superseded in 2015 by the Every Student Succeeds Act (ESSA), whichrelaxed directives on regular assessment.

in the USA in 2001 Congress passed a new law—the ‘No ChildLeft Behind’ Act (NCLB)—which sought to increase accountability fromschools through an array of measures, one of which instructed teachers on thekind of ‘scientifically proven’ practices they must adopt in their work.

A rhythmically propulsive dance music,marabi was forged principally by unschooled keyboard players who were a notorious part of theculture and economy of illegal slumyard liquor dens. Harmonically, it rested upon a cyclical patternstretched over four measures, with one measure per chord: I–IV–I6-4–V. The cyclical nature of this styleclearly derived from indigenous sources, repetitive harmonic patterns being typical of traditionalAfrican musics

Der Zettelkasten weiß tendenziell immer weniger als man selbst, aber immer soviel, wie wir auf Karten geschrieben haben.

Neben den Methoden herkömmlicher Recherche werden daher normalerweise Zettelkästen angelegt, in denen auf Karteikarten notiert ist, was ständig zitiert werden muss. Studenten erproben dies oft zum ersten mal intensiv für ihre Diplomarbeit. Nach Schlagworten und Autoren werden Notizen alphabetisch geordnet und was in den Kasten einsortiert wurde, kann man auch wieder herausholen.

Well, I don’t think there is any question about it. It can only be attributable to human error. This sort of thing has cropped up before  and it has always been due to human error.

Hyperauthorship:APostmodernPerversionorEvidenceofaStructuralShiftinScholarlyCommunicationPractices?

Historically, authorship implied writing and this associ-ation with the act of writing remains the core of the standardmodel of authorship acknowledgment. H

Whilecriticaltheoristsmayquestionthe“prestigeofauthorship”and“allmanifes-tationsofauthor-ity”(Birkerts,1994,pp.158–159)—whichhelpsexplainthepredilectionforpostmodernistandeschatologicaltitlessuchasTheDeathoftheAuthor(Bar-thes,1977),WhatisanAuthor?(Foucault,1977),andTheDeathofLiterature(Kernan,1990)—thereislittledoubtthatboththesymbolicandmaterialconsequencesofauthor

ome kind of ontolog-ical reassessment of authorship is called for to ensure thatauthority, credit, and accountability, currently apportionedin confused fashion across authors, acknowledgees and con-tributors, are henceforth distributed appropriately, parsimo-niously, and unambiguously. I

what is implied by a byline in thesecases is typically a very precise, often specialized input to acomplex, multidisciplinary project. The classical idea ofauthorship cannot credibly accommodate the legions ofcoworkers associated with large-scale collaboration, nor canit adequately reflect “the epistemic role of support person-nel” in the conduct of science (Shapin, 1995, p. 359).

day’sbiomedical journal article is the progeny of occasionallymassive collaborations, the individual members of whichmay have minimal involvement in the fashioning of theliterary end-product itself, with the act of writing beingdelegated to a subgroup or designated spokespersons. I

he “modern scriptor” (Bar-thes, 1977, p. 146) is no longer the sole conceiver, fabrica-tor, and owner of the published article.

hmed, Maurana, Engle,Uddin, & Glaus (1997)

In biomedicine, authorship has irrevocably shed some ofits craft associations:

However, a diverse body of work on thesocially situated nature of scientific communication alreadyexists which points the way. This ranges from Crane’s(1969) pioneering analyses of invisible colleges throughLatour and Woolgar’s (1979) classic study of laboratory lifeat the Salk Institute to Traweek’s (1992) richly texturedethnography of the HEP community. In addition, the workof Schatz and colleagues on the Worm Community Systemproject, which was designed to capture the full range ofknowledge, formal and informal, of the community of mo-lecular biologists who study the nematode worm C. elegans(see: http://www.canis.uiuc.edu/projects/wcs/index.html)can provide useful insights; so, too, research into the mate-rial practices and social interactions of scientists working incollaboratories, such as the Upper Atmospheric ResearchCollaboratory (see: http://intel.si.umich.edu/crew/Research/resrch08.htm) or the Space, Physics & Aeronomy ResearchCollaboratory (see: http://intel.si.umich.edu/sparc/) at theUniversity of Michigan

is wiseto avoid generalizations and to concentrate instead on show-ing how interactions between coworkers, specifically theorchestration of information exchange and coauthorship, aregrounded in local culture.

