Pelr

ILL:

TN: Len( Duo Nol ! Spc

SUNY Series in Philosophy and Biolog! David Edward Shaner, Furman University, editot

HISTORYAND EVOLUTION

lnslrr
Reng Libra

edited by

Matthew H. Nitecki

tI
For

and

Do s V. Nitecki

thip

lach
1300

\u!tl
State Universiry oI New York Press

What's So Special about the Past?

Rachel Laudan

During the past half-dozen years, a number ol biologists have become intrigued by ana.logies between history and their own dhcipline. SPurred by the problems of reconsttucting phylogenies, they have asked whether biolo' gists might find solutions to these prcblems in the historical litemture

(O'Ham 1988a, 1988b). For centuries, after all, histonans have been prcfes_ sionally concemed with reconstructiflg the past and have done so witl con_ siderable success. Surely biologists could benelit from this experience, or so
the argument goes.
shall counter by suggesting that t]le iflitially plausible program of turnirrg to history to learn how to deal wilh the biological past is miscon_ ceived, resting as it does on a couple of tacit assumptions that on caleful examina[on tum out 10 be problematic. The program assumes first, that all

investigations of the past share crtain conrmon features that set them off from other kinds of inquiry, and second, that philosophy of history is tbe best guide to these features. Because it can be more quickly dealt with I shall start wilh the second of these assumptions. Tuming to the philosophy of history to find out what has been dh_ covered about ways to reconstruct the past seems only natural, IJ nothing else, what has been written on th philosophy of history far outweighs what
has beefl written on all the other historical sciences combined. Indeed, apart

from history the only historical scienc io have leceived any metlodological attention at all has been geolog/ (Kitts 1977;I-audan 1987). But the lessons to be leamed from philosophy of history are much more limited tlan mrght at fkst siglt appear. To begin witlt, philosophe$ of history struggle with a number of prcbleris that axe quito irelevant io biologists. Prominert among these are how to handle humar intentions and how to cope with the absence of a good causal theory of human behavior, problms that history shares wlth

thc social sciences but not with biolos/. Biology after all does

rct

deal witb

56

Rathel Laudan

Wha.'s So Special aboul the Patt?

it already possesses some rather good causal hypotleses. More important tlough philosophers of history are
intentions and evolutionary theory preoccupied with the nature of historical explanation, particularly narrative explanatiorl Begindng in the 194G Carl Hempel claimed that historical explaflations were sketchy versions of explaDations in the natural scieflces (Hempel [1942] 1965, 1962). According to his deductive-nomoloSlcal model, scientific explanations worked by deducing what was to be explained from general laws, and historical explanations differed chiefly in the degee ol explicitness of the laws invoked. This sparked an intense debate amongst philosophers of history. Many, perhaps mos! cogdzant of history's roots iD

ir

neglected other active lesearch areas in ihe Philosophy of science such as realisrq theories of screntific change, and validatiorL that are equally rclevant to history The latter is particularly importaflt because what is driving the biologists' inlerest in philosophy of history is, I believe, Dot a question of explanation but of epistemology, not a question of the form their theories take but of whether they ale ieliable and well_founded, that is to say, ques_ tions of validatiorl Biologrsts wan! strategies for recomtructing the past as

literaturc and the humadties, and its fierce independence from the social sciences in spite of t}re recoDciliatory moves of social and economic historians, rejected the idea that history was nothing more than embryonic science

