citation classics -- four years of the human side of … · curremcomments citation ckmrics—four...

12
curremcomments Citation Ckmrics—Four Yearn of the Human Sfde of Scfence Number 22 Most readers of Current Contents@ (C@ ) are by now familiar with the Cita- tion Classics feature which began in 1977. It seems to me that after four years of thm weekly feature an evalua- tion is now in order. In reviewing Citation Classics, one is immediately struck by the fact that our goals have changed somewhat in the past four years. Originally, we planned to draw our classics from a “group of 500 papers most-cited during the years 1961-1975.”1 By 1979, however, we realized the consequences of this nar- row approach to selecting papers. Most of the 50U most-cited papers came from the life sciences. There are many rea- sons for this. Consequently, in 1979, in order to make our coverage more com- prehensive and representative of all the fields of science, we began to publish six different classics each week—one for each edition of CC,Z except Arts & Humanities. And in early 1981, to eliminate our backlog we began publishing two classics in CC/Life Sciences per week. While citation frequency is a prime indication of a paper’s impact, we have not rigidly adhered to any particular level of citation. A paper in one of the basic engineering sciences may be a classic even if it has been cited only 30 or 40 times. This would still be orders of magnitude greater than the number of citations received by millions of average junel,1981 papers even in the life sciences. And in the social sciences, as in engineering, books often are as important as journal articles so we must apply separate crite- ria for them. Apart from expanding the definition of a Citation Classic so that it comes closer to what scientists perceive as a classic, our basic goal in publishing Citation Classics has remained the same: to present the human side of sci- ence whale paying tribute to d]verse ad- vances in science and scholarship. rhese commentaries were designed to allow scientists to talk about their major works from a personal standpoint, re- vealiig what prompted the research, the :ontnbutions of coauthors, and obsta- cles that were encountered in both research and publication-in short, those details that are rarely revealed in !ormal scientific publication. The more than 750 classics published so far have :xceeded my expectations. I think it is unfortunate that scientific Ioumals do not have the flexibility or ;he sense to provide some of this back- ground. But in the context of a scientflc ioumal, it may not be particularly rele- ~ant to learn, for example, that Martin ieligman’s paper on the laws of learning “esulted from an illness he contracted kfter eating b6amaise sauce.j In thk >articular instance, although the illness was not related to the eating experi- mce, it so conditioned Seligman that he 123

Upload: hakhanh

Post on 25-Sep-2018

219 views

Category:

Documents


0 download

TRANSCRIPT

curremcommentsCitation Ckmrics—Four Yearn of the

Human Sfde of Scfence

Number 22

Most readers of Current Contents@

(C@ ) are by now familiar with the Cita-

tion Classics feature which began in1977. It seems to me that after fouryears of thm weekly feature an evalua-tion is now in order.

In reviewing Citation Classics, one isimmediately struck by the fact that ourgoals have changed somewhat in thepast four years. Originally, we plannedto draw our classics from a “group of500 papers most-cited during the years1961-1975.”1 By 1979, however, werealized the consequences of this nar-row approach to selecting papers. Most

of the 50U most-cited papers came fromthe life sciences. There are many rea-

sons for this. Consequently, in 1979, inorder to make our coverage more com-prehensive and representative of all thefields of science, we began to publish sixdifferent classics each week—one foreach edition of CC,Z except Arts &

Humanities. And in early 1981, toeliminate our backlog we beganpublishing two classics in CC/Life

Sciences per week.While citation frequency is a prime

indication of a paper’s impact, we havenot rigidly adhered to any particularlevel of citation. A paper in one of thebasic engineering sciences may be aclassic even if it has been cited only 30or 40 times. This would still be orders ofmagnitude greater than the number ofcitations received by millions of average

junel,1981

papers even in the life sciences. And inthe social sciences, as in engineering,books often are as important as journal

articles so we must apply separate crite-ria for them.

Apart from expanding the definition

of a Citation Classic so that it comescloser to what scientists perceive as aclassic, our basic goal in publishingCitation Classics has remained thesame: to present the human side of sci-ence whale paying tribute to d]verse ad-vances in science and scholarship.rhese commentaries were designed toallow scientists to talk about their majorworks from a personal standpoint, re-vealiig what prompted the research, the

:ontnbutions of coauthors, and obsta-cles that were encountered in bothresearch and publication-in short,those details that are rarely revealed in!ormal scientific publication. The morethan 750 classics published so far have:xceeded my expectations.

I think it is unfortunate that scientific

Ioumals do not have the flexibility or;he sense to provide some of this back-ground. But in the context of a scientflcioumal, it may not be particularly rele-~ant to learn, for example, that Martinieligman’s paper on the laws of learning“esulted from an illness he contractedkfter eating b6amaise sauce.j In thk>articular instance, although the illnesswas not related to the eating experi-

mce, it so conditioned Seligman that he

123

still (15 years later) cannot eat bc$arnaise

sauce, The paper grew out of his con-jectures on why this is so.