iomedical collaborations are moreheterogeneous and socially diffuse in character and do notappear to have the same degree of multilayered, internalreview as HEP research collaborations. T

TheHEP research community is thus characterized by highlevels of internal scrutiny, mutual trust—witness, for in-stance, the institutionalized practice of relying upon, andciting, preprints—and peer tracking, such that it is notsusceptible to systematic fraud. Contrary

The answer probably has to do with the relative intensityof socialization and oral communication (Traweek, 1992,pp. 120 –123), along with the character of the organizationalstructures and value systems, which define collaborations inlarge-scale, high-energy physics and biomedical research.

ights (1, 3, 5) are summed to(a) estimate each individual’s overall contribution, whichcan range from 1–35 (735 being the maximum possiblescore), and (b) determine the sequence in which coauthors’names are listed on the resultant publication.

owever, it isimportant to distinguish between generic job categories andthe specification of tasks performed; the contributorshipmodel is designed to record each individual’s actual input(e.g., experimental design, data collection, statistical analy-sis, final article revision), not job title (e.g., coprincipalinvestigator, technician, systems analyst), since the lattermay on occasion mask or inflate the former (Stern, 2000).

The standard model accepts that authorship is linkedinextricably to writing. But writing is no longer a necessarycondition of coauthorship in certain cases. Thus, an alter-native to authorship is required to accommodate the manyother contributions that shape the published byproducts ofcollaborative activity, be they research reports, journal ar-ticles, conference papers, or technical reports. C

To date, more than 500 journals have adopted theICMJE’s (1997) principles of authorship as laid out in the5th edition of theUniform Requirements for ManuscriptsSubmitted to Biomedical Journals(Klein, 1999; Stern,2000). According to these concrete guidelines, candidateauthors must satisfy three conditions. They must make:“. . . substantial contributions to (a) conception and design,or analysis and interpretation of data; and to (b) drafting thearticle or revising it critically for important intellectualcontent; and on (c) final approval of the version to bepublished.” Laudable though these guidelines are, it is un-likely that they will solve the problem. A study by Hoen etal. (1998) in the Netherlands found that authors and theircoauthors did not always agree with one another’s assess-ments that the ICMJE criteria had been met. In the UK,Bhopal et al.’s (1997) survey of medical researchers dis-covered that, although most respondents concurred with thethree criteria (more than 80% in each case), a majority(62%) did not feel that all three conditions should have to besatisfied to warrant author status. F

Rennie, Yank, andEmanuel (1997) that the distinction between the two modesof credit allocation is inherently artificial. Consequently,they have argued for explicit description of all individualcontributions as a means of eliminating ambiguity. Such aproposal would remove both authorship and acknowledg-ment from the frame, a really quite significant break withscholarly publishing tradition. This alternative amounts to aradical model of authorship attribution in contrast to thestandard model

he problem with arbitrary capping, whetherof authors or acknowledgees of one kind or another, is thatsome individuals’ potentially important contributions, bethey clinical investigators (Carbone, 1992) or telescopeoperators (Cronin, 1995), may be erased. This could, con-ceivably, have negative downstream implications in termsof remuneration and promotion prospects for those—the“invisible technicians” (Shapin, 1995, p. 355)—whose ef-forts have been withheld from the public ledger. It mightalso reduce, in line with theories of reciprocal altruism(Nowak & Sigmund, 2000), potential collaborators’ will-ingness to ‘donate’ their services.

Some journals place a limit on the numberof coauthors; for example, theDutch Journal of Medicinedoes not publish articles with more than six authors (Hoenet al., 1998).

Kassirer and Angell (1991, p. 1511)of theNew England Journal of Medicinewere bemoaningnot only “ambiguous authorship” but “lengthy acknowledg-ments,” inflated by the inclusion of “everybody who hadanything to do with the study, including those who weremerely carrying out their jobs, such as technicians.”

ocial significance of acknowledg-ment practices in a variety of disciplines, including astron-omy (Verner, 1993), genetics (McCain, 1991), biology(Heffner, 1981), chemistry (Heffner, 1981), psychology(Heffner, 1981; Cronin, 1995), information science (Cronin,1995, 2001), sociology (Patel, 1973; Cronin, 1995), politi-cal science (Heffner, 1981), and philosophy (Cronin, 1995).

