Culture History v. Cultural ProcessKent V. Flannery
10J
102
Archaeology and Ethnology at Harvard Uni-versity-has written a monumental synthesis~fNew World prehistory (19660). ThereIS nothing like it. Recently we have badseveral edited volumes on the ew Worldwith contributions by regional specialists,but this book is written cover to cover byone man. Thus the inevitable lack of first-~and familiarity with certain areas is par-tlall~ offset by the advantage of having oneconsistent approach and writing stylethroughout. Although aimed at the student,the book's costly format almost prices itout of the student range. I t is a centerpiecefor the coffee table of the archaeologicalfraternity, at least until an inexpensive pa-perb~ck edition can be produced.. Willey's archaeological career is reflectedIII monographs and articles on every majorl~nd mass of the New World, from the re-£on of the Woodland culture in the U.S.
ortheast to the Maya area, the shellmounds of Panama and the coastal borderof th~ Andean civilization. He is a perennialfavorite who for a variety of reasons hasnev~r come under attack. One reason is hisavo~~ance of anyone polarized theoreticalposition, the other is his adaptability in theface of continual change. While other mem-bers of the establishment have clenched theirfists and gritted their teeth when their for-ffi:rly useful theories dropped from favor,Willey has shown no such hostility' youngerarchaeologists sense be would rafuer join
them than lick them. And he is always freeto join them as long as he maintains novested interest in an) comprehensive theorythat needs defending.
This book, well organized from the pri-mary literature and from constant conversa-tions with Willey's COlleagues, is no excep-lion. II is unlikely to stir up controversyexcept where Willey commits himself toone of a series of possible theories proposedby others-for example, siding with EmilW. Haury rather than Charles C. Di Pesoon the interpretation of the U.S. Southwest,or with Henry B. Collins rather than Rich-ard S. Mac eish on the American Arctic.It is not Willey's aim to intrude his owntheories into the synthesis. Indeed, he tellsus that he is "not demonstrating or cham-pioning anyone process, theory or kind ofexplanation as a key (0 a comprehensiveunderstanding of what went on in prehistoricAmerica." Clearly Willey feels that it wouldbe misleading to do more than present thestudent with the facts as most of his col-leagues agree on them in 1966. Hence "theintent of Lhis book is history-an introduc-tory culture history of pre-Columbian Amer-ica."
This statement by Willey makes it ap-propriate to consider one of the current theo-retical debates in American archaeology:the question of whether archaeology shouldbe the study of culture history or the studyof cultural process. In view of this debateit is interesting LO note that in practicallythe same paragraph Willey can brand hisbook "culture history" and yet argue thathe is "not championing anyone point ofview."
Perhaps 60 percent of all currently am-bulatory American archaeologists are con-cerned primarily with culture history; thisincludes most of the establishment and nota few of the younger generation. Another10 percent, both young and old, belong towhat might be called the "process school."Between these two extremes lies a substan-tial group of archaeologists who aim their
Culture History v. Cultural Process:A Debate in American Archaeology
A DOMIN ANT characteristic of Americanarc~aeology has been its long history of re-action to American ethnology. When eth-nology was little more than the collecting ofspears, baskets and headdresses from theIndians, archaeology was little more thanrecovery. of arti~acts. Whcn ethnologv in-creased Its attention to community structure,archaeology responded with studies of set-tlement pattern-an approach in which Gor-don Willey was an innovator. Publicationof works by Julian H. Steward and otherson "cultur~ ecology" was answered by greatarchaeolo¥,lcal emphasis on "the ecologicalappro~ch. When the concept of culturalevolution ~merged triumphant after yearsof sup~resslOn. archaeology showed great in-teres~ 10 .evolutionary sequences and in theclassification of "stages" in the h. uman ca-reer. The lOteraction of these two di . I". ISC1Pmeshas been mcreasen by the fact that in theU.S. both are housed in departments ofanthropology; as Willey remarked some 10years ago. "~~erican archaeology is an-thropology or It ISnothing."
And now, in 1966, Willey - Bowditch Pro-fessor of Mexican and Central A .mencan
fire freely at both history and process. Andalthough Willey himself belongs to thisgroup, his Introduction to American Archae-ology also constitutes a massive restatementof the accomplishments of the culture-history school.
