The Image of Objectivity Author(s): Lorraine Daston and Peter Galison Source: Representations, No.

40, Special Issue: Seeing Science (Autumn, 1992), pp. 81-128 Published by: University of California Press Stable URL: http://www.jstor.org/stable/2928741 . Accessed: 27/06/2011 12:02
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=ucal. . Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

University of California Press is collaborating with JSTOR to digitize, preserve and extend access to Representations.

http://www.jstor.org

LORRAINE

DASTON

AND

PETER

GALISON

The Image of Objectivity

The Talismanic Image oply of visual methodsin the sciences,concluded: "There is no doubt thatgraphical expressionwillsoon replace all otherswheneverone has at hand a movement or change of state-in a word, any phenomenon. Born before science,language is often inappropriate to express exact measures or definiterelations."Others the of dialectic," the "powerof arguments," mightcryout to salvage the "insights of language," but theirprotestations "insinuationsof elegance," or the "flowers were loston Marey,who dreamed of a wordlesssciencethatspoke insteadin highspeed photographs and mechanicallygenerated curves; in images thatwere, as he put it,in the "language of the phenomena themselves."' of a new brand of scien"Let nature speak foritself"became the watchword At issue half of the nineteenth in the latter that tific century. emerged objectivity as well: the all-too-human scientists was not onlyaccuracybut morality must,as a themselves fromimposingtheirhopes, expectations, restrain matterof duty, generalizations,aesthetics,even ordinarylanguage on the image of nature. Where human self-discipline flagged, the machine would take over. Wary of human between nature and representation, intervention Marey and his contemporaries turned to mechanicallyproduced images to eliminatesuspect mediation. They enlistedpolygraphs,photographs,and a hostof otherdevicesin a near-fanatical effort to create atlases-the bibles of the observationalsciences-documenting and flowers in images thatwere human bodies, elementary birds,fossils, particles, freeof human interference. certified in the late nineThis essay is an account of the moralizationof objectivity teenth and early twentiethcenturies as reflectedin scientific image making. atlases from diverse fields (anatomy,physiology, We will use scientific botany, paleontology,astronomy,X-rays,cloud-chamber physics)and from a span of several centuries (eighteenthto twentieth) to chart the emergence and nature and subjectivity. of new conceptions of objectivity We do not intend anything of thegenreand history of scientific atlases; approaching a comprehensivesurvey willbe primarily ratherour attention focusedon the latterhalfof the nineteenth century,when atlases proliferatedin number and kind, purveyingimages of fromspectrato embryos,2 and when atlasesbecame manifestos forthe everything new brand of scientific In order to the of this form objectivity. highlight novelty of objectivity, we shallcontrast itto theideals and practices of earlieratlas makers.
REPRESENTATIONS

IN 1878

FRENCH PHYSIOLOGIST

E.J. Marey, surveying the pan-

40

* Fall

1992 ?

THE REGENTS OF THE UNIVERSITY

OF CALIFORNIA

81

or "mechanical"objectivity is onlyone What we willcall "noninterventionist" intoour current, of severalelementsthathistorical pressureshave fusedtogether Modern mixes rather than intenotion of objectivity objectivity. conglomerate and which are historically conceptuallydistinct. grates disparate components, in has its own addition to the collectivehistory of these Each history, components that explains how all of them came to be amalgamated into a single,if layered, confused concept. This layeringaccounts for the hopelessly but interestingly from which can be to of the term objectivity, applied everything present usage to As hisemotional detachment. to correctness empirical reliability procedural we will not be concerned with recent controversiesover torians of objectivity, existsand, ifso, whichdisciplineshave it.We believe,however, whetherobjectivity these debates by revealingboth of scientific that a history may clarify objectivity of the componentsthatmake up the currentconand contingency the diversity cept. Withoutknowingwhat we mean and whywe mean it in asking such quesit is hard to imagine what a sensible tions as "Is scientific knowledgeobjective?," answer would look like. of onlyone componentof objectivity, In what followswe address the history but we believe thatthiscomponent revealsa common pattern,namelythe negaas wax is relatedto subjectivity tivecharacterof all formsof objectivity. Objectivity to seal, as hollow imprintto the bolder and more solid featuresof subjectivity. formof subjecEach of the several componentsof objectivity opposes a distinct of defined some no means is each all) (by aspects the personal. bycensuring tivity; be The historyof the various formsof objectivity might told as how, why,and came to be seen as dangerously when various forms of subjectivity subjective. to thesubjectivity was indifferent Mechanical objectivity of,forexample, personal and aesthetic of scientific idiosyncrasies;rather,it combatted the subjectivity on a moral It took and anthropomorphism. building, judgment, dogmaticsystem to control were amenable of these because thought subjectivity aspects aspect it centered on the scientific image because images were through self-restraint; charmsagainst thoughtleast vulnerableto such subjectiveintrusions-protective and bad faith, system building. ambiguity, between atlasmakerswas nota mismatch The problemfornineteenth-century but for had it been world and mind, as seventeenth-century epistemologists, The moral remediessoughtwere those rathera strugglewithinwardtemptation. of self-restraint: images mechanicallyreproduced and published wartsand all; textsso laconic thattheythreatento disappear entirely. Seventeenth-century episobjectivity temology aspired to the viewpoint of angels; nineteenth-century effaces of saints. Although mechanical objectivity aspired to the self-discipline it demands othertraits;it has a positiveas well as a some featuresof the scientist, to eliminate attempts negativesense. In itsnegativesense, thisideal of objectivity the mediating presence of the observer: some versionsof this ideal rein in the judgments thatselectthe phenomena, whileothersdisparage the senses thatreg82 REPRESENTATIONS

isterthe phenomena, and stillothersward offthe theoriesand hypothesesthat distort the phenomena. In its positive sense, mechanical objectivity requires perseverance,prepatience,unflagging painstakingcare and exactitude,infinite ternaturalsensoryacuity,and an insatiableappetite for work. The phenomena never sleep and neither should the observer; neitherfatigue nor carelessness thatsmearsa measurementor omitsa detail; the vastexcuse a lapse in attention be endlesslyrepeated. of nature require thatobservations ness and variety is heroic What unitesthe negativeand positivesides of mechanicalobjectivity on the one side, the honestyand self-restraint required to foreself-discipline: of one's own senses; on and even the testimony swearjudgment, interpretation, and measurefor observation the other,the taut concentration precise required is of work that It a vision scientific around the clock. ment,endlesslyrepeated of thebourgeoisratherthanthemoodybrilliance the plodding reliability glorifies of the genius. It is also a profoundlymoralized vision, of self-commandtriof fleshand spirit.Like almost all umphing over the temptationsand frailties albeitof a highly itpreaches asceticism, formsof moral virtuosity, specialized sort. and other had less to do withenvy,lust,gluttony, and frailties The temptations withwitting and unwitstandard sins than withseeing as ratherthan seeing that; scienBut in the viewof late-nineteenth-century tingtamperingwiththe "facts." to as the seven as difficult combat sins were almost these tists, deadly professional conscience. ones, and required a sternand vigilant withmoralizedscience, Mechanized scienceseems at first glance incompatible but in factthe twowere closelyrelated.Whilemuch is and has been made of those and moral-that distinguishhumans distinctive traits-emotional, intellectual, from machines, it was a nineteenth-century commonplace that machines were human virtues. of Chief certain among these virtueswere those assoparagons ciated with work: patient, indefatigable, ever-alert machines would relieve human workerswhose attentionwandered, whose pace slackened, whose hand trembled. Scientistspraised automatic recording devices and instrumentsin much the same terms. As the photograph promised to replace the meddling, instrument so the self-recording promisedto replace the meddling, wearyartist, weary observer. It was not simplythat these devices saved the labor of human observers; they surpassed human observersin the laboring virtues: they produced not just more observations,but better observations.Of course, strictly speaking, no merit attached to these mechanical virtues, for their exercise involvedneitherfreewillnor self-command. But the factthatthe machines had of theirown powers of no choice but to be virtuousstruckscientists distrustful of as a distinct Instead freedom of will, machines self-discipline advantage. offeredfreedom fromwill-from the willfulinterventions that had come to be If the machine was ignorant seen as the most dangerous aspects of subjectivity. of theoryand incapable ofjudgment, so much the better, for theoryand judgment were the firststeps down the primrose path to intervention. In its very The ImageofObjectivity 83

the machine seemed to embodythe negativeideal of noninterventionist failings, withitsmorality of restraint and prohibition. objectivity, In thisessay,we argue thatthisformof scientific objectivity emerged onlyin the mid nineteenthcenturyand is conceptuallydistinct fromearlier attempts to be "true to nature" in its methods (mechanical), its morals (restrained),and its can be found in metaphysics(individualized). Although mechanical objectivity almost everyscientific our attentionto atlases endeavor,we shall largelyrestrict the strongassociationbetween the (and related volumes) for two reasons: first, visual and the factual made atlases prime bearers of the new objectivity; and betweenthe missionof all atlases to characterize(not simply second, the conflict on the inventory)phenomena on the one hand, and the ban on interpretation were prepared to pay forthatobjectivity. other,shows how high a price scientists we use earlier,alterThe remainderof the essayis divided intothreeparts.First, nativeapproaches to creatingpicturesthatwere trueto nature,but not objective and objectivity. in the mechanical sense, to pryapart the ideals of verisimilitude that techniques of mechanical reproduction Second, we examine the attraction held foradvocates of the new objectivity, and theirultimatedisappointment with we with some on how and these techniques. Third, conclude reflections why came to be moralized. objectivity Truthto Nature use objecHistoriansand, especially, philosophersof science routinely as a panhistoricalhonorific, awardingit to thisor thatdisciplineas it comes tivity little attention to when objectivity itself but of scientific developed, or paying age to whatservedas itssource. Before theseveralcomponentsof objectivity emerged otherideals guided and earlytwentieth and merged in the nineteenth centuries, scientific practice.Among these was the preceptof truthto nature,and nowhere atlas makers,who believed was this precept more revered than among scientific theirimages to be the closest possible renderingof what trulyis. The choice of is"engaged atlas makersin ontological whichimages bestrepresented"whattruly laterforbade. In thissection, and aesthetic judgments thatmechanicalobjectivity of scientific atlases and then surveythe nature and functions the we first explain atlas makers reconciled these funcvarious means by which eighteenth-century tionsto theirmandate to create images true to nature. of the sciences of the eye have From the sixteenthcenturyon, practitioners prepared editions of their designated phenomena in the form of atlases, profuselyillustratedvolumes of carefullychosen observables-bodily organs, constellations,floweringplants, instrumentreadings-depicted from a carefully The purpose of these atlases was and is to standardizethe chosen point of view.3 observingsubjectsand observed objectsof the disciplineby eliminatingidiosyn84
REPRESENTATIONS

crasies-not only those of individualobserversbut also those of individual pheto the nomena. Because we moderns habituallyoppose our brand of objectivity their about we fret most of subjects: "peridiosyncratic subjectivity individuals, sonal equations," their theoreticalbiases, their odd quirks. But idiosyncratic objectspose at least as greata threatto communal,cumulativescience,fornature and individuality seldom repeats itself, variability being the rule ratherthan the safe for make nature The aims to atlas science; to replace raw expeexception. of individualobjects-with rience-the accidental,contingent experience specific digested experience. All sciences must deal with this problem of selecting and constituting and too variousnaturalobjects. objects,"as opposed to the too plentiful "working This is a problem anterior to the problem of reference,and posterior to the of study.The problem of problem of selectingdomains of phenomena worthy referencedeals withhow conceptsadhere to theworld.If working objectsare not raw nature,theyare not yetconcepts,much less conjecturesor theories;theyare the materialsfromwhich concepts are formed and to which theyare applied. The problem of selectiondeals withwhichphenomena are keyto the essence of afterthischoice of phenomena has been things.Workingobjects are constituted mechanicianshad selected penmade-after, for example, seventeenth-century or twentiethdula as more revealingof the essence of motionthan air currents, as more banded snails had selected revealingof biologists centuryevolutionary the interplaybetween natural selection and random driftthan chimpanzees. Workingobjects can be atlas images, type specimens,or laboratoryprocessesof the sectorof nature under invesany manageable, communal representatives do such standardized without No can science workingobjects,forunretigation. fined natural objects are too quirkilyparticularto cooperate in generalizations and comparisons. Atlases supply workingobjects to the sciences of the eye. For initiatesand neophytesalike, the atlas trainsthe eye to pick out certain kinds of objects as exemplary (e.g., this "typical"liver ratherthan that one with hepatitis)and to regard themin a certainway(e.g., using the Flamsteedratherthan the Ptolemaic celestialprojection).To acquire thisexperteye is to winone's spursin mostempirical sciences; the atlases drillthe eye of the beginner and refreshthe eye of the such old hand. In the case of atlases thatpresentimages fromnew instruments, in the field addressed as the X-rayatlases of the earlytwentieth century, everyone bythe atlas mustbegin to learn to "read" anew. Because atlases habituatethe eye, theyare perforcevisual, even in those disciplineswhere other sensationsplay a into as many role (e.g., texturein botany,which refinesthe lay hairy significant distinct termsas Eskimos allegedlydo snow).Whateverthe amount and avowed of thetextin an atlas,whichvariesfromlong and essentialto nonexistent function and despised, the illustrations command centerstage. Usually displayedin giant drawn and engraved,and expensivelyproduced, theyare format,meticulously The ImageofObjectivity 85

