Download Cultural Transmission Theory and its Implications for Hunter-Gatherer Archaeology and more Papers Cultural Studies in PDF only on Docsity!
Cultural transmission theory and hunter-gatherer archaeology Jelmer W. Eerkens, Robert L. Bettinger, and Peter J. Richerson
Keywords Cultural transmission, hunter-gatherers, archaeology, temporal variation, spatial variation
Word count 7161
Abstract Cultural transmission (CT) describes the myriad processes whereby information is transmitted from a transmitter to a recipient. Much empirical and modelling research suggests that there are many different pathways and biases through which such information is transmitted. In particular, the content, context, and mode of the transmission can structure variation over time and space in resulting information at a population level. These ideas can be applied to the archaeological record of all societies, but we explore how these ideas might inform specifically on hunting and gathering groups. CT theory suggests that hunter-gatherers, due to their smaller population sizes and simpler social structure, should be characterised by a more limited range and generally more simple material technologies, but display broader variation about a particular mean type, than other more densely populated societies. We exemplify some of these processes with two case studies from the North American Great Basin.
Author biographies
Jelmer W. Eerkens is a Professor in the Department of Anthropology, University of California, Davis. His work focuses on the archaeological record of hunter-gatherers and the effects of cultural transmission over long periods of time as manifested in material culture. His fieldwork is based in Western North America and Peru.
Robert L. Bettinger is a Professor in the Department of Anthropology, University of California, Davis. He is an expert in hunter-gatherer studies and is author of Hunter-Gatherers: Archaeological and Evolutionary Theory (1991) and Hunter-Gatherer Foraging: Five Simple Models (2009). His work focuses on hunter-gatherer expression in marginal environments (alpine and desert) and connections with early agriculture. His work seeks to generate quantitative models of hunter- gatherer behavior that can be applied in a wide range of archaeological and ethnographic contexts, and to make theoretical contributions to cultural ecology, quantitative methodology, and evolutionary theory.
Peter J. Richerson is a Distinguished Professor Emeritus at the University of California Davis. He is one of the major figures in the development of the theory of cultural evolution and has written widely on the subject. His first book with Robert Boyd in 1985, Culture and the Evolutionary Process , is a classic in the field. His work emphasizes the development of quantitative models that illuminate the evolutionary properties of human culture and animal social learning, and the processes of gene- culture coevolution.
Introduction In this chapter, we highlight cultural transmission (CT) theory as a framework for understanding change over time and space in human material culture, with particular emphasis on how this approach can help archaeologists interpret the record. CT theory seeks to place the evolution of culture within a rigorous and scientifically-defined context, and describes the broad range of mechanisms that humans (or non-humans, in some cases) use to acquire, modify and retransmit cultural information. Such information can represent rules about eligible marriage partners, instructions for how to produce fishing nets, lyrics associated with a particular song, oral histories, gossip, or even misinformation or maladaptive information (e.g., the consumption of brains of dead relatives or the use of lead glazes on drinking cups). Many applications of CT attempt to define these mechanisms mathematically and to model their effects over time (e.g., Bentley 2005; Bentley and Shennan 2003; Boyd and Richerson 1985; Cavalli-Svorza and Feldman 1981; Henrich and Boyd 1998; Mesoudi and Lycett 2009).
By definition, CT applies to all humans over all time, hunter-gatherers, agriculturalists, and members of industrial societies alike. However, we attempt to highlight below more particular
through cultural means, archaeology is uniquely situated and has an important role to play in understanding long-term (i.e., macro) CT processes.
CT defined CT should be thought of as a specific model within the more general class of inheritance systems, where information, be it genetic, language, or material culture, is passed between sources and destinations. The specific inheritance systems differ in their details, for example, the form and context in which information is transmitted (e.g., as DNA or as verbal instructions), how information is received and stored (e.g., encoded within chromosomes, or remembered within the brain), the fidelity of such information transfer (e.g., copying error), and how individuals subsequently act upon information received (e.g., modification, transformation). These dynamics greatly influence the overall evolutionary trajectories of genes, material culture, language, and the like, but at their most fundamental level, all inheritance systems share some similarities in how they operate over time.