Once we have determined throughvarious citation analyses that a paper isa milestone paper in its field, how do weget the essays, or vignettes, written?First, we contact the author and ask hnor her to prepare a commentary. Theauthor also receives an author’s guidewhich explains the feature and the kind

of information we’d like included in thecommentary. The author also receivessamples of published Citation Classics.We encourage the author to stress the

personal factors involved in getting theoriginal paper published and we ask forinformation about coauthors, and howand where the work took place. One ofthe most important questions is why he

or she thinks the publication has beencited so often. In order to make theessay topical for readers interested infollowing up on the subject, we ask theauthor to cite a more recent review arti-cle or publication. We also ask eachauthor to mention any awards or honorsthat resulted from the research. This isfurther confirmation that it is indeed a

classic. I take particular pleasure in

noting that the authors of several clas-sics went on to be recipients of the Na-tional Academy of Sciences award forscientific reviewing! 4.5

Since we began Citation Classics in

1977, we’ve extended invitations to over2,000 authors. It is somewhat disap-pointing tome that only about one thirdhave accepted. Of course if we had abetter response we’d have written fewer

letters, but eventually we would have

covered all of them. Another 100 au-thors have promised to send manu-scripts, but have yet to do so. We

telephone most to reinvite them. Nearlyhaff of those asked, however, havesimply never acknowledged our letters.We can’t even be sure they were de-

hvered. “1’hw IS to be expected since

many have moved.I had been concerned about the

possibility that an inordinate percentageof refusals and nonresponses involvedforeign authors. What if some of theseauthors are unable or unwilling toprepare a commentary in English? Butin fact about 30 percent of nonrespon-dents are from outside the US, closelycomparable to what one would expectfrom the number of classics pubfished.It also turns out that the refusal rate of

about 50 percent for authors in English-speakhg countries is almost identicalfor the other countries. This woulddispel the notion of an English-languagebias,

The average age of the classicscovered today is 19 years, and the vast

majority of classics (454) were publishedin the 1960s, One hundred sixty-one

were published in the 1950s, 73 in the1970s, 35 in the 1940s, and five in the1930s. Of course, some authors haved~ed since writing their classic articles,but we do accept surrogates if a coau-thor is not available. About 10Q authorsresponded but refused our invitation.

I am delighted to report that 14 Nobel

prizewinners, not to mention dozens ofother prizewinners and academy mem-bers, have taken the time to write abouttheir Citation Classics. These positiveresponses make the varied reasons forrefusal hard to comprehend. Some au-thors claimed a lack of time, others alack of interest in their original papers,and still others expressed a feeling thatit just wasn’t worth the effort, especially

if they have left the field. I think it is un-

fortunate because these authors deny

their colleagues and society a perspec-tive on their work that they alone can

provide. In the future, we intend to ask

one of their colleagues or students ifthey will prepare essays. These com-mentaries may become a part of our

124

forthcoming Encyclopedic Atlas of

Science. And we hope to publish collec-tions of these essays for use by graduatestudents and others interested in theway science actually works.

From the time the first Citation

Classics was published on January 3,1977, until the last classic of 1980 onDecember 29, 1980, we covered 728classics in CC. By the end of 1981 thisfigure will exceed 1,000! Figure 1 pro-vides a breakdown for the 24 countriesrepresented at the time the authors

published their classic papers. Mostauthors came from English-speakingcountries. The US accounted for S02papers, over two thwds the UnitedKingdom, 93; Canada, 29; and Austra-lia, !9. All but three of the originalpublications were in English. Of the re-maining three, two were publiihed inFrench, and one in German. Oneauthor claimed that when he fiist

FfgIIra Ii Countries wldch produced original Ci&r-tion (2amics articles. with the number of papersfrom each,

Numberof

country Papem

usUnitedKingdom

England 04Scotland 8Wales I

CanadsAustraliaSwedenFederal Republic of Germany (FRG)FrsnceDenmarkIsraelJapanSouth AfricaSwitzerlandArgentiuaBelgiumFinlandGerman Democratic Republic (GDR)IndmMexicoThe NetherlandsNew ZealandSpainUSSRNot available

S&?93

29191165332221111111111

41

published his results in French journalsthe work received little attention.bWhile this was not necessarily true for

papers pubfished in the 1950s, it is clearthat publication in any foreign languagetoday can delay recognition of signifkcant work.7 Not alf scholars accept thisexplanation alone, however. Derek J.de Solla Price, Yale University, for ex-ample, suggests that the preponderanceof papers from English-speaking coun-tries may be due in part to what he feelsis an English-speaking bias in our cita-tion index.a I would argue that our biasis towards the high impact journals,

regardless of their language. And it isquite possible that a few Russian au-thors are overlooked because citationsto their papers may be fragmented invernacular and translation journals.

Nearly 2S0 ddferent institutions arerepresented in Figure 2 which lists ad-dresses for the original publications.However, since many authors havemoved, they now work at over 315 dif-

ferent institutions which are listed inFigure 3. If overlaps are eliminated,there are over 425 institutions. To savespace we have not repeated the namesof institutions named in Figure 2. Thefact that so many new institutions showup in Figure 3 illustrates the growth ofresearch worldwide and the migrationof classic authors to other institutions.While most worked at academic institu-tions, industry and government are well-

represented. These figures are based onthe use of the addresses for fwst authorsonly. I doubt that including secondauthors would change much in thm case.

As could be expected from the ex-amples set by our other citation studies,a few institutions dominate the lists. Thecombined campuses of the University of

California, for example, top both lists.Forty classics were written at California,but 55 authors are now currently atwork there. The migration to the west-

125

Ffmwe 2; The institutional affihations of Cifa/ion C/m$icJ au[hors, at {he time they wrote their clas$icpa~er% with the number of authors from each

University of CaliforniaBerkeley 17Davis 3Irvine 1Livermore 1Los Angeles 7Riverside 2San Diego 3San Francisco 5Santa Barbara 1

Natl. Inst. HeakhNat]. Cancer Inst. 10Natl. Heart, Lung and Blood Inst. 6Natl, Inst. Afkergy and Infectious 3

DiseaaeNatL Inst. Arthritis, Metabolism 2

and Digestive DisordersNatL Inst. Chifd HeaIth and 1

Human DevelopmentNatl. Inst. Dental Res. 3

University of WiaconainHarvard UniversityBell Labs., Murray HII1and Holmdcl, NJUnivemity of London