of the atlases. Indeed, to call them "illustrations" the raisond'etre at all is to belie forit suggeststhattheirfunction is merelyancillary, theirprimacy, to illustrate a textor theory. Some earlyastronomical atlasesdo use the figures as genuine illusto explicate rivalcosmologies.4 But in mostatlases fromthe eighteenth trations, the are the and centuryon, pictures alpha omega of the genre. of standardizing In addition to theirprimary function objectsin visual form, atlas picturesserved other purposes in the natural sciences. In part,theyserved forthe scientific the cause of publicity what is ephemcommunity, bypreserving what is rare or inaccessible to all who can purchase the eral and distributing volume, notjust the luckyfewwho were in the rightplace at the righttimewith the rightequipment.5In part, picturesserved the cause of memory, for,as the atlas makers never tire of repeating,images are more vivid and indelible than words. And in part, especially for nineteenth-century authors, picturesserved would the the cause of incorruptibility: check they impulse to infuseobservation and endure as facts for tomorrow's a researchers long after with pet theory, had gone the wayof phlogiston. today'stheoriesand systems Consider forexample Frenchpathologist Jean Cruveilhier's pioneeringatlas, du humain Cruveilhier was at pains to argue Anatomie (1829-35). pathologique corps however"graphic," his colleagues out of theirpreferenceforverbaldescriptions, in whichthereexistabundant to normalanatomy, of diseased organs. In contrast time," opportunitiesto observe thisor thatorgan "a second, a third,a twentieth the opportunitiesfor the pathologistare rare and fleeting: "A lost occasion may perhaps never recur."Even an observerwiththe eyes of a lynxand the memory ifhe does not engrave themas if of an elephant cannot "fixthe fugitive features, in bronze, so as to be able to representthemat will,to put themintorelationwith analogous facts."Simplypicklingthe anomalous organ in alcohol is a poor subfora faithful stitute drawingof a freshcadaver,since changes of formand color "denature them" and in any case such a specimen "profits onlya small number" of normal the patholoof observers.Lacking the repetitive anatomy, experience reprogistis doubly aware of the psychologicaltruththata picture"constantly duces the same image," creatinga vivid,indelible memory.Finally,the faithful drawing, like nature, outlives ephemeral theories-a standing reproach to all a factto fita theory.6 This twist who would, whether"bytheirerroror bad faith," was new to the nineteenth and last functionof the atlas image as sentinel century, to come. was a portentof the mechanicalobjectivity atlases was the dictum "truthto What was not new to nineteenth-century on itsaccuracy,on nature": there is no atlas in any fieldthatdoes not pique itself to fact.But in order to decide whetheran atlas pictureis an accurate its fidelity decide what nature is. All atlas rendering of nature, the atlas maker must first makers must solve the problem of choice: Which objectsshould be presentedas In the late the standard phenomena of the discipline,and fromwhichviewpoint? of and a these crisis choices nineteenth denial, forthey anxiety century, triggered 86
REPRESENTATIONS

but in earlier periods atlas makersfaced up to to subjectivity, seemed invitations theirtask withconsiderablymore confidenceand candor. This is not to say that in the pejorativesense of rendering theyabandoned themselvesto subjectivity, On the contrary, whims decreed. their as theywere well nigh personal specimens maniacal in theirprecautionsto ensure the accuracyof theirfigures, according to theirown lights.However,theydid conceive the exercise ofjudgment in the selectionof "typical," "characteristic," "ideal," or "average" images to be not only inevitablebut laudable, the essence of the atlas maker's mission. In their view, whatevermerittheiratlases possessed derived preciselyfromthesejudgments, and fromthe breadth and depth of experience in theirfieldupon which those interresisted to mechanicalobjectivity judgmentsrested.Atlasmakerscommitted solelyto truthto nature,relishedit. pretation;theirpredecessors,committed These early atlas makers,while proud of theirinterpretive skills,did not all same The words nature" the of "truth to the notion ideal, way. typical, interpret all even not and are characteristic, average though they fulpreciselysynonymous, will filledthe same standardizingpurpose. Examples fromthe earlier literature on how to be true to nature more vivid. A schematic make their differences of earlier atlases will show thattruthto nature was both a possible and typology The categories variegatedideal long beforethe advent of mechanicalobjectivity. the nineteenth cento and instantiating early examples span the late seventeenth thatmovesfromtypesand ideals to characteristic turies,witha rough chronology a tendency individualsto individualstout court. However,thischronologyreflects and severalof thecategoriescoexistin time. ratherthan a clear-cutperiodization, to show that concern for These alternativeways of being true to nature suffice for not concern the twocondoes further, necessarily objectivity; accuracy imply threatenedto undermine when mechanical objectivity cerns came into conflict nature. the primarygoals of atlases in representing In eighteenth-century atlases, "typical"phenomena were those that hearor "archetype," and fromwhichindividual kened back to some underlyingTypus be at least The could derived, conceptually. typicalis rarelyif ever phenomena embodied in a single individual;nonetheless,the researchercan intuitit (see the of fig.1) fromcumulativeexperience.As Goethe wroteof his archetype Urpfjanze of the animal skeleton: willbe suggested conhere,a generalpicture Hence, an anatomical archetype [Typus] which will of all as one us to an the forms animals potential, guide orderly descriptaining in generalimpliesthatno tion of each animal.... The mere idea of an archetype theparticular animalcan be used as our pointofcomparison; can never serve particular for the whole.7 as a pattern [Muster] This is not to say that the archetypewhollytranscendsexperience, for Goethe claims thatitis derived fromand testedbyobservation.However,observationsin search of the typicalmustalwaysbe made in series,forsingleobservationsmade The ImageofObjectivity 87

FIGURE

.the
The UrpJnze,

a
Goethe.s ". : . .

of a higher-4 sketchof the "typus meant to and insect," plant representno plant in particular but ratherthe morphological prototypefromwhichall higher derived. Reproduced fromK. Die Lothar Wolfet al., eds., Goethe: zurNaturwissenschaft Schriften (Weimar,1947-1986), vol. 9a, Zur ed. Dorothea Kuhn, Morphologie,

FIGURE

i.

?.

si..

/ ,d

?7&

By

'4 i(c i

> ..W

"For the observerneversees the pure by one individual are highlyidiosyncratic: with own his eyes; rather,much depends on his phenomenon [eine Phnomen] mood, the stateof his senses, the light,air,weather,the physicalobject,how it is Thus forGoethe,writing before handled, and a thousand othercircumstances."8 the act of "definition" the advent of mechanicalobjectivity, required to distillthe but rather typicalfromthe variable and accidentalis not a slide into subjectivity a precaution againstit. Typical images dominate the anatomicalatlases of the seventeenththrough mid nineteenthcenturies,but not always in the unalloyed form celebrated by whichwe shall call the "ideal" and the "characGoethe. Two importantvariants, atlas of thisearlierperiod. Briefly also illustrations teristic," stamp put,the "ideal" but the perfect, while the "charimage purportsto render not merelythe typical in an individual.Both ideal and characteristic acteristic" image locates the typical of both insistedupon images standardize the phenomena, and the fabricators 88 88
REPRESENTATIONS

pictorialaccuracy. But the ontologyand aestheticsunderlyingeach contrasted as a fewexamples willshow. sharply, With the collaborationof Dutch artistand engraverJan Wandelaar,9 Bernhard Albinus, professorof anatomy at Leyden, produced several of the most influential anatomical atlases of the idealized sort,includingthe Tabulaesceleti et In hominis the to musculorum this Albinus described work, (1747). preface corporis his goals and workingmethodsin considerabledetail,and in termsthatseem selfhe is at once committed contradictory bythe standardsof mechanicalobjectivity: to the mostexactingstandardsof visualaccuracyin depictinghis specimens,and to creatingimages of "the best patternof nature"(see fig.2). To the formerend, he goes to lengthsuntilthenunheard of among anatomists to meticulously clean, reassemble, and prop up the skeleton,checking the exact positionsof the hip bones, thorax,clavicles,and so on, bycomparisonwitha veryskinnyman made to stand naked alongside the prepared skeleton. (This test cost Albinus some anxietyas well as timeand trouble,forthe naked man demanded a fireto ward offthe winterchill, greatlyacceleratingthe putrefaction of the skeleton.) Still worriedlestthe artist err in the proportions, Albinuserectedan elaborate double

s:.i-.::' .:e,.....~..:..

.?...

-FIGURE

2. Albinus, "Fourth and Muscles Human Skeleton ofthe Body(London, 1749). The London publishersapparentlyhad Wandelaar'soriginalillustrations reengravedbyseveral hands; this is signed C. Grignionand G. Scotin.Albinus permitted his artistto make "ornaments"to fill the page, preservethe light and shade of the figures, and to "make them [the tables] more agreeable." This two-and-a-halfviewed in year-oldrhinoceros, was inserted the into 1742, on these latter background grounds: "We thoughtthe rarity of the beast would render these figuresof it more agreeable than any otherornament."

:::' ...? ?.* ' ....~7"".:'....one ? of?' ..up;...-. i;

*...i

'~

The ImageofObjectivity

89

grid,one mesh at 4 rhenishfeetfromthe skeletonand the otherat 40, and positioned the artistpreciselyat thatpoint where the struts of the gridscoincided to the eye, drawingthe specimen square by square, onto a plate Albinus had ruled witha matchingpatternof "cross and streight [sic]lines." This procedure, sugAlbinus's the W. 'sGravesande,is strongly gested by Leyden colleague, physicist of Alberti'sinstructions reminiscent fordrawingin perspective, and amounts to a kind of remote tracingof the object. The fixedviewpointof the artistand the mapping of visualfieldonto plane of representation bymeans of the gridssubject the artistto an exacting discipline of square-to-squarecorrespondence in the name of naturalism.We are not surprisedto read thatAlbinus,like the Renaisof perspective,also prescribedhow the finishedengravings sance practitioners should be viewed as well as drawn.'1 But we may be surprisedto read thatthese remarkablefigures, whichoccasioned three monthsof "an incredibledeal of trouble to the ingraver," are not so of skeleton Albinus the particular actually prepared. Having painstakingly thus taken every ordinary and several extraordinarymeasures to ensure the of object and subject,Albinus'spronouncements integrity aboutjust whatitis the finishedpicturesare picturesof comes as a distinct shock to the modern reader. They are pictures of an ideal skeleton,which may or may not be realized in nature,and of whichthisparticularskeletonis at bestan approximation.Albinus is all too aware of the atlas maker'splight:nature is fullof diversity, but science cannot be. He must choose his images, and his principle of choice is frankly normative: differ from notonly as totheage,sex,stature Andas skeletons one another, and perfection inthemarks ofstrength, and makeofthewhole;I made ofthebones,butlikewise beauty ofboth and agility; thewhole ofitelegant, choiceofone that discover signs strength might so as neither to shewajuvenileor feminine roundnottoodelicate; and at thesametime in an unpolished and nor on thecontrary nessand slenderness, clumsiness; roughness and pleasing to theeye.Foras I wanted to shewan all of thepartsof itbeautiful short, ofnature, I chusedtotakeitfrom thebestpattern ofnature."' example Albinus selectsa skeleton"of the male sex, of a middle stature, Accordingly, and verywell proportioned; of the most perfectkind, withoutany blemish or (For Albinus it went withoutsayingthata perfectskeletonwas perdeformity." an "ideal"-and forcemale; his followerSamuel Soemmerringlaterconstructed ideology-laden-female skeleton.)12 But stillthe skeletonis not perfectenough, and Albinus does not scruple to improvenature by art: "Yet howeverit was not but somethingoccurred in it less compleat than one could altogetherso perfect, wish. As thereforepainters,when theydraw a handsome face, if there happens to render the likenessthe to be any blemishin it mend it in the picture,thereby more beautiful; so those thingswhich were less perfect,were mended in the and were done in such a manner as to exhibitmore perfectpatterns;care figure, taken at the same timethattheyshould be altogether just."'3 being
90
REPRESENTATIONS

"Perfect"and "just" (i.e., exact): these were Albinus'spolestarand compass, and he saw no contradictionbetween the two. Albinus could hold both aims and an attitudetowardjudgment and because of a metaphysics simultaneously with those of the late nineteenthcentury. contrasted that sharply interpretation Albinus believed thatuniversalssuch as his perfectskeletonhad equivIn effect, and that the universalmightbe reprealent ontological warrantto particulars, sented in a particularpicture,if not actuallyembodied in a particularskeleton. That universalcan onlybe knownthroughminuteacquaintance withthe particno matterhow precise, ular in all its details,but no image of a mere particular, can capture the ideal. That requiresjudgment informedbylong experience. Nor was anatomyanomalous in itsidealizingtendencies.Untilwellintothenineteenth and "perfectedtheirfossilspecimens,"'4a century, paleontologistsreconstructed a few decades later,who prided their successors criticized by practice sharply as themselveson "represent[ing]actual specimens withall theirimperfections, and anatomists have been."'5 Late-nineteenth-century theyare,not whattheymay from were that to that and believed real, stray only particulars paleontologists in theinterests of dubious theoriesor systems. was to invitedistortions particulars Like all atlas makers,theystillhad to choose theirimages fromnature'sembarrassmentof riches,but the choice now filledthemwithanxietylesttheysuccumb In contrast, Albinus and other idealizing atlas of subjectivity. to the temptations of makersdid not hesitateto offer pictures objectstheyhad neverlaid eyes upon, of truthto nature ratherthan in violationof it. but in the interest Idealizers of Albinus's stamp were not unaware of what we mightcall the to portray thisparticular alternative-that is,theattempt "naturalistic" objectjust of verisimilar art.(Such objectswere generally, as itappeared, to thelimits though of a largergroup.) Ludwig Choulant,the great not always,deemed characteristic and champion of idealhistorianof anatomical illustration nineteenth-century statedthe naturalistic alternative izers such as Albinus and Soemmerring, onlyto
reject it:

and instruction oftheanatomist, undertheguidance theartist Whenever alone,without will be theresult, a purely individual andpartly takes thedrawing, arbitrary representation thisindividual's is exeevenin advancedperiodsof anatomy. however, Where, drawing of an expert it becomes effective and underthe supervision cutedcarefully anatomist, notonly itsharmony with for ofinstruction, itsindividual nature, truth, purposes through norm[Mittelform], which is no ofanatomic sincethis butalso forthedevelopment science; an exactknowledge buthas becomeideal,can only be attained individual through longer itis thesummation.'6 ofwhich ofthecountless peculiarities of the naturalistalternativein There were eighteenth-century representatives but itwas largelyaesthetics ratherthananxietythatdeteranatomicalillustration, mined their quite explicitchoice. The BritishanatomistWilliam Hunter's The in which the object is representedexactlyas it was seen" as opposed to portrait, The ImageofObjectivity 91

Anatomyof the Human Gravid Uterus (1774), for example, opted for "the simple

FIGURE 3. Plate 2 fromWilliamHunter, Anatomy of the HumanGravidUterus (Birmingham,1774), depictingthe womb of a woman who died suddenlyduring the ninth monthof pregnancy.Hunter used thirteen different subjectsin his atlas,at various stagesin pregnancy. as individual Althoughhe emphasized theirportrayal of objects,he clearlyintended themto be characteristic the anatomyof pregnantwomen in general.