A nice feature of such inheritance systems is that they can be modelled using recursive equations or as Markov Chains, facilitating the analysis of change (or stasis) over time. Such equations have been well-studied by mathematicians (e.g., Ewens 1979; Meyn and Tweedie 1993) and as a result, many Markov equations can be solved analytically (e.g., with mathematical equations with known solutions). Furthermore, with modern computers, extremely complex Markov processes can be analysed empirically, and equations that do not have defined analytical solutions can be examined and solved through computer simulation. From these efforts, anthropologists and others have been able to describe some of the regularities and patterns that should describe the transmission of cultural information, in other words, allowing researchers to predict the short- and long-term effects of CT. Thus, these two approaches, mathematical modelling and computer simulation, facilitate a scientific approach to CT studies, where hypotheses drawn from theory can be modelled and tested against actual data from the archaeological or anthropological record.
Commonly, mathematical modellers use the tools of population biology to study cultural evolution. The general form of such models is:
Where is some property of the population (say the frequency of a particular stone tool type) at time , is the property of the population at time , and ∑ represents the combined effects of the processes acting to change the population over time. For example, suppose we are interested in the change over time in a stylistic feature of stone points, for example ribbon pressure flaking (i.e., where flake scars are long, parallel, and run across the entire surface of a flaked tool). An aesthetic force might favour new variants as knappers become bored with the common design, while utilitarian forces might favour a variant with the least drag and best penetration. A theoretical investigation might focus generically on the way aesthetic and function biases interact to generate evolution by picking forms for the two processes and iterating the model algebraically or by computer simulation (Boyd and Richerson, 1985; Cavalli-Sforza and Feldman 1981). An experimental study might ask participants to iteratively design projectile points on a computer, simulate their performance, and allow performance information to be exchanged between participants, in a time-transgressive manner (e.g., Mesoudi and O’Brien 2008), or have participants listen to, and subsequently retransmit, a narrative, observing where information is modified, innovated, or deleted (e.g., Mesoudi and Whiten 2004). A field study would attempt to use the archaeological record to infer the forces in operation in a sequence of change within a particular social and physical setting (Eerkens and Bettinger 2008; Eerkens and Lipo 2005).
Genetic transmission has been well-studied and many, though certainly not all, of the details have been worked out. By comparison, there has been much less research on CT, though the literature is rapidly expanding (for recent works see McElreath et al. 2008; Mesoudi and Lycett 2009). As a result, much of the scientific study of inheritance systems has focused on GT instead of CT. One of the main reasons for this, we believe, relates to the fact that many social animals use simple forms of transmission. Humans are unique in the sheer quantity of information transmitted culturally. For example, it has been argued that humans have highly evolved cognitive and communication skills to quickly and efficiently acquiring generic cultural knowledge, such as recognising eye contact and the use of social referencing to interpret emotional displays (e.g., Csibra and Gergely 2009; Gergely et al. 2007). These mechanisms allow humans, but apparently no other animal, to evolve extremely complex cultural adaptations by successive innovations and transmission (Boyd and Richerson 1996; Tomasello 1999).
discussed biases, though others are certainly possible. Much research in CT examines the evolution of the biasing mechanisms themselves (e.g., is copying a prestigious person hard-wired or learned behaviour; e.g., Efferson et al. 2008), their universality among different cultures (e.g., Henrich et al. 2006), and their cumulative effects on culture over time. Cognitive biases, too, affect the rate and accuracy of information transmission. The biases listed in Table 7.9.2 affect primarily the recipients of cultural information. Less theorising has been focused on cognitive biases that might affect the transmitters, though some archaeologists (e.g., Spencer 1993), have considered how aspiring leaders might use biased transmission systems to increase their power base and/or authority.
Information acquisition Information packaging Information modification Frequency-dependent (e.g., conformity, pro-novelty)
Independent-biases (e.g., hitchhiking / info. bundling)
Copying error
Model-based (e.g., prestige, success, similarity)
Acquisition of traits one at a time from a range of models
Filtration through individual worldview
Purposeful innovation (e.g., directed vs. undirected) Table 7.9.2. Cognitive biases in CT
The exact nature and universality of such biasing mechanisms has been debated. In any case, the interaction of the different biasing mechanisms with different content (e.g., auditory vs. visual; complex vs. simple), learning contexts (e.g., ritual vs. apprenticeship vs. domestic), and modes create a myriad of different transmission pathways. This complexity makes application to the archaeological record, where we may not be able to reconstruct all the different factors, challenging. However, as others have shown (and we hope to highlight below), there is interpretive ground to be gained by such applications.