Bazerman (1984, 1988) has chronicled the evolutionof the acknowledgment during the 19th and 20th centuriesin the journal literature of experimental physics, showinghow it became, to paraphrase Grafton (1997, p. 233), anintegral part of the rhetoric of narration and annotation

“sub-authorship collaboration” (Patel, 1973, p. 81),

twe should not assume too much in terms of common un-derstandings: the dividing line between the two classifica-tions, author and acknowledgee, is neither universally ap-preciated nor consistently applied. Cronin (1995, pp. 85–86), for instance, has shown that interpretative disputes arenot uncommon, and that some researchers feel that theyhave been denied their just deserts by being downgradedfrom coauthor to acknowledgee. A

The explosion of coauthorship naturally raises the ques-tion why authorial surplus could not be accommodated inthe acknowledgment sections that accompany the great ma-jority of scientific articles.

I have chosen not to hide behind the cloak of anonymity, orbypass the rigors of peer review by posting a version of thispaper on my Web site;

ut before considering theICMJE’s guidelines for authorship, it may be instructive toreflect on the role of acknowledgments in the primarycommunication process, and the relationship between au-thorship and acknowledgment.

ublinCore and REACH (Record Export for Art and CulturalHeritage) prefer the terms ‘creator’ and ‘maker’ to the moretraditional ‘author’ (Baca, 1998). T

In rare cases, a questionable, published paper may acquire“orphan” status (Rennie & Flanagin, 1994, p. 469), as allconcerned try to wash their hands of it, invoking hyperlaborspecialization as grounds for exoneration. Such a scenario isinconceivable under the standard model, where authorshipand accountability are isomorphic. But when authorship/ownership of a study is distributed across multiple contrib-utors, many of whom may have zero or weak relation-ships—whether personal or institutional—with their myriadcoworkers (Katz & Martin, 1997), the practical (i.e., en-forceable) allocation of accountability may pose intractableproblems

Additionally, Ducor (2000) investigated asmall set of patents in molecular biology and their concom-itant publications in the scientific literature. Of the 40patent-article pairs examined, all but two listed more au-thors than inventors, which raises interesting questionsabout the relative stringency of the criteria employed forconferring authorship and inventorship.

Slone (1996), in a survey of “major research” articlespublished in theAmerican Journal of Roentgenology, foundthat undeserved coauthorship rose from 9% on papers withthree coauthors to 30% on papers with more than six coau-thors.

Flanagin et al.(1998) developed a multivariate logistic regression model totest the hypothesis that coauthored articles (operationalizedas papers with six or more authors) were increasing at a rategreater than would be expected when confounding vari-ables, such as the number of centers, were taken into ac-count. They found that 19% of original research reports hadhonorific authors, individuals who were garnering phantomfodder for their curricula vitae. They also discovered that11% of articles had ghost authors, which means that quite afew individuals were not receiving due credit for theircreative or material contributions to the research process—“the ghostly inferred hosts of unnamed actors who shiftedinstruments about and exerted their muscular labor’(Shapin, 1995, p. 379). Their findings, based on surveys ofcorresponding authors, are in keeping with other estimatesof honorific authorship in the biomedical literature.

here is com-pelling evidence that many individuals receive unwarrantedcoauthor status (variously referred to as ‘guest’, ‘gift,’ or‘surprise’ authorship) while others are denied legitimatelyearned author status (‘ghost’ authorship). As Slone (1996, p.578) notes, authorship “cannot be conferred but must beearned.” I

ichard Horton (1998, p. 688) editor ofThe Lancet, speaks, in fact, of “the shattered system ofacademic reward and its symptom, broken rules of author-ship,” a view which seems neither extreme nor marginaljudging by the tenor of the debate being conducted in thepages of the global biomedical literature (Klein & Moser-Veillon, 1999)

eceiving credit under false pretenses,are cause for grave concern (Anderson, 1991).

mass of editorial commentary andcorrespondence in the letters pages of major journals (e.g.,Constantian, 1999; Rennie, Yank, & Emanuel, 1997). Ho

norificauthorship and data integrity, seem to be of especial concernto the biomedical community, given widespread media cov-erage of, and speculation about, fraudulent practice, theeffects of which, in both career and personal terms, can bedevastating (e.g., Kevles, 1998).