Most culture historians use a theoreticalframework that has been described as "nor-mative" (the term was coined by an ethnolo-gist and recently restressed by an archaeolo-gist). That is. they treat culture as a bodyof shared ideas, values and beliefs-the"norms" of a human group. Members of agiven culture are committed to these normsin different degrees-the norm is really atthe middle of a bell-shaped curve of opin-ions on how to behave. Prehistoric artifactsare viewed as products of these shared ideas.and they too have a "range of variation"that takes the form of a bell-shaped curve.
In the normative framework cultureschange as the shared ideas. values and be-liefs change. Change may be temporal (asthe ideas alter with time) or geographic (asone moves away from the center of a par-ticular culture area. commitment to certainnorms lessens and commitment to othersincreases). Hence culture historians havealways been concerned with constructing"rime-space grids"-great charts whose col-umns show variation through the centuries.Some have focused an incredible amount ofattention on refining and detailing thesegrids; others have been concerned with dis-covering "the Indian behind the artifact" ~reconstructing the "shared idea" or "mentaltemplate" that served as a model for themaker of the tool.
While recognizing the usefulness of thisframework for classification, the processschool argues that it is unsuitable for ex-plaining culture-change situations. Membersof the process school view human behavioras a point of overlap (or "articulation")between a vast number of systems, each ofwhich encompasses both cultural and non-cultural phenomena-soften much more of thelatter. An Indian group, for example, may
THE THEORETICAL BASE Culture History v. CullufBI ProcessKent V. Flannery
104 105
participate in a system in which maize isgrown on a river floodplain that is slowlybeing eroded, causing the zone of the bestfarmland to move upstream. Simultaneouslyit may participate in a system involving awild rabbit population whose density flue-tuates in a In-year cycle because of preda-tors or disease. It may also participate ina system of exchange with an Indian groupoccupying a different kind of area, fromwhich it receives subsistence products atcertain predetermined times of the year; andso on. All these systems compete for thetime and energy of the individual Indian; themaintenance of his way of life depends onan equilibrium among systems. Culturechange comes about through minor varia-tions in one or more systems, which grow,displace or reinforce others and reachequilibrium on a different plane.
The strategy of the process school is~herefore to isolate each system and studyIt as a sepa~ate variable. The ultimate goal,of course, IS reconstruction of the entirepattern of articulation, along with all re-lated systems, but such complex analysishas so far proved beyond the powers of theprocess theorists. Thus far their efforts havenot produced grand syntheses such as Wil-ley'S.but only.small-scale descriptions of thedetailed workings of a single system. Bythe~e methods, however, they hope to ex-~laIn,. rathe~ than merely describe, varia-nons 10 prehistoric human behavior.
S? far the most influential (and contro-versial) member of the process school habeen Lew.is R. Binford of the University o~Ne."' MeXICOat Albuquerque, and it is inter-estmg to note that Binford's name is con-fin.ed t~ a single footnote on the last page ofWilley s text. It is Binford's contention thatculture historians are at times stopped shortof "an explanatory level of analysis" bthe norm~tive framework in which they co~struct their classifications. Efforts to reco _struct ~he "shared ideas" behind artifa:tpopulatIOns cannot go beyond what B" f d
II " I III orca s pa eopsychology" -they cannot cope
with systemic change. And where Willeysays that "archaeology frequently treatsmore effectively of man in his relationshipsto his natural environment lhan of otheraspects of culture," Binford would protestthat most culture historians have dealt poorlywith these very relationships; their model of"norms," which are "inside" culture, andenvironment, which is "outside." makes itimpossible to deal with the countless sys-tems in which man participates, none ofwhich actually reflect a dichotomy betweenculture and nature. The concept of cultureas a "superorganic" phenomenon, helpfulfor some analytical purposes, is of little util-ity to the process school.
As a convenient example of the differ-ence in the two approaches, let us examinethree different ways in which American ar-chaeologists have treated what they call"diffusion"-the geographic spread of cul-tural elements. It was once common to in-terpret the spread or such elements by ac-t~al mi~rations of prehistoric peoples (aView, still common in ear Eastern archae-ology, that might be called lhe "Old Testa-ment effect"). The culture historians at-tacked this position with arguments that itwas not necessary for actual people to travel-just "ideas." In other words, the normsor one culture might be transmitted to an-other culture over long distances, causinga change in artifact styles, house types andso on. A whole terminology was worked outfor this situation by the culture historians:they described cultural "traits" that bad a"center of origin" from wbich they spreadoutward along "diffusion routes." Along theway they passed through "cultural filters"that screened out certain traits and let otherspass through; the mechanics of this processwere seen as the "acceptance" or "rejection"of new traits on the part of the groupthrough. whose filter they were diffusing. Atgreat distances from the center of originthe traits were present only in attenuatedform, haVing been squeezed through so manyfilters that they were almost limp.