CT and archaeology Many of the basic ideas about the transmission of cultural information have been with anthropologists since the late 1800s. For example, Franz Boas (1896, 3-4, quoted in O’Brien and Lyman 2002, 229) suggested a general algorithm for understanding similarity in human cultures: the closer people lived to one another the more similar cultures should appear because similarity in
culture was most likely due to dissemination of cultural information (as opposed to independent origin). Boas later suggested that most of the cultural inventory of a society was the cumulative result of ‘diffusion’ from neighbouring cultures (Hatch 1973).
Anthropologists such as Holmes (1886), Petrie (1899), and Kroeber (1916) were quick to pick up on these ideas, and to develop them further, to help explain similarity and/or difference in the archaeological record. For example, Kroeber (1940) considered how small errors during diffusion affected the distribution of culture-historical types over space and time. This view of culture included most of the important elements of a Darwinian evolutionary system, including the transmission of information (cultural in this case), mutation of that information (through copying errors), and retransmission of this information in the modified state. The only major Darwinian element lacking from this view is the notion of various global selective forces that winnow out some of these mutations. Today, such refinement of CT continues, and has focused on more rigorous and mathematical definitions of the specific processes involved. CT is being used by an increasing number of researchers in a diverse array of academic fields including anthropology, archaeology, biology, economics, and psychology, among others.
What is particularly relevant to archaeologists is not the ability to see and reconstruct every transmission event, but to deduce how transmission content, context, mode, and cognitive biases affect patterns in information and hence material culture, especially at a societal level (Eerkens and Lipo 2007; see also chapters in O'Brien 2008). Theory suggests that such patterning is likely to be evident especially in measures of information diversity or variance (e.g., Bettinger and Eerkens 1999; Eerkens and Bettinger 2001; Eerkens and Lipo 2005; 2007). In short, CT theory makes few predictions about modal or average tendency in assemblages of artefacts, but is more specifically concerned with measures of dispersion or variation. Testing CT theory then, requires population- level measures of dispersion or variation and archaeology is one the few, perhaps the only, discipline that can generate the necessary data to do so, especially for long-term outcomes. For example, direct extrapolation of evolutionary processes measured in the short run by direct observation are often very different from long run averages (Gingerich 1982). This is probably because evolution frequently meanders and reverses direction. Evolution observed over longer spans of time depicts long-term trends that average out much of the small-scale fractal variation that a direct observer of human transmission might see. Furthermore, many of the cultural traits transmitted by animals, such as birdsong, leave little material evidence behind, again, leaving
Shennan (2001) and Powell et al. (2009), which shows that population size and connectedness have a positive relationship with the maintenance of complex cultural information.
Thus, CT theory predicts that, due to their smaller population sizes, hunter-gatherers in general ought to be characterised by a more limited range of material technologies than agricultural and industrial ones (e.g., Powell et al. 2009; Richerson et al. 2009). This result is not especially surprising, but might explain the long periods of apparent stasis witnessed in Lower and Middle Palaeolithic sites in the Old World. In those sequences, a narrow range of tools and technologies characterise tens to hundreds of thousands of years. As well, this line of reasoning may explain the explosion in the diversity of material culture observed during the Mesolithic and Neolithic periods as global climates warmed in the Holocene and populations appear to have grown (Shennan 2001). Likewise, the rapid evolution of very complex artefacts beginning with the industrial revolution involved both large numbers of communicating innovators and the expansion of formal education that made the transmission and storage of complex information much more efficient. Donald (1991) ranks the development of literacy and numeracy one of the major human cognitive advances, although they are cultural not genetic innovations.
