What Can Archaeology Do With Boyd and Richerson's Cultural Evolutionary Program? more

30 The REVIEW ofARCHAEOLOGY Vol. 26, Number 2 Evolutionary Archaeology Invited Review: What Can Archaeology Do With Boyd and Richerson's Cultural Evolutionary Program? By Benjamin Marwick Australian National University Not By GenesAlone: How Culture Transformed Human Evolution (2005) PETER ]. RICHERSON and ROBERT BOYD. University of Chicago Press, Chicago. xi + 342 pages, 3 halfrones, 5 line drawings. $30 .00 ISBN 0­ 226-71284-2 In a famous letter, the economist Alfred Mar­ shall outlines a method for economic theorising: "(1) Use mathematics as shorthand language, rather than as an engine of inquiry (2) Keep to them till you have done (3) Translate into Eng­ lish (4) Then illustrate by examples that are important in real life (5) Burn the mathematics (6) If you cannot succeed in 4 then burn 3." (Pigou 1925). If Marshall's method is relevant to the way Boyd and Richerson work, then their new book is evidence that their theorising has reached an advanced stage. In their new book Not by Genes Alone (hereinafter, NBGA) there are none of the dense maths that distinguished their influential book Culture and the Evolution­ ary Process (1985), instead there are numerous examples drawn from the human sciences. The main point of this new book is to show that Darwinian evolutionary theory and methods are essential and productive tools for the analy­ sis of human culture. This is a theme that Boyd and Richerson have been promoting since the late 1970s, but NBGA presents a more accessi­ ble account of their cultural evolutionary pro­ gram and outlines a manifesto for future research. The book is aimed at readers in social science and humanities departments, with no graphs, only a single equation buried in the end­ notes, axiom-like chapter headings, and case studies drawn from across the human sciences. The publication of this new synthesis of their ideas provides a good opportunity to review the main arguments of Boyd and Richerson's work as described in NBGA and evaluate the impact their program has had on archaeological research. In their invitation to the social sciences to see wha t evolutionary biology has to offer, Boyd and Richerson build the conceptual framework of NBGA around two ideas that are the core of their program. The first idea is population thinking. Amongst biologists this is a funda­ mental ontology that has made all the difference since its introduction by Darwin. Before Dar­ win's publications, most biologists considered a species as an immutable, discrete, Aristotelian kind and variation as illusory. After Darwin, this idea was replaced with the view that species are populations of individuals showing continuous variation and individual variation within a species is the result of ongoing processes of mutation. This shift from essentialist to materialist thinking has thoroughly permeated biology and Boyd and Richerson argue that a similar shift would be advantageous for the social sciences. Their key point is that within and between cul­ tural groups there is a continuous range of vari­ ation in the cultural information in any individ­ ual's head. Culture is defined as 'any kind of mental state, conscious or not, that is acquired or modified by social learning and affects behav­ iour' (p, 5). This definition of culture distances Boyd and Richerson from sociobiological approaches to cultural evolution that tend to define culture as the expression of naturally selected adaptive genes (Ehrlich and Feldman 2003). Thinking about the population proper­ ties of culture is not unique to Boyd and Rich­ erson and ~o the foundation for a variety of models of evolution and human behaviour (Bryant 2004). The uniqueness of Boyd and Richerson's program comes from the second major idea: that of the many forces shaping cul­ tural evolution, some of the most powerful are those arising from the psychology of individuals that makes them more likely to adopt some variants of culture than others. CULTURAL TRANSMISSION BIASES AND ARCHAEOLOGICAL ApPLICATIONS These forces occur during the social learning of culture and are described as cultural trans­ m ission biases, occurring when people choose (consciously or otherwise) to adopt certain cul­ tural variants rather than others. In N BGA Boyd and Richerson describe three kinds of transmission biases using examples from previ­ ously published social science literature. The first bias is the 'content-based bias' that occurs whe n individuals are more likely to learn, remember, and teach some cultural variant because of its content. Content-based bias can Fall 2005 The REVIEW ofARCHAEOLOGY 31 result from (not necessarily conscious) calculations of costs and benefits associated with cultural variants or when the content of the variant more easily lends itself to learning, memorisation, and teaching than alternative variants. The second bias is 'frequency-based bias', occurring when the commonness or rarity of a cultural variant is the criteria determining transmission. When the common variant is more frequently transmitted this bias is called 'conformist bias'. The third bias is 'model-based bias' that occurs when a variant is transmitted because of its association with a suite of other attributes associated with individuals exhibiting the variant. Finally, there is 'guided variation', where individuals acquire new cultural variants by copying ~ existing behaviours and then modifying them through trial and error to suit their own needs. These concepts were first presented by Boyd and Richerson in the 1980s and are often invoked in archaeological studies but less frequently used to systematically investigate archaeological problems. Good examples of a thorough employment of these concepts are found in the work of Bettinger and Eerkens (1997; 1999) who use metric analyses of late Holocene stone artefact assemblages from the Great Basin of western North America to explore the consequences of different cultural evolutionary processes. In their 1997 study they analyse a large sample of projectile points (n = 5285) to test predictions about the effects of different types of cultural transmission on metric variability. They expect that guided variation and content-based bias will be important during times of low population densities and technoorganizational complexity because competing variants are easily compared by field testing and individual experience. When