and insisted tlat history differed from science in fundamental ways (see papers in Dray 1966). In the years since, philosophers of history have continued to debale whether the structure of explanation in history is or is not the same as the structure of explanation in th natural sciences, frequently identirying narrative explanation as characteristic of history and questioning whether it can be reduced to deductive-nomological or statistical explanation (Danto 1985; wlite 1965; Richards this volume). This debate was imiiated fifty years ago and since theD philosophers of history, with a few honorable exceplions (Mandelbaum 1977; Roth 1988, 1989) have failed to keep pace with change in cognate disciplines. For example, philosophers of history still tate recelt political ard diplomatic history as their prime e\amples, oblivious of the fact that much of the best arld most methodologrcally sophisticated histo cal scholarship in the last quarter century has beeD in other areas such as economic ard sooal history, This would not matter were it not for the fact thal, as we shall shortly see, political and diplomaiic history are perhaps the branches of history furthest removed from the problems biologrsts encou er when they try to recoDstruct the past. Furthermore, philosophers ol history have not incorporated the results of the extensive research on tleory ol explanation carried out since Hempel's pioneering work, research that might significa[tly alter tleir analysis of history (surveyed in Kit.her and Salmon 1989). Nor is that all, for ir thcir cnthusiasm for qucstions of cxplsnation, philosophcrs of history have

reliably ar possible in situations of incomplete information. Of course, theories of explanation demand that tle components of th er?lanation are well warafied. But most discussions of historical explanation are much ftore coflcemed with form and with the differences in form between histori cal and rcnhistorical explanation than they are with warrant. In sunl there are reasons to be suspicious about tuming to philosophy of history in its present state for guidance about reconstructirg tle biological past. But even if this were not so, and philosophy of history once again became a sigruficant branch of philosophy, I would have qua]ms about assuming its relevarce to biologists interested in the past without further ques_ tiorl This is nol because I think history is unscientific. On the contmry [ am completely convi ced that history is (or should be) scientific. Rather, it is because I doubt that t]rc Nstorical sciences share conmon features that distinguish them from the other sciences. To explain why I shall broaden the discussion beyofld history to include all tle scierces that deal with the past - the 'historical sciences" al I shall term them since tle word "history" is generally restricted to human history. Historical sciences are far commoner thaD is generally rccognized, Perhaps because the accidents of the organiza_ tion of modem u.dversities mask tlei ubiquity. Cosmogony, for example, has always found its home in depa.rtments of physics and aslronomy. Isrge parts of paleoDtolog/ have been housed in geolos/ departments, not in biology departments, because through most of the nineteenth and twentieth centudes geologists fouDd their results oucial while biologists found tbem peripheral. Archaeolog/ aIId antkopolosr are relatively new disciplines without well-established Diches h the uaiveruity slructure. But once we appreciate the range of historical sciences, then it makes sense to ask uhe_ ther biologists might not leam as much aboul reconstrucling the past from gaologists, arahaeologists, or cosmogonists as from historians.

What's So Special abou rhe Past?

59

To sEucture my discussion of the historical sciences, I shall bonow a way of analyzing them f.om the great Victorian philosopher of science. Wiliam Whewell, aware that doing so opens me to the charge that my philosophy of the historica.l sciences is even more outdated than that of the philosophers of history, I believe rhe choice is justified, however. on a number of grouDds. Wlewell was the first, and thus far the las! philosopher of science to pay attention to all the historical sciences, not just history itself, As a former President of the ceological Society of Irndon and colleague at T.inity College, Cambridge of the leading historians of the day, he was intimately familiar with the best contemporary practice in the historical sciences. And he was much morc interested in vatidation than in explanation. So while his ana.lysis of the historical sciences (or as Wlewell termed them, the palaetiological sciences) will doubtless need to be modified, it provides a good starting point. Among them he numbered geologr, paleontology, cosmogony, philolo$/, and wha! we would term archaeologr aIId history. He took geolog/ as his main example, a move that I shall copy, and distinguished thee task for such a historical science (1837, 3: 402-3): ,,the Descdption of the facls and phenomena; - the general Theory of the causes of change ap_ propriate to the case; - and the Application ofthe tbeory to the facts.,,These he called respectively 'descriptive geology,,, ,,geological dynamics,,, and
"physical geology."