92

REPRESENTATIONS

as were not actually "the representationof the object under such circumstances of "the on in the but conceived seen, elegance and harimagination" grounds bore "the mark a that asserted He natural of the "simpleportrait" object." mony but itself" as the object of truth,and becomes almostas infallible acknowledged that,"being finishedfroma viewof one subject,[it]willoftenbe somewhatindistinctor defectivein some parts,"whereas the figure"made up perhaps froma of studies afterNATURE, mayexhibitin one view,whatcould onlybe seen variety and of of abridgement, in several objects; and it admitsof a betterarrangement, in Hunter'spreferencefor of distrust There was not a whiff greaterprecision."17 the portraitof the individualobject, for he admittedthatconsiderationsof preNor did he regard the aesthetic cision may favorthe ideal or typicalalternative. withsuspicion,as being opposed to scientific Hunter, accuracy.On the contrary, and be like Albinus, considered the beauty of the depiction to parcel of part achievingthataccuracy,not a seductionto betrayit. did indeed It would be a mistaketo take Hunter at his word-that his figures has Ludmilla As was seen." as it the shown, Jordanova object "exactly represent and limbs their with Hunter's deeply unsettling preternatuamputated figures, of the day of naturalism in conventions the artistic outlines, participate crisp rally violencewroughtupon the femalecadaver (see fig. and also in a none-too-subtle Hunter's figurescarrythe 3).18Like the photographsof the nineteenthcentury, the conventions(e.g., sharp have been that to taught stamp of the real only eyes outlines versus softedges) of that brand of realism. Moreover,Hunter's speciwere already objectsof art even before mens, like all anatomical "preparations," wax or with were drawn, dyes to keep vesselsdilated and "natural"injected they Hunter claimed to have moved"notso much even after death.'9 Although looking of his specimens,he considered it part of truthto nature as one joint of a finger" to injectthe womb with"some spiritsto raise it up, as nearlyas I could guess, to was first the figureit had when the abdomen opened."20 For our purposes, his naturalismand the thatscientific because it shows,first, naturalismis instructive cult of individuatingdetail long antedated the technologyof the photograph;21 and, second, that naturalismneed not be coupled withthe anxietyof distortion and the rejectionof aesthetics. Even the naturalismof thecamera obscura did notobviatethe need forselecon the part of the atlas maker. The tivejudgment and extended commentary William Cheselden anatomist persuaded his two Dutch artistsVanderEnglish gucht and Shinevoetto use "a convenientcamera obscura to draw in" (see fig.4) forhis Osteographia so thattheycould accomplishtheirfigures (1733) "withmore obscura was of the camera the mechanical labour." Yet and less precision accuracy who chose his specimens,carefully forthe learned anatomist, no substitute posed stances (e.g., an arched cat skeleton facing off against a them in true-to-life growlingdog skeleton),and could vouch for everydrawn line as well as every printedword: "The actionsof all the skeletonsboth human and comparative,as The ImageofObjectivity 93

well as the attitudesof everybone, were my own choice: and where particular there I expressed on account of the anatomy, parts needed to be more distinctly in always directed; sometimes the drawingswiththe pencil, and oftenwiththe needle upon the copperplate, and where the anatomistdoes not take thiscare, he will scarce have this work well performed."22 The camera obscura, like the took in that its the nineteenth largely place photograph century, guaranteed an but almost effortless atlases accuracy, eighteenth-century required more than mere accuracy.What was portrayedwas as importantas how it was portrayed, and atlas makers had to exercisejudgment in both cases, even as theytried to withgrids,measurement, eliminatethe wayward or the judgmentsof theirartists camera obscura. Atlases of "characteristic" images mightbe seen as a hybridof the idealizing and naturalizingmodes: althoughan individualobject (ratherthan an imagined composite or correctedideal) is depicted,it is made to stand fora whole class of similarobjects. It is no accident thatpathologicalatlases were among the first to for of neitherthe Typus the "pure phenomenon" nor use characteristic images, the ideal, withitsvenerableassociationswithhealthand normality,23 could propNor could the diseased black-and-white erlyencompass organ. engravings:Cruto the necessity of new dimensionsof veilhier'sexquisitelycolored plates testify in depictingthe pathological.His as well as of greaterspecificity, representation, subjects are individuated, poignantly so under the circumstances-"Benoit
E R x --;.X... ... ... ...::

FIGURE 4. Title page illustration of WilliamCheselden, Osteographia; or, TheAnatomy Bones(London, 1733), showingan artist seated beforea ofthe camera obscura drawinga half-skeleton (whichis suspended upside down, since the camera obscura image is inverted).The camera obscura, the camera lucida, tracing,and othermechanicalmeans of renderingscientific images show thatthe photographwas an innovationin degree, not kind.

94 94

REPRESENTATIONS REPRESENTATIONS

(Esther), laundress, 25 years old"-and Cruveilhieris voluble on the clinical details surroundingeach sad case. Althoughhe is, as we have seen, waryof systems and persuaded that picturescan do what words cannot, he does not share the viewof lateratlas makersthatthe textshould be limitedto tersecaptions,and tabu. Cruveilhiernot onlyinfersdiagnoses from thatall speculationwas afortiori visiblesigns; he does not scruple to hazard conjectureson the basis of his observations,preservedin his figures:"Several factslead me to believe thatthe inflammation of the lymphaticvessels is primitive,and that that of the synovial These broader claims make sense [membrane]could onlybe the consequence."24 of a whole class is characteristic case and discussed if individual the figured only his figures forall his warningsagainstsystems, of such ailments.For Cruveilhier, are evidence musteredin the serviceof explanationsand causal conjectures,not brutevisual factsshorn of text,much less theory. Atlases of characteristic images presentedindividualcases as exemplaryand of broader classes and causal processes. For example, B. A. Morel's illustrative de l'espece etmorales humaine Traite desdegenerescences intellectuelles, (1857) physiques, insistedthat constantcauses "tend to create typesof a determinateform,"and that these pathologies would prove as "distinctive, fixed, and invariable" as could be subsumed Thus even normal types. by type: two young girls people of theirphyssimilarin heightand symptoms were "so perfectly (each illustrated) and intellectualstate,that one common descriptionis equally ical constitution Individual depictionbyno means precluded essenapplicable to both of them."25 tialisttypologies,even in pathology.Botanists,zoologists,and paleontologists the characteristic individual that stood for an entire species in institutionalized individualdiscoveredof thatspecies: "By the the typespecimen,usuallythe first 'type,'we understand thatexample of any natural group whichpossesses all the leading charactersof thatgroup."26 Even averaging,withits emphasis upon precise measurementof individual Gottlieb Gluge, professor objects,could be made to servetheends of essentialism. of anatomyin Brusselsand discipleof thestatistician Adolphe Quetelet,delivered a paean to measurementin his AtlasofPathological (1853), inveighing Histology of and set about the errors estimation by eye, weighing organs to the against his hundredthof a gram. As usual in characteristic atlases, subjectswere individand his measurementsindividuuated by description("Male. Baker. Suicide"), of even normal organ size. ated them stillfurther, by displayingthe variability But also as usual in characteristic atlases,and in complete accordance withQuethese individuals nonetheless telet's own brand of statisticalessentialism,27 to an underlyingtypeof which theywere characterpointed beyond variability to deviations:"Already, fromthe fewinvestigations of this strict limits istic, setting the mostfrequent kind whichhave been made, an average is presentedindicating variationsof the disease."28 atlases of the mid nineteenthcenturymark a transition The characteristic The ImageofObjectivity 95

between the earlier atlases that had sought truthto nature in the unabashed ideal, characteristic depiction of the typical-be it Typus, exemplar,or averageand the later atlases thatsought truthto nature throughmechanicalobjectivity. Like the latter, the characteristic atlases presented figuresof actual individuals, not of types or ideals that had not and/or could not be observed in a single instance. But like the former,these individuals stillembodied types of whose the atlas makerwas firmly convinced.The typical mustnow be instantiated reality in the individual,but the typicalnonethelessexists,to be discerned byjudgment and long acquaintance with the phenomena. Like the atlas makers of the late nineteenthcentury, the makers of characteristic atlases voiced uneasiness about but theyalso expressed confidence the baleful effects of hypothesis and system, thatimages would suffice to fendoffsuch distortions, and saw no reason to subject In effect, themselvesto a ban on interpretation. theyrecognizedthe existenceof an enemywithin, but theywere not yetsufficiently alarmed to combat it withthe asceticismof noninterventionist objectivity. However,later atlas makerswere considerablymore anxious about the subConflicts between truthto type and jectivityimplicitin judgments of typicality. truthto the individualspecimenbroughtthisnew anxietyoverjudgmentintothe illustrator of Victorianfloras,wryly open. For example, WalterFitch,the prolific warned of the dangers of excessive accuracy in an 1869 article addressed to would-be botanical artists: variation inform tothegreat iftheartist render presented Owing bysome[orchid] species, he isliabletohavehisveracity calledinto correctly anyspecimen putinhishands, question, he had better and ifanyabnormal comehisway, notbe rashenoughto represent growth as some who what be hisspecialty. by authority hasmadeOrchids may regarded impossible It might or possibly a petgenus-an actof tendto upsetsomefavourite theory, destroy a proper wanton which no artist endowed with forthedictaof men impertinence respect ofscience wouldeverbe wilfully of!29 guilty Fitch'srueful advice was more than the artist's revenge against the overbearing who had for generationspeered over theirillustrasupervisionof the scientists, tors' shoulders in schoolmasterishfashion. He also pointed an accusing finger themselveshad begun to suspect their worst enemies just where the scientists within themselves. lurked, namely Long accustomed to monitorthe vagaries of theirartists, the atlas makershad begun to scrutinize themselves. Fitch knew how easily a reasonable concern with typicality (he himself of their normal "some structure" was essen[orchids'] general knowledge thought could degenerate intoan unreasontial to drawingthese polymorphousflowers) In illustrations a transition such as figure5, he registers able partiality. between to obligatoryrestraint fromsuch judgments. judgments of typicality obligatory In 1851, his patrons,the botanists were stillcapable of Joseph Hooker,pereetfils, from of "characteristic" "faithful" Rhododendron arboreum, figures distinguishing both.30 the and while sm., By 1869, however, characteristic, aforrecommending
96
REPRESENTATIONS

Hoke' dains

ee

cualycmpste

f evrl

lssmsi

sgese

by hisreark aproos of ths figure "It [thespecies]seems tobe shyo

Hook. lu.,from Joseph Dalton Hooker, 5. Rhododendron argentum, (2nd ed., London, 1849). That ofSikkim-Himalaya TheRhododendrons Hooker's drawingswere actuallycompositesof severalblossomsis suggested by his remarkapropos of thisfigure:"It [the species] seems to be shyof that flowering, thisseason at least (1848); foritwas [onlywith]difficulty specimensto completemydrawing."In the laterhalf I could procure sufficient
FIGURE

Objectivity The Image ImageofObjectivity

97

tiori the ideal, renderingcould become an object of scientific suspicion,opening the as Fitch of bias and distortion. up, only half-jokingly suggested, possibility in this same a to take dim view of artists Similarly, paleontologistsbegan period who "made up the great imperfections of the type-specimen withan ideal representation."31 This is the climatein whichimages of individualscame to be preferred to those of types,and in which techniques of mechanical reproduction salvationfromtheirown worstselves. seemed to promise scientists

Objectivityand Mechanical Reproduction it mightbe thoughtthatthe shiftfromideal types On first reflection, of the introduction of photography. to individual depiction came about because above a of a For "straight" is, all, signature photography particularscene, a specificand localized representation only awkwardly adaptable to a mosaic compoindividuals. But as we have noted, resistanceagainst the sition from different or ideal began long before photographic representationof an abstractedTypus in the pages of medical atlases afterthe 1870s. Even when evidence proliferated it by no means stabilized the the photograph dominated atlas representation, the debate over objectivity. Quite contrary-photographicdepictionentered the camera obscura drawings, with X-rays, lithographs, photoengravings, frayalong and ground glass tracingsas attempts-never whollysuccessful-to extirpate selecbetween object and representation.Interpretation, human intervention all to and itself came as appear subjectivetemptations tivity, artistry, judgment requiringmechanicalor procedural safeguards. of using the authors'invective Once again we can use our strategy againstthe and its complement, senses of subjectivity subjective to unravel the different of how the scientist the the we discuss artist, First, subjectivity objectivity. to and the the means mechanical artist, among atlas police growingshift deployed as well as surveillanceof the artist.We next use makers toward self-surveillance the question of the photograph; we argue the great X-raydebate to face directly that while photographyplayed a centralrole in the continuingdevelopmentof the debate over how to it neithercreated nor terminated mechanical objectivity, of mechanical objecthe establishment we come to full-fledged depict. Finally, What we findis thatthe image, as as the ideal of scientific representation. tivity tied to a relentlesssearch to replace is inextricably standard bearer of objectivity, individual volition and discretion in depiction by the invariable routines of mechanical reproduction.This mechanizingimpulse is at once a part of the disof moral vision; the two were course of natural philosophy and a constituent inseparable. Nothing in the works of Albinus or his contemporaries quite

98

REPRESENTATIONS

I~

4.. ....

withthe auxiliarymechanismsof dissection(pins, nails,ropes, ties,and blocks), he achieved an effect thatshocked readers in the late nineteenth and early centuries.These two printsare fromhis Anatomia humani twentieth corporis (Amsterdam,1685), plates 30 (back dissection)and 87 (skeletonwithhourglass). moralism that animates the lateprepares us for the fervorof self-denying briefsformechanized scientific nineteenth-century representation. Consider William Anderson's 1885 "Introductory Address" to the Medical and PhysicalSocietyof St. Thomas's Hospital. His subjectwas the history of the relationof art to medical science,and his message was clear: medicineno longer could employ the great artists of the ages, as Vesalius had, but theirserviceshad in any case become obsolete. Scientific understandinghad not only made artistic it had shown thatthe artistcould prove to be a liability. insightsupererogatory, The seventeenth-century AmsterdamanatomistGodfried Bidloo, for example,