Birbeck CoI], 1Imperial CoIl. Science and 3

TechnologyInst. Cancer Res. 2Inst. Chifd Health 1Inst. Psychiatry 1Liater Inst. Preventive Med. 1Miidlesex Hosp. Med. Sch. 4Royal HoUoway COIL 1Royal Postgraduate Med. Sch. 3University Cofk. 1

Johns Hopkins Univ. and Hosp.Stanford UniversityCaliomia Inst. TechnologyColumbia UniversityUniversity of CambridgeUniversity of ChicagoUniversity of IUinOisUniversity of PennsylvaniaUniversity of Washington, Seattle, WACornell UniversityUniverai~y of TexasGeneral EIecu-ic Co., Schenectady, NYMaaaachuaetts Inst. TechnologyUniversity of Miine90taYale UniversityCamegi&MeUon UniversityOsford UniversityUS Department of Agriculture

Agriculture Res. Ctr., Beltsvilke, MD 3Cereal Science and Foods Lab., 1

Peoria, nAgricultural Marketing Serv., 1

Beltsvifle , MDRegional Poultry Res. Lab., 1

Eaat Lansing, MfUniversity of MichtganUniversity of Rncheater

40

25

24221818

14131111111111101099777‘16b6

66

~rookhaven NatL Labs.. Umon. NYDuke UniversityMcGill University, Montreal, CanadaYatl. Bureau of Standards,

Washington, DC~atl, Res. Councif of CanadaState University of New YorkUniversity of ColoradoWashington University, St, Louis, MOcommonwealth Scientific and Industrial

Res. Organization (CSIRO), AustraliaNew York UniversityNorthwestern UniversityPennsylvania State UniversityLJniveraity of EdinburghUniversity of Lund, SwedenUniversity of MelbourneUniversity of OregonLJniveraity of TorontoEJM Corp., Yorktown Heights, NY[tsdkma UniversityIowa State UniversityMayo Clinic and Foundation,

Rnchester, MNMicidgan State UniversityNASA, Green belt, MDPurdue UnivemityRCA Labs., New York, NYRockefeffer UniversityUS Naval Res. Lab., Waahmgton, DCUniversity of AdelaideUniversity of PittsburghUniversity of Reading, UKWeizmann Inst, Science, Rebovot, IsraelAlbert Einstein COIL Med., New York, NYAvco-Everett Res. Lab., Everett, MABaylor University, Waco, TXBrown University, Providence, RIBucknefl University, Lewi.sburg, PACanadian Department of Agriculture

Case Western Reserve University

Colonial Sugar Refining Co.,

Indonroopilly, Australia

Dow Chemical Co., Midland, Ml

E. 1. Du Pent de Nemours and Co.,Wifsnington, DE

Haakins Labs., New Haven, CTHumble Od Co., Houston, TXInst. Paateur, Paria, FranceMaaaachusetts General Hosp., Boston, MAMedical Res. Councif, UKMount Sinai Sch. Med., New York, NYNatf. Inst. Med. Res., Mdl HiU, UKNat]. Physical bb., Middleaex, UKOhio State UniversityOregon State UniversityPeter Bent Brigham Hmp,, Bnaton, MAPrinceton UniversityPublic Health Res. Inst. of the City of

New YorkRothamsted Experimental Station,

Haqrenden, UK

5555

s5554

444444443333

3333333333

;222222

22

2222222222222

2

126

State University of IowaUK Atomic Res. Estab., HarwefJ,UKUS Air ForceUS ArmyUS Public Health Serv.US Steel Corp., Pittsburgh, PAUniversity of BminghamUniversity of BristolUniversity of GeorgiaUniversity of KansasUniversity of MiamiUnivemity of MontrealUniversity of StockholmVeterans Admin. Hosp., Bronx, NYVkginia Polytechnic Inst. and

State UniversityWalter and Eliza Half Inst. Med. Res.,

Melbourne, AustmfiaWelfcome Res. Lab., Kent, UKWnnds Hole Ocemographic Inst.,

Wnods Hole, MAAarhus University, Risakov, DenmarkAbbott Labs., Chicago, JLAcademy of Natural Sciences,

Phiidelphia, PAAcademy of Sciences of the GDRAKied Chemical Corp., Morristown, NJAmerican Cyanamid Co,, Stanford, CTArgonne Nat]. Lab,, Argonne, JLAustralian Department of Science and

Industrial Res.Australian Nat]. ObservatoryAustralian Nat]. UniversityBonneville Power Admin., Portland, ORBowling Green State University, OHBritish Museum, London, UKBritish Postgraduate Med. Sch. ,

London, UKCanadkm Department of FisheriesCardiff Royal Infiiary, UKChester Beatty Res, Inst., Lmndon, UKChddrm’s Asthma Res. Inst. and Hnap.,

Denver, COChifdren’s Hosp., Boston, MAClinkal Res. Ctr., Harrow, UKCofJege of Veterinary Med., FinfandCmokes Labs., Ltd., fmrdon, UKDefense Res. Board, CanadaDefense Res. Estab., CanadaDenniaen University, Granviffe, OHDistiller’s Co., Epsom, UKEastman Kodak Co,, Rnchester, NYEssex Univemity, Colchester, UKFairchfid Camera and Instrument Corp.,

Palo Alto, CAFels Res. Inst., YeUow Springs, OH

Free Hosp. for Womm, Brcmkfine, MAGeephyaics Corp. of America, Bedford, MAGlynn Res. Labs., Badmin, UKGrasslands Res. Inst., Hurley, UKHeyden Chemicaf COW., GarfMd, NJHoffmann-La Rnchc, Nutley, NJHosp. Jnfmtif de Mexico,