be a sophisticated task involving evidence from litholog/, paleontology and oth.r disciplines. Similar diffiqrlties were encountered irl oiher historical sciences. Scientists lecognrzed fossils as the remarns of living beings only after centuries of argumeDts and countetuguments, GeoPhysicists found globa.l geomapetic reversals so improbable and so dependen! on sophisticateal instrumentation that tley subjected both the concept and the instru_ meDts to severe scnrtiny befole accePtirg tlem irl the mid'1960s. Scholars regarded chipped stones wiih skepticism for yea$ before they were prePaxed to gla.nt that they had been manulactured by prehistoric man. Historians
now, as iD the past, se as one of their tdckiest problems fiDding appropiiate

units to periodize the Past. Onc geologistr had identified their units, developed means of individuating ihem, and found ways of detecting them in the impeifectly pre_ selveal record, they laced their second problem, how to detemine the relations between them, in this case relations of earliet and later' They relied on the theory ihat formatioDs had been deposited sequentially, with younger rocls supenmposed on older rocl(s (the principle of superposition). They recognized the limitatioDs

of this pdnciple. Igneous rocks, intruded from

belov/ into the sequence of lotmations, might be under, not over, oldel rocks; structual disturbances subsequent to deposition could overturn the sequence of formalions, again resulting ifl younger rock under, not over, older rocks. CoDsequently, they developed ancillary lines of evidence to supplement the pdnciple of superpositiorl They argled that fossilized worm casts and rrpple

As afly scientist knows, description of a domain is not easv. As I se ir. descriprive geolos/ faces rwo major problems: tirst. how ro ijenril, and individuate its basic u lj or entities; and second, how to detemine th rclations between those uaits. Nineteenth_century stratigraphers,

all the rock laid do\{rl in a given period - was the appropriate unit for reconstructing tle past (Iaudan 19g7). They knew full well that most weste r European rock wete inaccessible to observation over most of their extent. Thy recognized that since litholo$/ also reflected environmental
conditions, it was not an cntircly rcliablc guidc to thc pcriod in which a roc* was dcpositcd. Idcntirying and lndividuadng formations thus turncd out to

ample, had to establish the existence and significance of their units ard devise rEeans of detecting and observi[g them, particdarly in the mary cases wherc the record is impe ectly preserved, even before determining iheir relatioruhips. Only after long debate did they decide that the ,,formation, _

for

ex-

marks indicated which side of a sfatum had originally faced up even if the rock was novr reversed, and they checked which sedimentary rocks were intersected ty i8leous rock. Gradually they Put together an increasingly Iine-$ained and retiable chronolos/ of the dePosition of rock on the earth's

contine s, a chronolos/ that they had every reason to believe had sound warant. Thus, it is clear that descriplive geolory, and other descrlptive historical scieDces, too, ilvolve tricky epistemic problens, problems that saicntists have found strategres for solving in many cases, so that they have
Sood reasol to believe their conclusions to be well-founded.

this is cofiect, we can retum bdefly to the philosophy of history to concct what I see as a widesPread misconception in that literaiure. A de' scrlptlvc historical sci.ncc is nothinS other than what Philosophe$ of history

lf

Ra.hel L@dan

What's So Special abou the Past?

61

such as Isaac NeMo& devoted interr[iDable hours to reconciling thie records

ogy tas not a.lways bee[ easy becausc many events were eithea not dated or dated otr diffe.ent time scales in different pa s of the world. From the fiftee h through the eighteenth century, historians, including leading scholars

chegpal examples of history from reccrt political or diplornatic history where ihe problerns ol establishing chronolog, arc minor or nonexistent. Wlewell knew better. He recog zed the epistemic difriculties geologists encountered in constructing the stratigmphic column and racing tle history of life. We caE see that these difriculties occur in many of th; historical sciences. ReceDtly geomagnetists have struggled to track the record of geomagnetic reversals. paleontologists who have done a magnificent job tracing tle sequence of major life forms, nonetheless, still face problerns with an imperfecdy presrved recoral. Archaeologistj, at least until thc advent of thc "new archaeology,, a couple of alecades ago, were almost exclusively Preolcrpied sorting out piehistoric chronoloS),. Even within history chronol-

