Technology as Human Social Tradition outlines a novel approach to studying variability and cumulative change in human technology—prominent research themes in both archaeology and anthropology. Peter Jordan argues that human material culture is best understood as an expression of social tradition. In this approach, each artifact stands as an output of a distinctive operational sequence with specific choices made at each stage in its production. Jordan also explores different material culture traditions that are propagated through social learning, factors that promote coherent lineages of tradition to form, and the extent to which these cultural lineages exhibit congruence with one another and with language history.
Drawing on the application of cultural transmission theory to empirical research, Jordan develops a descent-with-modification perspective on the technology of Northern Hemisphere hunter-gatherers. Case studies from indigenous societies in Northwest Siberia, the Pacific Northwest Coast, and Northern California provide cross-cultural insights related to the evolution of material culture traditions at different social and spatial scales. This book promises new ways of exploring some of the primary factors that generate human cultural diversity in the deep past and through to the present.
Technology as Human Social Tradition Cultural Transmission among Hunter-Gatherers
Understanding Technology as a Human Social Tradition
One definitive feature of the human condition is reliance on highly sophisticated technological solutions. These physical objects are termed material culture and include elaborate tools for capturing, processing, and storing resources, technologies for travel, vernacular architecture, as well as all the other objects used by people in all spheres of social life. In general, however, people tend not to invent such objects and technologies for themselves through personal trial-and-error learning but predominantly acquire existing designs and cultural ideas from other people. Nor is this a relatively new phenomenon, linked only to the rise of modern urban and industrial life. Even in small-scale hunting and gathering societies, people primarily learn how to make useful things from other individuals during childhood and adolescence. And, of course, they may also add their innovations and improvements to these designs later on in life, passing these changes on to later generations. People actively participate in the reproduction of cultural knowledge, and most technologies used by humans form long-term historical tradition that are passed on to others through exactly this kind of social learning.
As anthropologists and archaeologists frequently document, these enduring lineages of cultural tradition can extend in recognizable formats over many, many generations, in some cases persisting for millennia. But if these material culture traditions are reproduced through social learning, then whom people learn from, what they learn, why, and when can all have major cumulative effects on larger patterns of cultural diversity and change. In one way or another, exactly how such material culture lineages are reproduced, and why they are subject to continuity or transformation, have been the focus of debate for well over a century.
This book therefore examines three interlocking topics that are central to all archaeological and anthropological inquiry: the role of technology and material culture in social life; the reproduction of social traditions; the factors that generate and sustain cultural diversity. In fact, the overall aim of this book is to outline a new kind of approach for researching variability and change in material culture. This can be summarized as "Technology as human social tradition." The main argument is that human technological traditions exhibit heritable continuity: they consist of information stored in human brains and then passed on to others through social learning; people born into specific cultural settings acquire, participate in, and thereby reproduce these material culture traditions, passing them on to future generations. This system of inheritance involves both the reproduction and modification of cultural information, but also the expression of these ideas and skills in production of material objects over time. Such technological traditions can therefore be understood as material manifestations of a complex transmission system in which cultural information is inherited, reproduced, and cumulatively transformed by individuals and their communities.
Applying this perspective to human technology builds on, but also largely transcends, much of the earlier work conducted by archaeologists and anthropologists into the significance, function, and social meanings associated with patterns of continuity and change in material culture. In the nineteenth and early twentieth centuries anthropologists collected objects and artifacts from different cultures either to assess the cultures' relative levels of progress or to simply catalogue global cultural diversity in terms of the artefacts and objects used by different tribes and ethnic groups. But for most of the twentieth century, anthropological interest shifted to other themes such as kinship and religion, with material culture largely ignored. More recently, there has been renewed interest in material culture in general (Tilley et al. 2006) and in the anthropology of technology, that is, how individuals acquire and practice embodied craft skills within the daily routines of specific sociocultural settings (Ingold 2000; Leroi-Gourhan 1993; Mauss 1979; Stark et al. 2008a).
Archaeologists have also built most of their discipline on the recovery and documentation of human technology, but they have done so in a range of different ways, generating contrasting and at times contradictory insights into material culture variability and change. Until the early twentieth century, major changes in material culture were also used to map either general stages of human progress, and in the days before radiocarbon dating was developed, artifact lineages were used to construct relative chronologies and to trace the traditions and histories of different archaeological "cultures" as well as the prehistoric origins of modern ethnic groups and nation-states (Lyman et al. 1997; Shennan 1989; 2002a:67; 2009a:2; Trigger 2006:211-313). By the mid-twentieth century, archaeological interests had shifted toward investigating the functional roles played by artifacts, objects, and toolkits, especially among prehistoric hunter-gatherers. In this period, interests in the archaeology of cultural traditions was largely eclipsed by an overarching concern with studying the adaptive dynamics of human cultural systems, which required identification of cross-cultural regularities in technology across different environmental settings (Lyman et al. 1997:217, 230; Shennan 2002a:72, 184; Trigger 2006:480).
By the end of twentieth century, however, theoretical fashions were beginning to swing in the opposite direction, deliberately highlighting the local, idiosyncratic, and historically contingent nature of cultural phenomena, as well as the social and symbolic dimensions to material culture, the active role that objects play in social life, and their significance in expressions of identity, religion, and ideology (Hodder 1982, 1986). Some general interest in researching cultural traditions returned, but it was generally limited to examination of small-scale social settings and the description of the microroutines of daily practice. Less effort was directed to linking these small-scale processes to the deeper mechanisms of long-term culture change (Trigger 2006:444-78).
Most current anthropological and archaeological work on technology now tends to be united by a shared interpretive interest in the contextual significance of material culture (or "materiality') and its general historical contingency. The broad consensus is that creation of material culture through day-to-day practice forms part of the wider process of cultural and social reproduction and that this involves creativity and subjective cultural choices, and hence the agency and history of individuals and their social collectives (Apel 2001; Dobres 2000; Gosselain 1998; Ingold 2000; Killick 1994; Leroi-Gourhan 1993; Lomonnier 1993; Mauss 1979; McEachern 1998:246; Sillar 2000; Sillar and Tite 2000; and see Tilley et al. 2006).
This book takes a different tack. It draws on the substance, content, and focus of many of these older debates about the relative functional versus symbolic roles performed by material culture but takes research into new directions by employing some of these older ideas as useful points of departure. The main starting point is the absolutely central and now consensus idea that the practice of craft traditions form part of general cultural reproduction. However, the book argues that this renewed emphasis on understanding technology as a fundamentally social tradition-that is, as cultural information reproduced through social learning-generates some broad analogies with the ways in which evolutionary biologists have investigated the transmission of genetic information. Both genetic and cultural inheritance systems can therefore be argued to exhibit evolutionary properties of "descent with modification" (Boyd and Richerson 1985; Cavalli Sforza and Feldman 1981; Ellen et al. 2013; Lycett 2011; Mesoudi 2011; Mesoudi et al. 2006; O'Brien et al. 2013; Shennan 2002a, 2009b; Richerson and Boyd 2005).
On one level, exploring these analogies between cultural and genetic inheritance systems is important for archaeologists and anthropologists because, on a heuristic level, they generate new frameworks within which to think about the social reproduction of material culture traditions, especially the links between individual actions and long-term outcomes, including the broader patterns of variability and long-term culture change that result. On another, and perhaps more important empirical level, these analogies also provide a bridge for applying some of the powerful analytical methods developed in the biological sciences to material culture data sets, thereby tackling long-standing questions about the factors that generate continuity, diversity, and change in material culture traditions across an interlocking range of different social, spatial, and temporal scales.
One of the most important challenges is generating suitable data for implementing this kind of approach. Building on key arguments in the anthropology of technology, it is argued in this book that the composition of particular material culture traditions can be defined and documented in terms of distinctive sets of "cultural traits." This is because all craft traditions can be studied in terms of their unique production sequences, or "design grammars," which consist of different stages of production and the associated choices by human practitioners as to what materials or methods to deploy within the different production stages. Each of these choices can be defined as traits, and specific combinations of these design traits can then be argued to make up particular kinds of artifacts and their associated craft traditions. Defining and documenting these traits is therefore about understanding the inherent creativity and historical agency central to the practice of craft production. Moreover, large data sets recording variability in material culture can be generated by the same trait-based approach and then subjected to further analysis, hypothesis testing, and contextual interpretation in order to understand how technological traditions have diversified and changed over time.
If this general approach holds fast-and if material culture variability and change can be productively approached in terms of dynamic social traditions-then three overarching research themes emerge:
1. Propagation of Cultural Traditions. How are material culture traditions reproduced through social learning; how do individuals acquire knowledge of specific design traits and how best to combine them; what factors promote them to maintain or adjust these traditions; what patterns of cumulative change are generated?
2. Coherence in Cultural Traditions. To what extent do material culture traditions consist of particular combinations of design traits; is there just a rapid and relentless mixing of traits, or do specific combinations form coherent designs endure in recognizable formats over generations? At what social scales do such coherent lineages of tradition emerge? To what extent can the deeper history of these coherent lineages be reconstructed, and what forms do these histories take? Do they undergo repeated splitting processes, with the branching away of new descendent traditions, all of whose genealogies can be traced back to a common ancestor?
3. Congruence among Cultural Traditions. Does each lineage of tradition have its own independent history, or are technological traditions propagated in ways that ensure that a number of different traditions eventually become bundled together?
This book systematically addresses these three themes, applying a "descent with modification" perspective to the study of material culture traditions across a range of concrete empirical case studies. These are united by a shared focus on understanding the technologies of different hunting and gathering societies across a range of different cultural settings.
This research is neither limited to anthropology nor directed just at an archaeology readership-its themes and approaches span both disciplines: the case studies employ anthropological information and ethnographic data, but many of the questions addressed, and the insights that are produced, are of equal relevance to archaeologists seeking to understand the significance of spatio-temporal patterning in prehistoric material culture. In this sense, the book can best be understood as an interdisciplinary exercise in hunter-gatherer ethnoarchaeology (David and Kramer 2001; Lane 2014), one that undertakes a sustained contextual and comparative analysis of material culture evolution, integrating it to the newly emerging science of cultural diversification and change that now spans archaeology, anthropology, biology, evolutionary psychology, and historical linguistics (Ellen et al. 2013; Henrich and McElreath 2003; Lycett 2011; Mesoudi 2011; O'Brien et al. 2013; Rogers and Ehrlich 2008:3416; Shennan 2009b; Steele et al. 2010).
This introductory chapter aims to contextualize the aims of the research, its general approach, and the debates and questions it addresses. It starts from first principles by examining the unique sophistication of human social learning and the ways in which it underpins the maintenance of cumulative cultural traditions. It then examines how cultural transmission theory, which was inspired by exploring some of the broad analogies between cultural and genetic inheritance, can be used to provide a general framework for examining the reproduction of cultural traditions within different populations, and also how some of these specific processes of cultural propagation link directly into long-standing anthropological and archaeological debates about large-scale patterning in the coherence and historical congruence of different material culture traditions. Chapter 2 outlines the central methodology, chapters 3, 4, and 5 contain the main case studies, and chapter 6 undertakes a cross-cultural comparative analysis of the overall results, linking these insights back to general debates about variability and change in material culture.
What Makes Human Culture Unique?