The Image ofObjectivity 99 Imageof Objectivity 99 The

both for art and science, but the man who struckAnderson as "too naturalistic was usually almost Zolaesque in his superfluousrealism could not always resist to pictorialallegory"(see figs.6 and 7).32If even Bidloo had fallen the temptation what was needed to restrainthose impulses was a machine prey to temptation, and forcibly thatwould automatically exclude such imposed meaning.John Bell, and Muscles Anderson granted artistic the to whose 1810 Anatomy Bones, Joints, of "he was above a man of because all was saved science, and as he did not merit, of to sacrifice care risk any accuracy by trustingthe unaided eye of the draughtsman,he had each specimendrawn under the camera obscura."33 Drawing closer to the present,Anderson creditedartwithan unprecedented artfromthefineartsof the past into but he did so onlybyredefining importance, schematic form of illustration, diagram, photograph,or model. In thisnew any aids could "serve as a new language that speaks withstrength sense, the artistic or spoken wordswould conveytheirmeaning slowly and clearnesswhere written was but This new,science-directed and imperfectly."34 art,of whichphotography and byso itssubject"withthe eye of the understanding," a part,would scrutinize of anatomicalor pathdoing might"provide us witha more useful presentation ological factsthan we could hope to gain fromthe pencil of Botticelli."35 of the past, according to The secretto the displacementof the titanicartists of in the the control Anderson, lay representational process itselfby automatic be avoided, whetherthese temptameans. Only in this way could "temptation" of tionsoriginated(as in Bidloo's case) in artistry or, as in othercases, in systems art: and would over mechanization could triumph thought.In an age of science, but the modern or Berrettini, We have no Lionardo [sic]de Vinci,Calcar,Fialetti, inartistic allthat helacks oftheneedsofscience makes up incomprehension draughtsman of as those of the as effective broadsheets no boast We can Vesal,or engravings genius. that newprocesses butwe are abletoemploy evenoftheplatesofBidlooand Cheselden, that error and others without oftheoriginal thedrawings ofinterpretation, object reproduce effects ofcolourat small useful expense.36 giveus very Such photomechanicaleliminationof the engravercut one handworkerout of the reproduction cycle and therefore,Anderson believed, contributedto the realistic even slavishly eradicationof interpretation. ones, agreed thatthe Artists, an artists'verypresence meant that images were mediated.Jules Champfleury, in for the realist movement France, ally of Gustave Courbet and spokesman insistedthat"thereproductionof naturebyman willneverbe a reproductionand ... since man is not a machineand is incabut alwaysan interpretation imitation, Of course Champfleurylauded the of rendering objects mechanically."37 pable lambastedit. where Anderson artist's intervention, interpretive bythe partialapplicationof photographictechnology Policingof subjectivity even where the was widespread in the last decades of the nineteenthcentury, actual use of photographs in the album was impractical,too expensive, too
100
REPRESENTATIONS

detailed. For example, a quite common use of the detailed, or even insufficiently Typicalof photographwas to interposeitin the drawingstage of representation. was the careful selectionof photographsby the authors ofJohnsuch a strategy AtlasofBonesandLigaments. ston's Students' Only aftersuch a selectiondid theyturn the traced who artist an to the image over photograph as the basis for the final E. Ponfick's magnum opus of 1901, an atlas on medical surSimilarly, drawing.38 rules had confined gical diagnostics,reassured the reader thatboth he and strict the artist'sactions. Outlines of organs were recorded on a plate of milk-glass to transparent mounted over the body; the image on glasswas transferred paper; on inscribed was the fromthe transparentpaper paper destined for the image observedthe workof the artist fullwatercolorpainting."As I [Ponfick] constantly the copy of colours the and distances the and carefully, comparing remeasuring of every withthose of the originalsection,I can justlyvouch forthe correctness Along the same lines,JohannesSabotta,one of the greatGerman anatomists advertisedthe use of photographyin the preparation of the turnof the century, of his atlas-even though the images themselveswere multicolorlithographs. "No woodcuts have been employed,since the failureof the lattermethod to proshown by several of the newer true to life has been distinctly duce illustrations of the woodanatomical atlases. It leaves entirelytoo much to the discretion of method the whereas reproduction depends photomechanical engraver, entirely upon the impressionmade upon the photographicplate by the original power of the illustrator, drawing." As a furthercontrol on the discretionary taken and enlarged to section Sabotta had a photograph of the designatedbody the size of the intended drawing.40 Sabotta followed much the same method when he turned to histologyand microscopicanatomy in his 1902 treatiseon that subject. Readers mightworry of livingtissue; the doctor reassured that the samples were not representative two came from themthatthe vast majority hanged men, severalothersfromtwo of the gallows,so the "material"was still"fresh."It seems that additional victims hours some two-and-a-half the corpses were stillwarmwithlife(nochlebenswarm) afterdeath. Again Sabotta had photographsmade thatwere used as the basis for should not be drawings. Here, however,he noted that precision (Genauigkeit) accidentalfeatureof the preparation pushed too far-for then everydisturbing would enter the representation.Instead, some figureswere actuallymade from two or threedifferent perhaps anticipating preparations.Somewhatdefensively, his readers thatthe combinationwas not made arbireassured Sabotta criticism, of the camera to eliminatevariationin but withthe carefulrepositioning trarily perspective,and the photographicenlargementswere cut and reassembled to reproduce a mosaic photograph against which the drawing would be judged. no possiThis, the author tells us, "would give the draughtsman[demZeichner] forsubjectivealterations."41 bility The ImageofObjectivity 101
line."39

the Typus, thus crossed the categoriesof the characteristic, Sabotta's strategy and the ideal. By invokingspecificphotographsas controlson the mechanicsof routeto the charglance to followthewell-worn reproduction,he appears at first detail meant to stand in for the acteristic-the individual depicted in striking class. But by amalgamatingfractionalpartsof different microscopicindividuals to constructthe basis fromwhich drawingswould be made, Sabotta leaves the Is the finaldrawingmade fromthe mosaic domain of the purelycharacteristic. an ideal-the pictureof a perfectsample one mighthope one day to find?Is it a a kindof limiting case? pictureof an ideal thatmightwellnot existbut represents Or does Sabotta expect his routinizedprocedures to give rise to diagrams that would stand in for a Typus, lyingaltogetheroutside the collectionof individuals and future, yetexpressingan essentialelementof all of them?Such past,present, is devoted,instead,to are pushed aside; Sabotta'sattention ontologicalquestions the procedure of controlledreproductionas a means of squelchingthe subjecof interpretation. For Goethe, Albinus,and Hunter,the atlas maker bore tivity to resolve-one way or another-the problemof how an essential responsibility an entireclass of natural phenomena. Sabotta's could exemplify single pictures cobbled photographs form an apt metaphor for his uneasy authorial position

FIGURE

criminal. 8. The synthetic By superimposingseveral projectedimages of malefactors, FrancisGalton hoped to achieve a typethatwould be produced independentlyof any artist's subjectiveimpulses. From Galton, Nature18 "Composite Portraits," 97. (1878): 97-100,

102

REPRESENTATIONS

to standaside, to keep betweenthe older desire to perfectand the newerstricture and hands offthe machine-generatedimage. large this fear of interpreBy fromtheamalgamatedimage towardtheindividual;thevery tationfueled a flight act of combining elements from differentindividuals appeared to lateobserversto leave fartoo muchjudgment to the artist. Some, nineteenth-century however,would not give up so easily. Francis Galton shared none of Sabotta's uncomfortableacceptance of the need foramalgamation.Galton,in colslightly embraced the possibilityof laboration with Herbert Spencer, enthusiastically judgment and of capturing,in one image, the vivid simultaneouslyeliminating to exploitphysimage of a group. Indeed, Galton was persuaded thatall attempts failurewithout to were doomed to iognomy grasp underlyinggroup proclivities was His a mechanized abstracting simple. Each remedy disarmingly procedure. on transdrawn member of the group to be synthesizedhad his or her picture parent paper. By exposing a photographicplate to each of these images, a composite image would arise. Such a process would free the synthesisfrom the vagaries of individual distortion;even the time of plate exposure given to each individual could be adjusted on scientific grounds,such as degree of relatedness writesGalton, in the case of family averages. "A compositeportrait," of themind's thatwouldrisebefore thepicture eyeofa manwhohad thegift represents evenofthehighest in an exalted power degree.Buttheimaginative imagination pictorial and is so aptto be biasedbyspecialcasesthatmayhavestruck artists is farfrom precise, of the forms. The merit thatno twoartists theirfancies, agree in anyof theirtypical to no errors is itsmechanical beyond beingsubject precision, composite photographic to all photographic thoseincidental productions42 the artist's than merelyconstraining Galton'swas a scheme thatwould go further depiction of an individual; the device would remove the process of abstraction from the artist'spen. No longer would even patternrecognitionbe leftto the artists.Murderers or violentrobberscould, for example, be broughtinto focus so that the archetypicalkiller could appear before our eyes (see fig. 8). The problem ofjudgment, forsomeone like Galton,arose withthe artists. for"subjective alterations," Policingthe artists-containingtheirpredilection or artistic ... "discretion," realism," superfluous "Zolaesque "pictorialallegory," of a vastlymore momentin the construction "bias" by "fancy"-was but the first Indeed, what characterizedthe creation of lateencompassing set of restraints. a form of selfnineteenth-century pictorial objectivismwas self-surveillance, For in thisperiod, the scientific controlat once moral and natural-philosophical. as a means of hemmingin theirown authors came to see mechanical registration aesthetic to norms, hypotheses,language, even impose systems, temptation Whatbegan as a policing anthropomorphicelementson pictorialrepresentation.

The ImageofObjectivity 103

directedboth of othersnow broadened into a moral injunctionforthe scientists, at themselves.Sometimes the control of the scientists' at others and reflexively subordinatescould be accomplished routinelyby invokingthe "personal equatermthatwould be used to adjust each workerror-correction tion,"a systematic forinstance,one needed to record the precise timeat er's results.In astronomy, whicha star or planet crossed a wirein a viewingdevice. This was accomplished bypressinga button.But the procedure was more complicatedthanitlooked, for different willshowthatthiswillrequire of character a very periodsof knowledge slight be but there will in of a second be but a fraction It will fordifferent time case, any people. manwho theeager, between a constant a distinct difference, difference, impulsive quick, and the whenhe seesstarand wiretogether, theinstant as itwere, habitually anticipates, that the contact has till he is sure waits man who slow-and-sure carefully quite phlegmatic, it. These differences are so truly records and firmly takenplace and thendeliberately and after a given forthem, to correct thatit is quitepossible personalto the observer times to thoseofsomestandard to reducehistransit habithas becomeknown, observer's observer43 own proclivity to bythe scientist's Adjustingforthe more subtleinterference aesthetics,or theories was a more complex affair.But impose interpretation, of the attemptabound, both in machine-dominatedrepresentational examples schemes that used some typeof photographyin one fashionor another,and in those that did not. The opthalmoscope, for example, provided the basis for a whole genre of atlases of the eye. One rathertypicalone, by Hermann Pagenand extraordinary stecherand Carl Centus in 1875, exhibits clearlythe necessity "The authorshave endeavoured,in these [pictures], of self-surveillance: difficulty as possible. It cannot be hoped thattheyhave to representthe object as naturally in this succeeded attempt:they are but too conscious, how often in its always of the investigatorhas delineation the subjective idea [subjective Anschauung] escaped his hand."44Or elsewhere, before theconditions itpurely havekept them, authors] only describing objective, They[the of prevailing ownviews and theinfluence itboththeir to excludefrom and endeavoring and and to add theoretical It wouldhavebeen easyto extendit considerably, theories. a to be this considered the authors but avoided,if conclusions; carefully thing practical to the readerthe value and to preserve their workwas to possessmorethana passing view and unbiased ofunprejudiced judgment45 advantages Consider some of the waysin whichthe photograph-made by visiblelight or X-ray-was deployed. In his microscopicstudiesof nervecells (1896), M. Allen "The method of Golgi has described the object under investigation: Starr first its branches and subbranches of shown that each cell is an independent entity, ... with fromoriginto ending,interlacing theiridentity bothvarietiespreserving those of other cells, as the branches of trees in a forestmay interlace,but really and separable fromeach other as are those treeswiththeirtwigsand as distinct leaves." He then pointed to the inadequacy of artisticportrayal:"In the most
104
REPRESENTATIONS

FIGURE 9. Skull X-rayfromRudolf Grashey, Rontgenbilder Typische vomnormalen Menschen classi(Munich, 1939). Grasheytransferred ficationfromauthor to reader by publishinga seriesof "wanted thatillustrated the farreaches of the normal posters"(Steckbriefe) and thereby demarcated the normal fromthe pathological.

The ImageofObjectivity 105

recent text-booksof neurology and in the atlas of Golgi these factshave been shownbydrawingsand diagrams.But all such drawingsare necessarily imperfect It has seemed to me, therefore, and involvea personal elementof interpretation. thata series of photographspresentingthe actual appearance of neurons under but also of serviceto students."46 the microscopewould be not onlyof interest By Starr the virtue to eliminate very expunged "personal interpretation," striving thatsomeone like Albinus had lauded in the use of drawingsand diagrams. formthe drawingwas never seen as a competitive In X-rayphotography, invisibleto the radiogram or "skiagram"took centerstage as a new visual reality, human draftsmenin a way thateven cells under a microscopewere not. By the but also turnof the century, (mostwidelyin Germany, X-rayatlases proliferated in the United States, France, and Britain) as is evidentin the famous tomes by and Typische Rudolf Grashey,Chirurgisch-Pathologische Rintgenbilder Rintgenbilder Menschen vomnormalen (fig. 9), the latterof which went through six printings between 1905 and 1939 and whichcontinuesto be a standardreferenceworkin the field. Like his colleagues in anatomy,Grashey signaled his aversion to the avoided artistic artistic aids; in those few early in his volume: "I have vigorously on the cases where, because of the uneven coveringof the emulsion [Deckung] I had be added have to so a visible contours few afterwards, explicitly negative, of artistic aids was onlythe But Grashey'scaution beforethe pitfalls indicated."47 analysisof the precautionsnecessaryin beginningof a rathermore sophisticated the use of the X-ray to produce reliable images. By 1905, when Grashey was itwas clear thattherewere systemcompletinghis workon normalRontgenbilder, atic mismatches between macroscopic anatomy and the X-ray image of the human body.There were elementsof thebody thatdid not produce image traces elementson the X-raythatdid not and therewere representational on the X-ray, knife.As a result, under the anatomist's characteristics identifiable to correspond the diagnostician had to learn-through a study of an atlas such as this-to qualify the mechanical procedure of X-rayingwith a knowledge of systematic deviationsbetween anatomyand its Rontgen representation. Secondly,the atlas the observer for the to of the could, by prepare multiplication examples, help of the X-raytube from a movement that resulted of enormous variation image Such a displacementof the camera, or a rotationof the body partunder scrutiny. tube, or body could easily make certain contours disappear and other ones forminto twodimenappear. Third, bycollapsinga complex,three-dimensional mislead. could itself sions, the projective process easily Macroscopic anatomy alone cannot rescue us here; onlythe systematic Rontgenphotographyof a skeleton, in which parts are marked with metal tags, can reveal the characteristic distortions photographicprocess. produced in the dimension-reducing of using Grashey points to the immense difficulty Finally,and most subtly, normal from the individual photographs to demarcate the pathological. The