sa.rDe or similar kinds. Thei. misconception, I suggest, is their tendenca to dismiss chronolos/ a6 "mere. ahronolosr. just a preparation for tha narrative they assume to be the characteristic product of historicat inquiry and on which they focus most of their attention. This rnay bc because they ta.ke their ar-

call a chronicle or chronolos/. They defirc it as on arangement of past objects or evenls in a temporal sequence (White 1965:222) _ not just a[y objects or eve[ts, but those that we have reason to bIieve are the

quite well in other histoncal sciences that deal with obiects deposiied in laye , such as paleontolos/ alld archeology. ln history of lanSrage it is rct so easily applied: when the distinguished nineteenth-century linguist, Max Muller, struck by the apparent analogies between geolos/ and language, attempted to adopt geological methods, he met wiih sorry results. In written

history it is inappropriate sirce documents are not arranged in layers: his_ todans trace anachronisms, references from one text to another, and other clues to establish sequence. And in astto omy, it is quite useless: aslronomers establish the relative ages of stals by their colols. I1l biology, which is the chief concem of this volume, Paleoniologrsts can use these methods to reconstnrct the outlines of the history of life on earth while finding them absolutely useless when it comes to the details simply because the preserva' tion of the record does not support such a fine-grained analysis. Thus the possibility of using a melhodolog/ developed in one hhtorical dhcipline to rcsolve prcblems in anotler depeflds on how similar the two domains are. That both deal with the past is not enough. That leads to t]le second question. Are the epistemic problems we have discussed ard the stralegies for overcoming them unique to the histori_ ca.l sciences? The answer must be no. Problerns of idettifying and individu_ ating umts and of determining the relations between them occur in all tle
scienaes, not just in the historical sciences. The chronologres of the hislorical sciences are simply a speoal case of the general class of descriptive theories.

of the Hebrcws, Greeks, and RoEans with the records of other peoples of the Near East. To this da, demographiq social ard economic historians
have to determitre chronologies in their domaim, In sum, we should Dot be misled by the rich written record of modem political ard intellectual

Scientists develop descriptive tleories of their domains, whether or not these domains have a diachroric dimension. Doing so involves more thatl simply

reading olf the evidence,

if for no other

reason than that the evidence is so

itrto underestimating the difriculties of constructing chronologies. Except in rare cases of well-prese.ved records (and even there problems of individuatiDg udits caD occur), descdptive historical science oI chronologr faces, and in many cases has overcohe, severe epistemic difficulties.

hisbry

This leads to two fu.ther questions. First, can the strategies de_ veloped to overcoma
exportcd to aEothcr histodcal dornain, in thc way that soDe biologists hop; to axport solutions from history lo biolo&r? Thc arrswcr is that it alt dcpcnds
ihese problms in one historical domain be suJessfully

often imperfecl. Constructing descriptive theories in nonhisto cal sciences involves difficulties simitar to tlose encounlered in the historical sciences and calls lor resort to the same kinds of methods to dea.l witl them. When Kepler propounded his theory of planelary motioq for example, he used a unit - the planet - the defi tion of which had been forged over centudes of
astronomical inquiry. He could no more easily read off &e description of planetary motion from tables of astronomical obsrvatiolts than could geolo_ Sists easily read off the sfatigaphic column from tecords of Iield obse a_ tions. Thc same is lrue of descriptive theories in natural hhtory (most sys-

oll tha naturc of thc dom8ifi. Th. prlncipla of superpositioq ray, works

t.matics), crystallography (crystal structure), biolog (cell theory), and a host

Rachel Laadan

Whd's So Special about the Past? struct causal theodes also. And they have in fact regularly done so. Charles Lyell (183G33), for example, deftly adapted a methodolog/ established in the nonhistorical scierces to dea.l with the past (Isudan 1982). This was t}te so_ called wra ca.$a fie]J.lrc,d, initially articulated by Newton in his famous rules

of other scienccs. Ibe imperfectly preserved traces of past objects and events that historical scientists oflen bemoan as if they faced speciat prob,
lerns do not differ in sigdificant epistemic respccts from the imperfect traces of objecb t}lat are too la.rgc, too small, or too distant to be easily observed,