Since process theorists do not treat agiven tool (or "trai t" ) as the end productof a given group's "ideas" about \\ hat atool should look like but rather as one com-ponent of a system Ihat also includes manynoncultural component, thc:) treat diffusionin different willllo. The process theorist is notultimately concerned \lolth "the Indian be-hind the artifact" but rather \\ irh the s)'Stembehind both the Indian and the artifact: whatother components dot'S the system have,what energy source keeps it going, whatmechanisms regulate it and so on? Oftenthe first step is an attempt to discover therole of the trail or implement by determiningwhat it is funcLiona.J1l associated with; someprocess theorists have run extensive linear-regression analyses or muhivariant factoranalyses in order to pick up clusters of ele-merus thai vary with each other in "nonran-dom" ways. When such clustcrings occur,the analyst postulates a sysrem-aools X, Y,and Z are variables dependent on one an-other, can tituting a functional tool kit thatvaries ncnrandomly with some aspect of theenvironment, such as fish, wild cereal grains,white-tatted deer and so on. By definitionchange in one part of a system produceschange in other parts; hence the processtheorists cannot view artifacts X, Y, and Zas products of cultural norms, to be ac-cepted or rejected freely at way stationsalong diffusion routes. When such clementsspread, it is because the systems of whichthey are a part have spread -often at theexpense of other systems.
Thu the archaeologist James Deetz re-cemly presented evidence that the spread ofa series of pottery designs on the GreatPlains reflected not the "acceptance" of newde5igns by neighboring groups but a break-down of the matrilocal residence pattern ofa society where the women were potters.Designs subconsciously selected by thewomen (and passed on to their daughters)ceased to be restricted to a given villagewhen the matrilocal pattern collapsed andmarried daughters were no longer bound to
reside in their mothers' villages. In this case,although each potter obviously did have a"mental template" in her mind when shemade the pot, this did not "explain" thechange. That spread of design could only beunderstood in terms of a system in whichdesigns, containers and certain female de-scent groups were nonrandomly related com-ponents. The members of the process schoolmaintain that this is a more useful explana-tory framework, but even they realize thatit is only a temporary approach. They arebecoming increasingly aware that today's hu-man geographers have ways of studying dif-fusion that are far more sophisticated andquantitative than anything used by contem-poraryarchaeologists.
One other example of the difference inapproach between the culture historian andthe process theorist is the way each treatsthe use of "ethnographic analogy" in ar-chaeological interpretation. The culture his-torian proposes to analyze and describe aprehistoric behavior pattern, then searchthe ethnographic literature for what seemsto be analogous behavior in a known ethnicgroup. If the analogy seems close enough,he may propose that the prehistoric behaviorserved the same purpose as its analogue andthen use ethnographic data to "put flesh onthe archaeological skeleton."
The process theorist proposes a differentprocedure. Using the analogous ethnic group,he constructs a behavioral model to "pre-dict" the pattern of archaeological debris leftby such a group. This mod~1 is then testedagainst the actual archaeological traces of theprehistoric culture, with the result that athird body of data emerges, namely the dif-ferences between the observed and the ex-pected archaeological pattern. These differ-ences are in some ways analogous to the"residuals" left when the principal factorsin a factor analysis have been run, and theymay constitute unexpectedly critical data.When the archaeologist sets himself the taskof explaining the differences between the ob-served archaeological pattern and the pattern
THE THEORETICAL BASE Culture History v. Cultural ProcessKent V. Flannery
106 107
predicted by the ethnographic model, he maycome up with process data .not obtainedthrough the use of analogy alone.
Willey is certainly alert to the current de.bate, and although he summarizes the NewWorld in a predominantly culture-historyframework, he concludes Volume I with adiscussion of the hopes and promises of theprocess school. These he leaves for the fu-ture: "1 shall be less concerned with processor a search for cultural 'laws,'" he says,"than with at times attempting to explainwhy certain cultural traditions developed,or failed to develop." Certainly the processschool would argue that he cannot explain,within a culture-history framework, why suchtraditions developed or failed to develop;yet, as he explicitly states, explanation is notthe purpose of this volume but rather his-tory.