Mexico C]ty, MexicoHoughton Poultry Res. Station,

Houghton, UK

222222222222222

2

22

111

11111

111111

1111

11111111111

11111111

1

Ifhnois Inst. Technology, Chicago, JLIllinois Wesleyan University,

Bloomington, JLImperial Cancer Res, Fund, London, UKImperial Chemical Indust.,

Welwyn Garden City, UKIndian State Institution, New Defhi, In&iInst, Advanced Study, Princeton, NJInst. Cancer Res., Philadelphia, PAInst. Fisheries Investigation, SpainInst, Psychiatric Res., Indianapolis, JNInst. Res. Cancer, Vfflejuif, FranceInst. Investigaciones Bioquim{caa,

Buenos Aires, ArgentinaJacksnn Lab., Bar Harbor, MEJewish Heap,, St. Louis, MOJohannes Gutenberg Univerairy, MainzKanaas State University, Manhattan, KSKarolinska Inst., Stockholm, SwedenKing Gustav V. Res. Inst.,, Stockhohn, SwedenEcole Normale SupfMeure, Paris, FranceLund Inst. Technology, SwedenM.D. Anderson Hosp. and Tumor

Institution, Houston, TXMartin Marietta Corp., Baftimore, MDMassachusetts Mental Health Ctr.,

Boston, MAMaudsley Hosp., London, UKMay Inst. Med. Res., Cincinnati, OHMead Johnson & Co., Evansville, fNMedical COU. Georgia, Augusta, GAMethndist Hosp., Indianapcdia, JNMinneanta Mining & Manufacturing Co.,

St. Paul, MNMixing Equipment Co., Rochester, NYMobile Oil Co., Princeton, NJNatl. Inst. Mental Health, Rnckvifle, MDNatl, Vegetable Res. Station,

WeUesboume, UKNatl. Women’s Hosp., Auckfmd,

New ZealandNew York Bloecf Center, NYNobel Medical Inst., Stnckhofm, SwedenNorth Caroliia State UniversityOak Ridge Natl. Lab., TNOhio Agr’iculturaf Experimental Station,

Columbus, OHOntario Cancer Inst.Ontario Res. FoundationPneumoconiosis Res. Unit,

Johannesburg, South AfricaPolytechnic Inst. New York,

Farmingdale, NYPortsmouth & Isle of Wight Area Pathology

Serv., Portsmouth, UKQuartermaster Res. and Engineering Ctr,,

Natick, MAResearch Board of CanadaRijksuniveraiteit JAden, The NetherlandsRnckweJl Park Memorial Inst., Buffalo, NYRowett Jnst., Bucksbum, UKRoyal COIL Science, London, UKRoyal CoIl. Surgeons of England,

London, UK

11

1t

1111111

111111

111

1I

111111

11I1

1

11111

111

1

1

1

111111

127

RoyaI Danish Sch. Pharmacy,

Copenhagen, Denmark

Royal Radar Estab., Worcester, UKRoyal Signals & Rndar Estab., Mrdvem, UKRutgers University, New Brunswick, NJSt. Mark’s Hosp., London, UKShell Development Co., EmeryviUe, CASinclair Res. Labs,, Harvey, fLSloan-Kettering Inst. Cancer Res.,

Rye, NYSorbcmne, Paris, FranceSquibb Inst. Med. Res., NJStandard Oil Co., Whiting, NJSwiss Federal Inst. Technology,

Zurich, SwitzerlandTechnische Hochschule Stuttgart, FRGTohoku Pharmaceutical Sch,, Sendai, JapanTony Res. Station, Aberdeen, UKTufts UniversityTulane University of LouisianaUS Department of CommerceUS Department of Health, Education, and

WeffareUS Department of InteriorUS Geological SurveyUniversity College, SwanSea, UKUniversity College of Wales, UKUniversity of AberdeenUniversity of AlaskaUniversity of ArkansasUniversity of Bern, SwitzerlandUniversity of Bradford, UKUniversity of British ColumblaUniversity of CincinnatiUniversity of ConnecticutUniversity of CopenhagenUniversity of DurhamUniversity of DitsseldorfUniversity of FloridaUniversity of Freiburg

1

111I111

1111

1111111

11111I1111111t111

University of GenevaUniversity of GlasgowUniversity of HawaiiUniversity of LeedsUniversity of LiverpoolUniversity of Lcmvain, BelgiumUniversity of ManchesterUniversity of ManitobaUniversity of MarburgUniversity of MarylandUniversity of MunichUniversity of Natal, South AfricaUniversity of NebraskaUniversity of Newcastle-upon-TyneUniversity of North CarolinaUniversity of OttawaUniversity of SydneyUniversity of TennesseeUniversity of TokyoUniversity of UppwdaUniversity of UtahUq!versity of VermontUmversity of Waterloo, CanadaUniversity of Western OntarioVanderbilt UniversityVarian Associates, Palo Alto, CAVeterans Admin. Hosp., Durham, NCVeterans Admin. Hosp., Lhtle Rock, ARVetemns Admin. Hosp.,

White River Junction, VTWake Forest University,

Winston-Salem, NCWestinghouse Electric Corp.,

Pittsburgh, PAWestminster Hosp., London, UKWheeler Labs., Great Neck, NYWistar Inst., Phlladelphki, PAWorld Data Ctr., Moscow, USSR

111111I1111I11111111I1I1I1111

I

1

11I1

Ffgrrre 3# Citation C/rrmics authors’ current institutional affiliations which were not listed in Figure 2. Thenumber of authors from each is also shown.