Ovcrcoming observational difficutties is always a challenge to scientists: tle fa.t that in the historical sciences many of these difficulties arise hom imperfect preservation of the past does not establish th epistemic uniry

ard

uniquencss of lhe historical sciences. Even lhe claim that records of the past may be completely destroyed and tlus forevet inaccessible to observation does flot make the historical sciences unique, fo. it is quite possible to imagitre ihat therc are contemf,oEry features of tle world that we cannot observe directly or indirectly. HeDce if we ask,Do alescriptive theories in the historical sciences differ epistemically from descriptive theories in the nonhistorical sciences?', the ans\rer is "No.,' So far as descriptive theory goes tleD, the hisrorical scieoces look jusr like any orber sciences.

of philosophizin& By the early nifleteenth century, the v?m causa method was gercrally assumed to involve two strictules. First, scientists should invoke oDly causes for which they had evidence indeperdeDt of the eflect they were explaining; second they should also have independent evideDce that these causes were adequate to produce the effect. Modirying these strictures to the geological c{se, Lyell argued that the only way geologists
could obtarn independent evidence Ior the existence of postulated past causes

Having spelled out my case for the descriptiv historical sciences at soEe length, I car now briefly deal with geologicat dynamics and phlsical geolory. It might be argued that it is in ,geological alynamics,, (or causal tleory) that geologists (or historical scientists) really find themselves at an epistemic disadvaltage compared to their peers in physics or chemistry.
Physicists and chemists routinely perform expe.iments, systematically varying
causes and examining the rcsultant effects. But historica.l scientists frequently investigate causes ihat operate too slowly for their effects to fall within the ken of contemporary observation. The dse and lall of mountains and the movernent of plates take place ovI millions of years. The selective prcs-

sues on living beings opemte over many generations, Human populations

and economic systems change over hundreds ofyeaN. Causal processes that act so slowly as to dwarf the human time scale obviously cannot be inves-

tigated experimentally.

In this context, though, it is worth remembering tlat once again his_ torical scieltists are not in a uDique positio[ many nonhistodcal scientists study phenomena on which they cannot perform experiments. yet, far from
languishin& they have produced some of our proudest causal theories, such as NewbniaE mechadcs, relativity thcory, and the tleory of evolution. If ihis is so, then therc seem6 no rcason why historical scicntirts cannot con-

wrs to observe these causes acting at present, and that the or y way they could obtai[ evidence that th cause was adequate to prcduce the effect was to assume that the present intensity of the cause was no different from its past irtensity. On this basis he consEucted a causal theory of long_tem climatic change. Not everyone agreed with the theory or with his choice of the veru cawa melhod in particular, but nobody questioned the appropriate_ ness of his strateg/ of modirying a methodologt developed in a nonristorical science to overcome related epistemic difficulties in a historical scienc. Yet here agai[ philosophers of both history and geolog/ may mislead us, for they routinely deny lhat historical scientists construct causal theories, suggesting inslead that they rely on theories developed in the nonl storical sciences. For reasons of their o\IT r, historians, at least in public, have usually eschewed tlle suggestion that tley formulate causal iheories, msisting that they resort to intuition, cornmofl sense, or latterly to the social sciences for their repertoire of causes. The philosopher of history Maurice Mandelbaum, while distancing himself ftom t}Ie position, could state that it was "a com' monplace in tle literature of our subject that histo.i.urs are coDcerned with particular events that occurrd at specific times and places, and not with them oily iD so far as they represent events of a given type" (1977:4). The philosopher of science, Ertst Nagel, took the same line: ". . a geologist scek to ascertain . . . the sequential order of Seologic formations, and he is ablc to do so irl part by applying various physical la\Is to his materials of study" (196f:550). For him, the key difference between the historical and the nonhistorical scienccs was that the hhtorical sciences rely on the nonhistod_

64

Ra.hel Laudarl

Wat's

So Special about the Past?