The human species is unique. What makes it so special is the highly developed capacity for maintaining cumulative cultural traditions through the practice of teaching, imitation, and other forms of social learning. But biologists are quick to point out that the enormous species gulf between humans and other even closely related animals is only a relative one. Understanding some of these relative similarities and differences forms a useful starting point for examining what is so distinctive about human social learning and the elaborate cultural traditions that it can sustain.
A broad distinction can be made between individual learning and social learning. According to the former, individuals learn directly from their own experiences of the surrounding world. However, when this individual dies, these accumulated understandings are lost, and each new offspring must embark anew on his or her own process of environmental learning (Shennan 2002a:38, 2002b:185-86). In contrast, social learning involves acquiring information from other members of the same species, which can lead to replication of that information over time. This distinction rests on the source of theinformation learned and is not about the specific information content (Shennan 2002b:186). Thus, "social learning is learning that is facilitated by observation of, or interaction with, another individual" (Hoppit and Laland 2013:4).
Although learning from other individuals is central to the reproduction of human culture, for example, in language or in craft traditions, social learning in itself is not unique to humans, and transmission of this kind of nongenetic information between individuals is surprisingly common among a wide range of organisms (see Bentley et al. 2008; Danchin et al. 2010; Humle and Newton-Fisher 2013; Krützen et al. 2005; Laland et al. 2013; Lycett 2010, 2011; McGrew, 2004; Van Schaik et al. 2003; Whiten, 2007, 2010; Whiten et al. 1999; Whiten et al. 2005; and see Hoppit and Laland 2013 for a general overview
However, the mechanisms by which such traditions are maintained are very variable and span a wide range of cognitive complexity (Shennan 2002b:188). An insightful way to examine the central features of human cultural transmission systems is to run through a short comparative analysis of social learning among humans and chimpanzees, their closely related sister genus (see general summary by Whiten 2011; table 1.1).The goal here is to distinguish which features of social learning and cultural transmission can be ascribed to common ancestry, and which features reflect changes since the ancestral divergence around 6-7 million years ago (Whiten 2011:998). Exploring exactly what makes human social learning capable of sustaining the intergenerational transmission of so many rich and diverse forms of cultural information quickly becomes central to understanding the deeper evolutionary dynamics of their technology and material culture.
Understanding cultural transmission involves studying social traditions. Thus, it is important to understand exactly what constitutes a social tradition; following Fragaszy and Perry (2003:xiii), this can be defined as a "distinctive behaviour pattern shared by two or more individuals in a social unit, which persists over time and that new practitioners acquire in part through socially aided learning." This definition is important because it renders culture as a community-level phenomenon, minimally defined by a tradition shared by at least two individuals, but typically many more (Whiten 2011:999). It also includes the requirement for persistence, which has two implications: (1) traditions can become more substantial as they spread from a minimum of two individuals, and then potentially across larger communities and populations; (2) these traditions are enduring and can potentially persist over multiple generations (Whiten 2011:997). It is these features of culture that enable each new generation to build on the innovations of the previous one, meaning that human traditions involve transmission, but also the more selective capacity for accumulation and editing of this cultural information. It is these combined features of human cumulative culture that make it possible to argue that their cultural traditions can evolve according to principles of descent with modification.
Working through the details summarized in table 1.1, it is clear that three broad themes make the attributes and capacities associated with human cultural traditions quite different than those of even closely related primate species: (1) specific social learning processes, (2) unique cultural content, (3) distinctive population-level patterning.
In particular, it is the sophisticated social learning processes that appear to underpin the distinctive features of human cultural traditions (table 1.1: 1 a-e). Human copying includes both emulation (reproducing end results) and imitation (reproducing actions); in fact, the highly developed capacity for imitation among humans may be key to the phenomenon of cumulative culture, and it certainly seems to be important in the emergence and long-term stability in cognitively opaque and often essentially arbitrary human technological traditions such as stone-tool making, basketry, and pottery (Tehrani and Riede 2008; Want and Harris 2002; but see Caldwell and Millen 2009).
This highly developed human capacity for social learning is important, because it underpins (1) "copying sophistication and fidelity"(Whiten 2011:1001)-there has to be fidelity, otherwise traditions cannot persist in recognizable format. In fact, this kind of high-fidelity transmission appears to be one of the key drivers of human cumulative culture (Lewis and Laland 2012) and is underpinned by a package of other sociocognitive processes-including teaching through verbal instruction, imitation, and also prosociality (Dean et al. 2012), such that teaching, language, and cumulative culture can all work together to reinforce one another.
Humans also engage in (2)"rational" copying" (Whiten 2011:1002)-this means that humans' imitation mechanisms include a substantial element of selectivity. There is also (3) a striking degree of "conformity"in human cultural traditions (Whiten 2011:1002-3)-the apparently deep and enduring human motivation to be like others in a small-scale social group has long been a phenomenon studied by social psychologists (and see Pagel and Mace 2004). Another important feature of human social learning is (4), "ratcheting versus conservatism"-in human culture, progressive accumulation and improvement plays important roles, probably because humans tend to imitate, whereas chimpanzees rely to a greater extent on emulation (Whiten 2011:1003; and see Tomasello et al. 1993; but also Caldwell and Millen 2009).
Humans also engage in (5), explicit "teaching"(Whiten 2011:1003). The subject of teaching is now, in fact, a major interdisciplinary research field spanning cognitive psychology, comparative biology, anthropology, and archaeology (Csibra and Gergely 2006, 2011; Richerson and Boyd 2005; Tehrani and Riede 2008; Tomasello et al. 1993; Want and Harris 2002). Many human cultural traditions pose enormous learnability problems, and simple imitation alone cannot account for high-fidelity human copying; it must be reinforced through other mechanisms (Tehrani and Riede 2008:318). These range from direct teaching and explicit linguistic interaction (e.g., Tomasello et al. 1993) through to the broader concept of "natural pedagogy," which is "a particular kind of social learning in which knowledge or skill transfer between individuals is accomplished by communication" (Whiten 2011:1150).
At a general level, this involves experienced individuals modifying their behavior in ways that include implicit communicative gestures in order to facilitate the learning of a novice. Such gestures might include simple expressions of approval or disapproval, which help replicate arbitrary traditions (Castro and Toro 2004); a pedagogical instinct that expresses itself in "motherese," for example, when distinctive tone, vocal modulations, eye contact, and infant-directed speech are used when parents address children (Csibra and Gergely 2011:1150), as well as in a range of other cues that enable parents and tutors to increase the efficiency of social learning. These forms of communication are important in focusing attention on functionally important aspects of behavior, skill, or sets of tasks, and this kind of relevance-guided instruction is particularly essential for transmission of difficult-to-master skills (Csibra and Gergely 2006, 2011).
The acquisition of complex cultural traditions-such as crafting and tool-making skills-is generally based on acquiring combinations of different kinds of information, including routinized motor patterns that are eventually enacted automatically without much conscious thought. These skills are not so much taught as discovered anew via "progressive teaching" or "scaffolding" (Tehrani and Riede 2008:320). This often includes repeated cycles of demonstration of the complex motor tasks and then self-practice for cumulative reinforcement and correction, all of which make human cultural learning an extended and very complex cognitive process (Shennan 2002b). In general, there is huge global diversity in child-rearing practices, but all these practices appear to be united by at least some form of communication, either explicit, or implicit, or a mix of the two, all of which enable more experienced practitioners to transmit to novices a wide variety of cultural knowledge, ranging from how to make artifacts, conventional norms and behaviors, arbitrary referential symbols, and a range of other cognitively opaque skills and knowhow (Whiten 2011:1152-54). In fact, it exactly this kind of communication-aided teaching, broadly defined, that marks out human cultural transmission as being very different (Csibra and Gergely 2011; Whiten 2011:1150).
The next important theme is to examine how these uniquely human social learning mechanisms make the "cultural content" of their traditions so distinctive (Whiten 2011:1004; table 1.1, part 2 a, b). First, human social culture also has symbolic reinforcement of systems of rules and institutions that regulate actions, including language itself, through to ceremonial traditions, dance, music, and religion. Second, human technological traditions embrace an enormous range of complex subsistence-related tools and equipment that are used even by highly mobile hunter-gatherers, including hafted and multicomponent weaponry and tools, leather clothing, knots, lashings, mats, basketry, and other woven fabrics (Whiten 2011:1004).
Third, what are the outcomes of unique human capacities for cultural traditions? (See table 1.1, part 3 a, b, c.) Many of the cultural traditions passed on by social learning in fish, bird, and other mammal populations concern only single patterns of behavior. In contrast, (1) human culture differs profoundly in encompassing countless traditions that span a huge range of behaviors. No species even comes close to this breadth and diversity. Interestingly (2), social learning in chimpanzees also sustains "local cultures incorporating, and differentiated by, multiple traditions," such that chimpanzees also live in communities able to display unique cultural profiles for a defined subset of such traditions (Whiten 2011:999-1000). However, this does not really compare to the ways in which distinctive cultures are expressed by humans in vast numbers of ways. Finally, and perhaps most crucially, humans also exhibit (3) "cumulative cultural evolution" (2011:1001).This capacity enables cultural achievements developed in one generation to be retained by the next one and further refined, a phenomenon reflected in the vast complexity and variety of human cultures today, for example, in relation to technology, language, and social institutions. In other species, cumulative culture is at best rudimentary, though the reasons for this remain poorly understood, but it must be linked to the evolved human capacity for sophisticated, high-fidelity, and selective social learning.
Implications for Human Culture: "History Matters-Pass It On"
Thus, humans are strikingly different from even closely related primate species in requiring social learning mechanisms to become such an enormous and essential part of their behavioral repertoire. For example, the oldest form of human subsistence adaptation is hunting and gathering, and yet even the daily practice of foraging requires skills, strategies, and bodies of knowledge that have accumulated and been refined over many, many generations. In addition, the wide range of ethnographically documented forager societies would not be able to inhabit such utterly diverse ecological settings as Arctic sea ice and tropical rainforests without complex multicomponent tools such as toggling harpoons, blow guns, poison-tipped arrows, sleds, kayaks, warm winter clothing, snow knives, and knowledge of how to build basic shelters, as well as baskets, pottery, and other containers for transporting, processing, and storing foods and fuels.
Each of these items of technology embodies stocks of knowledge and skill that couldn't be learned in an entire human lifetime of simple, individualized trial-and-error experimentation. Instead, the objects represent bodies of cultural knowledge that are reproduced, refined, and adjusted over many generations in the form of enduring lineages of tradition. These diverse material culture traditions are complemented by other forms of nonmaterial culture, such as deeper understandings of how to hunt seal through the winter ice or to detoxify seeds and obtain medicinal plants from a tropical rainforest. These bodies of cultural knowledge are also passed on in the form of socially learned rules, heuristics, and techniques, often with little or no understanding of how or why they work in practice (Henrich and McElreath 2003).
To summarize, human existence has several strikingly different features. One is the maintenance of large numbers of cultural traditions. The second is the widespread use of complex technologies-these are the elaborate multicomponent tools, artifacts, built structures, and other items of material culture that people learn to construct and use as part of their daily engagements with the world. Since earliest prehistory, material culture traditions like stone-tool making, basket weaving, and pottery manufacture have all been developed and maintained through the evolved human capacities for sophisticated social learning. This capacity for high-fidelity replication and selective improvement appears to underwrite the human capacity for cumulative culture, which is as much an evolved characteristic of human biology as being able to walk upright (Henrich and McElreath 2003:1, 27).
What Gets Passed On?