Brigham Young University, Provo, UT 3 AU India Inst, Med. Sci., New DeUd, India 1Tel Aviv University, Israel 3 American University, Washington, DC 1Temple Univemity, Phfladelphla, PA ARCO, Harvey, IL 1

1University of Arizona : Arizona State University 1University of Delaware 3 Bispeberg Hosp., Copenhagen, Denmark 1Brandeis University, Waltham, MA 2 Bolt Beranek & Newman, Inc., Cambridge, MA 1CERN (European Organization for Nuclear Res.) 2 Boston University I

Geneva, Switzerland Bruce Lyon Memorial Res. Lab., 1City University of New York 2 Oakland, CAColorado State University 2 Bnrnel University, Uxbridge, UK 1Exxon Corp., Houston, TX 2 Brussels University, Belgium IHahnemann Med. COU.and Hosp., 2 CNRS, Strasbourg, France 1

Phiiadelphla, PA Caliiomia Department of Health 1University of Alberta, Edmonton, Canada 2 Canadian Wikffife Sew. 1University of Guelph, Ontario, Canada 2 Cancer Control Agency of British Columbkr, 1University of Oklahoma 2 Vancouver, CanadaUniversity of Paris 2 Cardiovascular and Chest Surgical Associates, 1University of Southampton 2 Boise, IDYork University, Downsview, Ontario, Canada 2 Carlsberg Foundation, Copmrhagen, Denmark 1Adderrbronke’s Hosp., Cambridge, England 1 Ctr. Studies of the Pemon, La Jolla, CA 1Aeronautical Res. Lab., 1 Chalmers Inst. Technology, G6teborg, Sweden 1

Melbourne, Australia

128

Chiidren’s Hosp. Med. Ctr,, Cincinnati, OHCity of Hope Natl. Med. Ctr., Dum-te, CAClaremont Graduate Sch., CACfirricalRes. Inst., Montreal, CanadaCoUege of Wilfhn & MaryCommunication Res. Cm,, Ottawa, CanadaDecision Res., Eugene, ORDenver General Hosp.DickinsonCoJfege,CarJiale,PAkcole National Sup&ic.re des Mines,

Paris, FranceEcoIogy and Environment, Inc., Decatur, GAEducational TestingServ., Princeton, NJFlorida Medicaf Entomology Lab.,

Vers Beach, FLFreernan Hosp., Newcastle-upnn-Tyne, UKFunk Seeds Intematl., Bloomington, JLGeorgetown University, Washington, DCHampshire District Pathology Serv.,

Portsmouth, UKHarris Corp., Melbourne, FLHazeltine Corp., Greenlawu, NYHt3p. de l’ErrfantJesus, Quebec, CanadaHughes Aircraft Co,, Fuffertorr, CAHungarian Academy of Sciences, BudapestIbaraki University, Mete, JapanIrrd}anInst, Science, Bangalore, IndiaIndiana Sch. Med.Inst. Animal Physiology, Cambridge, UKInst. Bkxbemistry and Technology,

Munster, FRG

1111111111

111

1111

1111111111

PortJand State UniversityQueenEJizabcthMed.Ctr., Birmingham,UKQueen’s University, Kingston, CanadaQueen’s University of Beffast, UKResearch & L.a.wrTechnology, Inc.,

Rockpot-t. MARheinische-Westfnfi$che Techniache

Hochschule, Aachen, FRGRice University, Houston, TXRoyal Jnfmary, Glasgow, UKRoyal Liverpnol Hosp., Liver-pool, UKRutgers Med. Sch., Piacataway, NJSan Diego State UniversitySoroka Med. Ctr., Beer-Sheba, IsraelSouthern IUiioia UnivemiryStrangeways Res. Lab., Cambridge, UKSwedish Department of Occupational

HealthSyracuxe UniversityTokyo fnst. Technology, JapanTropical Prcducts Institution, London, UKUS Fnod and Drug Admin.USSR Academy of Science, MoscowUniformed Servkces University of the Health

Services, Bethesda, MDUnilever Res. Lab., Sharnbrnok, UKUniversity of AlabamaUniversity of AmsterdamUniversity of AucklandUniversity of Barcelona

Inst. National de Cierrcias v Technolorna. Mexico 1I University of Brsel

Intel Corp., Santa Clara, C-A -Jntemational Lab. Res. Animal Dkeases,

Nairobi, Ken yaIsrael Irrst. Technology, HaifaJ.st.s-L1eblg University, Geis.sen, FRGLaTrobc University, Bwrdnora, AustraliaLehigh University, Bethlehem, PALetterman Army Inst. Res., San Francisco, CALewisham Hospital, London, UKLlandough Hosp., Penarth, UKLoma Linda University Med. Ctr., CALong Island University, Greenvale, NYL4misiana State UniversityLoyola University, Maywoud, JLMcMaster University, HamiJton, CanadaMeat Res, Inst., Langford, UKMichigan Technological University,

Houghton, MIMidwest Med. Lab., St, Louis, MOMifl Hill Labs., London, UKMississippi State UniversityMonash University, Clayton, AustrafiaMondfiore Hosp. and Med. Ctr., New York, NYMt. Holyoke CoUege, South Hadley, MANatL Jewish Hosp. and Res. Ctr.,

Denver, CONew York Med. COJJ.,NYOak Ridge Assxiated Universities, TNOptical Sciences Co., Placentia, CAPace University, New York, NYPeter MacCuUum Hosp., Melbourne, AustraJiaPlttman-Moore, Glenorie, AustraliaPolish Academy of Sciences, Wamaw, Poland