Let us hope, as Willey seems to, that thereis a place in American archaeology for bothapproaches. Certainly we can use both thehistorical synthesis and the detailed analysisof single processes. By no stretch of the im-agination do all process theorists proposeto reject history, because it is only in the un-folding of long sequences that some proc-esses become visible.
In fact, what does the difference betweenthe two schools really amount to? In termsof the philosophy of science, I believe theprocess approach results in moving "de-cisions" about cultural behavior even fartheraway from the individual. It is part of atrend toward determinism that the culturehistorians began.
It was once common to hear human his-tory explained in terms of "turning points,"of crucial decisions made by "great men."This view proved unacceptable to the culturehistorians, with their normative frameworkof shared i?ea.s, values, and beliefs. Theyargued convlDclOglythat this body of sharednon~s ~~termined the COurse of history-nott~e mdiVldual, who was simply a product ofhl~ . culture. PO~Si~l~ the most devastatingCntlque of the mdlvldua[ as decision-maker
was due to Leslie A. While, who in onebrilliant polemic concluded Ihal the courseof Egyptian history and monOlhe.ism wouldhave been the same "even had IthnaloDbeen a bag of sand,"
Now the process school would lite to movecrucial decisions still farther from the indi-vidual by arguing thai S)'SICms, once set inmotion, are self-regulaling to the point wherethey do not even necessarily allow rejectionor acceptance of new traits by a culture.Once a system has moved in a certain direc-tion, it automatically see up Ibe limitedrange of possible moves it can make at thenext critical turning point. This view is 001
original with process-school archaeologists-it is borrowed from Ludwig \00 Bertalanf-Iy's framework for the de\'eloping embryo,where systems trigger behavior at criticaljunctures and, once they bave done so. can-not return to their original peuem. The prcc-ess school argues thai there are systems sobasic in nature that they can be seen cperat-ing in virtually every field-prebistory notexcepted. Culture is about as powerless todivert these systems as the individual is tochange his culture.
Obviously individuals do make decisions,but evidence of these individual decisionscannot be recovered by archaeologists. Ac-cordingly it is more useful for the archaeolo--gist to study and understand the system,whose behavior is delectable ever and overagain. ObViously this approach is 100 deter-ministic for some purposes, but for others itis of great theoretical value.
But then if both historical and processualapproaches are useful, why should there bea debate at all? r believe the debate existsbecause of two basicaUy different altitudes to-ward science.
The previous generation of archaeologists,who did mostly culture history but also laidthe foundations for the process school, wereoften deathly afraid of being wrong. Manyo~ them felt (and many still feel) that if wew~ll only wait until all the facts are in theyWill speak for themselves. They spoke in awe
of the incompleteness of the archaeologicalrecord and of the irresponsibility of specu-lating on scanty data. Somehow they seemedto feel that if the) could get together a fewmore potsherds, a few more: pr~jectile. pointsor a few more architectural details, their con-elusions would be unshakable. There has notbeen, however, any convincing correlationbetween the quentiues of deta ~ey amassedand the accuracy of their conclusions.
The process theorists assume that "truth"is just the best current hypoth~sis, a~d thatwhatever they believe now will ~Iu~a~elybe proved wrong. either within their hfetl~eor afterward. Their ..theories" are not likechildren to them, and they suffer less ~raumawhen the theories pro\e "wrong." Their con-cern is with presenting developmental modelsto be tested in the field, and they have notedno consistent relationship between the use-fulness of a given model and the absolutequantity of data on which it is based. To be
< body ofuseful a model need only organize adisorganized data in such a way that hypothe-
ses can conveniently be tested, accepted,modified or rejected. Thus the process schoolwill continue to present model after modelon the basis of returns from the first fewprecincts, and at le.ast some of the culturehistorians will continue to accuse them ofbeing "hasty," "premature" and "irrespon-sible." And the issue will be settled yea:sfrom now by another generation that willprobably not belong to either school.
Willey's synthesis sums up nearly 100years of American archaeology, and !t.comesat the start of one of the most excltlO.g .ar-
< t begun My predictionchaeological eras ye .for the next decade is that we shall see gen-eral systems theory, game theory and loca-ticnal analysis all applied successfully toAmerican archaeology in spite of the loudestmuuerings of the esta?lishm.e?1. I also pre~diet that, in spite of his decIsIOn. to co~cen
hi own efforts on producing reliableirate IS bculture history, we shall hear all these su -versive approaches applauded by GordonWilley.