the population grows and/or technology becomes more complex they expect frequency-based and modelbased biases to be more important because individual field testing of variants is inefficient compared to relying on social transmission of pre-tested variants. These expectations allow them to hypothesize that complex point shapes will have less metric variation than simple forms, that arrowheads will be less variable than dart points and that long-lived forms will be more variable than briefly appearing forms. These expectations receive only equivocal support from the data. As predicted, complex point forms show less metrical variation than simple forms, and arrowheads show less variation than dart points. However, 81 % of metric variation is strongly correlated with artefact size (r = 0.899), meaning that differences in types of cul­ tural transmission have only a small effect. Inspired by Bettinger and Eerkens, Shott (1997) undertook a similar test of these social transmission models using metric attributes of stone points from late Holocene contexts of the North American Midwest and found that dif­ ferences in cultural transmission have an even smaller effect on lithic variation than observed by Bettinger and Eerkens. More encouraging results come from Bet­ tinger and Eerkens' (1999) study of changes in metric variables of stone points during the introduction of bow and arrow technology to eastern California and central Nevada at around AD 300-600. Using an approach similar to their earlier study, they equate guided transmission with high metric variation and low correlation between different metric attributes (such as mass and basal width). Lithic assemblages with metric variables that are less variable and more highly correlated are equated with model-based bias. Their analysis shows that assemblages from eastern California have poor correlations of basal width and mass, suggesting that bow-and­ arrow technology was probably introduced and spread by guided variation. Conversely, in cen­ tral Nevada, the new technology was probably introduced and maintained by indirect bias because the metric attributes are strongly corre­ lated. Bettinger and Eerkens interpret this data to mean that eastern Californian groups acquired the bow and arrow from distant and unfamiliar neighbours, and with limited contact they had to develop and customise the bow and arrow technology largely by trial and error. The opposite seems to be true for central Nevada where the appearance of bow and arrows was probably a result of faithful copying, suggesting closer social contacts with the donor group. They further conclude that these differences in social transmission indicate substantial differences in social organisation and hunting behav­ iours in the two regions. Despite their innovative statistical approaches and the success of their second study, Bettinger and Eerkens' method has not been widely adopted in other lithic studies, nor is it cited in 32 The REVIEW ofARCHAEOLOGY Vol. 26, Number 2 NBGA. This is probably becau·se stone artefact assemblages in general (especially non-point types) are somewhat lim ited in shape and size variation, as well as limited in the potential for decoration that is independent of the artefact's physical performance. This is because stone artefact production involves a one-way and con­ tinuous decrease in size and often substantial changes in morphology over the life of the arte­ fact by removal of mass from the original piece (Shott 2005). More often, evolutionary studies of stone artefacts follow the evolutionary ecolo­ gy approach (also known as human behavioural ecology) that provides models to associate tech­ nological attributes of lithics with behaviours relating to individual fitness in given contexts (e.g., Bamforth and Bleed 1997; Bousman ~005; Kuhn 2004). This approach interprets akmblage change as evidence of adaptive vari­ ation in behaviours responding to changing social and ecological conditions. It is based on the assumption that ' natural selection has designed individuals to respond to changing conditions in a way that yields the greatest pos­ sible benefit for the individual's survival (Boone and Smith 1998). This is an application of evo­ lutionary thinking that is distinct from Boyd and Richerson's program because evolutionary ecology generally attributes archaeologically observed changes in behaviour to phenotypic plasticity (O'Brien and Lyman 2002) rather than the ongoing selection processes that inter­ est Boyd and Richerson. We might expect that Boyd and Richerson's approach would have grea ter influence in archaeological studies of artefacts that preserve information about learn ing and information transmission and are relatively unaffected by loss of mass and major changes in morphology during thei r use-lives. T hese kinds of artefacts can be made with additive technologies such as ceramic, metal, and clot h production where variation unrelated to the artefac t's physical per­ forma nce is possible through the use of moulds, patterns, and decorations. Ethnoarchaeological studies of ceramic production suggest that suc­ cessful transmissio n of ceramic traditions requires a lon g-lastin g relationship betwe en teachers and learners (Shennan and Steele 1999), providing an ideal context to examine the effects of tra nsmission biases. Indeed, stud­ ies of cultural evolutionary processes using ceramic assemblages have been much more suc­ cessful than the lithic studies. Although not explicitly framed in Boyd and Richerson's terms, Neiman's (1995) sophisticated and influential study of variations in decorations on cooking pots across 35 Hopewellian assemblages in Illi­ nois identified changes in the levels of inter­ group social transmission that relate to changes in the level of long-term residential movement of potters between groups. Neiman's study drew on a specifically archaeological adaptation of evolutionary theory proposed by Dunnell (1978; 1980) that is quite independent of Boyd and Richerson's program. Despite this different intellectual heritage, Neiman's work is signifi­ cant because his approach is basically similar to Boyd and Richerson, and his methods have been used by others who have more closely fol­ lowed Boyd and Richerson's scheme for under­ standing cultural evolution and transmission. Neiman's work is important because he showed, like Boyd and Richerson, how methods derived from population genetics can be used to generate