106

REPRESENTATIONS

to themechanicalregistration as was Grashey, problemis this:ifone is committed, of images of individuals, then how would one distinguishbetween variations and thattransgressed the bounds of the "normal"and variations within normalcy of the pathological?Grashey'sown solutionwas to eleentered into the territory Abweiselteneren vate the most strikingof such rare deviations (augenfalligen the in honor of to a They would Rontgen laboratory. (Ehrenplatz) place chungen) thenserve as boundarypostsof the normal,guidingthe diagnosticianawayfrom of pathologywhen a patientarrivedwitha subjectivedifficulty. a falseattribution By 1900, the metaphysicalposition underlyingGrashey'sview was widespread thathad prevailedin the early fromthe implicit different and utterly metaphysics did representsomenineteenthcentury.For Goethe, the depiction of the Typus that this or from not nature in individual). For Albinus, apparent (though thing of a subjectreferredto nature not onlybecause it borthe "true"representation rowed fromseveral individualsbut because it improvedabove any single one of them. For Hunter, the link to the general occurred through a particularindividual, chosen preciselythat it mightrepresent(in both senses) a whole class. as theywere, all threeviewstook it for grantedthata single represenDifferent tationcould stand in forthe myriadof variationsfound in nature. Grasheyand his contemporariesdisagreed. For them,the link to the multibe it ideal, tude of variantscould not be contained in any single representation, was to serve as a could do a the most Instead, picture typical,or characteristic. stands signpost,announcing thatthisor thatindividualanatomicalconfiguration in the domain of the normal. Many such instanceswere needed to convey the extentof the normal,as the normalspanned a space thateven in principlecould from the rest. each differing not be exhausted by individual representations, Thus when W. Gentner,H. Maier-Leibniz,and W. Bothe published theirAtlasof Chamber theyincluded multipleexamples of alpha Photographs,48 Expansion Typical fromdifferent substances,and particlesionizing a gas, beta particlesscattering it was Each electrons individual, (see 10). hoped, would fig. positronsannihilating evoke a class of patternsin the mind of the reader. This is the essential point. was supposed theburden of representation While in theearlynineteenth century, of pattern now it fellto the audience. The psychology to lie in the pictureitself, claims of the author. recognitionin the audience had replaced the metaphysical fear of their themselves, by transferring subjectivity theyassuaged Mistrusting ofjudgment to the audience. the necessity faced the same problem.It seems thatat the turnof the twenPaleontologists tieth century,students of British graptoliteswere confounded by conflicting descriptionsof theirfossils.Accordingto one expert,the problemwas thatmany of the best figureswere "more or less unconsciouslyidealised" or inadequate in some other way. Better reproduction(see fig. 11) would defer responsibility by the interpretive eye to the reader. It was, one author wrote, shifting

The ImageofObjectivity 107

FIGURE 10. Cloud chamber photographof an electronlosingenergyin a gas as

in the it spiralswithina magneticfield.Because the magneticfieldis stronger drift centerof the cloud chamber,the spiralsslowly upwards. The reader is in the spacing of the spirals(attributed to instructed to note 1) irregularities multiplecollisionsof the electronwithgas atoms) and 2) the sudden change in to the emissionof a photon orbit(attributed curvaturearound the seventeenth by the electron). Readers were supposed to studythese images and so to learn Source: W. Gentner,H. Maierfromthe ordinary. to separate the extraordinary Chamber Leibnitz,and W. Bothe, An AtlasofTypical (New Photographs Expansion York, 1954), 51. the paleontologists FIGURE 11. Graptolite.Like Grashey, hoped to shift fromauthor to reader. By photographinggraptolites, theyhoped responsibility to avoid the "unconscious"alterationof images imposed bythe authorFrom Gertrude Lillian Elles and Ethel M. R. Wood, A Monograph illustrator. of British (London, 1901), detail frompl. 4. Graptolites
REPRESENTATIONS

108

of the fossilsthemselves-by mechanicalmeans if obviouslynecessaryto give such figures as possible-as should agree withthe originalsin all respects,showingtheirimperfections of thatthe reader mightbe in a positiontojudge of the fidelity well as theirperfections, and mightalso be able, should the need arise, the descriptionsby the figuresthemselves, the actual fossilor typespecimenrepresentedon the plates.49 to identify Casting a brief glance forward in time, we see that mid-twentieth-centuryatlas makers took exactly this path. Consider the frank admission of the necessity of the capacity to recognize patterns in this excerpt from the preface to an atlas of electroencephalography published in 1941: in the hope thatit willhelp the reader to see at a glance what This book has been written it has takenothersmanyhours to find,thatitwillhelp to trainhis eye so thathe can arrive at diagnoses from subjectivecriteria.Where complex patternsmust be analyzed, such criteriaare exceedinglyserviceable.For example, althoughit is possible to tellan Eskimo froman Indian by the mathematical relationshipbetweencertainbody measurementsat than can be obtained fromany a glance and can often arrive at a betterdifferentiation a of indices. It would be wrong,however,to index even from or group quantitative single disparage the use of indices and objectivemeasurements;theyare useful and should be employed whereverpossible. But a "seeing eye" which comes fromcomplete familiarity can whichan electroencephalographer withthe materialis the mostvaluable instrument competentuntilhe has acquired it.50 possess; no one can be truly For the electroencephalographer, the acquisition of the "seeing eye" permitted the recognition of distillation of the pathological from the normal. For the particle physicist, exploiting the cloud chamber or the emulsion method, the seeing eye separated the novel from the known. P. M. S. Blackett put it succinctly in the foreword to one of the most successful of the particle atlases, published in 1952: An importantstep in any investigation using [the visual techniques]is the interpretation to recognise of a photograph,oftenof a complex photograph,and thisinvolvesthe ability of events. To skill in a sub-atomic different acquire interpretation, many types quickly kinds preliminary studymustbe made of manyexamples of photographsof the different of knownevents.Only when all knowntypesof eventcan be recognisedwillthe hitherto unknownbe detected.5' Neither electroencephalographer nor particle physicist could simply point to a picture and instruct the reader to find identical occurrences in the pictorial output of their own instruments. Earlier generations of atlas makers chose "truth to nature" as their slogan: their pictures would depict the designated phenomena as they were, as they ought to be, or as they existed beneath the variation of mere appearances. Byt the late nineteenth century,however, the atlas makers no longer could make such unproblematic claims for the general applicability of their images, and by the early twentieth century, they had shifted responsibility to the reader. Caught between the infinite complexity of variation and their commitment to the representation of individuals, the authors must cede to the psychological. The Image of Objectivity 109

Selectionand distillation, among the atlas writer's previously principletasks,now were removed fromthe authorial domain and laid squarelyin thatof the audiof individualrepresentation ence. Such a solutionpreservedthe purity at the cost of acknowledgingthe essentialroleof the readers' response: the human capacity to renderjudgment, the electroencephalographers allow,is "exceedcheerfully inglyserviceable."For Grashey,the problemoccurs in the shadows of bone, not but the weightof nature'sdiversity is similarly ink tracings, felt:"One mustknow these variations,"Grashey insisted. "We need an all-pointsbulletin issued for them. A series of pictures in this atlas is devoted in part to spreading widely forthem."52 wanted posters[Steckbrief] Grashey'spolice metaphorwas entirely appropriate.Not onlywas the history bound up withthe history of of late-nineteenth-century photography thoroughly its into crime control,the X-rayphotograph itselfwas increasingly finding way court.53 As it did, the difficulties evidence and legal evisurroundingscientific the problemswere diagnostic, and could be attacked dence merged. For Grashey, with compensating techniques like any other form of scientific instrumental instrumental error.For others,however, problemsrapidlyexploded intojuridical disputes with profound professionaland pecuniaryconsequences for the cliniborder betweenjudgment cian using them. At issue was, once again, the shifting of human intervention and mechanization,betweenthe possibility (or necessity) of the technology. The disputes over and the routinized,automaticfunctioning that new rather evidence show shifted photomechanical techniques photographic Of all the audiences who than eliminated the suspected sources of subjectivity. addressed the "medico-legal"concept of evidence, perhaps the mostactive (and distressed)was the assemblyof clinicalsurgeons,who saw in the new X-rayphotographya potentiallegal weapon thatcould be turnedagainstthemin malpractice suits. The resultingfracas,recorded in (among other places) the American Journal of photographs as characteristics captures the distorting of theMedical Sciences, Above all, critics evidence in both senses-legal and scientific. challenged the vullocation of in the relative the camera, the Xof to the nerability image changes A Dr. Ames that "I have under lamented the and object investigation. ray tube, and we cannot alwaysbelieve to mysorrowlearned thatthe rayhas manytricks, what we see, or ratherfail to see, and a picture,to tell the truth,must have the plate, the object to be photographed,and the tube in proper relationduring the exposure."54Insistingon the same point,a New York doctor wanted swornwitand patient.55 nesses in court to attestto the exact placementof the instruments and legal evidence is reiterated characterof scientific The interwoven byanother doctor, who warned his colleagues that theyshould not forgetthat "the X-ray or negligencein fastening the plate and makingthe operator eitherbywilfulness and an unprejudiced artist exaggerate any existing deformity exposure may be insisted should upon."56
110 REPRESENTATIONS

For some of the clinicians,the veryformof X-rayphotographywas a threat radiated authority, even whilepractitiobecause the photographicmedium fairly of Xconfrontedits deceptiveness.In part the ability ners of the art frequently not on medium an could bestowed the vision where to ordinary rays penetrate aura of superhuman power.But in addition,byitsverynature,X-raytechnology was parasitic on the widespread assumption that the photograph does not lie. How could it,designed as itwas to eschewthe dangers of subjectiveintervention? So while a moderate clinicianmightwant to use the new device as a supplement to other practices,the image of the X-rayappeared (in courtat least) to preempt and displace all other formsof knowledge. One doctor commented plaintively are not identical.In a thingwhichpurportsto be that"usefulnessand infallibility a representation analogous to a photograph, showing only what exists and is sure to be made bysome of exact accuracy, nothingelse, the claimof infallibility, bysomejudges and juries."57 lawyersand listenedto approvingly The doctor was right.By 1900, the photograph did wield a powerfulideological force as the very symbolof neutral, exquisitelydetailed truth.Even if there remained in the photograph an ineradicable people by then knew better, Edgar Allan Poe's homage to the daguerreotypeseems to capglow of veracity. "If we examine a workof ordinary ture the dream of such perfecttransparency: art,by means of a powerfulmicroscope,all tracesof resemblanceto nature will of the photographicdrawingdiscloses onlya disappear-but the closestscrutiny more more absolute truth, perfect identityof aspect with the thing represented."58What sustained Poe's dream of perfectidentitywas faith,not techelectronmicroscopists was nology.One of the greatestof mid-twentieth-century knowscrispnessand truthare not blunt on the subject: even when the scientist coextensive,the photographicdream is stillcompelling. forthemorphologist, thana matter ofdemonstrated itis morean article offaith Perhaps is sharp, fine-textured and that an imagewhich aesthetcoherent, fact, orderly, generally be is is more to true than one which and indistinct. coarse, likely disorderly ically pleasing of faith, this butithas proven tobe Likeothermatters maynotwithstand logical analysis oftheprogress soundand hasbeenresponsible for much that hasbeenmade operationally level.To accept in descriptive at theelectron anyother prinmicroscope cytology guiding and technical carelessness ineptitude.59 cipleis toencourage Or, as Charles Rosen and Henri Zerner put it, "It is not that a photograph has more resemblance than a handmade picture (many have much less, and what could be more likesomethingthan a successfully painted trompel'oeil?),but that .... We tend to trustthe camera more than our belief guaranteesitsauthenticity our own eyes."60 Against this deeply ingrained trust,though in different ways, both the expert and advocate Grasheyand the ratherdefensivedoctorswere in the same position: both sought to challenge the manifesttransparencyof the meaningof the X-rayphotograph.But whilethe medicswantedabove all to check The ImageofObjectivity 111

the hastyconclusions of an alleged jury of peers, Grasheyhoped to reestablish controlof the modalitiesof distortion. transparency throughthe systematic A more radical critiquethan thatadvanced bythe frightened surgeonscame fromthose who doubted thatthe X-rayphotograph,or any photographforthat could ever become a stable piece of evidence. Photographslied. And in matter, withX-rays, severaldoctorsfound it essentialto puncthe climateof infatuation ture the inflatedclaims made for the medium. One respondent pointed to the should theybe presentedwithoutinterpretation; of photomicrographs obscurity of information with takes place he added, so, X-rays.6'Here the encryption just in the technology itself: as had introduced new just photomicrography previously into pictorialrepresenvisual conventions(over and above straight photography) tation,so now X-raysdid the same. As one bluntcommentator put it: "Everybody and thata likeunstable knowsthata skiagraphmaybe easilymade to telluntruths Preciselybecause of their conclusion that veracitymay attach to skiagrams."62 not did a carry transparentmeaning,the American Surgical Assophotographs ciation unanimouslycounseled its membersto use theirmedical knowledgeand learn to read what might otherwisebe misleading. In an environmentwhere and frequently interpretationswere going to be rife, sometimes arbitrary, to their professionalexistence,surgeons had to join the extremelydangerous ranksof the expertsto defend themselves.63 For these doctors,danger lurked in the impositionof individual interpretation; the photograph promised freedomfromthe singlewill,but in and of itself was insufficiently powerfulto wrestcontrol from an individual photographer, doctor,or lawyer.The photograph,in otherwords,did not end the debate over it entered the debate. One response was to demand witnessesto the objectivity; productionof the image, another to require expertsto mediate between picture and the public, and the thirdwas to recommend that the surgeons themselves learn the techniques necessaryto eliminatetheir dependence on intermediate was less hopeful, and advocated a rejection of the readers. A fourthcriticism controls.As one doctor commented, method itselfrather than more stringent and positive indicationsis far more obtained by experience long "Knowledge visiblealone to the eye."64 valuable than any representation The clinicians'fear of evidence "visiblealone to the eye" remindsus both of betweenthe positionof researchersand docthe commonalitiesand differences made without tors. Both groups had a profound fear of a willfulinterpretation law-where the law could be laid down throughlaboratoryprocedure, solemn But whilemedical docor the physicallaws of automaticenregistration. witnesses, scientific from researchers' tors worried about deception without, anxietycentered around the more insidious threat from within.Consequently,while the doctors could stilllook to judgment as an aid against the deception of skewed the researcherscame to fear theirownjudgments as another,more danfigures, form of interpretation. gerous
112
REPRESENTATIONS

By the late 1920s, polemics against the danger of individualjudgment had reached a crescendo. Erwin Christeller, a research scientist, used his Atlas der und erkrankter to caution the scientist Histotopographie gesunder Organe againstproducing his own drawings-tempting as that mightbe.65Instead, he counseled handing the task to technicianswho could produce pictures withoutpassing mechanthroughthe stage of using a model; the procedure could be made "fully ical and as far as possible,forcibly this direct guided by reproductionprocedure of the art department."Such forcibleself-restraint fromintervention removed the possibilityof the scientist's own systematic beliefs or commitmentsfrom blocking the passage fromeye to hand. This desire to extricateeveryone,even counsel to his himself,fromthe exertionofjudgment extended to Christeller's fellowanatomists:turnover your manuscript to the publisherwithyouroriginal anatomical preparationsso the lattercan be reproduced "purelymechanically" At the same timethe scientist's controlwas necessaryto block (reinmechanisch).66