ow, for instance, should a promotion andtenure committee view the contribution of the 99th listedauthor on a particle physics paper or the 36th author on agenome sequencing study? What may seem to constitute aminiscule portion of a single journal article may, in fact,have consumed a significant amount of that individual’sprofessional time and energy. T

owever, multipleauthorship and hyperauthorship are not problematized byphysicists as they are by the biomedical community.

t this point, the notion of authorship,literally interpreted, is effectively rendered meaningless.

ennieand Yank (1998, p. 829), when “the number of collaboratorsgrows arithmetically, it becomes exponentially harder toaffix responsibility.”

he practice of promiscuous coauthor-ship puts considerable stress on this tried and tested model.

Under the standard model, the rights and responsibilitiesof authorship are clearly apprehended by all parties: authors,editors, referees, and readers. In appending my name to thisarticle I am nailing my colors to mast; if the article attractscritical approval, is discussed, quoted, and, in due course,cited in the scholarly literature, I shall be happy to bank thesymbolic capital which accrues to me as author and origi-nator. If the paper is challenged because of exiguity oftheoretical, historical, or empirical heft, I shall simply haveto face the music: there are no coauthors to help deflectcriticism. Likewise, if I am challenged for drawing toosparingly, selectively, or generously on the ideas and workof others, I understand the possible consequences. However,I have chosen not to hide behind the cloak of anonymity, orbypass the rigors of peer review by posting a version of thispaper on my Web site; rather, I want to publicize my ideasamong my peers, and the best way to do that, and signify mytrustworthiness, is to pursue publication in an accreditedforum. As a serial author, I am fully cognizant of the rightsand responsibilities of authorship. I understand the norms ofscholarly publishing, and I am aware of the sanctions thatmay be invoked if infractions occur. Should the argumentsin this paper prove flawed, no one but myself is to blame,and that includes those whom I have named in the acknowl-edgments section. If the paper attracts attention, I shall behappy to bask in the glow.

e need toconsider how multiple authorship, in extremis—what I havechosen to term ‘hyperauthorship’— undermines commonlyheld assumptions about the nature and ethical entailments ofauthorship, and how, in exceptional cases, it can lead tofundamental questions about the integrity of the researchcommunity as a whole. Unfortunately, little effort is madein the biomedical literature to distinguish systematicallybetween what might be termed acceptable levels of multipleauthorship and unacceptable levels of hyperauthorship.Studies of coauthorship trends, as will become clear,

n the realm of periodical publications,the sovereignty of the standard model is being most hotlycontested in biomedical research fields, where intense levelsof professional collaboration and coauthorship are common-place (Croll, 1984; Rennie & Flanagin, 1994; Rennie, Yank,& Emanuel, 1997; Rennie & Yank, 1998; King, 2000).Proposals for reform, which seek to retire the concept ofauthorship and replace it with a scheme for the allocation ofspecific, task- or job-related credits (e.g., Squires, 1996;Smith, 1997) are not only being debated by editors andothers, but are being adopted by leading scientific journal

Rennie’s and Flanagin’s (1994, p. 469) beguilingly simplequestion: “. . . how many people can wield one pen?” S

he standard model of scholarly publishing, inwhich it is assumed that a work is written by an author, t

Some medical commentators (Rennie, Yank, &Emanuel, 1997, p. 582) have advocated abandoning theconcept of author altogether in favor of ‘contributors’ and‘guarantors,’ thereby freeing us “from the historical andemotional connotations of authorship.”

ouble-blind) peer review became an established component of thepost-war scientific bureaucracy (Chubin & Hackett, 1990,pp. 19 –24)

e well-estab-lished “conventions of impersonality” in scientific writing(Hyland, 1999, p. 355) and resulting in the erasure of styl