Effects of information content and complexity CT theory also predicts that the form in which information is transmitted (e.g., written, verbal, visual instructions) and the complexity of this information (the length and regularity of instructions; see Gell-Mann and Lloyd 2003) should affect fidelity and the preferred mode of transmission. For example, written instructions are likely to be transmitted with greater fidelity than verbal or visual ones, and given a particular transmission mode, complex information is subject to greater copying error (lower fidelity). Likewise, while simple technologies are subject to more tinkering and experimentation, the attributes of complex technologies are often transmitted together as part of information packages. This leads to greater linkage or correlation between the attributes or components of a complex technology.
Lacking writing, hunter-gatherers transmit information via verbal and visual means. Hence, one prediction of CT theory is that within a particular technology, there should be greater variation in the size and shape of tools than among societies where technological information is transmitted primarily by writing. Similarly, because hunter-gatherers tend to have simpler tools (as per the reasoning above), all other things being equal, we expect more tinkering and experimentation with
these technologies, per capita, than more complex ones that characterise agricultural and industrial ones. The net result of these effects is that hunter-gatherer material technologies should have fewer technologies and artefact ‘types,’ but a broader range of variation about a particular mean, and less covariation between attributes within a type. By contrast, material technologies among agricultural and industrial societies should be much more diverse overall (e.g., more types), but with less variation about means within a technology (or artefact type) and greater covariation between the attributes of those technologies.
Testing these hypotheses derived from CT theory will require collecting large sets of data across diverse ecological and temporal scales, and examining diversity and variation in artefact types, and covariation among attributes within types. Limited data sets examined by Eerkens and Bettinger (2008; Bettinger and Eerkens 1999) for just hunter-gatherer groups of different population densities generally support these notions. However, additional comparisons using other data sets are clearly necessary.
Effects of social organisation Many, though certainly not all, hunting and gathering groups are characterised by simpler social organisations, generally lacking formalised and/or institutionalised positions of leadership (e.g., Fried 1967; Service 1962). In such settings, social and political organisation tends to be much smaller in scale and localised. As a result, individuals tend to have greater interaction with the members of their immediate group, and more limited contact with people living in more distant locations. In more complex societies, individuals from a wider range of social and political settings are often brought together (e.g., in markets or urban living), where they may acquire and transmit cultural information.
Individuals in smaller and simpler groups are apt to acquire information through social learning (e.g., copying) and individual learning (e.g., experimentation) in frequencies similar to their counterparts in more complex societies. However, the low degree of political organisation in hunting and gathering groups will tend to limit the spatial extent over which material culture will spread (e.g., Powell et al. 2009; Richerson et al. 2009). For material culture that contains a high degree of stylistic information (as conceived by Dunnell 1978; see also Rogers and Ehrlich 2008), that is, where environmental or other ‘global’ conditions do not greatly influence performance characteristics (see Eerkens and Bettinger 2008), this should result in smaller and more localised
collection. CV and related measures (see Eerkens and Bettinger 2001) are salient in archaeological studies of cultural transmission, which has fairly straightforward implications about artefact variation. As noted above, for example, cultural transmission will produce greater trait variation in smaller populations than larger one.
Myriad other forces, many directly connected with environment, also affect cultural transmission and trait variation (e.g. Eerkens and Bettinger 2008). For example, three particularly well-defined Great Basin point types – Desert Side-notched, Large Side-notched, and Elko Eared – show a small but nevertheless consistent decrease in mean CV at higher latitudes (rDesert Side-notched = -.20; rLarge Side-
notched = -.16; r^ Elko Eared = -.10). There are many potential explanations for this, the most likely being the greater dependence on hunting among Great Basin groups at higher latitudes (rlatitude-hunting = .34, ngroups = 39; Binford 2001). That projectile point mean CV decreases along with this suggests the hypothesis that as hunting becomes more important, hunters are increasingly inclined to standardise the projectile points they use, perhaps by more carefully copying the points knapped by good hunters, Desert Side-notched, Large Side-notched, and Elko Eared all showing this basic environmental relationship. Figure 7.9.1 shows these relationships for these three point styles, graphing the mean CV by latitude.