11

111111I1111111

1111111

111111t

University of ExeterUniversity of Giiteborg, SwedenUniversity of Haifa, IsraelUniversity of IowaUniversity of KentuckyUniversity of Laval, CanadaUniversity of Pierre and

Marie Curie, ParisUniversity of Rhnde IslandUniversity of Saarland, FRGUniversity of Southern AlabamaUniversity of Southern CaJifomiaUniversity of SussexUniversity of ViennaUniversity of VirginiaUniversity of Wiirzburg, FRGUpjohn Co., Kalamazoo, MIVarian/Extrio, Gloucester, MAVeterans Admin. Ctr,, Laiwifle, KY

Veterans Admin. Hosp., Miami, FLVeterans Admin. Hosp., San Diego, CAWashirrgton State University, Pufkrrran,WAWestern Michigan UniversityWest Park Hosp., Epsom, UKWittenberg University, Springfield, OHYMCA Tribal Development Project,

Tamif Nadu, IndiaRetired

11111

1

111111111

1I1I11

1111I11I11111

111111111111111111

34

“One or more of thexe authors represents a secondiffdiation.

129

em US is significant. Also high on bothlists are National lnsthutes of Health, 25classics; University of Wisconsin, 24;Harvard, 22; Bell Labs., 18; University

of London, 18; Johns Hopkins, 14; andStanford, 13. Also prominent are Cor-nell, Columbia, University of Cam-bridge, University of Chicago, Universi-ty of Pennsylvania, University ofWashington, University of Texas, andState University of New York.

If we eliminate those institutions

which account for only one paper, only

98 account for 523 classics. Similarly,

on] y 105 institutions are involved in cur-rent affiliations. Seventy-three of theseinstitutions appear on both lists. InFigure 4 there is a table showing the

Ftgure 4: The number of institutions which ac-counted for one or more classic papers

Number Number01 Pmpem Oi Iustitutlons

40 1

25 124 122 118 214 113 111 510 29 27 46 55 84 93 142 42I I59

number of institutions which accountedfor one or more classic papers.

We also studied the number ofauthors per paper. It is significant thatthere are so many single-authoredpapers—302. Nevertheless, multi-au-thored papers dominate. Of these,however, 245 had only two authors.

Papers in clinical medicine had anaverage of three authors, more authorsthan the other disciplines, We alsoknow that high-energy physics papers

have many authors. Figure 5 shows thenumber of papers for each group. Theaverage number of authors per paper inthe classics is 2.06, This is signflcantlylower (20 percent) than the 2.S6 averagefor papers covered by the Science Cita-

tion Zna’exQ (SCF ) in 1980.Which types of papers become Cita-

tion C[assics? We know that superstarmethodology papers, especially in bio-chemistry, are to be expected. Never-theless, less than one fourth of thepapers involved are methodological.Almost one third were theoretical or ex-perimental. The remaining large catego-ry was review papers—about ItX). And

another large group was papers orbooks which provided oft-cited tablesand data of one kind or another or“tests.” These are only crude measures,but it should dispel the notion thatmethodology papers dominate citationstudies. But once a well-known pro-cedure is adopted it may be citedthousands of times. The interestingquestion is why a smalf number of suchclassics fail to succumb to the oblitera-tion phenomenon.

Why are articles cited? The answer

may seem, at first glance, quite simple,yet no one really knows—least of alf thecited authors themselves! A surprisingnumber of authors, upon receiving our

request to comment upon their Citation

Classic, have themselves questionedwhy their papers were cited. An equallylarge number of authors assert in theircommentaries that they don’t reallyknow why they’ve been cited. It’s amaz-ing how many were not really awarehow often or by whom they were cited.Only a small number are interestedenough to go back to the citing papers

through the SCI to find out exactly why

they’ve been cited. In correspondencewith some authors, I’ve pointed out thatonly a content analysis of the citing

papers can reveal the “why” of citation.

130

Ffgure 5: The number of authors per paper.

Nusnbar Numberof Authors of Papers

1 3022 2453 924 475 256 107 38 19 110 1

11 1

I also send them copies of a few ISF’papers in which we have done contentanalysesg, 10 and recommend they getstudents to help with the library workinvolved.

Of course, many authors maintain ex-tensive reprint collections related totheir discoveries or have an intense in-terest in the way their work has been ap-plied. Hans Krebs, for example, in hisCitation Classic, 11 quotes from anhistorical evaluation of his classic paperon urea fortnation. But most authorsmerely make educated guesses as towhy they’ve been cited so often. I urgethem to do these content analysesbecause they can lead to importantreview papera.

Authors frequently assert that

timeliness was the main reason theirpapers had so much impact. Many feel

that if their work had appeared earlieror later, it might not have receivedmuch notice. Others maintain that theirpapers were cited simply because theydeveloped an often-used formula orprocedure. Others believe that theirpapers were highly cited simply becausethey included a comprehensive reviewof the literature. But in my studies Ihave found that the influence of the

cn”tical review is significant-it doesn’tsimply serve as a convenient way to cite

the earlier literature. Equally impor-tant, some of these authors correctlybelieve that the new or surprising or

even the startling results published intheir papers stimulated more work onthe same topic. Sometimes the citingauthor makes thk self-evident, but moreoften than not this is only implied in thework.