12. Tatteredtissue.By notomitting the fibrous edges of his sections,Erwin Christeller made visuallyexplicithis abstinencefromintervention-thefiguretherefore wore its so to speak, on itssleeve. This particularfigure authenticity, depicts a low-gradetumorin the passagewaybetweenthe stomachand the beginningof the small intestine (polypoid adenoma of the pylorus).Source: Christeller, Atlasder underkrankter Histotopographie gesunder Organe(Leipzig, 1927), table 39, fig.79.
FIGURE

The ImageofObjectivity 113

withthe productionof images. or ignorancefrominterfering others'inclinations "I do not want to neglect to mention that through the whole conduct of the printingprocess, I maintained continuous control of the photographers and and puttingat theirdiscolor engravers,even givingthem detailed instructions posal myown instruments."67 Once so policed, and presumablyonlythen,could the photographicprocess be elevated to a special epistemicstatus,puttingit in a categoryof its own. As Christellerput it: "It is obvious thatdrawingsand schematahave, in manycases, many virtues over those of photograms. But as means of proof and objective und objektive documentation to ground argumentation [Beweismittel Belegefur was The far are superiority photographic superior."68 Begrunde]photographs With of individual removal to the attached respect to judgment. inextricably that no method was Christeller for color, perfect:drawings, thought example, however,carried withthem an inalienable subjectivity. Photograms,by contrast, but onlybythe crudeness imposed by the limwere tarnishednot by subjectivity ited palette of the color raster.Given the choice, the author clearlyfavored the on the crude but mechanicalphotographicprocess.Accuracyhad to be sacrificed altar of objectivity. So rivetedwas Christeller bytheideologyof mechanizationthathe was deterin the photograph as a literalmark of objectivity: mined to leave imperfections orcrack bodies[such ofanyforeign oftheelimination theexception With as] dustparticles so thatthetechnically havebeen undertaken, to thereproductions lines,no corrections in some places.For example,thereare smallintrusions errorsare visible unavoidable is alsoan] on theedgeofthesections; tissue the fibrous of [there fringes [Uberschlagstellen] these ... [I displayed tissue ofsoft absence because]I believed imperfections components. of the thelimits with to display it myobligation also,at thesametime, greatobjectivity
technique.69

therole of thedeliberateand humtissueedge servedforChristeller The tattered bling fault in a Persian carpet. But while the carpet maker seeks to avoid the torntissuesamples,such as the ones Christeller's hubrisof attemptedperfection, to the disto objectivity, a as forward were in 12, testimony put displayed figure atlas was in the Their to the of self-denial presence temptation perfect. ciplined ideal. of the a rebuke to the aestheticizedimprovement Such a rejectionof subjectivetemptationpermeated atlases of the time. In AtlasoftheSpinalCord(1901), the author spared Alexander Bruce's Topographical to regularize the presentationof each picture,enlargingeach by preno effort cisely the same amount, and reproducing the photographs by photogravure. "Everycare," he insisted,"has been takento secure thateach figureshould be an of itscorrespondingsection,and it has accurate, unsophisticatedrepresentation been thought advisable to leave the Plates to speak for themselves,and not to interferewith them by lines or marks to indicate the position of cells or other could corruptan individual,so such artifacts structures."70 Justas "sophistication"
114
REPRESENTATIONS

13. Portablepolygraph.This device was designed to measure, inter alia, the pulse of the heart,the pulse of the arteries, and muscularcontraction. For E.J. Marey,these and respiration, other automatically were produced figures doublyimportant:first, the divisiveness transcended and of language, they incompleteness and second, theycaptured,without the human what interpretation, senses never could. From Marey, La Methode dans lessciences graphique 457. (Paris, 1878), experimentales
FIGURE

of custom could spoil the honest plainness of an unmolestedplate. Left to their own devices the platescould and would "speak forthemselves," untranslated and If thismeant deletingthe lines or marks untrammeledby human intervention. thatmightaid the reader in understandingthe plate and relatingwhat was seen to other samples, so be it: the cost of interference was too high to pay. For both Bruce and Christeller the search forobjectivity was not merelysurfacecommentary;it penetrated into fundamentaldecisions about laboratoryprocedure and representational strategy. The moral narrativesurroundingthis mechanical construction of pictorial took manyforms.As we have argued, pictures(properlyconstructed) objectivity served as talismanicguards against frauds and systembuilders, aesthetes and idealizers. But picture-producing instruments carried a positiveas well as negative moral weight;theycould do thingsthathumans could not, and avoid what humans could not help but do. Joined withthe virtuesof machines in general, writers like the Frenchphysiologist (see fig. Mareysaw in the imaginginstrument of the both an of ideal scientific work and a more 13) possibility realizing general ideal of a universal pictoriallanguage. Here Marey transcended the standard The ImageofObjectivity 115

atlas rhetoricabout pictures and languages by extending the mystiqueof the and graphs. But the privisual to the dense symbolicpresentationof functions instruments called of what inscription Marey (appareils inscripteurs) maryimpact of different was this:byautomatingthe registration typesof force(throughelecresearcherscould reformthe veryessence tricalcurrents, weight,temperature), research and evidence. "The graphical method translatesall these of scientific of forces into an arrestingformthat one could call the changes in the activity as it is superior to all other modes of of the phenomena themselves, language for a Such Marey,universalin twosenses. Graphical language was, expression."7" boundaries of natural languages to the artificial across could cut representation could cut across discireveal nature to all people, and graphical representation diverse as the as to boundaries pulse of a heart and capture phenomena plinary the downturnof an economy.Picturesbecame more than merelyhelpfultools; theywere the words of nature itself. Sometimes the "words" of such mathematicalpictures were warnings: in research, the false correlationof variables would be reined in by the sudden, advance unmistakableappearance of"incoherent"graphicalcurves.A legitimate would be rewarded bythe opposite-a graphicalversionof a numericallaw that Like his anatomical and "luminous"(lumineuse). would be "arresting"(saissisante) thejob of policing not contemporaries,Marey set the graphical representation to admonish the sciauthors scientific but the technicians themselves-here, only entistto go no further. As oracles speaking nature's own language, the inscriptioninstruments function. They could actuallybecome acquired a second, even more far-reaching and exact observers, "Patient the ideal observers science had always sought: blessed withsenses more numerous and more perfectthan our own, theywork by themselvesfor the edificationof science; theyaccumulate documents of an which the mind easilygrasps,makingcomparisonseasy unimpeachable fidelity, and memoryenduring."72 Echoing thesesame themes,the Frenchpopularizerof science Gaston Tissandier celebrated the mechanicalobserver,both more exact and exacting than its human counterpart.Who, he asks, could resist"a certain emotion"of awe at seeing the topographyof the moon reproduced photographexactitude.The camera would neithertireat the microwithmathematical ically, and scope, nor fail in the repetitiousbut essential readings of thermometers barometers."That which man cannot do, the machine can accomplish,"Tissanmechanicalaids had offeredassisIn the eighteenthcentury, dier proclaimed.73 be "true to nature." With Marey that would of in the tance production images the machine-madeimage replaced the variegatedideals and his contemporaries, of truthto nature witha moral order of objectiverepresentation. to emphasize thetwo-sidedgoal here quite deliberately, We use thetermmoral of using mechanizationto achieve "truthto nature."True, the rhetoricof everincreasingprecision is used to celebrate the technicalprogressionfromcamera
116
REPRESENTATIONS

lucida to photographicreproduction.But as many atlas writers indicated,phoeven or not was mechanical tochemical, usually the means always reproduction findmostsimilarto a bird,a to make an image thatreaders would automatically Burdened withdetail not found in the reader's own dissectedcorpse, or a cell.74 specimens,produced in black and white,oftenblurredto boot, therewere many cases where the photograph was unable to provide the audience with a guide What the photograph(along withtracings, byan illustrator. equal to thatoffered was a path to other mechanicaldevices) offered and smoked glass,camera lucida, truthful sort,one thatled not by precisionbut by autodepiction of a different willfromthefieldof discourse.On this the scientist's of exclusion the mation-by of resemblancewas more thanjustifiedbythe immediacyof the viewany sacrifice of machine-madeimages of naturethateliminatedthe meddlesome intervention of for this rendition The search before mere similarity. humans: authenticity was a moral,as much as a technical, quest. objectiverepresentation

ObjectivityMoralized in the serviceof truth was nominally Although mechanicalobjectivity When forced of self-restraint. to nature,its primaryallegiance was to a morality chose the often makers the atlas to choose between accuracyand moral probity, as we have seen: betterto have bad color,ragged tissueedges, and blurred latter, The discipline earlier atlas boundaries than even a suspicion of subjectivity. of truthto nature had been in the interests makers had imposed on theirartists had but deemed nature as as itself), (construed variously they judgment and Later atlas or characteristic. of the truly selectionessentialto the portrayal typical the eschewed their as of artists, makers,as fearfulof themselves typicalbecause judgment and selection were needed to detect it, and judgment and selection of interpretation. bordered on the dread subjectivity of presenting maker could atlas no Nonetheless, figdodge the responsibility the the how to reader teach the ures thatwould typical, ideal, the charrecognize the average, or the normal. To do so would have betrayedthe mission acteristic, of the atlas itself.A mere collectionof unsortedindividualspecimens,portrayed subversive.Caught would have been positively in all their intricatepeculiarity, and the scyllaof irrelevance,some atlas between the charybdisof interpretation makers worked out a precarious compromise. They would no longer present typical phenomena, or even individual phenomena characteristicof a type. Rather,theywould presenta scatterof individual phenomena thatwould stake what out the range of the normal,leavingitto thereader to accomplishintuitively to distinto an dared to do the atlas maker no longer ability explicitly: acquire guish at a glance the normal fromthe pathological,the typicalfromthe anomalous, the novel fromthe unknown. The ImageofObjectivity 117

Not only did mechanical objectivity prune the ambitionsof the atlas; it also transformedthe ideal character of the atlas maker. At the veryleast, the atlas maker of yore had been a person qualified by wide experience and soberjudgmentto selectand presentan editionof interpreted phenomena forthe guidance and so on. An exalted fewhad been of otheranatomists, botanists, entomologists, universaltruthfromflawedparticuatlas makers of genius, capable of intuiting lars, even when scientific knowledgewas meager: Choulant praised Vesalius not methodsbut also forhis "eye artistically trained,"whichled only forhis scientific As him to anatomical truthby "pursuing beauty in all the works of nature."75 bestwhen artand science self-consciously mightbe expected, the genial image fit and botanists. of anatomists But even atlas in the the as case idealizing converged, and less idealizingtendencieswere emphatically makersof lesser gifts presentin and theirworks,selectingand preparing theirspecimens,alternately flattering for the the best with publisher engravers,all bullyingtheir artists,negotiating to theirknowledgeand withthe aim of publishingatlases thatwere a testimony after their titleto authority and and all, were, judgment judgment. Knowledge a or untutored artist could scienotherwise publish any greenhorn authorship; between essentialand accidentaldetail; failure tific atlas. Failure to discriminate to amend a flawedor atypicalspecimen; failureto explain or commentupon the of an image-all of these would have been taken as signs of incomsignificance not virtuousrestraint, bythe earlieratlas makers. petence, in However,already in the earlydecades of the nineteenth centuryscientists diverse fields,and of verydiverse methodologicaland theoreticalpersuasions, especially judgment and imagibegan to fidgetuneasilyabout the perilswithin, nation. Scientistssometimessought,not alwayswithsuccess, to discipline these "innerenemies,"as Goethe called them,76 byrules of method,measurement,and the But more oftendisciplinecame fromwithin, work discipline.77 confronting It is thisinternalconflict thatimpartedto "innerenemies" on theirown territory. its high moral tone. Imaginationand judgment were susmechanical objectivity but ratherbecause theywere because theywere personaltraits, pect not primarily of lack sufficient and Moreover, disciplinepointed required discipline. "unruly" forone's own prettiest to characterflaws-self-indulgence, impatience,partiality ideas, sloth, even dishonesty-which were best corrected at their source, by even in the heat of discovery. assuming the persona of one's own sharpestcritic, The British physicistMichael Faraday described this supreme act of selfof the spirit: "The world little discipline in the language of the mortification which have passed through the and theories of the how knows many thoughts have been crushed in silence and secrecyby his mind of a scientific investigator own severe criticismand adverse examination; that in the most successful instancesnot a tenthof the suggestions,the hopes, the wishes,the preliminary was of course the or self-control conclusionshave been realized."78 Self-discipline Samuel Smiles as "the cardinal Victorian virtue,celebrated by the homelysage
118
REPRESENTATIONS

man is self-governing primaryessence of character.... The most self-reliant, the the higher will be discipline, always under discipline; and the more perfect his moral condition."Scientists among Smiles'sheroes: Farfiguredprominently aday "was a man of excitable and even fierynature; but, through high selfdiscipline,he had convertedthefireintoa centralglowand motivepower of life"; labors during the nightor in Alexander von Humboldt "pursu[ed] his scientific the early morning,when most other people were asleep"; the mark of the sucScience now and painstaking was "sedulous attention cessfulscientist industry."79 will. to a titanic demanded self-discipline, grafted the machine atlas makers of the later nineteenthcentury, For the scientific and a At once failed. will aided where the polyvalentsymbol,the powerful First,the machine was fundamentalto the veryidea of mechanical objectivity. capacityof a machine to turnout thousandsof identicalobjectslinked itwiththe standardizingmissionof the atlas,whichaimed, afterall, bothto standardizeand to reproduce phenomena. The machine also provideda new model forthe scale and perfectionto whichstandardization mightstrive.Echoes of the popular fasof manufactured cinationwiththe ubiquityand standardizedidentity goods crop folClerk of this literature in the scientific Maxwell, elsewhere period.James up disto be too similar atoms for a as them used metaphor lowingJohn Herschel, Reichsanstaltin the then-recently The Physikalisch-Technische tinguished.80 on scientific unifiedGermanysoughtto impose the same level of standardization had set forcommercialwares,81and wares thatthe Customs Agency (Zoll-Verein) all overEurope and NorthAmericaconvened to estabcommissions international Rudolf Virchow and other physicalquantities.82 lish standard units of electricity when he standardization with associated luster of cultural the some caught und Naturforscher of extolled "geistigeEinheit"to the 1871 meeting Deutscher Artzte shortlyafter German unification:"The task of the future, now that is to establishthe innerunity... externalunityof the Reich has been established, the true unificationof minds, puttingthe many members of the nation on a common intellectual footing."83 embodied a instruments, Second, the machine,in the formof new scientific ever alert,probing beyond positiveideal of the observer: patient,indefatigible, tooktheircue frompopular of the human senses. Once again, scientists the limits and machine. Charles Babbage, mathematician rhetoricon the wonder-working labor of mechanical the muse of manufacturing, advantages rhapsodized over for tasks that required endless repetition, great force,or exquisite delicacy.Sciof about the possibilities entistthat he was, Babbage was especiallyenthusiastic sevcould counteract and for to machines record, observe,measure, they using weaknesses: "One greatadvantage whichwe mayderive from eral all-too-human the idleness, is fromthecheck whichitaffords againstthe inattention, machinery admonished their or the dishonestyof human agents."84 Justas manufacturers workerswith the example of the more productive,more careful,more skilled The ImageofObjectivity 119