These general insights into the cumulative nature of human culture are important, but they immediately generate some deeper questions: What exactly gets passed on during human social learning? Earlier work on this theme had sought to identify discrete "units" or "memes" of cultural inheritance that were able to replicate themselves (Dawkins 1976). However, this simple "meme-as-replicator" model is problematic because the process of cultural inheritance is far more complex (Shennan 2002a:47). Similarly, social and cultural anthropologists are quick to emphasize that cultural knowledge is not passed on in a ready-made formats via a simple process of information transmission, akin to a kind of telegraphic transfer system, but that knowledge undergoes a process of continual regeneration through the social contexts of interaction that link novices and instructors (Ingold 2007:17).
More recent work on cultural inheritance has therefore shifted toward broader and more fluid concepts such as the generic term cultural information (Boyd and Richerson 1985; Richerson and Boyd 2005; Shennan 2000, 2002a; Smith 2001:96), which can be defined as skills, beliefs, values, and attitudes that are stored in human brains and acquired from others by teaching, imitation, and other forms of social learning (Richerson and Boyd 2005:61). In fact, there are many reasons to believe that what is transmitted is neither discrete nor faithfully transmitted, so the flexible and more general term of cultural variant is probably more appropriate (2005:63), at least when discussing cultural; transmission in more generic terms.
There is also growing interest in understanding how these cultural variants are reproduced as social traditions. Noting that Kroeber and Kluckholm (1952) once listed 168 different definitions of culture, Whiten et al. have argued that the term tradition is less controversial (e.g., Whiten et al. 2011:940). In this sense, a tradition can be defined as "a distinctive behaviour pattern shared . . . and acquired by social aided learning." They key point here is that new traditions can emerge and and persist over multiple generations, so the importance of social learning is embedded within the concept. In other words, it is the enactment of the enduring tradition that is social, rather than the constituent cultural variants that they reproduce.
Luckily, more precise definitions of "cultural information" or "cultural variant" have long been deployed in the study of material culture traditions (see chapter 2). Here, the choices made at different stages in the production of complex technologies like basketry, skis, or tailored clothing can be defined as "cultural traits," with the presence and absence of such traits documented across different artifacts. Of course, the knowledge, insights, and skills associated with the reproduction of these traits are acquired through observation, imitation, and other forms of social learning, are stored as information in human brains, and are then deployed strategically and creatively in further craft production activities-this creates heritable continuity. Variability in such traits, and the ways in which they are combined across different traditions and social groups, can then be recorded through systematic ethnographic survey. Importantly then, the defining, recording, and studying of distribution of such traits involve understanding human decision-making processes and the ways in which this creativity and historical agency generates broader patterns of cultural diversity and longer-term transformation (see Ingold 2007:16) Together, these combined features of heritable continuity on the one hand, and the cumulative addition of novelties and change on the other, also mean that these human cultural traditions can also be argued to exhibit evolutionary properties of descent with modification (e.g., Lycett 2011; and see following).
Researching Cultural Evolution
If the propagation of human cultural traditions can be examined from this kind of ' social learning' perspective, then what is the best theoretical framework for investigating the deeper, cumulative dynamics of this historical process? Specifically, what creates different spatial and temporal patterns of variability and change in different cultural traditions?
Exploration of this kind of descent with modification perspective on the history of human cultural traditions builds on the foundations of dual inheritance theory (or cultural transmission theory) associated with the work of Boyd and Richerson (1985) and Cavalli Sforza and Feldman (1981). Its main argument is that humans possesses two distinct systems of information transfer, one genetic and the other cultural (see also Mesoudi et al. 2006; Mesoudi 2011; O'Brien et al. 2013; Richerson and Boyd 2005; Shennan 2002a, all with references). The biological inheritance system involves replication of genetic information via sexual reproduction and is shared by humans and other animal species. Simply put, human parents pass their genes on to their biological offspring, but to no one else.
In contrast, the inheritance system involved in the replication of cultural traditions is purely cultural; it involves the transfer of cultural information between individuals via social learning, as examined previously. And although people can and do acquire substantial information through their own trial-and-error experimentation, they most often tap into the enormous bodies of accumulated cultural information by learning socially from others. Having acquired this information, they replicate and modify it through their own actions and practices before passing it on to others, all of which can generate a kind of dynamic heritability in long-term cultural traditions. This means that these cultural traditions, just like genes, can also be understood as an information transmission system that exhibit evolutionary properties (Lycett 2011; Mesoudi et al. 2006:329).
Before examining this relatively new perspective in some greater detail, it is important to define what is meant by an evolutionary approach to culture in order to clarify how it relates to other kinds of evolutionary thinking. This also serves to situate the current descent with modification approach within earlier streams of research and thinking.
Defining Cultural Evolution
The concept of evolution has had a long history, and many of the central ideas have changed significantly during the past hundred or so years, with some undergoing fundamental revisions in the past fifty years. In archaeology and anthropology, by far the most influential strand of evolutionary thinking has been the idea of "progressive" social evolution. This emerged in the later nineteenth century and was associated initially with Herbert Spencer and Lewis Henry Morgan, charting general cultural developments from the stages of "savagery" through to "barbarism"and eventually to "civilization"(see Pluciennik 2005). In the mid-twentieth century, a more multilinear understanding of cultural evolution emerged through the work of V. Gordon Childe and Julian Steward; others, like Leslie White, continued to focus on understanding more general patterns of cultural evolution (Childe 1951; Steward 1955; White 1949, 1959). These perspectives had emerged in the 1930s and 1940s and influenced archaeology and anthropology in the 1950s and 1960s. These now neoevolutionary approaches were eventually combined with ecological neofunctionalist thinking in the work of Elman Service, Marshall Sahlins, and Marvin Harris (Shennan 2009a:1-2; Trigger 2006:386-444). However, just as these two general strands of evolutionary thinking (social evolution and neoevolutionism) came under sustained critique within each discipline throughout the 1970s and 1980s (Trigger 2006:444-78), the evolutionary branch of anthropology began a process of intellectual renewal, largely because of integration of earlier developments that had taken place outside the discipline some decades earlier (Shennan 2009a 2).
These previous developments were all associated with the early to mid-twentieth century neo-Darwinian "modern evolutionary synthesis" in biology, which successfully integrated Mendelian theories of genetic inheritance with Darwinian processes of natural selection (see Huxley 1940, 1942). This provides a widely accepted view of biological evolution in which transmission, mutation, selection, and drift combine within in the general process of descent with modification. This process works at different scales, generating changing frequencies of genetic traits within populations (microscale evolution) but is also linked to evolution of traits across populations, where phylogeny (a branching tree diagram) reconstructs the general tendency for new evolutionary lineages, such as species, to split away from ancestral forms (macroscale evolution).
Central concepts of the modern evolutionary synthesis were slowly incorporated into anthropology, initially spawning the controversial field of human sociobiology, which explored links between patterns of human social behavior and the process of natural selection (Dawkins 1976; Shennan 2009a:2, 11-12; Wilson 1975), or how different human cultural and social institutions affect individuals' survival and reproductive success (Chagnon and Irons 1979). This field eventually fragmented into evolutionary psychology (Bentley et al. 2008:120-22; Fuentes 2009:44-52; Shennan 2002a:15-16, 2009:3; Smith 2000) and human behavioral ecology (Bentley et al. 2008:117-20; Fuentes 2009:38-44: Shennan 2002a:15-16; 2009:3-4; Smith and Winterhalder 1992; Winterhalder and Smith 2000).
Meanwhile, a third neo-Darwinian approach was emerging: dual inheritance theory (or cultural transmission theory). This was developed through the work of Cavalli Sforza and Feldman (1981), and initially involved experimentation with modified mathematical models developed in population genetics to understand how cultural attributes (as opposed to genes) were passed on from one person to the next by social learning, and how these processes could eventually have population-level outcomes by affecting the frequencies of cultural attributes carried within populations over time. A first full statement of the dual inheritance approach was eventually set out in their Cultural Transmission and Evolution: A Quantitative Approach (1981) and established a basis for modeling cultural changes within a modern evolutionary framework (Shennan 2009a:3). This was followed by publication of Boyd and Richerson's Culture and the Evolutionary Process (1985) and later by Durham's Coevolution: Genes, Culture and Human Diversity (1991).
Elements from this broader evolutionary approach to cultural transmission provide the main theoretical framework for the book. Genes and culture are approached as two analytically distinct systems of inheritance, each of which creates historically contingent patterns of cultural and genetic diversity (Boyd and Richerson 1985; Cavalli Sforza and Feldman 1981; Durham 1979, 1982, 1990, 1991; Richerson and Boyd 2005; and see Collard et al. 2008). Therefore, cultural evolution, in this very specific sense, refers to change over time in the nongenetic information possessed by human societies as it is affected by transmission and innovation processes.
On the one hand, dual inheritance theory drew analogies between the parallel mechanisms for inheritance, mutation, selection, and drift as they operated on both cultural information and on genes. For example, the genetic system is based on biological reproduction, while the cultural one involves transmission of cultural information via teaching, imitation, and other forms of social learning.
But on the other hand, dual inheritance theory also served to highlight a range of powerful forces in cultural evolution that have no real analogue in genetic inheritance systems-these features make cultural evolution unique. For example, people cannot decide on their own genetic parentage-they are literally born with it, but over the course of their lives, they can-and generally do-choose what kinds of cultural information to copy, when, and from whom. These sources of cultural information can range from biological parents and close kin through to friends, strangers, or influential leaders; moreover, people can modify, adjust, or even discard the kinds of cultural information they acquire, basing these judgments on a wide range of factors (see following).
These distinctive aspects of cultural transmission mean that new ideas can spread rapidly through populations within a single generation, ensuring that cultural evolution can potentially be much faster than genetic evolution; in addition, the routes of cultural transmission may not always be in step with the parent-to-offspring routes of genetic transmission. Together, these two unique features of cultural transmission enable rapid adoption of highly adaptive traits among unrelated individuals and can also enable the spread of nonadaptive or even highly maladaptive traits. In fact, "culture is interesting and important because its evolutionary behaviour is distinctly different from that of genes" (Richerson and Boyd 2005:7).
More generally then, dual inheritance theory involves analysis of social learning and intergenerational cultural inheritance, but it is also a theory about history; and for humans, it is about the unique histories of specific social traditions and how the cultural content of these traditions evolves through time (Bentley et al. 2008:112-13; Boyd and Richerson 1985; Cavalli Sforza and Feldman 1981; Collard et al. 2008:204-5; Durham 1990, 1991; Fuentes 2009:52-57; Richerson and Boyd 2005). Thus, the three key implications of this descent with modification approach to cultural evolution are worth highlighting:
1. Cultural evolution has broad similarities with biological evolution, and thinking about culture within a descent with modification framework can generate new perspectives and questions about how best to explain patterns and processes of historical change in human cultural traditions, such as language and complex technologies, as well as a range of other behaviors (Richerson and Boyd 2005:58-98). Cultural evolution is cast as a historical process that involves the cumulative decision-making processes of human agency, which affect consciously and unconsciously how social traditions are replicated, as well as the specific cultural content of those traditions. How these cultural attributes evolve within populations (microscale cultural evolution)-and also across different populations (macroscale cultural evolution), societies, or cultures-then becomes a question for empirical research. This serves to ground the approach within long-standing debates in archaeology, anthropology, and related disciplines about what factors generate continuity, diversity, and change in human cultural traditions.