Environment, however, is unlikely to account for differences across these same three point types in the relationship between mean CV and sample size. In theory, mean CV should vary independently of sample size; it will vary more across small samples than large ones drawn from the same population, but the mean of many small samples and many large samples, both measuring mean CV, should be the same. This expectation holds for Large Side-notched and Elko Eared points, where the effect of sample size on mean CV is weak and conflicting, slightly increasing in Large Side-notched points (r (^) Large Side-notched = .06), slightly decreasing in Elko Eared points (rElko Eared = -.13). In Desert Side-notched points, by contrast, mean CV is very strongly, and inversely, correlated with sample size (rDesert Side-notched = -.52). While sample size is jointly determined by a host of factors, including sampling intensity, that Desert Side-notched mean CV varies so strongly with sample size suggests that in this case sample size is closely tied to assemblage size, and potentially, population size. While this is not a particularly strong test of the hypothesis that cultural transmission reduces trait variation in larger populations, it demonstrates the kind of data and statistics appropriate to testing the implications of cultural transmission theory in the archaeological record.
Figure 7.9.1. Variation in means, as measured by CV, with latitude in three projectile point types in the North American Great Basin. The graphs show a weak but consistent effect where points become less variable with increasing northerly latitude.
Discussion and conclusions At its most basic, CT is simply a means to explain how cultural information is transmitted between individuals. This is how CT theory was applied in the early part of the twentieth century. However, in the last several decades CT theory-building and modelling exercises now provide archaeologists
with rich explanatory framework to interpret and make predictions about the archaeological record, especially at different temporal and spatial scales.
In our view, the most powerful aspects of CT relate to predictions about variation and diversity in material culture, and changes over time and space therein. Although the majority of these predictions relate to other aspects of human societies, some of these predictions are of special relevance to hunter-gatherers. Thus, the population size and density of most hunter-gatherers, the types of tools they tend to produce, and the means by which they transmit culture (verbally and visually), is likely to structure the material culture left behind by such peoples. In particular, CT theory predicts a more limited range of tools (i.e., less diversity) for hunter-gatherers, but greater variation within a particular tool type, compared to subsistence strategies that support higher population densities.
One interesting aspect of these predictions regards the rate of evolutionary change within such material technologies. Rates of change in the diversity of technologies should be strongly and positively linked to population size. On the other hand, rates of change within a particular technology should not, but may be more strongly tied to the amount of time a technology has been used within a particular socio-economic setting. Thus, hunter-gatherer ceramic or bow-and-arrow technologies should change just as rapidly (and we specifically refer to change in within-type variation), if not more so owing to higher copying error, or drift-like effects, than agricultural or industrial ones. Of course, such changes may degrade the functionality of certain implements, which may cause them to be abandoned.
Owing to the nature of the data we collect, namely material culture as it appeared at different points in space and time, archaeology has much to contribute to CT studies. Testing hypotheses derived from theory with actual data is one aspect of this, but we hope that future research will also contribute to CT theory-building. As we have shown, certain aspects of hunter-gatherer culture, especially long-term evolutionary change, can be informed and understood by CT. We hope that others will agree and continue to apply these ideas to other parts of the archaeological and ethnographic record.
References Bentley, R. A. 2005. Academic copying, archaeology and the English language. Antiquity 80, 196-201.