Most authors correctly assume thattheir papers were cited for positivereasons. Only a few authors feel thattheir work was also cited for negativereasons, although in at least one case,

the paper was often cited for negativereasons.lZ,lJ

Understanding the “why” of citationis an area of research that interestssociologists of science. In fact, SusanCozzens of 1S1 has recently summarizedvarious theories about citations, 14 andin particular Mike Moravcsik, OregonUniversity, has examined the types ofcitations in physics in detail. 15 Theirstudies may lead to a better understand-ing of the processes of scientflcdiscovery. In The Force of Kno wledge,

John Zitnan, our salty colleague fromthe University of Bristol and editor ofthe Philosophical Magazine (a physicsjournal), points out the importance ofcitations to scientific knowledge:

A typical scientific paper is full ofreferences or citations to the exper-iments, calculations, observations, ortheories of other people. It does notstrike out on its own into theunknown, but timidly takes one littlestep forward from the base secured byprevious research. In other words,modem research is highly collabora-tive, despite all the competition.Everything we do is deeply indebtedto, and embedded in, the achieve-ments of our predecessors and con-temporaries in our InvisibleCoUege.16(p. 1OO-I)

In a 1971 article on citation indexes

prepared by 1S1 for the Encyclopedia ofLibmry and Information Science, we

listed 15 reasons that authors cite otherworks. Among them were to pay hom-

~ge, to identify methods or equipment,

131

to provide background reading, and togive credit for related work. IT

What some authors fail to recognize is

that their “simple” discoveries make itpossible for others to go on to do studiesthat were previously impossible. Inevahsatirtg the relative impact of highlycited papers the authors themselves cantake the simplistic approach and assumethat a paper was cited simply because acertain method or theory is used. Butjust as often the methodology is inter-twined with other complex ideas thatmay never have been developed had not

the methodology been employed. Thiswas certainly the case recently when we

examined why the work of R.M. Camp-bell et al. was cited.lg

Perhaps the most fascinating aspectof Citation Classics is the perception ofthe author in evaluating the importanceof the work in question. There is often asense of irritation that we were not wiseenough to reafize that his or her most-cited work was not his or her most im-

portant. While we never make that as-sumption, it is somewhat distressing thatwe are resented because there is not aone to one correlation between citationfrequency and the author’s evaluationof importance. Clearly many authorsfeel, and correctly so, that the classicarticle we have selected is their most imp-ortant work. But the opposite reaction

of Heinz Fraenkel-Conrat and his coau-

thors is typical of many other classicauthors. “All three authors have at leastten other papers to their credit whichthey would list above thk one in impor-tance. And what this paper is quoted foris not its intrinsic point (which had some

importance) but for the fact that it con-tains a paragraph describing the methodof washing and suspending commercialbentonite clay. “19 What Fraenkel-

Conrat and others might have added is

that some of their other articles were

not only important papers but also

qualify as Citation Classics.

Harriet Zuckerman, Columbia Uni-versity, in her book Scientific Elitezo

discusses the phenomenon of the scien-tist who doesn’t feel hk highly ac-claimed work is his most important. Asignificant number of Nobel prizewi-

nners have these feelings. She offerssome possible explanations. One, shesays, is the fact that often the highly ac-claimed work was the result of chanceor serendipity y. To some scientists, thisseems less important, or valid, thansomething that was well thought-out orplanned.zo

Drawing on the work of Robert Mer-

ton, Columbia University, on the behav-ior of scientists, which demonstrated acommon scientific drive for “recognizedonginality,”zl Zuckerman further

points out that since often an acclaimedwork represents an accidental discov-

ery, for many scientists this is not con-sidered particularly original. Thus suchwork is not as meaningful as a work thatwould “deepen scientists’ understanding

of large problems. ”zo (p. 211) ManyNobelists seem to feel that if it hadn’thappened to them, it would have hap-pened to someone else. We have herethe idea that many discoveries are “in-editable. ” One wonders how much thiskind of thinking pervades the halls ofCongress these days.

Finally, says Zuckerman, there may

often be a disparity between the personalsignificance of a work to its author, andthe scientific significance of the samework, 22 that is, the acceptance of his orher peers. It’s one thing to correlate anauthor’s personal perception of impor-tance to that of his or her pews andanother to match this with actual impactas measured by citation frequency.

Still another explanation is offered byLewis Goldberg, University of Oregon.In writing about his classic on the

human judgment process, he notes thatpeople frequently employ a techniquehe calls “availability” in decision mak-

132

ing. “When people have to estimate thefrequency of an event,” he writes, “theytypically rely on the ease with which in-stances spring to mind. Use of such atactic is not unreasonable, but undersome circumstances it can lead us astray

(e.g. the frequency of more st~~ng ormemorable events gets overestimated).So it is with me: had someone asked meto estimate which of my publicationswas the most frequently cited, I’d cer-tainly have selected another!”~

Taking a slightly different tack, Pricehypothesizes that a failure of communi-cation may actually be a major part ofthe problem. Many authors, he feels,are simply unable to evaluate their ef-fectiveness at communication. An au-thor, says Price, may welf have dis-covered “the discovery to end all dis-coveries,” but if he or she doesn’t com-municate that to others, it has littlevalue. In other words, an author maybe

aware of the scienttilc quality of hii orher work, but have no idea of the quali-ty of communication. And, Price con-tinues, citation frequency must be aproduct of the inherent quality of awork and the effectiveness of its com-munications He concludes thatwhether we liie it or not, the citationrecord is the most accurate index of the

world scient~lc community’s opinion ofa scientist we have available.zd

The reader is, of course, the ultimate

judge of whether the space and energydevoted to Citation Classics is worth theeffort. I continue to read each and everyclassic before it is published. The quali-ty has improved enormously over theyears and in many instances we havebeen able to call attention to important

work that has otherwise not been ac-knowledged. That many of the papersdid elicit awards confirms our choice,

but I take special satisfaction in servingas a public relations catalyst for manyscientists who receive inadequate recog-nition for their work. Indeed, a large

number of individuals turn up here whodo not show up on our lists of most-cited authors. This illustrates that notonly the superstars make important con-tributions to progress. I also hope it em-phasizes the need for scientists and ad-ministrators to concentrate on publish-ing fewer papers of higher quality andhopefully higher impact.