2. The analogy between genetic and cultural transmission also has important methodological implications. Despite some key differences, the two inheritance systems share enough general similarities that some of the powerful analytical tools developed to study biological evolution and genetic transmission can be adapted to undertake empirical study of cultural evolution (Collard and Shennan 2008; Collard et al. 2008; Mesoudi et al. 2006). Creative exploration of both the positive and negative analogies between these genetic and cultural inheritance systems then provides a useful starting framework for the wider cross-disciplinary application of models, methods, and theory developed in the biological sciences to analysis of cultural data (Collard and Shennan 2008; Lipo et al. 2006; Lycett 2011; Mace et al. 2005; Mesoudi et al. 2006; Mesoudi 2011; O'Brien and Lyman 2003a, b, O'Brien et al. 2013; Steele et al. 2010).
3. Finally, genes and culture can be argued to form separate components of a wider "coevolutionary" system. This means that independence can be maintained between the study of cultural and biological evolution (Bentley et al. 2008:112). However, humans clearly inherit both genes and cultural information, generating a "swirling dance" of genetic information and cultural traditions that are reproduced among individuals and across generations (Boyd and Richerson 1985; Durham 1990, 1991. 1992; Richerson and Boyd 2005:191-95; and see chapter 6). In other words, biological histories of people and the histories of cultural traditions are linked to one another, but they are certainly not the same thing (Shennan 2004:25).
The next sections of this chapter examine some of the productive conceptual issues that arise from identifying these general parallels between genetic and cultural inheritance systems. These can be explored under three main headings: (1) How is cultural information reproduced by social learning within specific populations? This relates to the theme of cultural propagation. (2) Do coherent cultural lineages form, and what kind of historical patterns emerge? This relates to the theme of cultural coherence. (3) Humans simultaneously maintain large numbers of enduring cultural traditions, but to what extent does the history of these cultural lineages become bundled together? This relates to the theme of historical congruence. At this stage, the focus is on understanding generic cultural evolution. The more specific ways in which technology and material culture traditions might start to evolve are examined toward the end of the chapter.
In contrast, methodological implications arising from the positive and negative analogies drawn between genetic and cultural transmission are considered more fully in chapter 2 and are applied to empirical data sets in chapters 3, 4, and 5. The third implication-genes-culture coevolution-also of great importance, remains beyond the scope of the book but is considered briefly in the conclusion in chapter 6.
Propagation of Cultural Traditions
The main argument here is that the propagation of information in cultural traditions has some broad similarities, but also some important differences. This creates a useful framework for exploring how cultural information (attributes, variants, or traits) evolve within populations (microscale cultural evolution).
Biological organisms reproduce after their own kind, and they are fundamentally genetic systems that carry information. Each organism contains a distinctive genome, a collection of nucleic acid molecules that contain genes. Genes carry information; they also replicate. When an organism reproduces it makes copies of itself, passing on this information to another generation of similar creatures. In this way, organisms are self-reproducing structures that operate on the instructions of a genome-this encodes a genetic program that specifies the basic structures of the organism and its general mode of operation. The genome also specifies the structure of a new replica of itself, so that each offspring will have its own copy of a genome. Changes are inevitable in a heredity system, and during the processes of replication, new organisms sometimes inherit genomes that contain either errors (mutations) or novel combinations of genes. These changes make offspring different from their parents in a stable, heritable way.
Evolution emerges naturally from this picture. The process is played out in local ecosystems because reproduction is generally so successful that each new generation tends to produce more offspring than can possibly survive. In local ecosystems every new generation of organisms has to find a way of making a living for itself within a specific ecological niche, but because of minor variations in their genomes, some of these organisms tend to be marginally better at this than others. These better-adapted individuals will be better nourished and tend to have more offspring, and they will thereby be selected as the parents of the next generation of offspring, which will inherit the features that made their parents better adapted. Through this general process, organisms will acquire new characteristics generation by generation.
Biological evolution has four essential features: (1) transmission: the inheritance of genetic information via sexual reproduction that links parents and their direct offspring; (2) mutation: this is the ultimate source of all variation and the key to evolution. Genetic mutation and recombination is a random, low-frequency, and nondirectional process that creates new variants and injects novelty into the inheritance system; (3) selection: these processes change the frequencies of the genetic information being passed on; selection on different variants is directional, that is, certain variants within the wider pool of variants will have characteristics that enable them to adapt, survive, and reproduce better in new environmental conditions; (4) drift: this relates to random changes in the frequency of genetic traits over time, which is not a result of selection. In small populations these random patterns of change can have major impacts on what gets passed on to future generations. The dynamic interplay of these four factors feeds into an evolutionary processes of descent with modification that affects the frequency of genetic traits over time.
Biological evolution is a historical process, and changes injected at every new generation are cumulative. However, new features must be designed by slowly modifying what is already present, and each new pattern depends on what has gone before. This means that each organism has a unique combination of genes and is a unique historical object that carries with it genetic information derived from its ancestors. The implication is that diverse sets of creatures will have acquired their structures through evolution from common ancestors that were then shaped, generation after generation, by the evolutionary forces of descent with modification.
Every unique organism becomes what it is through a long evolutionary process, but this process can be studied at two different scales: (1) microevolution focuses on the smaller-scale changes that occur in genetic trait frequencies within a single species or population, but it can also be studied in terms of (2) macroevolution, which operates at the scale of separated gene pools and focuses on change that occurs at or above the level of species, including the process of speciation, which refers to the splitting away of different evolutionary lineages through a process of "cladogenesis" or "phylogenesis" (see following).
Cultural inheritance systems share some general similarities: (1) information (cultural) is passed on between individuals via social learning, generating heritable continuity (transmission); (2) change is also introduced into the system, potentially through simple copying error (mutation); (3) at any one point in time, people are able to choose from a pool of cultural variants, and this affects what gets passed on to the next generation (selection); (4) small-scale statistical anomalies in populations can lead to the disappearance of traits (drift) (see Richerson and Boyd 2005:69; Shennan 2002a:55; and see following). Viewed in these terms, cultural evolution can also can be viewed as a complex set of sampling and modifying process that affect what cultural information is passed on to subsequent generations.
"Population thinking" is also central to cultural evolution (Richerson and Boyd 2005:5, 59). For example, biological species are populations of individuals that carry a pool of genetically acquired information through time; some genetic variants are able spread through time, others diminish. In similar ways, human populations can be argued to carry a pool of socially acquired cultural information that they reproduce through imitation and other forms of social learning. Decisions made by individuals will work to increase or decrease the frequency of particular cultural attributes within that population, and over time, this affects the cultural attributes carried by future generations within the population. The classic exploration of this kind of population-scale phenomenon relates is Everett Rogers's work on the dispersal of innovations (1962). Some of the main features of a successful adoption cycle can be summarized in figure 1.1, where the bell curve records the rate of adoption. At first, no one in the population has the new attribute, but as it starts to catch on, more and more people in the population decide to take it up, and the rate of adoption rises sharply, but eventually peaks. Later in the adoption cycle, few people have yet to take up the innovation, and the rate drops until everyone in the population eventually has it.
This kind of adoption dynamic can also be represented as an S-curve, which charts the cumulative uptake of a new attribute, starting in the early stages when only innovators, early adopters, and then the early majority have it, followed by the late majority and, finally, the laggards, who are the last group to adopt it (figure 1.1). This is a fundamentally historical process because what has gone before directly affects what can happen next: at first no one in the population possesses the new cultural attribute, but gradually, more and more people choose to adopt it, until eventually, everyone in the population possesses it. Yet the elegant simplicity of these graphs can be deceptive: these changes are not an automatic process; they came about only through the decision-making processes of specific individuals (human agency). Over time, each of these individual decision-making events feeds into cumulative, and sometimes dramatic, population-scale outcomes, involving radical shifts in the kinds of cultural attributes reproduced by human communities. Understanding the persistence of traits within populations is therefore about understanding the intersections between historical processes and human agency.
So much for the general similarities between genetic and cultural inheritance. But what are the major differences? The rest of this section examines some of the important forces that are present only within cultural evolution, the most important being: (1) people can potentially have very many cultural "parents," not just their biological mothers or fathers; (2) humans have agency and are to a great extent able to make their own decisions about what kinds of information to adopt, when they adopt it, and also from whom.
Unique Modes of Cultural Transmission
In human populations, genetic information can only be passed "vertically" between parents and their biological offspring. Transmission of cultural information can also follow this vertical route but often does not, opening out other possible routes for cultural attributes to spread in populations. For example, "oblique"transmission involves learning members of the older generation who are not biological parents; "horizontal"transmission is furthest from the genetic route and involves individual learning members of the same generation. Cultural transmission may also be "one to many" (for example, a teacher instructing a class) or "many to one" or "concerted" transmission (for example, members of a conservative older generation insist on respectful practices during interactions with children). This might also tend to ensure uniformity in cultural practices, but also perhaps slower rates of change.
Shennan (2002a:50) provides a useful summary of these diverse modes of cultural transmission and also explores some the potentially different ways in which they might affect the acceptance of innovation, generate variations in difference between individuals within populations, and also create differences between groups and affect overall rates of cultural evolution (table 1.2). Cultural information transmitted from parents to offspring may be slower to change, whereas novel traits picked up from peers may lead to faster changes. Teachers instructing a school class can quickly pass on new information to a large group, and this may result in relative cultural homogeneity within that social group because the new information stems from a single source. In contrast, a powerful generation of conservative elders may reduce capacity for new information to spread easily. Of course, these are simple predictions, and their veracity would need to be examined in specific ethnographic settings, but they are useful here in sketching out some of the potential links between contrasting modes of transmission and resulting patterns of cultural diversity and change.
Distinctions between these different modes of transmission (vertical, oblique, horizontal, etc.) are useful in mapping out some of different routes along which cultural information can be transmitted. But in order for culture to evolve, there also has to be a mechanism that injects novelty in to the system. This is where some of the unique and potentially very powerful features of cultural inheritance really start to emerge, and a important distinction can be made between "random forces," which have some analogues in genetic inheritance; and "decision-making forces" which do not (Richerson and Boyd 2005:69; and see table 1.2).
Random Forces in Cultural Evolution
These forces range from a kind of "cultural mutation"through to "cultural drift,"both already noted. Mutations emerge through simple copying error or misremembering aspects of cultural information (Eerkens 2000; Eerkens and Lipo 2005). These kinds of novelties are nondirectional, and they may die out quickly if corrected by group consensus or peer group pressure, but in some cases they may catch on and eventually have substantial cumulative effects in the wider population.
Demographic factors are also important because these can affect patterns of cultural variation and change over time. The phenomenon of cultural "drift" is directly linked to such parameters (e.g., population size, density, and interconnectedness; see Henrich 2004). For example, when populations are relatively small, chance (i.e., stochastic) factors can play a much greater role in determining which cultural elements will be retained and transmitted to future generations (see Neiman 1995; Shennan 2000). This is because smaller populations are more likely to experience statistical anomalies that mean, for example, that information replicated by only a few specialists can disappear for chance reasons. For example, if an elderly master carver living in a small and isolated community on a very remote island happens to have no biological offspring and cannot find any apprentices in the local population, then his entire repertoire of carving skills will die out when he passes away.