expectations about human behaviour. Neiman used biological models describing neu­ tral variation in genetic populations to explain stylistic diversity in archaeological assemblages. First, he builds mathematical models to describe the effects of population size, drift, and innova­ tion rates on within- and inter-assemblage diversity. Drift is defined as 'sampling error' that occurs during social transmission and reduces variation, while innovation is defined as a con­ stant and selectively neutral source of variation that appears during social transmission. In brief, his models show that assemblage diversity is proportional to assemblage and population size, that assemblage diversity will increase during times of increased population and increased between-group interaction, and that under con­ ditions of drift and neutral variation, within­ assemblage variation is inversely proportional to between-assemblage variation. Second, Neiman analyses the diversity of the 26 different types of lip exterior decoration of ceramic cooking pots in 35 assemblages. Following Dunnell, lip dec­ oration is identified as an attribute that is 'sty­ listic' (attributes with variants that have equal value to an individual's reproductive fitness and change by drift, typically attributes unrelated to an artefact's physical performance, such as dec­ oration) rather than 'functional' (those that are Fall 2005 The REVIEW of ARCHAEOLOGY 33 under selection because they relate directly to the survival and reproduction an individual, such as performance related attributes). This distinction between function and style separates Neiman from Boyd and Richerson in two important ways. Firstly, Neiman does not con­ sider learning and social transmission biases to be relevant selective forces (Lipo et aI. 1997); and secondly, he does not allow for the possibil­ ity that style and function may have complex co-evolutionary relationships. For example, sty­ listic attributes may be signalling systems that influence an individual's reproductive success (Bird and Smith 2005). Neiman's third step is to group the assemblages into seven chronological units based on sherd thi ckness, which decreases over time. He then calculates the observed diversity values of lip types in each assemblage for each time period from the archaeological data and calculates expected values using the mathematical models. The results . show that within-assemblage diversity is low during the Early Woodland Peri­ od (before 200 BC) , increasing to a peak in the Middle Woodland Period (200 BC-AD 400) and declining again in the Late Woodland (AD 400-800) . Conversely, just as predicted, between-assemblage diversity is high during the Early Woodland, then becomes low, and is final­ ly high again in the Late Woodland. The simi­ larity between the observed results and the modelled results indicates that the neutral model accurately describes the archaeological evidence. Following his models, Neiman inter­ prets this as evidence that the highest levels of social interaction were occurring during the Middle Woodland period and the lowest levels occurring during the Late Woodland period. -Further, Neiman suggests that high levels of social interaction during the Middle Woodland relates to an increase in long-term residential movement of potters, a conclusion supported by evidence of gift exchange relations in the Middle Woodland that the n ceased in the Late Woodland. Given the elegance, robustness, and utility of Neiman's study, it is surprising his method has not become widespread. Koh ler et al. (2004) suggest the dogm atic separation of function and style by Neiman may have dis­ courage d ot hers who do not see such as clear distinction. There are two ceramic studies that have been inspired by Neiman's work and both reject the separation of functional and stylistic attributes as well as Neiman's conclusions that variation in diversity results only from innovation and drift. Instead, they adopt Boyd and Richerson's model of social transmission as a source of the variation that Neiman's neutral model cannot explain. Shennan and Wilkinson (2001) analysed chang­ ing patterns in the frequency of 35 different types of decorative bands on the bodies of ceramic vessels from two settlements of the early Neolithic Linienbandkeramik (5300 to 4850 cal BC) in western Germany. They use Neiman's model to generate expected and observed diver­ sity values from the archaeological data and find that the expected values of inter- and between­ assemblage diversity do not match the archaeo­ logical observations, so they reject the neutral model as the source of variation in LBK ceram­ ics. In the early LBK phases the diversity of ceramic assemblage decoration was less than would be expected under the neutral model, while in the later phases it was greater. To explain this they model ceramic variation as a cultural 'quasi-species' subject to mutation and selection. They propose that during early LBK phases assemblage diversity was limited by conformist bias resulting in the rejection of novelty. The increased diversity in the later phases was caused by a pro-novelty negative frequency-dependent bias (Bentley and Shennan 2003). They explain that this increase in novel decorations was an assertion of identities during increasing popula­ tion densities, suggesting a decline in inter-site interaction. Support for this interpretation comes from a cladistic analysis of relationships betwee n a larger samp le of sites in the same region showing that the diversities of late phase LBK assemblages are more closely related to their ancestral assemblages than interaction between neighbouring sites (Collard and Shen­ nan 2000). Despite the elegance of this study and its success in using transmission biases to explain archaeological data in behavioural terms, it on ly features in a footnot e in N BGA, perhaps due to Boyd and Richerson's preference to use more contemporary examp les from political sci­ ence, linguistics, sociology, and econom ics. Another instructive archaeo logical application of social transmission models is provided by Kohler et al, (2004) wh o examine changes in cera m ic diversity to explo re the effect s of 34 The REVIEW ofARCHAEOLOGY Vol. 26, Number 2 increasing settlement size and aggregation in the pre-hispanic (AD 1175-1315) Puebloan south­ west US. They measured changes in the diversi­ ty of 83 different styles of black-on-white wares from eleven sites ranging