admonished themselveswiththe example of the more attenmachine, scientists more honestinstrument. tive,more hardworking, forour purposes, the machine,now in the form Third, and most significant held out thepromiseof imagesunconof techniquesof mechanicalreproduction, This promise was never actuallymade good-neitaminatedby interpretation. ther camera obscura nor smoked-glasstracingsnor photograph could rid the atlases of judgment altogether.Nonetheless,the scientists' continuingclaim to is testimony to the intensity of theirlonging such judgment-freerepresentation In the this context machine stood for forthe perfect, authenticity: "pure" image. freefromthe inner tempit was at once an observerand an artist, miraculously or otherwiseinterpretnature. tation to theorize, anthropomorphize,beautify, the machine What the human observercould achieve onlybyiron self-discipline, achieved willy-nilly-such,at least, was the hope, often expressed and just as and symbolic functions of the machine blur,for oftendashed. Here constitutive the machine seemed at once a means to, and symbolof, mechanical objectivity. the scientific In this last interplayof machine and objectivity, image commanded centerstage. As we have seen, mechanicalobjectivity encompassed all of the theorist and in its science experimenter trafficking injunctions, admonishing in images. However,the atlas image in words as well as the atlas makertrafficking firstas its held a privileged position in the moralityof mechanical objectivity, enforcerand then as its purest realization.Cruveilhierhad hoped that images in the air; his succesto build systems would stand watchagainst the temptation sors ruefullyacknowledged images alone were not proof against the intrusions of the subjective,but theyhoped in theirturn thatmechanically produced images scientific armamentariumof machines would be. The late-nineteenth-century that spewed out images, both of visibleobjects and of invisibleforces,was a testimonyto the atlas makers'hope and ingenuity. One typeof mechanical image, the photograph,became the emblem for all "The photograph has acquired a symobjectivity: aspects of noninterventionist bolic value, and its fine grain and evenness of detail have come to implyobjecmetaphorforobjectivetruth."85 tivity; photographicvisionhas become a primary truerto nature than handThis was not because the photographwas necessarily made images-many paintingsbore a closer resemblanceto theirsubjectmatter than early photographs,ifonlybecause theyused color-but ratherbecause the not verisimilicamera apparently eliminated human agency. Nonintervention, and this is heart of mechanical at the tude, lay whymechanically proobjectivity, duced images captured its message best. Images had always been considered more directthan words,and mechanicalimages thatcould be touted as nature's were more immediatestill.Thus images were notjust the products self-portrait of mechanical objectivity; theywere also itsprimeexemplars. withitsstrongasceticovertone,also tapped roots But mechanicalobjectivity, deeper and older than the machine age. Self-disciplinecame hard, and the
120
REPRESENTATIONS

an aura of stoic nobility. struggleagainst the inner enemies took on, explicitly, Ernest Renan, the French apostle of science, chose the language of Christian to describethecreed of themodern scientist. asceticism and self-sacrifice Praising monothe "painstaking,humble, laborious" work required to write scientific more the "sweeter and Renan flatteringto temptation, recognized graphs, of and the fruit to pluck prematurely generalization theory.It was the vanity," forthemore giftedscientist, that all the stronger of thistemptation, verystrength resistance: conferredmoral dignity upon and to denyoneself is neededto brakethatfatal inclination virtue A profound scientific clamors forthedefinitive solution. whenthewholeofhumannature that haste, headlong can forbid The heroesof scienceare thosewho,capableof a moreelevated viewpoint, of their nature ... whenall theinstincts all premature themselves thought philosophical tofly tothehighpeaks.86 them off wouldhavecarried Christianlanguage, Though others may have hesitatedbefore Renan's frankly life permeated late-nineteenthhis moral and religious view of the scientific century visions. Pledged to depict the true objects of their world, scientists of demanded of themselvesa sleepless vigilanceagainst the several temptations molds. The into and evidence preconceived pouring theorizing,aestheticizing, calls to asceticself-discipline began in the earlydecades of the nineteenth rallying centuryand grew ever more urgent. They are the birth cries of mechanical objectivity. of forms, even when confinedto Asceticismcan take on a dizzyingdiversity asceticismof medieval women the Christiantradition.87 Justas the characteristic of was stamped withtheirspecial dailyconcerns,stalwart againstthe temptations so the asceticism of nineteenth-century scientists dealt food ratherthan power,88 withthe sins peculiar to theirtribe.St. Augustine had reproached himselfwith scientists withanthrolustand pride; nineteenth-century reproached themselves of and The self-command,of heroism interpretation. language pomorphism spoken through clenched teeth and born of not boldness but its opposite, was similar.This resemblancewas not loston severalnineteenth-century remarkably and as the reliwriters:Renan trumpetedscience as the "courage de s'abstenir," described the for the modern He self-consciously plight of the selfage. gion researcherin "the phrase of the Evangelist,to lose one's soul in order to effacing betweenscience and religion"in save it."89 James Martineau detected an affinity of everything even of the veryfaculties... bywhich a common distrust internal, the externalis apprehended and received."90 never existswithoutan Sociologistsof religion tell us that moral virtuosity scientific asceticism was no excepappreciativeaudience, and nineteenth-century of unto tion.Despite the formulaic scientific self-effacement, professions humility whose celibacy, asceticism was farfrommodestin itsaims. Like the priests fasting, and vigilspurifiedthem fordirectcontactwiththe godhead and made them fit The ImageofObjectivity 121

vesselsfordivine truthand worldly of the scientists power,the self-restraint purified them for direct contactwith nature and made them fitvessels for natural truth and worldly power. Noninterventionist was the professional objectivity but it was not for scientific ethos of scientists, consumptiononly.By ringingthe cultural of on the resonant themes self-purification changes through selfabnegation, scientistspersuaded themselvesand others of their worthinessto assume priestlyfunctionsin an ever more secularized society.Sometimes this ambitionto become the new clerisywas laid bare forall to see, as in the cases of Renan and Claude Bernard in France,T. H. Huxley and JohnTyndallin Britain, and ErnstHaeckel in Germany.91 Sometimessuch perceivedarrogance triggered predictableresistance. mustbe read in lightof thisresistance.When The trope of scientific humility French physiologist Claude Bernard harped on the modestyinstilledby experiment,he was counteringcharges thata new scientific dogmatismhad become as as medieval scholasticism: "The experimenter's mind imperiousand constricting in itsmodesty, or the scholastic's fromthe metaphysician's because experdiffers iment makes him, moment by moment,conscious of both his relative and his absolute ignorance. In teaching man, experimentalscience resultsin lessening scientific his pride more and more." For Bernard, pride takes on the specifically for "man is by nature metaphysicaland meaning of metaphysicalintervention, proud. He has gone so far as to thinkthat the idealisticcreationsof his mind, also representreality." whichcorrespondto his feelings, Againstsuch perversities is an insufficient of human nature,even experiment check,foritis alwayspossible The externalcheck of experimentmustbe supplemented to distortthe results.92 "must never answer for her the scientist by the internalcheck of self-restraint; fromtheresultsof an experto her answersbytaking, [nature]nor listenpartially his hypothesis."93 iment,only those whichsupportor confirm Humilityand selfthe thus define from without and other from one the within, restraint, imposed of the scientists. the pride-breaking morality the salient word is here,and withitcomes an apparent paradox. How Morality that seemed to insulate science from the could it be that the very objectivity distinctionas its motto-simultamoral-the creed that takes the fact/value of the highestorder? This apparent contradicneouslylayclaim to moral dignity It is an ethos of restraint, of the negativequalityof objectivity. tion is an artefact of and internalrestraints of method and quantification both external restraints of prohibiis a morality Otherwiseput, objectivity self-denialand self-criticism. but no less a morality forthat.Among those protionsratherthan exhortations, hibitions are bans against projection and anthropomorphism,against the insertionof hopes and fearsinto images of and factsabout nature: these are all forbidden.Seen fromthe standpoint and therefore subspecies of interpretation, there is of mechanical objectivity alone, nothingto distinguishthese formsof fromother formsnot so charged, such as system sociallycharged interpretation
122
REPRESENTATIONS

an acute building. It took an additional political awareness, more particularly awareness of ideology in the Marxistsense, to separate these particularformsof into a doctrineof fromthe others,and to convertnonintervention intervention "value-free"or "neutral"science.94 This vectorpointingtowardthe further developmentof the ideals and pracshould serve as a reminderthatthe emergence of mechanical ticesof objectivity in the latter half of the nineteenthcenturyby no means exhausts the objectivity as a whole. Other keyelementsof thathistory are of modern objectivity history atlas makers came to stillmissing. For example, we have shown how scientific as subjective,but not how the "subjective" brand judgment and interpretation as an epithetin science. As late as 1865, Berto be used se came exclusively per nard could classifymathematicsamong the "subjectivetruths. . . flowingfrom principlesof which the mind is conscious, and which bring it the sensation of absolute and necessaryevidence," in contrastto "the objective or outer world Nor can we explore, in the compass of truth"thatwould never attaincertainty.95 of objecbecame fusedwithothervarieties thisarticle,how mechanicalobjectivity makes with that the element such as objectivity synonymous metaphysical tivity, from with that identifies the element or the aperspectival truth, objectivity escape as well as history, any and all perspectives.Each of these elementshas a distinct them into a of a collective history binding singleconcept. partaking Because so much recent philosophical attentionhas been directed to aperinto the view from it is tempting to collapse all of objectivity spectivalobjectivity, This temptationto simplify nowhere.96 by conflationshould be resisted,for the in one mode may seem worthless whenjudged highestexpressionsof objectivity The the standards of another mode. that was the essence and photograph by cachet: at best it was emblem of mechanical objectivity carried no metaphysical an accurate renderingof sensoryappearances, whichare notoriously bad guides to the "reallyreal." Nor would ithave passed musterwiththe aperspectivalobjecthat eradicates all that is personal, idiosyncratic, tivity perspectival.The photoin was fact "look" radically perspectival-as many of our X-ray users graphic understandwhyphotographswear the halo neverceased to lament.We can fully of objectivity that beatifies only when we recognize that the kind of objectivity and not itsmetaphysical or aperspectivalkin. The themis mechanicalobjectivity, is a multifarious, moral of our storyis thatobjectivity mutable thing,capable of new meanings and new symbols:in both a literaland figurative sense, scientists of the late-nineteenth-century created a new image of objectivity.

Notes
whose and Simon Schaffer, Our greatestdebtsare to CarolineJones,Gerd Gigerenzer, are reflected extensivecommentsand criticisms throughoutthisarticle.In addition, The Image of Objectivity 123

we benefitedfrommanydiscussionswithand suggestions byNancyCartwright, Shelly Errington,Gerald Geison, Timothy Lenoir, Theodore Porter,and David Stump. as a researchassistanton the history of paleontology, Johanna Carr worked tirelessly and we are both very gratefulto her. This work was undertaken at the Center for withits support Advanced Study in the Behavioral Sciences at StanfordUniversity, and thatof grantsfromthe National Science Foundation. et particulierement en dans les sciences 1. E.J. Marey,La Methode experimentales graphique see FrancoisDagognet,Etienneeten medecine (Paris, 1878), iii-vi.On Marey, physiologie of clinical Jules Marey:La Passion de la trace(Paris, 1987); on the "objectification" and the medicine in general,see Stanley (CamReignofTechnology Joel Reiser,Medicine to mobilize allies and to secure polemical bridge, 1978). On the role of inscriptions and rhetoricalaims, see Bruno Latour's "Visualizationand Cognition: Thinkingwith and Society 6 (1986): 1-40, especiallyhis notionof "immutEyes and Hands," Knowledge able mobiles." This article also contains furtherreferences to articles relating to in contemporary debates withinthe sociologyof science. inscription 2. See, forexample,JosefM. Eder,Atlastypischer (Vienna, 1928); or WilhelmHis, Spektra I-VIII Atlas menschlicher Anatomie Tafel (Leipzig, 1880). One publisher Embryonen: alone, Lehmann Verlag,issued a series of seventeenatlases in medicine,and a much larger run of smallerhand atlases. used by Gerard Mercatorin 1596 for his map of the world: 3. The word atlaswas first withan introductionby B. Van 'T Hoff (Rotsee GerardMercator's Map oftheWorld, terdam, 1961), 17. The term spread to astronomicalmaps by the early eighteenth Cartography, century:see the titlesin Deborah J. Warner,The SkyExplored:Celestial of theseworks,theword 1500-1800 (New York,1978). Because of the oversizeformat atlas came in the eighteenthcenturyto designate a very large size (34" x 26.5") of and Encyclopaedia ofPaper and Paper-Making drawing paper: E.J. Labarre, Dictionary to all illus(2nd ed., London, 1952), 10-11. The term was apparentlytransferred when figureswere printed works in the mid nineteenthcentury, trated scientific separately from explanatorytexts,in large formatsupplements-hence atlases,derivingfromtheirsize: e.g., textin octo, atlas in folio.As textand figuresmerged into a single volume, often oversize,atlas came to referto the entirework,and atlasesto the whole genre of such scientific picturebooks. We shall use the termretrospectively to referto all such works,even those earlierones whichmay not use the word atlasin the title. 4. See for example Johann Gabriel Doppelmayr,Atlascoelestis (Norimbergae, 1742); or et novus (Amsterdam, seu atlas universalis Andreas Cellarius, Harmoniamacrocosmica 1666). 5. On the problem of limitedaccess to experimentalresultsand the solutionof "virtual TheLeviathanand theAir Pump: see Steven Shapin and Simon Schaffer, witnessing," and the Hobbes, Life(Princeton,N.J., 1985). Experimental Boyle, 2 vols. (Paris, 1829-35), 1:i-ii. du corps Anatomie 6. Jean Cruveilhier, humain, pathologique 7. Johann WolfgangGoethe, "ErsterEntwurfeiner allgemeinen Einleitungin die vergleichende Anatomie, ausgehend von der Osteologie" (1798; 1893), in Dorothea 14 vols. (7th ed., Munich, 1975), Kuhn and Rike Wankmuller,eds., Goethes Werke, ed. and trans.,Goethe: from trans. Miller, 13:172; modified) Scientific Douglas (slightly Studies (New York, 1988), 118. 8. Johann Wolfgang Goethe, "Erfahrung und Wissenschaft"(1792; 1823), in Werke, 24. 13:24; trans.in Miller,Goethe, of this period, see individual entries in Hans 9. On Wandelaar and other illustrators

124

REPRESENTATIONS

10. 11. 12. 13. 14.