In much larger populations, these kinds of specialized cultural attributes are generally maintained by a greater number of individuals, such that chance effects like this are less likely to have such dramatic outcomes-there are simply more master carvers, most have offspring, and many other youngsters may also be willing to learn the skills and then share them with others. This means that in a larger and better connected population the carving skills and attributes simply have more opportunities for longer-term persistence. In general then, sustained population growth combined with expanded social transmission networks will result in more effective instances of cultural transmission, serving to "buffer" against the loss of useful cultural traits via this kind of cultural drift process (Henrich, 2004; Kline and Boyd 2010; Lycett and Norton 2010; Powell et al. 2009; Shennan, 2000, 2001, 2002:55).
Decision-Making Forces in Cultural Evolution
These forces are more numerous and include both guided variation and biased transmission. The term guided variation refers to deliberate changes made by an individual to an existing cultural variant, generating changes that are subsequently transmitted to others. This would equate, in essence, to purposive innovation whereby an individual learned one way of doing something from a parent or other source and then modified the cultural variant they'd learned through practical experiences and trial-and-error experimentation. For example, a hunter might initially learn from his father how to haft knives with bone, which he initially followed, but after trying many different materials, eventually realized that willow was better because it had a superior grip and was easier to work into the required handle form. He then started to use willow-handled knives and eventually taught his sons to haft knives in the same way.
In many potential settings, and with much cultural information inherited from others, this kind of systematic experimentation among a range of different cultural variants may simply not be possible: it make just take too long to generate enough comparative information for a proper evaluation (because a hunter only hafts a new knife every ten or twenty years), or require too much effort (blades are costly and each hunter can afford to have only one knife at one time), or just generate "noisy" performance data that are hard to interpret (it doesn't seem to make much difference whether the handle is willow, beech, or bone-they all seem to do the job equally well). As a result, it may be easier to just stick with what's been learned during childhood and get on with other more important matters in life.
But another important alternative is to watch others, see what they do, observe the decisions they make, and then copy some of these alternative ways of doing things. This is termed "biased" cultural transmission and occurs when people preferentially adopt some preexisting cultural variants rather than others (Richerson and Boyd 2005: 68). This can have a very powerful cumulative effect on the distribution of cultural attributes within populations, for example, in the spread of new innovations (see figure 1.1). Richerson and Boyd liken this process to a process of cultural "shopping," whereby people are continually exposed to alternative ideas and values and can then evaluate and choose among them about which one to adopt-people are generally not stuck with what they have initially been taught, for example, by their parents (2005:68-69).
There are several different forms of biasedcultural transmission, and each is distinguished by key differences in the way that the decision about whether to adopt new a variant is reached. For example, content-based (or direct) biasinvolves careful comparison and evaluation of an initial cultural variant against an alternative variant, based on some kind of objective performance criteria. For example, the hunter's son described previously initially learned to haft knives with willow from his father, but then saw that his neighbor used a leather binding, and after trying out both variants, switched to leather because it provided a better grip in the cold conditions of the winter hunting season. In this way, directly biased transmission generally arises from some kind of generalized cost-benefit calculation (Richerson and Boyd 2005:69).
In contrast, indirectly biasedtransmission occurs when the choice to adopt a new variant as part of the wider "cultural shopping" process is based on other contextual factors rather than the intrinsic features of the rival variants. In fact, these include a number of unique forces in cultural evolution that can be grouped under the more general heading of "contextbiases." This is because it is some aspect of the wider social context that influences what is eventually transmitted, and choice is certainly not based on intrinsic features of the new variant but on nonrelated criteria. Here, there is no need (or no possibility, or time for) evaluation or learning by trial-and-error-judgments are based solely on the source of the copying.
The category of context bias can further be divided into "model-based bias"-here the choice to adopt a new cultural is based on the observable attributes of individuals who exhibit the new trait. For example, plausible model-based biases might include a predisposition to imitate successful leaders or persons thought to be prestigious in the local culture ("prestige bias"), or perhaps a desire to imitate individuals similar to oneself. However, a better indication of how good a new attribute is can be derived from just watching how many other people have it already; the more people who have it, the greater the odds that it might be more suitable. This tendency generates "frequency-based bias"-the most advantageous variant is thought likely to be the one already in most common usage, so copying this variant is likely to be a simple fast-track route to the correct attribute, or the variant that tends to perform best.
In a complex cultural worlds, often defined by limitless opportunities for acquiring or inventing new kinds of cultural attributes, these biases provide cheap and potentially useful heuristics or cultural "rules of thumb" that allow individuals to reach decisions quickly, and in ways that often require less investment in costly or time-consuming experimentation, which would involve systematic case-by-case evaluation strategies of the kind employed by scientists in materials testing labs.
As a result of these uniquely cultural transmission mechanisms, useful or potentially adaptive technologies and behaviors can spread rapidly within a single generation. In fact, this is one of the key evolutionary advantages of culture: behavioral responses can be much quicker than adjustments brought about by natural selection working only on genetic variation. But while allowing rapid culture change, neither frequency-based bias nor model-based bias includes any kind of objective reality check-the decision to adopt a trait is based purely on who already has it and also on what other people in the population are doing rather than the features of the trait itself. These features mean that cultural transmission mechanisms can exhibit runaway properties that lead to the rapid spread of nonadaptive or even maladaptive traits (Richerson and Boyd 2005:148-90).
Selection Mechanisms in Cultural Evolution
Building on these insights, the cultural attributes carried by individuals and their wider populations will also be subject to different forms of selection, but in contrast to genetic inheritance, this operates on several different levels due to variability in the range of transmission modes (e.g., contrast vertical, oblique, and horizontal transmission), as well as the other random and decision-making forces that affect cultural evolution (see, e.g., Shennan 2008, 2011). To summarize: (1) selection operates directly on the people replicating these cultural traditions via their genetically inherited dispositions, for example, via their susceptibility to specific diseases such as virulent new strains of an influenza virus (this is akin to natural selection in the narrow biological sense)-this might kill off some, most or even all, members of a specific population; if some cultural traits are carried exclusively by this population, these will also disappear when that population dies out (e.g., a particular language or way of making pots); (2) selection can also operate on populations via their cultural traditions-for example, farming communities may enjoy better general nutrition than foragers, leading to more offspring who live longer; in time, the cultural traits propagated by the expanding farming population will spread, and perhaps even replace those carried by the foragers; (3) complex processes of purely cultural selection that also affect the frequencies of cultural attributes in populations (table 1.3; Shennan 2002a:53; 2004:22-22; 2008; 2011:1070-71). It is these latter, purely cultural forces of selection that are the primary manifestations of human agency, that is, they directly reflect the conscious and unconscious decision-making processes that people employ throughout their lives when choosing whom to copy and what cultural attributes to retain, discard, adopt, or modify, as reviewed previously.
If microscale biological evolution is about studying the changing frequencies of genes in populations through time as a result of such processes as natural selection, then microscale cultural evolution refers to the changing distributions of cultural attributes in populations. This can also be seen as a kind of steady but cumulative copying and editing process, but one that involves social learning and human decision-making processes. Thus, distributions of cultural traits are likewise affected by certain processes such as natural selection, but also by a range of others forces that have no analogue in genetic evolution, all of which make the outcomes of cultural transmission much harder to predict (Shennan 2002a:65).
Parental Investments and Psychological Biases in Learning and Copying
The acquisition of cultural information through social learning does not necessarily come for free. In fact, it can be very costly for everyone involved. Passing on many complex cultural traditions can require sustained investment of time, concentration, and energy, both by the experienced practitioner and by the novice, especially in the acquisition of the many sophisticated but essentially arbitrary practices that humans devote so much time to maintaining. For example, many inherited cultural attributes are very hard to acquire, such as the complex motor skills required in basket weaving or flint knapping, or the the ability to play a musical instrument well, or to speak and write in a language with deep fluency. The same goes for all the skills and knowledge that humans have relied on over many millennia to undertake their routine subsistence activities like hunting, fishing, gathering, or the ability to process animal skins and make tailored clothing or store up resources for the lean season. In fact, becoming an accomplished practitioner of the many cultural traditions that individuals rely on in every sphere of their daily existence can require literally years of close pedagogic interaction, and during this time, infants must also be fed and sheltered until they can make their own active contributions to subsistence activities; only later will they become both independent members of the population and accomplished cultural practitioners. Perhaps for this reason, it often appears to be parents who invest most heavily in rearing their direct biological offspring, equipping them with the essential skills and cultural attributes that they need. For example, there is compelling ethnographic evidence that acquisition of traditional craft skills is mainly if not entirely vertical or oblique rather than horizontal-novices are generally more likely to learn from parents or other members of the older generation than horizontally from members of the same generation, which increases the heritable continuity in many cultural traditions (Shennan 2002a:40, 49; Shennan and Steele 1999; and see Hosfield 2009 for a broader discussion).
However, the general interest in studying children's acquisition of cultural attributes may have overemphasized the role of vertical transmission at the expense of other modes of transmission that become more important later in the life cycle (Tehrani and Collard 2009b:289): understanding how social learning over the entire human life history feeds into the evolution of specific craft traditions therefore remains an important topic for future research (chapter 3). Archaeologists, with their interest in longer-term histories of human cultural tradition, are particularly interested in the dynamics of social learning in the smaller-scale band societies that are typical of some modern foraging groups. However, ethnographic research on hunter-gatherer cultural transmission has been limited, though Hewlett et al. (2011), for example, have studied cultural inheritance in African forager communities. They concluded that cultural transmission is initially very rapid and predominantly vertical, but then switches to oblique routes between the ages of six and twelve and tends to involve broader general observation and wider imitation. Thus, it may be more useful to think of cultural transmission as a more extended process, whereby (1) infants initially acquire a repertoire of cultural skills and attributes vertically from their biological parents (and perhaps also obliquely from related members of the older generation), but then (2) older children, adolescents, and young adults eventually become much more selective in choosing from the range of other available cultural models, updating their cultural repertoire by innovating and imitating, thereby generating some of the wide range of cultural transmission biases explored previously (Whiten et al. 2011:944).
That said, many inherited cultural traditions may, in fact, require so much sustained social and material investment in teaching and learning that it may be difficult or even impossible to find the time and resources later in life to dedicate to such an endeavor. In contrast, other traits, like a catchy tune, fashionable new words, or expressions, small technical improvements, or even novel decorative motifs can easily be picked up at any stage of life from any range of sources and then built into existing traditions. Overall, this may mean that certain elements of a single cultural tradition like pottery making may be more stable, such as its basic method of coiling and shaping, which may have been acquired early in life and are harder to adjust, while other more ephemeral features like its surface decorations may be much more dynamic and variable over time, especially if they can diffuse easily and quickly among different groups of potters. These factors make it important to be specific about the kinds of cultural information being acquired and modified at different life stages.
A number of psychological factors also play an important role in cultural inheritance (see: Henrich and McElreath 2007). For example, even during their early phases of learning, children do appear cognitively able to make informed choices about whom to copy. Harris and Corriveau (2011) have argued that children tend to be less discerning about the kinds of information they learn but very aware of whom they learn from, including increased receptivity to information from familiar and reliable caregivers (with a tendency to reinforce vertical transmission from parents to offspring), and greater receptivity to informants from within the same cultural group. In general then, children may be less likely to endorse, imitate, and trust information from deviants from their own groups or from informants from a different group. Overall, this cognitive disposition may promote vertical transmission between parents and offspring, and where oblique or horizontal transmission does occur, it tends to be of a relatively conservative nature, primarily involving conformist members of their own cultural group (2011:1185).