in size from six to 65 rooms. Neiman's methods were used to examine the neutral model's predictions that the diversi­ ty of stylistic variants should increase as popula­ tion size increased from hamlets to villages. Like Shennan and Wilkinson, they find that the neu­ tral model does not explain the observed diver­ sity. Diversity is much less than the model pre­ dicts and Kohler et al. reject the possibility of increased specialization of ceramic production (fewer people making more vessels) because the~is no corresponding reduction in coeffi­ cients of variation of metric variables. Instead they suggest that the low ceramic design diver­ sity should be explained by differences in the social transmission of designs . They consider the possibility of model-based bias, where pot­ ters would actively imitate the products of one who was particularly prestigious, but they reject it because the relative homogeneity of architec­ ture and other materials at the sites does not suggest greatly differing degrees of prestige. Kohler et aI. conclude that conformist bias is the most significant force, where an individual selects the most common model to imitate. They view the conformity in ceramic design in the later phases as indicative of high levels of within-group cooperation and the punishment of defectors by ostracism. This conformity is argued to be an adaptation to increasing exter­ nal competitive pressures (for access to the best arable land, hunting territories, lithic raw mate­ rials, and probably defence) as population den­ sines Increase. These examples show how archaeologists have adapted the explanatory potential of Boyd and Richerson's cultural evolutionary forces. Appli­ cations to lithic studies have met with varying success but the ceramic studies have been very successful, largely due to Neiman's innovative modelling (although conceived for different purposes) and the intrinsically more plastic and variable nature of ceramics compared to stone. Given that social transmission biases have been shown to be important in prehistoric contexts, we might ask when these forces first became important in human societies. Unlike almost every other author who has written about the evolution of culture, Boyd and Richerson prefer not to equate transmission bias with selection. This is a fine point, but an important departure from cultural evolutionary models like merner­ ics (Blackmore 2000) and cultural virus theory (Cullen 2000). These approaches generally fol­ low the principles of biological Darwinian evo­ lution of descent with variation, equating cul­ tural variants with self-replicating entities such as genes or viruses that are directly subject to selection. Boyd and Richerson distance them­ selves from this kind of analogizing by following Sperber's (19%) idea of non-replicating, non­ particulate cultural variants that are very differ­ ent from genes. In Boyd and Richerson's pro­ gram transmission biases are unique evolutionary forces for selective retention of cul­ tural variants that originate from adaptations generated by natural selection operating on genetic differences between individuals. The adaptive advantage in these biases is that they often allow humans to quickly and reliably acquire adaptive cultural variants without costly and difficult evaluations of the advantages and disadvantages of different variants. Although the adaptive value of culture is generally intu­ itive, the specific reasons offered by Boyd and Richerson for why these biases are adaptive have been criticised for understating the costs of biased transmission (Sterelny 2006). THE ARCHAEOLOGY OF THE EVOLUTION OF TRANSMISSION BIASES Criticisms aside (it is only modelling after all), the ultimate derivation of the transmission bias­ es from natural selection on genes is the key to the dual inheritance programme of Boyd and Richerson. This program holds that humans, unlike other animals, inherit behaviour through two routes: genes and culture. In fact, the sepa­ ration of humans and animals is not as great as Boyd and Richerson suggest, with evidence for cultural groups, social learning, and even con­ formist bias in primate populations (Whiten, Horner, and de Waal 2005; van Schaik et aI. 2003). Similarly, Boyd and Richerson's evolu­ tionary narrative in NBGA has an unrealistical­ ly sharp break with primates and early Homo on one side and modern humans on the other. This distancing of Homo sapiens from extant relations and extinct ancestors does not agree with the available evidence. Their gloss of Homo habilis Fall 2005 The REVIEW ofARCHAEOLOGY 35 does not discuss significant evidence of cogni­ tive abilities that distinguish late Plioc ene hominids from primates, such as the diversity of technical behaviours in Pliocene stone artefact assem blages (Roche et al. 1999; de la Torre et al. 2003) or the patterns of raw material movement that involve much further transfers than demonstrated by wild primates and suggest greater planning depth (Marwick 2003). Mid­ dle Pleistocene hominids are similarly aban­ doned as 'only a bit brainier than the bipedal apes that preceded them' (p. 14 1). Boyd and Richerson more accurately and parsimoniously att ribute the long-term and wide-area similari­ ties of Acheulean handaxes as a result more of genetically transmitted psychology and raw material constraints than cultural transmission (McPherron 2000) although they omit any dis­ cussion of the social, cognitive, and behavioural significance of the geographic dispersal of Mid­ dle Pleistocene hominids out of Africa (Roe­ broeks 2001) qr the life history markers of these hom inids ~ indicate a social organisation more similar to humans th an pr imates (Leigh 2006; Connell, H awkes, and Blurton Jone s 1999 ; Wrangham et al. 1999 ). For Boyd and Richerson, most of the early and middle Pleis­ tocene is little more than a time of natural selec­ tion for larger brain sizes as an adaptation to the highly variable Pleistocene climates. Using com­ parative studies the y point out that larger brains are strongly associated with increased capacity for behavioural flexibility and social learning, suggesting rhat with an increasingly bigger brain early Homo could discover and learn novel behaviours that are adaptive to rapidly changing climates faster than genetic selection could sup­ ply them. They are on firmer ground in their discussion of the gradual appearance of signs of modern humans during the African Middle Stone