0authors—a leading indicator of hyperauthorship?—in-creased from 1 in 1981 to 182 in 1994 (McDonald, 1995)and that the average number of authors per paper in theScience Citation Index (SCI)increased from 1.83 in 1955 to3.9 in 1999 (personal communication with Helen Atkins,Director of Database Development, Institute for ScientificInformation, Philadelphia, 2000). To use a couple of ran-dom examples: a 1997 article inNature(cited almost 600times since then) on the genome sequence of a bacteriumhas 151 coauthors, drawn from dozens of research labora-tories scattered across twelve countries (Kunst et al., 1997).A recent (Daily et al., 2000). two-page article inScienceonthe economic value of ecosystems has no fewer than 17authors and five acknowledgees. I

s not still common prac-tice in fields such as philosophy or women’s studies (Cro-nin, Davenport, & Martinson, 1997). By way of example,Endersby (1996) has analyzed trends in, and reasons for,collaboration and multiple authorship in the social sciences.Patel (1973) has described the growth of coauthorship insociological journals for the period 1895 to 1965. Bird(1997) has found evidence of coauthorship growth in theliterature of marine mammal science (1985–1993), whileKoehler et al. (1999) found that the average number ofauthors per article in theJournal of the American Society forInformation Science(previouslyAmerican Documentation)rose from approximately 1.2 in the 1950s to 1.8 in the1990s

n a sample of 2,101 scientificpapers published between 1665 and 1800, Beaver andRosen found that 2.2% described collaborative work. No-table was the degree of joint authorship in astronomy,especially in situations where scientists were dependentupon observational data.

French science was much more profession-alized and institutionalized than was the case in either of theother European powers. Specifically, they found that morethan half of all the coauthored scientific articles in theirhistoric sample had been produced by French scientists.

Beaver and Rosen (1978) have shown how the differentialrates of scientific institutionalization in France, England,and Germany are mirrored in the relative output of coau-thored papers.

iking increase inrates of coauthorship, though the latter is only a partialindicator of the former: coauthorship and collaboration arenot coextensive (Katz & Martin, 1997, p. 1). This trend ismost noticeable in experimental high-energy physics(HEP), with its often very large teams and highly sophisti-cated collaborations (Kling & McKim, 2000). A similartrend, dating from the 1990s, can be seen in the biomedicalresearch literature, particularly with regard to publicationsarising from large, multi-institutional clinical trials (Rennie,Yank, & Emanuel, 1997; Horton, 1998). H

In some domains, path-breaking work is nec-essarily the outcome of collaborative activity rather thanindividualistic scholarship, a fact reflected in the modestproportion of federal research funds which is allocated toindividual investigators rather than teams. Collaborationsare a necessary feature of much, though by no means all,contemporary scientific research.

After World War II, collaboration became a defin-ing feature of ‘big science’ (Bordons & Gomez, 2000;Cronin, 1995, pp. 4 –13; Katz & Martin, 1997).

ncidentally, parallel,though not quite so dramatic, growth has been observed inthe number of individuals being formally acknowledged inscholarly journals for their multifarious contributions: whathas become known as subauthorship collaboration (Patel,1973; Heffner, 1979, 1981; Cronin, 1995, Cronin,2001)—an important, if underappreciated, indicator of in-formal scientific collaboration

o some extent, authorship has becomea collective activity, with numerous coauthors competingfor the byline, some of whom may not have written, strictlyspeaking, a single word of the associated work (McDonald,1995; Kassirer & Angell, 1991).

n general terms, the lone authorstereotype ignores the fact that a great deal of the scholarlyliterature is the product of a “socio-technical production andcommunications network” (Kling, McKim, Fortuna, &King, 1999),

The standard model of scholarly publishing assumes awork written by an author. There is typically a single authorwho receives full credit for theopusin question. By thesame token, the named author is held accountable for allclaims made in the text, excluding those attributed to othersvia citations. The appropriation of credit and allocation ofresponsibility thus go hand-in-hand, which makes for fairlystraightforward social accounting. The ethically informed,lone scholar has long been a popular figure, in both fact andscholarly mythology. Historically, authorship has beenviewed as a solitary profession, such that “when we picturewriting we see a solitary writer” (Brodkey, 1987, p. 55). Butthat model, as Price (1963) recognized almost three decadesago, is anachronistic as far as the great majority of contem-porary scientific, and much social scientific and humanistic,publishing is concerned.

here reputation, career suc-cess, and, ultimately, remuneration are tightly coupled withpublication salience and citation (Cronin, 1996), t

hapin (1995,p. 178) notes in his brilliant study of trust in 17th-centuryEnglish science,