In a similar way, general copying decisions may at all life stages relate not just to the specific cost or difficulty associated with acquiring new traits or attributes but may be affected by a range of other evolved psychological biases. These may include a tendency to preferentially imitate individuals perceived to be successful (Richerson and Boyd 2005:124-26), or simply to conform to the majority (Richerson and Boyd 2005:120-24; Henrich and McElreath 2007; McElreath and Henrich 2007, with references). In the latter case, such conformity bias would then serve to reduce general cultural variation within a population but tend to increase it between populations. Overall then, many of the ways in which people acquire and maintain cultural traditions through imitation and general social interaction tend to promote cumulative patterns of cultural diversification that can eventually become self-reinforcing (Boyd and Richerson 2005:379-96; Pagel and Mace 2004).
How Do Social Institutions Structure Propagation of Cultural Traditions?
Clearly, these microscale transmission dynamics are "agent centred" (Shennan 2011:1071); they focus on specific people and the ways in which their individual decisions lead to the propagation of cultural attributes and the cumulative effects of these processes at the scale of populations. In this sense, the analytical focus is on shorter time scales: from day to day, year to year and also along human biographic time scales. The interactions among all these factors are highly complex, and cultural traditions can potentially evolve along many different kinds of trajectories. However, simply highlighting complexity and potential variability does not address the more challenging question of what range of outcomes is more or less likely in any given historical setting and why? At this stage in the argument, it therefore becomes important to consider how other kinds of factors might start to structure opportunities for interaction, social learning, and skilled practice, and thereby, influence the deeper evolutionary patterns manifested by specific cultural traditions.
If a focus on the role of human agency within microscale cultural evolution highlights the role of individual decision making and personal choices, then it should also be tempered by an examination of other contextual factors that can structure opportunities for social interaction and cultural imitation. For example, one potentially important factor is the ways in which human populations organize themselves into kin groups and communities and how they move and interact within and between these groups over during different stages of their life. Borrowing from agency/practice theory, these enduring structures can be termed "social institutions" because they extend beyond the life spans of individuals and often persist over many generations (Giddens 1984:35-36; Pred 1990:123). For example, in many of the smaller-scale societies documented by ethnographers, these social institutions include settlement and subsistence patterns, territoriality and property rights, but perhaps most important, gender roles, kinship relations, and in particular, postmarital residence and resocialization patterns (Herbich and Dietler 2008). In fact, the structuring role played by these kinds of social institutions-and especially postmarital residence rules-starts to emerge as one of the primary factors that mediates between microscale propagation of cultural traditions and larger-scale patterns of cultural diversity (see chapters 3-6).
Coherence in Cultural Traditions
The previous section examined some of the enormously variable ways in which cultural attributes can be propagated. The interplay of these factors mean that cultural traditions can potentially, evolve in any number of different ways, which in turn affects larger-scale patterns of cultural diversity. Building on these insights, this next section asks an important follow-up question: To what extent are human cultural traditions coherent? This is a more controversial topic and raises some deeper questions about the likely patterns of human cultural evolution.
What is meant by cultural coherence? As attributes are propagated within specific social traditions, certain sets of traits may become integrated together and subject to strong heritable continuity within a given population, whereas other traits have a more individualistic history as a result of being borrowed from others or recombined in different ways.
A good example is the social tradition of having regular meals with family, friends, or members of the wider community. On a more abstract and theoretical level, it might be possible to speculate that this meal could just involve collective consumption of foodstuffs-during different meals, there could be any number of different foodstuffs and drinks, all from different sources, and in any order or combination, in any location, and using an infinite and perhaps even random range of serving vessels. In this scenario, the social tradition of having a daily, weekly, or monthly meal together endures, but it has little internal cultural coherence in terms of the way in which internal cultural elements of the tradition are being combined. The social tradition of having the meal has a history, but there is little coherence in its specific cultural content, which is more or less random, just a churning turnover of endless variability.
In contrast, and on a more realistic level, the social traditions associated with eating meals across most human cultures appear to have at least some structure and internal coherence, although the specific content is often arbitrary and subject to enormous cross-cultural variation: usually only certain foodstuffs are sourced and used; dishes are usually served in a particular order and combination, and often with certain kinds of vessels, specific seating arrangements, and even with formalized opening, serving, and closing ceremonies. In some cultural and ritual settings, these kinds of repeated events might have a highly structured and precisely defined internal cultural grammar. In this way, these general dining activities persist as a social tradition, but as an integral part of the reproduction of this tradition, a tightly integrated combination of specific cultural traits and attributes are also propagated. In these kinds of setting, the social tradition of dining together persists over time, but so too does this coherent combination of specific cultural traits and attributes, which together make up a distinctive lineage of cultural tradition, which will have its own history and genealogy. Theoretically, the deeper history of these coherent cultural lineages can also be reconstructed with the right kinds of concepts, methods, and approaches.
Social Scales of Cultural Coherence?
Specific combinations of cultural traits can therefore persist over time, but at what kinds of social scale might such coherent cultural entities emerge? Theoretically, distinctive cultural traditions can emerge within individual friendships or household units via their defining routines and microrituals, but they can also be shared by larger kin groups, settlements, communities, and small religious sects, as well as by large-scale political units or centrally organized religions. Perhaps one of the clearest examples of a coherent cultural tradition is what linguistics call a "speech community"-this a unique communication system propagated by a specific population, and consists of essentially arbitrary sets of traits, attributes, and collective rules (i.e., the words, syntax, and grammar that make up a particular linguistic tradition). Together, these cultural conventions and established combinations provide an effective means for communicating with others who are also participating in the same linguistic tradition. These acts of participation within a given linguistic tradition produce its distinctive cultural content within the population over time. Different groups often speak different languages; each language can form a distinctive and relatively coherent lineage of cultural tradition that generally outlives any one individual and persists in broadly recognizable format over many human generations.
History and Genealogy of Cultural Lineages?
Coherent cultural entities such as language clearly endure within populations over time. Other cultural lineages may endure in similar ways. But to what extent can the deeper history and genealogy of particular cultural lineages be reconstructed? What kinds of descent relationships emerge? Can different cultural lineages be traced back over time to to a single common ancestral entity, or do other kinds of historical patterns emerge?
These questions can also be examined by exploring some of the interesting parallels between cultural and genetic inheritance. If human culture is an inheritance system, defined by heritable continuity, then the argument can be made that cumulative process of social learning and cultural inheritance over generations might lead to the formation of phylogenetic descent relationships among different cultural entities, just as genetic transmission does at the scale of biological macroevolution (Foley 1987; Grancolas and Pellens 2005; Lycett 2011; Mace et al. 2005; Mace and Holden 2005; Lipo et al. 2006; O'Brien and Lyman 2003a, b; Mesoudi et al. 2007; O'Brien et al. 2001; O'Brien et al. 2013). In exploring how cultural traditions evolve at the scale of different populations, societies, cultures, or language communities, the most useful analogies here are with biological processes of macroevolution, which operate at the scale of separated gene pools and focus on change that occurs at or above the level of species (Ridley 2004). This change includes the process of "speciation," which refers to the splitting away of different biological lineages through an evolutionary process of cladogenesis or phylogenesis. This occurs when individuals in one population slowly become separated from others, and then remain apart long enough for the two new populations to evolve genetic differences. These genetic differences can emerge from, and are often sustained by, a wide range of factors, such as increasing habitat isolation (physical separation), temporal isolation (breeding at different times), sexual isolation (due to incompatible behavior or conduct, e.g., via different mating rituals), mechanical isolation (incompatible genitalia), together broadly termed as "hybridization barriers" or reproductive isolating mechanisms (RIMs).
The study of phylogeny involves reconstructing the history of a species and how it has changed into others; this pattern is usually represented by a branching tree diagram. Just as real-world trees grow from a single base, then steadily diverge into smaller branches, and then twigs and the final tips that spread out widely, so a phylogenetic tree is used by biologists to depict how descendants from one kind of ancestral organism have branched out and evolved into an array of diverse but historically related forms. Each line on the phylogenetic tree divides from an ancestral species into two or more descendant species, mapping out the process of speciation, which usually involves a pattern of cumulative divergence.
Each new evolutionary lineage will be characterized by shared set of mutations, and branches that divided earlier in evolutionary time will be differentiated by the large numbers of mutations that have occurred on independent branches since they split. In contrast, closer to the tips of the tree, the species that diverged much more recently are more closely related; they will share many more common mutations and be differentiated only by a few recent ones.
This treelike branching pattern is therefore a foundational model in evolutionary biology. The concept of homology is an essential tool for reconstructing phylogeny. A good example is the similarity in the forelimbs of mammals, for example, a human hand, a seal's flipper, or a bear's paw, all of which can be defined as homologous traits because they are all derived from a common ancestor somewhat further back in evolutionary history. Homologous traits can be at the morphological and also molecular level, and biologists can reconstruct phylogenies on the basis of the fossil paleontological record, through comparative anatomy, or increasingly, through analysis of DNA sequences.
Is the branching tree model applicable to the evolution of cultural traditions? Since the pioneering arguments of Sir William Jones in 1786, analysis of descent relationships between different cultural entities has also played an important role in historical and comparative linguistics, enabling the classification of languages into a structured taxonomy, made up of phyla, stocks, families, and subfamilies. Similar parallels were also drawn by Charles Darwin in Origin and also in Descent when he speculated that language evolution consisted of tree-like patterns of homologies due to shared communities of descent (Shennan 2011; Whiten et al. 2011). Like phylogenetic analysis in biology, rigorous comparative analysis has enabled linguistics to reconstruct how the ties of descent have created a hierarchy of differential similarity among particular language (Durham 1992:335; see, e.g., Lee and Hasegawa 2011 for examples; and see Heggarty et al. 2010; Mc Mahon and McMahon 2005 for a broader discussion).
But of the many different cultural traditions that are propagated by humans, linguistic evolution may represent a very special kind of cultural evolution (Gray et al. 2010:3928). This may be due to the distinctive ways in which languages are propagated and the unique social roles played by language, all of which may ensure that these traditions are more coherent and potentially evolve in more treelike ways. First, children mainly learn language vertically from their parents, and this enforced vertical parent-to-offspring transmission tends to maintain some degree of intergenerational consistency. Second, after acquiring initial knowledge of language from parents, most children, adolescents, and even adults then go on to acquire other linguistic traits and attributes from a vast range of other sources, often horizontally from peers in the same generation. But despite this, overall language change remains strongly constrained by the need to communicate effectively with others, injecting a strong "conformity bias." Languages can and do change, but they cannot change completely overnight; there is an inevitable conformist bias reinforcing fidelity of transmission because usage must be coordinated and errors corrected if mutual intelligibility is to be maintained. As a result, languages exhibit heritable continuity-they are passed on between generations, and in the process modified; but with enough time, one linguistic community may eventually start to diverge into two related languages, and this cumulative splitting process may eventually produce genealogies that can be depicted as branching trees mapping ties of historical relatedness (Gray et al. 2010; Steele et al. 2011:3783).
But what of other nonlinguistic cultural traditions propagated by humans? Do they share similar kinds of interindividual and intergenerational transmission routes and community-scale stabilizing mechanisms? Some anthropologists argue not. Kroeber (1948), for example, famously accepted a branching tree models for genetic descent and the evolution of species but argued that cultural evolution was characterized by a more reticulate pattern, with cultural lineages branching away but also rejoining again, producing a different kind of cultural tree that had a complex tangle of both splitting but also interconnecting branches (figure 1.2). More recently, other anthropologists have again argued that historical branching patterns in cultural traditions are nearly impossible to reconstruct because of the inherent nature of cultural evolution and the ways in which different cultural traits and attributes tend to be propagated (Dewar 1995; Hornborg, 2005; Moore, 1994a, b, 2001; Terrell, 1988, 2001, 2004; Terrell et al., 1997, 2001; Welsch et al. 1992; Welsch and Terrell, 1994).