Age. They note that archaeological evidence of cul­ tural modernity, or cumulative cultural adapta­ tion in their words, emerges from a thin spread of evidence over a large area and long time peri­ od during the Middle Stone Age (d'Errico et al. 2003) . For Boyd and Richerson, this is the time when natural selection for gradual increases in brain size finally culminates in the psychological machinery for social transmission biases. Disap­ pointingly, their discussion of the archaeological evidence for language evolution is overly-con­ servative and dated. There is also no mention of the impressive efforts made by lingui sts to understand the evolution of language (e.g., Jackendoff 2002), especially the important argument that language itself may be a distinc­ tive evolutionary system similar to biological and cultural systems (Brighto n, Smith, and Kirby 2005). Mention of other selection-based evolutionary systems, such as the immune sys­ tem (H ull, Langman, and Glenn 2001 ) and neurons (Edelman 1993) , might make Boyd and Rich erson's idea of cultural evolut ion appear more plausible to skeptics who regard evolutionary processes as exclusive to genetics. After the emergence of cumulative cultural adaptations in the African Middle Stone Age, Boyd and Richerson's theoretical and mathe­ matical modelling suggests that group selection of cultural variations becomes possible and results in cooperation, altru ism, and punish­ ment in small groups. From these group-select­ ed qualities follow the existence of ethnolinguis­ tic tribal institutions that facilitated periodic aggregation of small groups for communal activities. Although there is good archaeological evidence for this, especially in the later Pleis­ tocene (Conkey 1980; Vanhaeren and d'Errico 2006), Boyd and Richerson prefer to focus on ethnographic studies in NBGA. After the emergence of cumulative culture in the Mrican Middle Stone Age, th e next big thing in Boyd and Richerson's account of human histor y is the Holocene, when many human groups gave up the Pleistocene hunter­ gatherer way oflife to pursue domestication and organise themselves into more diverse, complex and larger social groups. In NBGA they address some of these changes in an engaging chapter on cultural behaviours that reduce our genetic fitness, or cultural maladaptations, which they view as inevitable byproducts of cumulative cul­ tural adaptation. In the Pleistocene, natural selection shaped our psychology so that it uses predictive cues-the transmission biases-as short-cuts to quickly generate adaptive behav­ iour, but in larger groups these biases, especially conformist and model-based biases, result in the spread of behaviour that can reduce our genetic fitness. As soon as an individual's behaviour is influenced by cultural variants transmitted by non-kin then non-adaptive cultural variants can spread. This focus on social learning separates 36 The REVIEW ofARCHAEOLOGY Vol. 26, Number 2 Boyd and Richerson from the most serious competing explanation for maladaptations, the evolutionary psychology paradigm. Evolution­ ary psychologists argue that culture derives from genetically-designed neural circuits selected to produce adaptive behaviours during Pleistocene conditions (Barkow, Cosmides, and Tooby 1992). According to the evolutionary psycholo­ gy paradigm, maladaptations result when these innate brain structures produce behaviours in social and ecological situations that differ from the Pleistocene contexts where they were forged. In their view, cooperation and altruism occurs because people erroneously believe they are dealing with kin and basic behaviours related to kin-selection are activated. THE ARCHAEOLOGY OF CULTURAL GROUP SELECTION Group selection is another concept that sepa­ rates Boyd and Richerson from evolutionary psychologists who gen,erally argue that selection only occurs at the level of an individual's genes. Group selection has important implications for cultural evolution with Boyd and Richerson suggesting that when individuals migrate into a new group they can change their behaviour t.o conform to new norms, thus reducing within­ group variance in norms, and maintaining between-group variance. The integrity of the cultural group is preserved because immigrants will preferentially adopt common practices, but the genetic integrity of the group is diminished because offspring acquire their genes from their parents, not from the group. This allows for the proliferation of distinctive cultural groups with long-term, stable cultural traditions. There are a number of empirical studies that apply phyloge­ netic analyses to anthropological and archaeo­ logical data to demonstrate this effect of group selection (Collard, Shennan, and Tehrani 2006). It is surprising that none of these studies are referred to in NBGA, given Boyd and Rich­ erson's commitment to using methods bor­ rowed from biological sciences and the increas­ ing significance of phylogenetic methods in the human sciences (Mace, Holden, and Shennan 2005; Lipo et al. 2005) . These phylogenetic studies focus on the debate about the relative importance of phylo­ genesis (where cultural similarities and differ­ ences among human, populations are primarily the result of a combination of within-group information transmission and branching away from ancestr al llopulations) and ethnogenesis (where cultural change occurs through the bor­ rowing and blending of variants among contem­ poraneous populations) in cultural evolution. Although the studies are not explicitly designed to test Boyd and Richerson's group selection concept, they provide good evidence of cultural group integrity. They follow cladistic methods used in biology and test a null model where new cultural variants arise from the bifurcation of existing ones and a tree diagram (cladogram) is drawn to show the ancestries of a number of variants. Statistical analysis of the relationships shown in the cladogram are used to test the null model by showing the importance of similarities resulting from shared ancestry (homologies) or those due to mechanisms other than shared ancestry (homoplasies). Tehrani and Collard (2002) used this method to examine 90 attrib­ utes on 60 woven artefacts produced between the l Sth and 20th centuries by five groups of Turkmen from Turkmenistan, northern Iran, and northern Afghanistan. A sub-sample was divided into two time periods based on the use of natural or artificial dyes to