6S

it is not the geologist,s taslg 4ua geologist, to establish the laws of mechari.s or of radioactive disinte$atioD which he employs in his investigations." The philosopher of geologr, David
Kitts, followed Nagel when he claimed that the goa.l of geology was
,,the

cal sciences for their causal theories, ". . .

construction of a chronicle of specific events occ-urring at specific times land aI the emphasis in recent yeaN upon ta(h scierlce' and the theoretical physical-chemical foundation of geolosr, the pdmary concem with

thatl wil]r

specific events
as my

[rct with causes] still domiDates our discipline,' (1977t5). But, brief disclrssion of causal iheory in geolory suggests, alrawing this sha4,

distinction between the historical and nonhistodcal sciencs may be premature, for there is a clear paaallel between the two. Both construct causal theories. With the exception of experiment, both have the same battery of

I have fudged the just issue of what makes a theory vrell-founded be it descriptive, causal or narrative. This is because we do not have a good account, let alone consensus, about what compdses a reliable, well-founded theory, though most scientists and philosophers of science would agee that at least a part of that warrant comes ftom the way we use evidence to support the iheory, Fortunately for my purposes though, we do not need to settle t}le vexed question of exaclly *hat constitutes a well-founded theory. All we ned is to show, as I have sought to do in this essay, is that evidential support for theories about the past is garnered and tested in wals similar to evidential support for
Readers may Dotice that throughout this discussion

theories about the present. If that can be done, the thesis that investigations into the past are epistemically unique can be dismissed because the problems of the historical sciences, like tlose of the nonlistorical sciences, boil down to the problems of warranting claims based on partial evidence. conclude that, if our aim is to devise means of acquiring reliable krcwledge, classirying the sciences into the hislorical and the nonhistorical is not particulaxly rclevant, a conclusion that others have also argued for (Ereshefsky, this volume). True, past objects and eve[ts cannot be directly ob-

for tleir tbeories. And the limits ol the use of experiment do not coincide with the boundary between tle historical and nonlistorical science.
tesls
Fina.lly w come to "physical geolog/. which Wlewell defined as the pairing of descriptive geolog and geological dynamics to produce a causal

history of t]re ea(h (a sense quite different &om tle otrrent mearing of physical geolos/, it should be noted). By this Wlewell meanr just what many
philosophers of history call naffative. W})ile studies of nax.ative are %dous atrd frequently arcaner one widely accepted meaning is the use of causal theory (generally believed 10 have been constnrcted by the rcnhisto cal
sciences)

to litrI

pasr eveds togerher (White 19651223). phitosophers of

rest cting the term "history" to mrative accounts of the past, excluding ,mere,, chronicle as at best protohistory Following their example a number of
scientists and philosophers of science have explored the concept of narrative and have made interesting use of the concept in dealing with the sciences (Ruse 1971; Hull 1975; O'Hara 19884 1988b). I place more emphasis on

history ard [terary theodsts have made much of narrative, frequently

ihe role of desoiptive and causal theories in the historical sciences, But whethor or not na.rative is the distinctive form of the historical sciences is irrelevant to my argument, because there is no reason to believe that tle warrant lor larrative is epistemically of a different kind from that for chronolog/ or causal theory albeit that it involves the additional step of deciding on the appropriateness of using particular causal thaodes to cooncct par_ ticular parh of chronologics.

served. But neither can many of the objects and events witl which the nonlistorical sciences deal. All the sciences have to work out tactics lor overcoming these difriculties. Whether or not a particular histo cal science can learn ftom some other science depnds on the nature of that inaccssibility, not oD whether it is past or present. In arguing that tlere is no significart epistemic difference between the historical ard lhe rcniistorical science, I am not forcinS the historical sciences back into the procrustean mold of philosophy of physics. If we have learned an)1hing over the past tweDty-five years it is that depending on their subject matter different sciences face different methodological problems and that their theories take differenl forms, It is simply to say that these differences do not map on to the distinction between historical and DoD] storical. The investigation of the past does not have a sei of common features that sets it ofl from other lofirs of iDquiry and, hence, there is no reason to thinl that biologists can learn more from the philosophy of history lban they can from the philosophies of other sciences witl cognate epistemic prcblems, be the sciences historical or nonhistorical. To return to tha quesliod motivating this chapter, "What's so

66

Rachel

In)non

What's So Special about the Pa.tt?