Perhaps one of the most importantcontributions our Citation Classics

series has made is that we now have ex-tensive examples of significant dis-coveries that were entirely unplanned,

and in fact, often accidental. This seren-dipitous theme is quite prevalent. As L.RL%%z of the Karotinska Institute ob-served: “Clearly, thii is another exampleof how futile it is to try to foresee thepath of fundamental research, to saynothing of governing it.”zs Citation

Classics can serve as an excellent model

of the actual discovery process. It is forthis reason that we hope to publiih themin a collection for use by graduate stu-dents and others.

In closing, may I urge every CCreader to inform us of any paper orbook which may qualify as a Citation

Classic. Each suggestion will be careful-ly evaluated and, if we agree, the au-thors will receive an invitation at theearfiest opportunity. We are not refer-

ring here to unrecognized but importantresearch. That is an entirely separatearea of research which we are studyingintensively. We will discuss these“sleepers” in the future. And when wecomplete the compilation of the .!iCZfor1955-1964, we will be better able toidentify certain classics which veryquickly after publication became a part

of the common wisdom of science.

● ☛☛☛☛

My thanks to Susan Fell Evans for herhe[p in the prepamtion of this essay.

0>9, %,

133

REFERENCES

1. GarJfeME. Introducing Citation Chr$ict: thehuman side ofsciendfic reports.Cuwent Contents (1):5-6, 3 January 1977. ”

2, . . . . . . . . . . . . . . . Expansion of Citotion Classics--25O unique commentaries each year.

Current C0ntents(l):5-12, lJanuary 1979.

3. Se#Jgmm ME P. Citation Classic. Commentary on F’sycho/. Rev. 77:406-18, 1970.

Current Contents/Social & Beha vioml Sciences (8):14, 25 February 1980.4, GariJef$E. The 1979NASIames Munay Luck award forexceflence inscienitiic retieting:

G. AlmIRobkon receivesthe first awardforhiswork on cyclicAMP.C.rrem Confenfs(18):5-9,N April 1979.

5. --------------- The 1980 NAS James Murray Luck award for excellence in scientific reviewing:Conyers Herring receives second award for hE work in solid-state physics.Current Confenzs (25):5-7, 23 June 1980.

6. Be.mabami W. Citation Classic. Commentary on Cancer Res. 20:712-27, 1960.Current Confent$/Li@ Sciences (10): 14, 3 March 1979.

7. Gmffeld E. Do French scientists who publish outside of France and/or in Engfish do better research?Currerrf Corrfenf$ (22):5-10, 29 May 1978. ”

8. Prfce D J D. Telephone communication. 20 ApriJ 1981.9. Small H G, Ched documents as concept symbols. Sot. .$md. Sci. 8:327-40, 1978.

10, GarJJeld E. Citation measures of the influence of Robert K. Merton.Tmns, NY Acad. Sci. 39:61-74, 1980.

11. Krebs H. Citation Cfmsic. Commentary on Hoppe-Seylers Z Physiol. Chern. 210:33-66, 1932.

Current Contents/Life Scjence. (52):12, 29 December 1980.12. Garffeld E. High impact science and the case of Arthur Jensen.

Current Contenh (41):5-15, 9 October 1978. ”13. Jensen A R, Chation Clasaic. Commentary on Harvard &duc. Rev 39:1-123, 1%9.

Currenf Confenls (41): 16, 9 October 1978.14. Cozr.mISS E. Taking the meaaure of science. (Restivo S P, cd.) New directions in the

xociology of science. lnt. Sot. Sociol. Xno w[edge Newsletter, March, 1981.

15, MoravcsJk M J, Murugeatm P & Shearer E. An analysis of citation patterns in Indian physics.Sci. Cult. 42:295-301, 1976,

16. Zfmmr 1. The force of krro wledge, London: Cambridge University Press, 1976.374 p,

17. Welmtnck M. Citation indexes. (Kent A & Lancour H, eds. ) .Errcyclopedia of library and

information science. New York: Marcel Dekker, 1971. Vol. S. p. 16-40. ”18, Campbe!f R M, Cutbbermon D P, Mattbaws C M & McFsrhe A S.

Citation Cfassic. Commentary on [nt. J. Appl. Radial. Isotop. 1:66-84, 1956.

Currant Con/errts/L@ Sciences (41):14, 13 October 1980.19, Frr&akel-Conrat H. Citation Classic. Commentmy on Virology 14:54-8, 1%1.

Current Conterrfs/Life Sciences (23):14, 4 June 1979.

20. Ztrckemrsm H. Scientific eli(e. New York: Free Prexx, 1977.335 p.21. Merton R K. Priorities in scientflc discovery: a chapter in the sociology of science.

Amer. Sot. Rev. 22635-59, 1957,

22. Znckermm H. Telephone communication, 19 February 1981.23. GobJbcrg L R, Citation Claaaic. Commentary on A mer, Psycho]. 23:48>96, 1968,

Currerrt Con/ents/Socia/ & Behavioral Sciences ( 10): 18, 9 March 1981,24. price D 1 D. Telephone communication. 28 January 1981,25. Rr4v6sz L. Citation Classic, Commentary on Cancer Res. .?044351, 19(K3.

CurrerrrContenf$/Life Scierrce, (41):10, 8 October 1979.

“Reprinted in: GarfJekJ E. Es.mys of an information scientist. Phdadelphla: 1S1 Press, 1980.3 VOJ.I

134