As Katzen (1980, p. 191) notes in heranalysis of early volumes of thePhilosophical Transac-tions:. . . no attempt is made to give prominence to the author ofthe article . . . there is generally no reference at all to theauthor in the heading that signals a new communication. Ifthe author is referred to in the title, it is likely to be in anoblique form . . . we are at the threshold between anony-mous and eponymous authorship

s Rennie and Flanagin(1994) remind us, there is no standard method for determin-ing order, nor any universalistic criteria for conferring au-thorship status:

much has changedin terms of the ways, both instrumental and stylistic, inwhich scientists communicate the results of their research totheir peer communities. However, in the intervening 300-plus years, certain symbolic and rhetorical practices, nota-bly the assertion and defense of authorship, and all thepresumptive rights associated therewith, have remainedcenter stage. In the 17th century, the business of authorship,as the business of science itself, was much less complicatedand contentious than today—which is not to say that prioritydisputes were unheard of, that egos were never bruised, orthat “the bauble fame” did not come into play in earliertimes.

etter writing continued as a medium for theinformal exchange of information and for requesting fellowscientists to replicate experiments (Manten, 1980, p. 8).

Before the precursors of today’s scholarly journals es-tablished themselves in the second half of the 17th century,scientists communicated via letters.

But while the traditional model of authorship persists, mostnoticeably in the humanities, it is no longer the sole ordominant model in certain scientific specialties.

uthorship (and therecognition that flows therefrom) is the undisputed coin ofthe realm in academia: it embodies the enterprise of schol-arship (Bourdieu, 1991; Cronin, 1984, 2000; Franck, 1999

o state the obvious, public affirmation of au-thorship is absolutely central to the operation of the aca-demic reward system, whether one is a classicist,sociologist, or experimental physicist.

Foucault,1977,p.125

Manguel(1997,pp.182–183)

Thisarticle(a)beginswithabrief,historicaloverviewofscholarlypublishing,focusingontheroleoftheauthorandtheconstitutionoftrustinscientificcommunication;(b)offersanimpressionisticsurveyandanalysisofrecentdevelop-mentsinthebiomedicalliterature;(c)explorestheextenttowhichdeviantpublishingpracticesinbiomedicalpublishingareafunctionofsociocognitiveandstructuralcharacteris-ticsofthedisciplinebycomparingbiomedicinewithhighenergyphysics,theonlyotherfieldwhichappearstoexhibitcomparablehyperauthorshiptendencies;and(d)assessestheextenttowhichcurrenttrendsinbiomedicalcommuni-cationmaybeaharbingerofdevelopmentsinotherdisci-plines

aproposaltoreplaceauthorsentirelywithlistsofcontributors(Rennie,Yank,&Eman-uel,1997)

thecentralissueisnotjustoneofmultipleauthorship,buttointroduceaneologismofhyperauthorship—

scaleofthephenomenonandassociatedethicalabuseshaveprovedtobesingularlyproblematicinthebiomedicaldo-main(e.g.,Houston&Moher,1996

Inthebiomedicalresearchcommunity,multipleauthorshiphasincreasedtosuchanextentthatthetrustworthinessofthescientificcom-municationsystemhasbeencalledintoquestion.Doc-umentedabuses,suchashonorificauthorship,havese-riousimplicationsintermsoftheacknowledgmentofauthority,allocationofcredit,andassigningofaccount-ability.Withinthebiomedicalworldithasbeenproposedthatauthorsbereplacedbylistsofcontributors(theradicalmodel),whosespecificinputstoagivenstudywouldberecordedunambiguously.Thewiderimplica-tionsofthe‘hyperauthorship’phenomenonforscholarlypublicationareconsidered.

lphabetization through weightedlisting to reverse seniority (e.g., Spiegel & Keith-Spiegel,1970; Riesenberg & Lundberg, 1990).

First of all, we can say that today's writing has freed itself from the theme of expression. Referring only to itself; but without being restricted to the confines of its interiority, writing is identified with its own unfolded exteriority. This means that it is an interplay of signs arranged less according to its signified content than according to the very nature of the signifier. Writing unfolds like a game [jeu] that invariably goes beyond its own rules and transgresses its limits. In writing, the point is not to manifest or exalt the act of writing, nor is it to pin a subject within language; it is, rather, a question of creating a space into which the writing subject constantly disappears.

Compell'd these skipping kerns to trust their heels,