According to this perspective, cultural evolution is argued to be a fundamentally different kind of process than macroscale biological evolution. It has a much faster tempo due to very high rates of cultural innovation, and it also has a fundamentally different mode-horizontal transmission-that creates reticulation. These features prevent deeper branching histories from forming and reduce the cultural landscape to little more than a blur of interrelated forms. This general process is often referred to as ethnogenesis, defined broadly as cultural evolutionary patterns that are produced by the borrowing and blending of ideas and practices, and the trade and exchange of objects, among contemporary populations (Borgerhoff Mulder et al., 2006:54).
As a result of these blending forces being so central to cultural inheritance, coherent entities tend not to form, and even when they do, any coherence is rapidly swamped by the combined effects of borrowing or rapid innovation, such that it becomes impossible to reconstruct any deeper historical or genealogical relationships between entities using the kinds of phylogenetic models commonly used by historical linguistics and biologists. Thus, if analogies between cultural and genetic inheritance have any validity at all, it is at the biological microevolutionary scale (see previous), that is, involving the ebb and flow of genetic information within a single interacting population, and not the process of branching macroscale diversification that involves largely separate gene pools.
More recent work has served to question now rather polarized debates about the extent to which cultural traditions evolve either by branching processes (phylogenesis) or according to a more reticulate or hybridized model characterized by rapid innovation and cultural mixing (ethnogenesis). In fact, rather than arguing for the absolute dominance of either branching versus mixing processes within overall cultural evolution, a more subtle and interesting set of questions can explored if the history of one particular cultural tradition is examined in relation to number of intersecting axes (Gray et al. 2010:3924).
Gray et al. (2010) therefore argue that when investigating the transmission of traits and the descent history of a single cultural tradition like language, it is more productive to ask where particular aspects of tradition are positioned along a three-dimensional continuum rather than to argue a priori for the exclusive validity of splitting processes (that can be mapped by trees) or mixing processes (that cannot). This culture-evolutionary continuum consists of three axes: Rv, Rh, and C. Rv indicates the rate of change in traits or attributes transmitted from generation to generation within each of the populations. If this rate is too slow relative to the time period studied, then there will be little character change to allow for the reconstruction of phylogenies. If it is too fast, then all historical signals will be erased, making phylogenetic reconstruction redundant. Rh is rate of horizontal mixing of traits and attributes between the different populations. At low rates of horizontal mixing, the estimated phylogenies will be good estimates of history, but at higher rates, the estimated phylogenies will be increasingly inaccurate and serve as poor summaries of the overall history. C forms the third dimension and measures cultural cohesion-the extent to which the different cultural traits are tightly coupled together as coherent entities. For example, in the case of languages, morphosyntactical traits will tend to evolve more slowly and are often tightly bound together, so that instances of horizontal borrowing tend to be rare. In contrast, a random sampling of the total lexicon would demonstrate that these traits tend to evolve more rapidly, often have lots of horizontal borrowing (e.g., loan words), and each trait (or loan word) may exhibit its own independent cultural history. The content of core vocabularies will tend to occupy a more intermediate position, with a slower rate of evolution and possibly a limited degree of horizontal borrowing.
This is an important conceptual development because it transcends the simple branching versus mixing debates examined previously, and argues that in fact, different elements of a single tradition like language or canoe making might evolve in different ways, some elements retaining a deeper historical signal, others influenced to a greater extent by rapid innovation, or by the horizontal mixing of traits between populations. This can be illustrated by further simple linguistic example: Swedish, English, Frisian, Dutch, German, Czech, Polish, Russian, and Japanese are all different languages, and all have features that define them as internally coherent and individually distinct cultural entities, such that each of these different cultural entities has a separate history. Some of these histories also exhibit higher levels of genealogical relatedness; Swedish, English, Frisian, Dutch, and German fall within the Germanic language family, and Czech and Russian fall into the Slavic language family, though Japanese falls outside these groupings and has a very different history. However, at another level, all these languages also share large numbers of loan words as a result of long histories of cultural contacts, interactions, mixing, and exchange among all the different linguistic communities groups. In this sense, different evolutionary patterns fit different sets of cultural traits and attributes even within a single and relatively easily defined cultural tradition like language. This is because the history of single traits (e.g., loan words) can be studied, as well as the history of the more coherent entity (e.g., the language).
Differential patterns of evolution have also been noted for other cultural traditions, for example, among functional and symbolic traits within ethnographic canoe designs in the Pacific (see Grey et al. 2010:29-31; Rogers and Ehrlich 2008; Rogers et al. 2009); in other words: "different aspects of culture can have quite different evolutionary histories" (Gray et al. 2010:3931).
These emerging insights into the complex evolutionary dynamics of particular cultural traditions therefore take future debates beyond the more divisive arguments about whether human culture evolves exclusively either by phylogenesis or by ethnogenesis. In fact, the key challenge in future research is to identify the extent to which all traits within a single cultural or linguistic tradition fit one overall pattern, or whether different sets of traits follow a range of different evolutionary histories. The same kind of reasoning applies to identifying contrasting evolutionary patterns among a range of different cultural traditions, a theme examined next.
Historical Congruence among Multiple Cultural Traditions
These discussions of cultural coherence in single cultural traditions (e.g., language, Pacific canoe making) are important and useful, but it should also be remembered that the evolved human capacity for high-fidelity social learning enables populations to simultaneously propagate multiple cultural traditions, ranging from language, hunting strategies, craft traditions through to religious rituals. As a result, each of these individual traditions may in turn have its own degree of coherence and unique evolutionary history (table 1.1, part 3a). Understanding this aspect of cultural evolution is particularly important because human groups tend to differ from one another in vast numbers of ways, with each group maintaining many unique sets of cultural traditions (table 1.1, part 3b). In fact, humanity appears to be extremely adept at organizing itself into distinct cultural groups and then signaling membership of, and loyalty to, these defined social units through a wide range of cultural media ranging from language, clothing, and adornment through to cuisine and religious cults (Mace and Pagel 2004). It has therefore become customary in anthropology to describe these units as "cultures," or "ethnolinguistic groups," with the implication that each of these groups has a distinct set of cultural attributes and internal features that serve to distinguish it from others.
How then do these entire human cultures evolve? Following early formulations of dual inheritance theory in the 1980s, there were renewed reconsiderations of the extent to which entire sets of cultural traditions might evolve through treelike branching processes similar to those observed in macroscale biological evolution biology (Durham 1991, 1992; and see Boyd et al. 1997; Foley 1987; Mesoudi et al. 2006:332-36; O'Brien and Lyman 2003b; O'Brien et al. 2001, 2013). For example, Durham (1991, 1992) presented the general argument that (1) socially transmitted information systems constitute entire "cultures," and that these provide human populations, in addition to the genetic information they carry, with a second important source of heritable variation; (2) these human cultural systems are historically interrelated by hierarchical patterns of descent; and (3) splitting was the dominant process in cultural macroevolution, generating cultural phylogenesis-this involves one entire culture branching away into two or more descendant cultures. This cultural splitting process was followed by cumulative cultural "transformation" within each of the descendant cultural lineages. In general then, the principle argument was that new cultures tend to originate from preexisting "parental cultures" via a branching process termed "diversification" or "culture birth," generally a result of "uniparental fissioning" Durham (1992:333).
Although acknowledging that there may be occasional instances of horizontal borrowing and hybridization of limited numbers of cultural traits and attributes between cultures, Durham (1992) argued that strong and pervasive barriers served to lock out the potential for horizontal cultural exchange between cultures. These barriers were termed TRIMS (transmission isolating mechanisms) and were linked to a range of ecological, psychological, linguistic, or cultural factors that would effectively dampen cultural exchanges between populations and served an analogous role to RIMs (reproductive isolating mechanisms-see previous) in biological speciation theory (and see Tehrani and Collard 2013).
More generally, they might also be linked to the effects of different kinds of social institutions such as strictly endogamous kinship systems, highly developed and strictly defended property right, that tie particular groups to exclusive territories but also enforce resocialization of individuals that move into a new community after marriage, compelling them to abandon the cultural attributes they have acquired from their original culture, and adopt entirely all the behaviors of the new group. Durham argued that TRIMs are important because they serve to canalize cultural inheritance within discrete populations; where these forces were strong and diversification was consistently uniparental, that is, one culture split into a series of daughter cultures, then there would be no mixed ancestries in a group of related cultures. Arguments for this kind of cultural phylogenetic process are supported by the observation that languages of genuinely mixed origin are relatively rare and that the same characteristics probably also apply to entire cultures (Durham 1992:334). Cultural phylogeny was therefore argued to underlay most patterns of cultural diversity-a distinct pattern of historical relationships that was generated by process of culture birth that led to the formation of a "systematic hierarchy of successive splits." Cultural phylogenies could then describe ancestor-descent relationships between these cultures, the family tree specifying a unique and unambiguous phylogeny like cladistic classification in biology.
More recently, a similar argument has been made by Foley and Larr (2011), who argue that cultural phylogeny played a dominant role in much of human prehistory. This is expressed by an accelerating trend toward treelike growth in regional cultural diversity, as lineage-based communities fissioned over time, resulting in cultural divergence and the formation of new groups and cultural boundaries. They note that local ecosystems served to structure general patterns of cultural diversification. For example, the general global diversity in human cultures is striking, but it is most pronounced in areas of high ecological productivity, which tended to encourage the dense bunching of populations around predictable and locally abundant resources that could be easily be owned and defended (e.g., in regions like California and the Pacific Northwest Coast; see chapters 4 and 5). However, the expansion and cumulative fissioning of groups into new generations of descendant cultures was also intermixed with frequent population extinctions and assimilations, followed by renewed expansions. Consequently, most regional patterns of cultural diversity consist of a series of historically related cultures interspaced with a few cultural isolates. Overall, these distinctive patterns are generated by the specific ways in which human populations have expanded and dispersed since the origin of the species, with the processes of phylogenesis ultimately driving cultural diversification (Foley and Larr 2011:1086-87).
Given these arguments, the main research mission of archaeologists and anthropologists was to "elucidate the patterns and processes of descent with modification in cultures" (Durham 1992:332), with the "family tree hypothesis" serving as the default hypothesis for all incidences of human macroscale cultural evolution (1992:333).
The key differences between competing phylogenetic and ethnogenetic models of human cultural macroevolution should now be clear. However, what tends to get lost in these debates is that all human populations simultaneously maintain a large number of different cultural traditions, ranging from language through to religion and cuisine, and that each individual lineage of tradition may be characterized by different degrees of coherence and entirely different historical patterns. In this sense, it is an oversimplification to seek to examine how entire cultures evolve; a more productive approach is to start by looking first at individual cultural traditions, trace how these evolve across populations, and only then examine the extent to which there is deeper historical congruence among a range of different cultural traditions. For example, do culinary traditions or religious practices form coherent lineages, and is the branching history of each of these lineages closely related to language history? If yes, this would suggest that multiple cultural lineages are evolving in tandem with one another. If not, this would suggest that each lineage has its own unique history.