Eerkens, J.W. and Bettinger, R.L. 2001. Techniques for assessing standardization in artifact assemblages: can we scale material variability? American Antiquity 66, 493-504. Eerkens, J.W. and Bettinger, R.L. 2008. Cultural transmission and the analysis of stylistic and functional variation. In M. J. O’Brien (ed.), Cultural transmission and archaeology: issues and case-studies , 21-38. Washington DC: SAA Press. Eerkens, J.W. and Lipo, C.P. 2005 Cultural transmission, copying errors, and the generation of variation in material culture and the archaeological record. Journal of Anthropological Archaeology 24, 316-34. Eerkens, J.W. and Lipo, C.P. 2007 Cultural transmission theory and the archaeological record: providing context to understanding variation and temporal changes in material culture. Journal of Archaeological Research 15, 239-74. Eerkens, J.W., Neff, H. and Glascock, M.D. 1999. Early pottery from Sun'gava and implications for the development of ceramics on Owens Valley. Journal of California and Great Basin Anthropology 21, 275-85. Efferson, C., Lalive, R., Richerson, P. J., McElreath, R., and Lubell, M. 2008. Conformists and mavericks: the empirics of frequency-dependent cultural transmission. Evolution and Human Behavior 29, 56-64. Ewens, WJ. 1979. Mathematical population genetics. Berlin: Springer Verlag. Fried, M.H. 1967. The evolution of political society: an essay in political anthropology. New York: Random House. Gabora, L.M. 2004. Ideas are not replicators but minds are. Biology and Philosophy 19, 127-43. Gell-Mann, M., and Lloyd, S. 2003. Effective complexity. Santa Fe, NM: Santa Fe Institute Working Paper 03-12-068. Gingerich, P.D. 1982. Time resolution in mammalian evolution: sampling, lineages and faunal turnover. Proceedings of the Third North American Paleontological Convention, Montreal 1, 205-10. Gergely, G., Egyed, K. and Király, I. 2007. On pedagogy. Developmental Science 10, 139-46. Gould, S.J., Raup, D.M., Sepkoski Jr., J.J., Schopf, T.J.M. and Simberloff, D.S. 1977. The shape of evolution: a comparison of real and random clades. Paleobiology 3, 23–40. Hatch, E. 1973. Theories of man and culture. New York: Columbia University Press. Henrich, J. 2004. Demography and cultural evolution: why adaptive cultural processes produced maladaptive losses in Tasmania. American Antiquity 69, 197-
Henrich, J. and Boyd, R. 1998. The evolution of conformist transmission and between-group differences. Evolution and Human Behavior 19, 215-42. Henrich, J. and McElreath, R. 2003. The evolution of cultural evolution. Evolutionary Anthropology 12, 123- Henrich, J., McElreath, R., Barr, A., Ensminger, J., Barrett, C., Bolyanatz, A., Cardenas, J.C., Gurven, M., Gwako, E., Henrich, N., Lesorogol, C., Marlowe, F., Tracer, D., and Ziker, J. 2006. Costly punishment across human societies. Science 312, 1767-70. Holmes, W. H. 1886. Ancient pottery of the Mississippi Valley. Fourth Annual Report of the Bureau of Ethnology, 1882-83, 361-436. Washington, DC: Smithsonian Institution. Kelly, R.L. 1995. The foraging spectrum: diversity in hunter-gatherer lifeways. Washington, DC: Smithsonian. Kerst, S.M. and Howard, J.H. Jr. 1984. Magnitude estimates of perceived and remembered length and area. Bulletin of the Psychonomic Society 22, 517–20. Kline, M. A. and Boyd, R. 2010. Population size predicts technological complexity in Oceania. Proceedings of the Royal Society B 277, 2559-64. Kroeber, A. L. 1916. Zuni Potsherds. Anthropological Paper, No. 18, Part 1. New York: American Museum of Natural History. Kroeber, A. L. 1940. Statistical classification. American Antiquity 6, 29-44. Krützen, M., Mann, J., Heithaus, M. R., Connor, R. C., Bejder, L. and Sherwin, W. B. 2005. Cultural transmission of tool use in bottlenose dolphins. Proceedings of the National Academy of Science 102, 8939-43. Lachnit, H. and Wolfgang, P. 1990. Speed and accuracy effects of fingers and dexterity in 5-choice reaction tasks. Ergonomics 33,1443–54. Laland, K.N., Odling-Smee, J. and Myles, S. 2010. How culture shaped the human genome: bringing genetics and the human sciences together. Nature Reviews Genetics 11, 137-48. Laland, K.N. and O'Brien, M.J. 2010. Niche construction theory and archaeology. Journal of Archaeological Method and Theory 17, 303-22. Lyman, R.L., VanPool, T.L. and O’Brien, M.J. 2009. The diversity of North American projectile-point types, before and after the bow and arrow. Journal of Anthropological Archaeology 28, 1–13. McElreath, R., Bell, A.V., Efferson, C., Lubell, M., Richerson, P.J, and Waring, T. 2008. Beyond existence and aiming outside the laboratory: estimating frequency-dependent and payoff- biased social learning strategies. Philosophical Transactions of the Royal Society B 363, 3515-