identify possible changes in the Turkmen material cultural evolu­ tion associated with the Russian colonisation of Central Asia in the 19th century. Tehrani and Collard (2002) aimed to determine if phylogen­ esis or ethnogenesis dominated the evolution of Turkmen textile designs prior to the Turkmen's defeat by Tsarist Russia, and to see if the contri­ butions of phylogenesis and ethnogenesis change following the Turkrnen's pacification and settlement by the Russian colonial authorities. Their results show that prior to Russian coloni­ sation about 70% of the similarities among the assemblages are homologous and approximately 30% are homoplastic, indicating that cultural phylogenesis is more important in cultural evo­ lution than ethnogenesis. After Russian coloni­ sation ethnogenesis becomes about 10% more important (about 60% of the interassemblage resemblances are homologous and about 40% are homoplastic) suggesting an increase in blending of designs and innovation, consistent with historical accounts of increased sedentism and increased production for market sale rather than domestic use. Tehrani and Collard's study is a good example Fall 2005 The REVIEW ofARCHAEOLOGY 37 of how phylogenetic methods demonstrate the integriry of group-level cultural phenomena. This might be expected, with the complexiry of carpet making and the time taken to learn the techniques limiting the potential for diffusion, but phylogenetic analysis of simpler cultural materials also shows the importance of group selection. Collard and Shennan (2000) analysed pottery data from seven early Neolithic settle­ ments in western Germany. Results from four settlements that were continuously occupied throughout the ten-phase period indicated that pottery decorations were generated both by phylogenesis and by ethnogenesis. On the other hand, the three assemblages that were newly founded in the ten-phase period derive from a single ancestral assemblage, suggesting phyloge­ nesis was most important. There is also evidence that phylogenesis is not always as important, or that its signals have been eroded by different rates of innovation (jordan and Shennan 2003), but a review of a-large number of quantitative studies by Collard et al. (2006) shows that in most cases phylogenesis is more than, or at least as important as, ethnogenesis. CONCLUSIONS Boyd and Richerson are not alone in the field of gene-culture co-evolution (Durham 1991; Cavalli-Sforza and Feldman 1981), but are probably the most prolific and influential. Their central thesis is that socially transmitted culture is crucial for a full understanding of human bdhaviour, while at the same time human cul­ t~re has a biological basis and can be considered an evolutionary system in its own right. This review has outlined the most important argu­ ments presented in NBGA and discussed a few examples of how these arguments interact with archaeological research. The contents of NBGA go beyond the scope'of archaeological inquiry and some of these other topics have been dis­ cussed by Mameli (in press) and Machery (2005). This review has shown that Boyd and Richerson have produced explanatory frame­ works that have helped generate substantial innovation in archaeological methods and robust anthropological explanations of past human behaviour. Boyd and Richerson have also suggested narratives of human evolution that are archaeologically testable but require revision to correct their underestimation of pri­ mate and early hominid abilities. Using the example of group selection and the debate about phylogenesis versus ethnogenesis, this review has also shown that archaeology has the potential to provide real-world tests of Boyd and Richerson's models. This lack of purpose­ built real-world tests and examples is a major weakness in Boyd and Richerson's work, and evolutionary approaches to the human sciences generally, which have often been accused of gen­ erating more programmatic statements than useful results (Bamforth 2003; Shott 1997). NBGA goes some way towards employing case studies from anthropology, history, linguistics, and psychology in the service of cultural evolu­ tionary models but lacks the empirical densiry and persuasive rhetorical force of popular writ­ ing in the biological sciences (e.g., Jones 1999; Ridley 2003). This is probably because biology has been so much more successful than anthro­ pology and many related fields of the social sci­ ences over the past 150 years. But why has this been the case? Mesoudi et al. (2006) suggest two answers to this problem. First, biologists rypically are more willing than social scientists to simplify complex systems down to workable assumptions and models that ultimately form the basis of sophis­ ticated explanations. Complex problems in the social sciences remain intractable because researchers object that human culture is too complex to apply simplifying assumptions and methods. Second, biology and its diverse sub­ disciplines are united by the theory of evolution but the social sciences have no such synthesizing framework showing how cultural anthropology, archaeology, psychology, economics, sociology, and history are studying complementary aspects of the same problems. Mesoudi et al. (2004; 2006) argue that cultural evolution has key Darwinian properties and suggest that is has the potential to synthesize the social sciences as it has the natural sciences. As outlined in NBGA, Boyd and Richerson's program is an essential contribution towards this evolutionary synthesis of the human sciences. Boyd and Richerson's work also points us in two important directions where progress is vital before a golden age of evolutionary synthesis can emerge. The first direction is experimental research to show how individual-level mechanisms for acquiring behaviours contribute to phenomena 38 The REVIEW ofARCHAEOLOGY Vol. 26, Number 2 at the population level. M icroevolutionary labo­ ratory studies have bee n the foundation of bio­ logical evolu tion beca use simple versions of com­ plex systems can be controlled and exami ned in ways th at are impossible in field conditions. Lab­ oratory studies of microsocieties are the equiva­ lent tools for examining many aspects of cu ltural tra nsm ission under controlled co nditions. Baum et al. (2004) have shown how traditions can appear and evolve in microsocieties of four peo­ ple under controlled conditions. More of th ese stud ies are needed to test the effects of the differ ­ ent tra nsmission biases identified by Boyd and Rich erson as well as th eorise the effects of power, conflict, and manipulation in social learni ng contexts. Most cultural evolutionary processes observed in th e archaeological record are just long-term pa tterns of behaviour by multiple agents (Flannery 1999), so a better un derstand­ ing of agent-level processes will improve our un d erstan ding of cultural evolu tion. The second direction where progress is necessary is about how the brain works. U nderstand ing the neural basis of learning, memory, and trans m ission is important for knowing how reliab ly Boyd and Rich erson's claim s abo ut learni ng and tra nsmis­ sion bias can be gene ralised. T his is obviously a field distant from m ost archa eologists and is still a long way from p roducing useful results (McGaugh 2000), but will reveal th e m echanics oflearning and ~ormation storage and hopeful­ ly help resolve problem s abo ut how cultural vari­ ants are pre served and replic ated. However, archaeology can continue to test the explanato ry power of cultu ral evolutionary th e­ ory without mi croevolutionary lab or atory stu d­ ies and brain scans ju st as ~arw in managed without molecular gen eti cs and g ame th eory. In fact, even the classical d efinition of th e gen e as a discrete unit of vertical inherit ance th at the modern evolutionary synthesis was founded upon is now old-fashioned. Pearson (2006) describes stu dies revealing complex extragenom ­ ic and horizontal modes of gen eti c inheritance and Mameli (2004) pr esents a len gthy catalogue of evidence for nongenetic selectio n and inheri­ tance in many species. This suggests th at cri­ tiques of analogies between cultural and genetic evolution need to be revised and the two trans­ mi ssion systems, by their di versity and com­ plexity, may have more in common th an previ­ ously suspected. In NBGA Boyd and Richerson confess that their work is similarly incomplete and in need cI 1 evision . This revision will only occur with l ontinued empirical testing and some of this can be done by archaeologists. To sum up, this new book is an accessible introduction to Boyd and Richerson's program of cultural evolution and is recommended to those interested in evolutionary approaches in the human sciences. They do more than just muddy the waters of the human sciences to declare them deep; they offer a method to find more interesting and productive shores. 0 REFERENCES CITED: Bamforth, D.B. and P. Bleed (1997) Technology, Flaked Stone Technology, and Risk. In, Rediscovering Darwin: Evolutionary Theory in Archaeological Explanation, edited by CM. Barton and G .A. Clark: Archaeological Papers of the American Anthropological Association Number 7. Barnforth , Douglas B. (2003) Review of Stephen Shen­ nan . Genes, memes and human history: Darwinian archaeology and cultural evolution (2002) Thames & Hudson, London. Antiquity 77 (298):875 . Barkow, ].H ., L. Cosmides, and] . Tooby, eds. (I992) The Adapted Mind: Evolutionary Psychology and the Generation ofCulture. Oxford University Press, New York. Baum, William, Peter ]. Richerson, Charles Efferson, and Brian Paciotti (2004) Cultural evolution in laboratory microsocieties including traditions of rule giving and rule following. Evolution and Human Behaviour 25 (5):305-326. Bentl ey, R.A. and S.]. Shen nan (2003) Cultura l transmis­ sion and stochastic network growth. American Antiquity 68 (3):459-485 . Betti nger, R.L. and]. Eerkens (I997) Evolutionary imp li­ cations of metrical variation in Great Basin projectile poi nts. In, Rediscovering Darwin: Evolutionary Theory and Archaeological Explanation, edited by C M. Barton and G. A. C lark. America n Ant hro pological Association, Arling­ ton , Virginia. _ _ _ _ (1999) Point Typologies, Cultural Transmission, and the Spread of Bow-and-Arrow Technology in the Pre­ historic Grea t Basin. American Antiquity 64 (2):2~ 1-242. Bird, Rebecca Bliege and Eric Alden Smit h (2005) Sig­ nalling T heory, Strategic Int eraction , and Symbo lic Ca pi­ tal. Current Anthropology 46 (2):22 1-248 . Blackm ore, S. (2000) The Meme Machine. Ox ford Press, O xford . Boone, ].L. and E.A. Sm ith (199 8) Is it evolution yet? A critique of evolutionary archaeology. Current Anthropolo­ gy 39 (Suppl cmenrl . l -i l-f S'Z. Fall 2005 The REVIEW ofARCHAEOLOGY 39 Bousman, C Britt (2005) Coping with risk: Later stone age technological strategies at Blydefonrein Rock Shelter, South Africa. Journal ofAnthropologicalArchaeology 24:193-226. Boyd, R. and P.J. Richerson (1985) Culture and the Evo­ lutionary Process. University of Chicago Press, Chicago. Brighton , Henry, Kenny Smith, and Simon Kirby (2005) Language as an evolutionary system. Physics ofLife Reviews 2 (3): 177. Edelman, Gerald M. (1993) Neural Darwinism: Selection and reentrant signalling in higher brain function . Neuron 10 (2):115. Ehrlich, Paul and Marcus Feldman (2003) Genes and cul­ tures: What creates our behavioural phenome? Current Anthropology 44 (1):87-109 . Flannery, KV (1999) Process and agency in early state formation. Cambridge ArchaeologicalJournal 9 (1):3-21 . Hull, David L., Rodney E. Langman , and Sigrid S. Glenn (2001) A general account of selection: Biology, immunol­ ogy, and behaviour. Behavioural and Brain Sciences 24:511-573. Jackendoff, R. (2002 ) Meaning, Grammar, 1. Oxford. Bryant, Joseph M. (2004) An Evolutionary Social Science? A Skeptic's Brief, Theoretical and Substantive. Philosophy ofthe SocialSciences 34 (4):451-492. Cavalli-Sfor;, L.L. and M.W Feldman (1981) Cultural Transmission and Evolution: A Quantitative Approach. Princeton University Press, Princeton. Collard, M. and S. Shennan (2000) Processes of Culture Change in Prehisrory: A Case Study from the European Neolithic. In, Archaeogenetics: DNA and the Population Prehistory ofEurope, edited by C Renfrew and K. Boyle. McDonald Institute for Archaeological Research, Cam­ bridge. Collard, Mark, Stephen J. Shennan, and Jamshid J. Tehrani (2006) Branching, blending, and the evolution of cultural similarities and differences among human popu­ lations. Evolution and Human Behavior 27 (3):169 Conkey, Margaret W. (1980) The Identification of Prehis­ toric Hunter-Gatherer Aggregation Sites: The Case of Altarnira. Current Anthropology 21 (5):609-630. Connell, J. E, K. Hawkes, and N. G. Blurton Jones. (1999) Grandmothering and the evolution of Homo erec­ tus. Journal of Human Evolution 36 (5):461-485. Cullen, Ben (2000) Contagious Ideas: On Evolution, Cul­ ture, Archaeology and Cultural Virus Theory. Oxbow Books, Oxford. d'Errico, E, C Henshilwood, G. Lawson, M. Vanhaeren, A-M. Tillier, M. Soressi, F. Bresson, B. Maureille, A. Nowell, J. Lakarra, L. Backwell, and M. Julien (2003) Archaeological evidence for the emergence of language, symbolism and music-an alternative multidisciplinary perspective. Journal ofWorld Prehistory 17 (1): 1-70. de la Torre, Ignacio, Rafael Mora, Manuel Dominguez­ Rodrigo, Luis de Luque, and Luis Alcala (2003) The Oldowan industry of Peninj and its bearing on the recon­ struction of the technological skills of Lower Pleistocene hominids. Journal ofHuman Evolution 44 (2):203. Dunnell, R.C (1978) Style and function: A fundamental dichotomy. AmericanAntiquity 43: 192-202. ____ (1980) Evolutionary theory and archaeology. In, Advances in Method and Theory, edited by M. B. Schif­ fer. Academic Press, New York. Durham, WHo (1991) Coevolution: Genes, Culture and Human History. Stanford University Press, Stanford. ....ms of Language: Brain, ution. Oxford University Press, Jones, Steve (1999) Almost like a Whale: The Origin of Species Updated. Anchor Books, London. Jordan, Peter and Stephen Shennan (2003) Cultural trans­ mission, language, and basketry traditions amongst the California Indians . Journal ofAnthropological Archaeology 22 (1):42. Kohler, Timothy A., Stephanie VanBuskirk, and Saman­ tha Ruscavage-Barz (2004) Vessels and villages: evidence for conformist transmission in early village aggregations on the Pajarito Plateau, New Mexico . Journal ofAnthropo­ logical Archaeology 23 (I): 100-118. Kuhn, Steven L. (2004) Upper Palaeolithic raw material economies at UcagIzII cave, Turkey. Journal ofAnthropo­ logical Archaeology 23 (4):431. Leigh, Steven R. (2006) Brain ontogeny and life history in Homo erectus. Journal ofHuman Evolution 50 (1): 104. Lipo, Cr. , M.J. O'Brien, S.J. Shennan, and M. Collard, eds. (2005) Mapping ourAncestors: Phylogenetic Methods in Anthropology and Prehistory. Aldine de Gruyter, New York. Lipo, CP., M.E. Madsen, R.C Dunnell, and T. Hunt (1997) Population structure, cultural transmission, and frequency seriation. Journal ofAnthropological Archaeology 16:301-333. Mace, R., CJ. Holden, and S.J. Shennan, eds. (2005) The evolution of cultural diversity. A Phylogenetic Approach. UCL Press, London. Machery, Edouard (2005) Review of Robert Boyd and Peter J. Richerson, The Origin and Evolution of Cultures, Oxford University Press, 2005. Notre Dame Philosophical Reviews. Mameli, Matteo (2004) Nongenetic Selection and Non­ genetic Inheritance. The BritishJournalfor the Philosophy ofScience 55 (1):35. _ _ __ (in press) Understanding Culture: A Commen­ tary on Richerson and Boyd's Not By Genes Alone. Biol­ ogyand Philosophy. 40 The REVIEW ofARCHAEOLOGY Vol. 26, Number 2 Marwick, B. (2003) Pleistocene Exchange Networks as Evidence for the Evolution of Language. Cambridge Archaeologicaljournal 13 (1):67-81 . McGaugh, James L. (2000) Memory-a Century of Con­ solidation. Science 287 (5451) :248-251. McPherron, Shannon Patrick (2000) Handaxes as a Mea­ sure of the Mental Capabilities of Early Hominids . jour­ nal ofArchaeological Science 27 (8):655. Mesoudi, Alex, A. Whiten, and Kevin N. Laland (2006) Towards a unified science of cultural evolution. Behavioral and Brain Sciences. Mesoudi, Alex, Andrew Whiten, and Kevin N . Laland (2004) Is Human Cultural Evolution Darwinian? Evi­ dence Reviewed from the Perspective of The Origin of Species. Evolution 58 (1):1-11. Neiman, ED . (1995) Stylistic variation in evolutionary perspective: implications for decorative diversity and inter-assemblage distance in Illinois Woodland ceramic assemblages. American Antiquity 60:7-36 . O'Brien, M .]. and R. Lee Lyman (2002) Evolutionary Archaeology : Current Status and Future Prospects. Evolu­ tionary Anthropology 11:26-36. Pearson, Helen (2006) Genetics: What is a gene? Nature 441 (7092):398 . Pigou , A.C., ed. (1925) MemorialJ-Pf Alfred Marshall. Macmillan, London. Ridley, Matt (2003) The Agile Gene: How Nature turns on Nurture. HarperColiins, New York. Roche, H ., A. Delagnes, ]. P. Brugal, C. Feibel, M. Kibun­ jia, V. Mourre, and P.]. Texier (1999) Early hominid stone tool production and technical skill 2.34 Myr ago in West Turkana, Kenya. Nature 399 (6731):57. Roebroeks, W. (2001) Hominid behaviour and the earli­ est occupation of Europe: an exploration. journal of Human Evolution 41:437-61. Shennan, S. and ]. Steele (1999) Cultural learning in hominids: a behavioural ecological approach. In, Mam­ malian Social Learning: Comparative and Ecological Per­ spectiues, edited by H . Box and K. Gibson. Cambridge University Press, Cambridge. Shennan, S. and]. Wilkinson (2001) Ceramic style change and neutral evolution: A case stud y from Neolith­ ic Europe. American Antiquity 66 (4):577-594. Shott, M .]. (1997) Transmission theory in the study of stone tools: A midwestern north American example. In, RediscoveringDarwin: Evolutionary Theory andArchaeolog­ ical Explanation. American Anthropological Association, Arlington, Virginia. ____ (2005) The reduction thesis and its discontents: Review of Australian approaches . In , Rocking the Boat: Recent Australian Approaches to Lithic Reduction, Use and Classification, edited by C. Clarkson and L. Lamb . Archaeopress, Oxford. Sperber, D. (1996) Explaining Culture: A Naturalistic Approach. Blackwell, Oxford. ~e. Mind Sterelny, K. (2006) The evolution and evolvability of cul­ & Language 21 (2):137-165. Tehrani, ]amshid and Mark Collard (2002) Investigating cultural evolution through biological phylogenetic analy­ ses of Turk men textiles. journal ofAnthropological Archae­ 010gy21 (4):443. van Schaik, Carel P. , Marc Ancrenaz, Gwendolyn Borgen, Birute Galdikas, Cheryl D . Knott, Ian Singleton, Akira Suzuki, Sri Suci Urarni , and Michelle Merrill (2003) Orangutan Cultures and the Evolurion of Material Cul­ ture. Science 299 (5603) :102-1 05. Vanhaeren, Marian and Francesco d'Errico (2006) Auri­ gnacian erhno-linguisric geography of Europe revealed by personal ornaments. journal of Archaeological Science 33 (8):1105 . Whiten, Andrew, Victoria Horner, and Frans B. M. de Waal (2005) Conformity to cultural norms of tool use in chimpanzees. Nature 437 (7059):737. Wrangham , R.W., ].H. Jones, G . Laden, D. Pilbearn, and N . Conklin-Brittain (1999) The Raw and the Stolen. Cooking and the Ecology of Human Origins. Current Anthropology 40 (5):567-594.