15. 16. 17.

Lexiconderbildenden Kiinstler vonAntike 36 biszur Gegenwart, Vollmer,ed., Allgemeines atlas illustrators were Dutch or vols. (Leipzig, 1927). Many of the eighteenth-century of descriptive art,see Svetlana Alpers, TheArt Dutch-trained;on the Dutch tradition Seventeenth DutchArtin the Century (Chicago, 1983). ofDescribing: and Muscles of theHuman Body Bernhard Siegfried Albinus, Tablesof the Skeleton (London, 1749), sig. cr. Ibid., sig. br. Londa Schiebinger,"Skeletons in the Closet: The First Illustrationsof the Female 14 (1986): 42-82. Skeleton in Eighteenth-Century Anatomy," Representations br. Albinus,Tables, sig. See for example James Sowerby,The Mineral Conchology of GreatBritain(London, Fossil Remains 1813), 101, 156, and passim; Daniel Sharpe, Description ofthe ofMollusca Foundin theChalkofEngland(London, 1854), plate 11, figs.la and lb. For idealizingof the late eighteenthand early tendenciesin geological illustrations cum-theoretical nineteenthcenturies,see MartinRudwick,"The Emergenceof a Visual Language for ofScience14 (1976): 149-95, 171. Geological Science, 1760-1840," History A andPermian Bowman (London, Brady, Monograph Henry ofCarboniferous Forominifera 1876), 7. and Bibliography Illustration (1852), ed. and trans. ofAnatomic Ludwig Choulant,History MortimerFrank (Chicago, 1920), 23. HumanGravidUterus WilliamHunter,TheAnatomy ofthe (Birmingham,1774), preface,
n.p.

18. See Ludmilla J. Jordanova, "Gender, Generation, and Science: William Hunter's Obstetrical Atlas," in W.F. Bynum and Roy Porter,eds., WilliamHunter and the MedicalWorld (Cambridge, 1985), 385-412, especiallyher percepEighteenth-Century of the femalecadaver and thatof the fetusin Hunter's tivecontrastof the treatment atlas. 19. On preparing the cadaver for illustration,see Robert Herrlinger and Marilene Geschichte dermedizinischen 2 vols. (Munich, 1972), 2:49. Putscher, Abbildung, 20. Hunter,Anatomy, plate 2, n.p. of "photographic" 21. On artistic vision,see PeterGalassi,Before anticipations Photography: and the Invention (New York, 1981). Painting ofPhotography Bones(London, 1733), "To the 22. William Cheselden, Osteographia; or,TheAnatomy ofthe Reader," n.p. Normal and the 23. Georges Canguilhem,On the (1966), trans.CarolynR. FawPathological cett(Dordrecht, 1978). vol. 1, book 17, p. 5. 24. Cruveilhier, Anatomie, de dege'nrescences etmorales de 'espece 25. B. A. Morel, Traits humaine: intellectuelles, physiques, Atlasde XII planches (Paris, 1857), 5, 19, 12. tothe Order MeroBritish FossilCrustacea 26. Henry Woodward,A Monograph Belonging ofthe "The Value of C. see also stomata O. Marsh, (London, 1866), 2; Type Specimens and Science 6 American Their of Preservation," Journal (1898): 401-9, 401; of Importance and Charles Schuchert,"Catalogue of the Type of Specimens of Fossil Invertebrates in the Departmentof Geology,United States National Museum,"Bulletin oftheUnited no. 53, part 1 (Washington,1905), 9-12. States NationalMuseum, essentialismin Quetelet and others,see Theodore M. Porter,TheRise of 27. On statistical Statistical (Princeton,N.J., 1986). Thinking trans. Joseph Leidy (Philadelphia, 28. Gottlieb Gluge, Atlas of Pathological Histology, 6. 1853),

The Image of Objectivity

125

Chronicle 29. Walter Fitch, "Botanical Drawing,"Gardeners' (1869); reprintedin Wilfrid Illustration Blunt, TheArtofBotanical (London, 1950), 276. ed. W. Hooker (2nd ed., 30. Joseph Dalton Hooker, TheRhododendron ofSikkim-Himalaya, London, 1849), n.p. A Monograph FossilTrigoniae British 31. John Lycett, (London, 1875), 134. ofthe 32. William Anderson, "An Outline of the Historyof Art in Its Relation to Medical Sciaddress deliveredat the Medical and PhysicalSocietyof St. Thomence,"introductory 15 (1886): 151-81, 170. as's Hospital, 1885, SaintThomas's Hospital Reports 34. Ibid., 175. 35. Ibid., 179-80. 33. Ibid., 172. 36. Ibid., 175. 37. Quoted in Linda Nochlin,Realism(Harmondsworth, Eng., 1971), 36. Students' AtlasofBones and Ligaments 38. Charles W. Cathcart and F. M. Caird, Johnston's (Edinburgh, 1885), preface,n.p. Atlasdermedezinisch-chirurgischen 39. E. Ponfick,Topographischer (Jena, 1901), Diagnostik "Methode" (i.e., methodologicalpreface),n.p. 40. Johannes Sabotta, Atlas and Text-Book (Philadelphia, 1909), 13 of Human Anatomy (emphasis added). Anatomie des der Histologie und mikroskopischen 41. Johannes Sabotta, Atlas und Grundriss Menschen (Munich, 1902), vi-vii. Nature18 (1878): 97-100, 97. 42. FrancisGalton, "Composite Portraits," A Glanceat Its History Greenwich: and Work 43. W. Walter Maunder, TheRoyalObservatory Simon for would like to thank Schaffer 176-77. We (London, 1900), bringingthis Mark Time: Discipline quotation to our attention;also see his article,"Astronomers in Context 2 (1988): 115-45. and the Personal Equation," Science Anatomie derAugapfels 44. Hermann Pagenstecherand Carl Centus,Atlasder pathologischen (Wiesbaden, 1875), vii,emphasis added. 45. Ibid., vii-viii,emphasis added. Cells(New York, 1896), v-vi. AtlasofNerve 46. M. Allen Starr, vol. 2 (Munich, 1924), from 47. Rudolf Grashey,Chirurgisch-Pathologische Rintgenbilder, first preface,iii-iv. 48. W. Gentner,H. Maier-Leibniz,and W. Bothe, An Atlasof Typical ExpansionChamber York, 1954). (New Photographs 49. Gertrude L. Elles and Ethel M. R. Wood, A Monograph (London, ofBritish Graptolites 1901), 2. 50. FredericA. Gibbsand Erna L. Gibbs,AtlasofElectroencephalography (Cambridge,Mass., 1941), preface,n.p. Photo51. P. M. S. Blackett,forewordto G. D. Rochesterand J.G. Wilson,Cloud Chamber Radiation(New York,1952), vii. oftheCosmic graphs Menschen vom normalen 52. Rudolf Grashey, (Munich, 1939), v. Rintgenbilder Typische as a means of social control,see 53. For a Foucauldian analysisof the role of photography (Amherst, Mass., 1988), and references John Tagg, The Burden of Representation therein. of theAmericanSurgical Association 54. Dr. Ames, citedin the "Reportof the Committee American on the Medico-Legal Relationsof the X-Rays," Journal oftheMedicalSciences 120 (1900): 7-36, 22. 7 May American MedicalAssociation, 55. Ibid., 29; citationto Dr. Samuel Lloyd,Journal ofthe 1898, 1111. American 56. R. Harvey Reed, "The X-Rayfroma Medico-Legal Standpoint,"Journal ofthe 35 (1898): 1013-19, 1018. MedicalAssociation,

126

REPRESENTATIONS

57. Ames, "Report on Medico-Legal Relations,"11. onAmerican 58. Cited in Richard Rudisill,Mirror Image:Influence Society ofthe Daguerreotype (Albuquerque, N.M., 1971); itselfcited in Allan Sekula, "On the Inventionof Photo(Houndmills, Eng., 1982), Photography graphic Meaning," in VictorBurgin, Thinking 86-87. Microsin Modern in Electron 59. Don W. Fawcett,"Histologyand Cytology," Developments B. M. ed. York, 1964). Siegel (New copy, and Realism 60. Charles Rosen and Henri Zerner,Romanticism (New York, 1984), 108. 61. Dr. T. S. K. Morton,in Ames, "Reporton Medico-Legal Relations,"24. 62. Lancet(1899), cited in ibid., 33. Skiagram (fromGreekskia,"shadow") was a termfrom medium. Skia1795 to 1805 for an image made by a shadow cast on photosensitized seem to have been used interchangeably. and radiogram skiagraph, gram, 36 (among eightconclusionsunanimously 63. Ames, "Reporton Medico-Legal Relations," adopted as expressingviewsof the AmericanSurgicalAssociation). 64. Reed, "Medico-Legal Standpoint,"1016. und erkrankter Atlas der Histotopographie 65. Erwin Christeller, gesunder Organe (Leipzig, 1927). 67. Ibid. 68. Ibid. 66. Ibid., 18. 69. Ibid., 19. Atlasofthe 70. Alexander Bruce, A Topographical SpinalCord(London, 1901), preface,n.p. iii. 71. Marey,Methode 72. Ibid., viii-ix. graphique, de la photographie 73. Gaston Tissandier,Les Merveilles (Paris, 1874), 226, 254-55, and 264. 74. The debate over the role of photographicversus hand-drawn representationcontinues. As John Law shows in his article,"Lists, Field Guides, and the Descriptive Organization of Seeing: Birdwatchingas an Exemplary Observational Activity," Human Studies11 (1988): 271-303. One opponent of photographic representation writes: "A drawing can do much more than a photograph to emphasize the field marks.A photographis a record of a fleeting instant;a drawingis a compositeof the artist'sexperience. The artistcan edit out, show fieldmarks to best advantage, and delete unnecessaryclutter.He can choose positionand stressbasic color and pattern unmodifiedbytransitory lightand shade..... Whereas a photographcan have a living a (279-81). On theopposing side, immediacy good drawingis reallymore instructive" the Audubon Societyguide advocated photographicrealism:"Photographsadd a new dimensionin realismand naturalbeauty.Fine modern photographsare closer to the way the human eye usuallysees a bird and, moreover, theyare a pleasure to look at" (286). 30. 75. Choulant,History, von Objekt und Subjekt" 76. Johann Wolfgang Goethe, "Der Versuch als Vermittler in in 14. trans. Werke, 13:14-15; Miller, Goethe, (1792; 1823), in Lorenz Kruger et al., 77. See ZenoJ. Switjtink, "The Objectification of Measurement," vol. 1, Ideas in History Revolution, eds., Probabilistic (Cambridge, Mass., 1987), 261-85; in Context Simon Schaffer, "Astronomers Mark Time .. .," Science 2 (1988): 115-46; Richard Yeo, "Scientific Method and the Rhetoricof Science in Britain,1830-1917," in J.A. Schusterand Yeo, eds., ThePolitics and Rhetoric Method (Dordrecht, ofScientific 1986), 259-97. 78. Quoted in Karl Pearson, TheGrammar (London, 1892), 38. ofScience 79. Samuel Smiles, Character York, 165, 167, 175; Lifeand Labor (Chicago, (New 1880), 1891), 58; Self-Help (New York, 1915), 144. We are gratefulto Simon Schafferfor to Smiles'sviewson scientists. drawingour attention

The Image of Objectivity

127

ed. W. D. Clerk 80. James Clerk Maxwell, "Atom,"in TheScientific Maxwell, PapersofJames Niven (New York, 1965), 445-84. 1871-1918 ThePhysikalisch-Technische 81. David Cahan, An Institutefor Reichanstalt, Empire: (Cambridge, 1989). 82. Simon Schaffer,"A Manufactury of Ohms: The Integrityof Victorian Values" (Unpublished MS). undArtzte Deutscher der44, Versammlung 83. Rudolf Virchow, address,Tageblatt Naturforscher (Rostock, 1871), 77. and Manufactures 84. Charles Babbage, On theEconomy (4th ed., London, of Machinery 1835), 54. 108. 85. Rosen and Zerner,Romanticism, de la science 86. ErnestRenan, L'Avenir (Paris, 1890), 235. in EarlyChrisand SexualRenunciation 87. Peter Brown, TheBodyand Society: Men, Women, xvii. York, 1988), (New tianity 88. Caroline Walker Bynum,HolyFeastand HolyFast: TheReligious Significance ofFood to MedievalWomen (Berkeley,1987). 235. 89. Renan, Avenir, and Religion:TheReactionto Scientific Science 90. Quoted in Frank Miller Turner,Between in Late Victorian Naturalism England(New Haven, 1974), 2. 91. Ibid., 9. of theexperimenter, and hypotheses 92. Claude Bernard countenanced theinterventions but he strictly distinguishedthe role of the experimenterfromthatof the observer withthe idea that the latterfunction who records the results.Bernard even flirted "an uneducated be best man, knowingnothingof theory";An performedby might Medicine(1865), trans. Henry Copley Green to theStudy Introduction ofExperimental (New York, 1957), 38. 93. Ibid., 28, 27, 22-23. TheOrigins Science?: Value-Free 94. RobertN. Proctor, ofan Ideal (Cambridge,Mass., forthund sozialpolicoming). See also Max Weber,"'Objektivitat'sozialwissenschaftlicher tischerErkenntnis"(1904), in Johannes Winckelmann,ed., Gesammelte fur Aufsdtze (3rd ed., Tiibingen, 1968), 146-214. Wissenschaftslehre Thomas Henry Huxley, 95. Bernard, Introduction, 28-29; compare Britishphysiologist ed. Gavin de Beer (Oxford, 1983), 95-96. Autobiographies, Nowhere 96. Thomas Nagel, The View (Oxford, 1986). Nagel givesas his source Berfrom The nard Williams'snotion of "pure" or "absolute inquiry"as developed in Descartes: (Harmondsworth, Eng., 1978), but itsmostlyricalphilosophical ofPure Inquiry Project exponentwas Charles Sanders Pierce; see forexample his "A CriticalReviewof Berkeof Chance: Selected ley's Idealism" (1871), in Philip Wiener, ed., Valuesin a Universe 81-83. Pierce S. C. York, 1958), (New of Writings

128

REPRESENTATIONS