67

special about the past?",

suggest that the answer, from an epistemic poiDt

wtM[,

lV.

1E3?.

Irino? of tL tnttll.tiw S.i.rc$.1Kls lanik\t Pdv6

of view, is "Absolutely rcthing."

References Dturo. A. C. 1985. N@an, o^.1
ropl,y

tvl;tt , M.

1965.

Fdn.lationt of

fi

qicd rrNLdge.

*s Ydt od L

on:

o/ri$o,y,
b.

tust Dubustd in j9641, N.w yort*

Ktukdq.. ttndtdc.

the

iDtesd

Coiubi, UnikKii pE<

otAna^ti.al phild

Dray,w.,el,
oI

Phitdophi.alAnot\is
E pt@ci@ @d

qd i ot!. Ne* yqk Harper &d Rw

HeDper. C. [1942] 1%5.

ki.ttil.

Th. tu ctio6 of s.leral I.\E ir hLtorv. 23142 R.dinted bA,Ncll Ok{ E sdrt k ftiroropit oI Sctei(?: N.w yort: Tt.

H.dpel,..C. I 062. Epbrar iotr h *iene dd hi.tory. I! Fd tie6 oI Sci.h.. @d phitorophr, ei_ R. c. Colodry. Pn$bust: UdErsiry of pnLsbugb pres.

Hu[, D. 19]5. central subj.cls dd historic.l nanarivs. Hiriory

Kn.her, P..

u.t

Th.ory ]'425174.
of

udw.saldon.&.199.Sci nlr EqluatiN. Mimqpoli$ uiikairy

snaefrte

Mia.

Ynts,D.197.7h.

ol cdrlosr. Daltd: soutt.m Methodisr Ulircrsiry prc$.

kudar
kudaa

R-

ol

s.iac.
1987

loaz The role of berbodolo6, in

13t?r5-50.

Lr.[r

gEolos,. Stu/ti!.

in Histry 6tt phitNophy

x.

I@

Uinetogt

Chjca8o: UDiw6ity of Chicaso

prss.

A@tW:

The

foundai@! ol a S.i.nc. b5ABJ1.

Lyeq C. 1&1G1833. P4}.irl.r a/ Ceolo&r. 3 voL.

ndoa: Mllrray.

Mddclb!@, M.
Nagel, E.-1961.

1977. Th.

UDiBsity Pres!,

Akatrdr ol Hilnoical(iorL4r. Aattidore Jotu

Hopkins

Ydk Hdmut.
os.

7h. Sr!.tuE oI5.1tu . poblew b be Lqic ol Sciennfu Erytbotioa.

Br,@ x'd

w.rl.l

N*

O'H
O

a,

R. 1988a. Hmag. to clio, or, tddd Syst matic Zoolry.371r42-55.

1o88b. Problens

hijtori..l philGophy fo. erclutiolary biol-

Hda, R.

in Oe

Z@lorist 28:744A,
Roth, P. 1988.

trdatiE pre*darion of ercturionary history..lua..d

of ti.tory. Itisto,! @tt Thzo,y
sd,t{.!D-ia_

Nd.atir expldatiotu: Ti. 6e

ntt3.
ol phila

Rott, P.

1989. How

Eraiives a\plain. sociol R.s.Mh

and rh. tb.ory of

Ru*, V. l97l.

NudiE .xplmlion

.!otuti6,

C@.dian

ldnol