifi-
cation, however, i s not the only motivation for
a s c i e n t i s t who w i s h e s to determine whether
h i s work h a s been applied or criticized by
others. Each s c i e n t i s t e v a l u a t e s the citations
xvi
to h i s works differently. One man’s work stimu-
lates another. The citation index facilitates
feedback in the communication cycle. Any
author may c h o o s e to ignore c i t a t i o n s t o h i s own
work ( l e s t h e feed h i s vanity) and s t i l l , as
described previously, wish to retrieve publica-
t i o n s which c i t e other works, by other scien-
t i s t s , in which h e i s interested.
 SOURCE COVERAGE AND STATISTICS
SOURCE JOURNALS
T h e journals processed for t h e first five voI-
umes of t h e Science Citation Index a r e presented
in several auxiliary lists. T h e first g i v e s the
source journal abbreviations used in the index
along with the full journal titles. The second
list is the same data arranged by full title. T h e
third list shows the source journals grouped by
country of origin. Several Soviet journals are
1iste.d under both the Union of Soviet Socialist
Republics a s well as the country of translation,
b e c a u s e some translation i s s u e s were processed
for convenience. In t h e s e cases, translation
journals a r e also asterisked.
In future volumes of the Science Citation index
the number of source journals to b e processed
will increase to several thousand. However,
journal and article statistics sometimes c a n be
quite deceiving. To process a single weekly
journal l i k e Nature requires far more effort than
t h e processing of fifty smaller quarterlies. One
large journal may produce s e v e r a l thousand
original articles per year, whereas a small
journal may contribute a few dozen.
The emphasis in our journal selection h a s been
t h e multidisciplinary journals. T h e s e are, gener-
ally, recognized as important in e a c h country,
but are only selectively and fragmentarily
covered by t h e discipline-oriented indexing
services. Our policy is to treat each journal
comprehensively. This policy will eliminate t h e
previous doubts as to whether o r n o t a particular
article m e t t h e s e l e c t i o n criteria of a specialty
index.
While many of the journal titles a r e i n the life
s c i e n c e s this also c a n b e somewhat deceiving.
T h e field of chemistry i s very well represented
but may be considered a physical o r l i f e s c i e n c e
according to one's viewpoint. Further, in chem-
xvii
istry a n d physics, in contrast to biology, the
main journals a r e generally quite large. It is
necessary therefore to cover more journals in
biology and medicine in order to cover a corn-
parable number of articles. More articles a r e
published in t h e Physical Review e a c h year
than i n dozens of smaller, but also significant,
biological publications.
While it is important for the u s e r to know t h e
s c o p e of t h e s e first five volumes, w e wish to
s t r e s s t h e basis for future s e l e c t i o n s of jour-
nals. The number of s o u r c e journals to be in-
cluded in future volumes of t h e Science Citation
Index is limited strictly by t h e economic factors
involved. It is our intention to expand s o u r c e
journal coverage as broadly a n d rapidly a s
economics permit. T h e Institute for Scientific
Information h a s now amassed considerable d a t a
which enables u s to make s o u r c e journal selec-
tions based on several criteria of importance.
T h e initial corps of journals will be those of
interest to the widest possible audience of
s c i e n t i s t s with primary emphasis on multidisci-
plinary and interdisciplinary journals as well
a s the most significant journals in every branch
of science. T h e subscribers will be consulted
a n d their interests must a l s o affect the selec-
tion criteria. T h e number of source journals
that w i l l be covered will be increased a s sup-
port for the service increases. However, i t
should be emphasized t h a t source journal cover-
age i n no way limits reference journal coverage.
In these first five volumes alone, approximately
20,000 different reference journals were cited,
T h e combination of a judicious selection of
source journals together with comprehensive
treatment of reference journals produces an in-
dex which extends far beyond t h e s c o p e of
ordinary indexes.
T h e non-journal material which is picked up as
references may prompt the u s e r to a s k whether