Boyd et al. (1997) undertook an insightful exploration of the kinds of broader-scale cultural patterning that might emerge through the evolution of multiple cultural traditions, which in turn affect the manner in which entire human cultures might evolve. These four models have played a central role in subsequent debates about macroscale cultural evolution.
Model 1: Cultures as Species
This where entire human cultures are either entirely isolated from one another (due to TRIMS-see previous discussion) or have a high degree of internal integration, or a combination of both (Boyd et al. 1997:365; see figure 1.3). As a result, they evolve in the same overall manner as vertebrae species do-each entire culture forms a coherent and tightly integrated system in which cultural information is only propagated internally within the culture. Horizontal exchange of cultural information between the different groups is minimal. Each culture is, however, made up of distinct cultural traditions. In this kind of setting, the full suite of the cultural traditions carried by each local population is therefore isolated from all outside influences, and the only pathways of information flow are within each local culture. Inheritance of all cultural information within each population therefore involves members of the older generation passing on traits to younger generations without any external cultural influences. If these essentially insulated populations then diverge and split, a new generation of descendant "daughter" cultures will then be formed, each of which may then split further into a new generation of related cultures. Hierarchical patterns of cultural descent with modification among these historically related daughter cultures can then be mapped as branching tree diagrams. According to this model of cultural diversification, phylogenetic methods of historical reconstruction should perform well for all the individual cultural traditions maintained by each population (e.g., for language, basketry, pottery, clothing), and the branching history of each of these cultural lineages will also be highly congruent with all the other cultural traditions.
Model 2: Cultures with Hierarchically Integrated Systems
Here, cultures consist of a core set of closely integrated traditions that are inherited from older to younger generations within each population, and perhaps canalized by the effects of TRIMS, which reduce overall cultural borrowings between populations (Boyd et al. 1997:365; figure 1.3). Beyond these core traditions, a greater degree of horizontal cultural mixing does take place between populations, but this only affects some of the more peripheral traditions, and not the cultural traditions that form the core elements of the system. In this scenario, phylogenetic methods will work well for all the integrated core traditions, but not for the traits that make up the more peripheral traditions, which tend to evolve via mixing and rapid innovation.
Model 3:'Cultures as Assemblages of Many Coherent Units"
Here, cultures still consist of many single coherent traditions, but each tradition is essentially independent and has its own independent branching history (Boyd et al. 1997:365-66). This may result from the fact that the specific packages of traits propagated within a highly coherent tradition-for example, a new religion, a novel way of preparing and serving food, knowledge of how to make iron-can sometimes "jump" horizontally between different populations, undergoing a process of descent with modification along the way. According to this model, cultures are made up of several of these distinct traditions, but each tradition will have its own independent source and descent history, each of which can be reconstructed with phylogenetic methods, but only on an individualistic case-by-case basis. For example, in a given world region, iron working and pottery traditions each have separate origins and independent branching histories, but neither of these histories maps onto the other or onto local language history. Each single tradition is therefore coherent, but there is no historical congruence between these separate traditions.
Model 4: Cultures as Collections of Ephemeral Entities
Finally, if rapid innovation and relentless horizontal mixing of information between populations dominates the propagation of cultural traditions, then there are no coherent cultural entities (Boyd et al. 1997: 366; figure 1.3). Instead, cultures will consist of little more than unstructured sets of highly diffusible traits subject to rapid exchange and recombination. According to this model, little historical signal is retained, and phylogenetic methods simply won't work, either at the scale of individual traditions, or at scale of the entire culture: all that can be detected is a blur of hybrid forms.
Boyd et al.'s (1997) four models succeed in sketching out positions along a potential continuum of evolutionary scenarios. Although models 1 and 4-cultures as species and culturesas collections of ephemeral entities plot opposing ends of this continuum, they appear unlikely to exist in reality. First, in one way or another, human groups have always interacted with one another despite speaking different languages or possessing different senses of identity. As a result, cultures have rarely, if ever, been hermetically sealed, and the widely documented range of contacts and exchanges between human groups suggest that it is highly unlikely that cultures could have evolved in ways analogous to biological species. Second, there is also little evidence to support the assertion that cultures are merely made up of a blur of hybridized traits and that their traditions have little or no historical continuity-archaeologists, for example, regularly identify artifact lineages that exhibit both internal coherence and long-term multigenerational continuity (Tehrani and Riede 2006).
A key research question therefore arises (Boyd et al. 1997: 386): In what kinds of historical settings are models 2 and 3 most likely to emerge? What is the highest scale of cultural coherence-is it possible to identify the existence of coherent traditions that are tightly bundled together and form enduring cultural cores (model 2), or is the highest scale of cultural coherence a series of distinctive but essentially independent traditions, each with its own unique history of decent (model 3)? Understanding which of these particular models fits a specific historical scenario requires first understanding how each of the individual cultural traditions has evolved. Only then does it become possible to examine the extent of historical relationships (congruence) between the different lineages of tradition that make up overall cultural history.
Studying Descent with Modification in Technology and Material Culture
The preceding section has examined three overarching themes: the ways in which general cultural information is propagated within human traditions in the form of social traditions, the extent to which these traditions are coherent, and the degree to which there is historical congruence between different lineages of cultural tradition. Clearly, the conceptual parallels between genetic inheritance and the inheritance of cultural information have been a useful framework for understanding these themes, as has exploration of the ways in which micro- and macroscale processes of biological and cultural evolution can productively be compared and also contrasted. Beyond these more heuristic frameworks, the analogy between genetic and cultural inheritance has also had important methodological implications, paving the way for analytical tools developed by biologists to study genetic inheritance to be adapted to for empirical analysis of cultural data (Collard and Shennan 2008; Collard et al. 2008; Lipo et al. 2006; Lycett 2011; Mace et al. 2005; Mesoudi et al. 2006; Steele et al. 2010).
As a result of an extensive body of research directly stimulated by Cavalli Sforza and Feldman (1981) and Boyd and Richerson (1985) pioneering formulations of dual inheritance theory has now accumulated, generating foundations for a rigorous science of cultural diversity and change that spans biology, evolutionary psychology, archaeology, and anthropology (Rogers and Ehrlich 2008:3416). To one degree or another, all of this is premised on the recognition and exploitation of these central analogies and conceptual parallels between cultural and biological inheritance systems examined previously (see Mesoudi et al. 2006 and Mesoudi 2011; Whiten et al. 2011). Across this new interdisciplinary research field, there has been particularly rapid progress in the development of theoretical models and simulation studies. However, empirical research has developed at a much slower pace; one major challenge continues to be the lack of large-scale high-resolution comparative cultural data sets of the kind that are typical in evolutionary biology and population genetics (Shennan 2008:3176). Building suitable new data sets therefore emerges as a major challenge (chapter 2).
How Do Material Culture Traditions Evolve?
So far, much of this discussion has revolved around the theme of generalized cultural evolution, often citing examples of the evolution of linguistic traditions. But language evolution may form a very unique kind of cultural evolution. As examined previously, it is heavily structured by particular modes of transmission, and though innovations and borrowing may occur, it still needs to solve the basic collaborative problem of effective group-scale communication, which injects a strong conformity bias. This, in turn, appears to encourage strong heritable continuity, the formation of coherent lineages of tradition, which seem to evolve via branching processes of diversification, making deeper linguistic histories amenable to phylogenetic reconstruction.
What about the evolutionary dynamics of other kinds of cultural tradition? This book aims to improve understandings of one relatively underdeveloped area of research: the evolution of technology and material culture. Understanding spatio-temporal patterning in tools, artifacts, and other surviving objects is absolutely central to archaeology and also important for anthropology, yet the exact range of processes that can characterize material culture evolution remain somewhat unclear. For example, do material culture traditions exhibit strong heritable continuity, and if yes, what patterns of diversity are produced? On the one hand, there are many reasons to suspect that macroscale material culture evolution may also exhibit tree-like properties. Active participation in many craft traditions often requires years of long-term investment by both experienced practitioners and novices, typically ensuring formation of distinctive artifact linages; community-scale propagation of these crafts may also be subject to strong conformity biases within a population, while TRIMs can insulate the craft from external influences, both of which would contribute to strong heritable continuity of distinct traditions within the population, and dampen the influence of external and borrowings. In fact, humanity does appear to be highly adept at forming high levels of cultural diversity (Pagel and Mace 2004), and material culture may-in some cases-play an important role in actively signaling different kinds of group-based identities (but see Shennan 1989; Trigger 2006:3009-10, 452-53).
On the other hand, other roles played by technology and material culture and technology may result in very different evolutionary patterns. Humans make tools and equipment to perform specific tasks, and some designs or entire technologies simply work better in certain functional and environments contexts than others. Thus, the set of traits combined in certain technologies can make them more efficient than other technologies, for example, if they can capture more energy or resources in relation to the time invested (a key insight to emerge from ecological and adaptive approaches in archaeology and anthropology). The deep attractions and functional advantages offered by some technological traits-or trait combinations-may simply override TRIMS, enabling knowledge of these to spread rapidly, effectively jumping horizontally between different populations (Dunnell 1978).
This book argues that the descent with modification approach provides a productive approach to investigating exactly these kinds of historical dynamics in technology and material culture traditions. At the microscale, a focus on cultural propagation highlights the roles of social learning, cultural context, and individual choice-this broadly aligns with current research being conducted within the anthropology of technology, with its focus on apprenticeship and intergenerational learning and the embedded nature of different material culture traditions (see Stark et al. 2008b). The descent with modification approach, however, also raises important questions about how lineages of material culture tradition evolve at the macroscale, that is, between populations and societies. In part, this requires understanding better how microscale processes of propagation actively feed into these macroscale patterns of material culture diversity. However, only limited attention is being directed at this issue because of the more popular interpretive interests in routine practice and day-to-day reproduction of cultural traditions within contingent local settings that typify much current material culture research in both archaeology and anthropology.
In fact, it is worth reflecting here on the fact that descent with modification approaches to material culture have a somewhat deeper history, especially in archaeology. Early ideas about cultural transmission-and reconstructing artifact lineages-can be traced back to North American cultural history, but this work was primarily directed at the challenge of developing relative chronologies prior to the development of radiocarbon dating (Lyman et al. 1997:231). A more explicitly evolutionary theory of artifact studies later arrived into archaeology from another direction, triggered by developments in palaeontology, the science of long-term evolutionary processes. Here, the development of phylogenetic methods gave palaeobiology new meaning: this was no longer about just cataloguing the diversity of life forms; it now represented a means of reconstructing macroscale evolutionary processes. It was not long before scholars interested in material culture began to take interest (Prentiss 2011), generating the unique offshoot branch of evolutionary archaeology (Dunnell 1978, 1980) within the new archaeology and its processual offspring (Trigger 2006:428-29). The core argument was that artefact variation through time and space should be studied and explained in the same way that palaeobiologists explain patterns in the temporal and geographic variations in bony morphology that are documented in the fossil record, and that result from natural selection and drift (Collard and Shennan 2008 :18-19; Trigger 2006:428-29). At a deeper level then, Dunnell's (1978, 1980) general approach was an important conceptual breakthrough because it attempted to reconstruct the process of cultural evolution by studying the characteristics and attributes of the surviving artifacts that made up the archaeological record, that is, from the perspective of the material culture itself (Shennan 2008a:3176; 2008B:78; 2011). This stands in contrast to the more agent-centered approach developed by early formulations of dual inheritance theory (e.g., Cavalli Sforza and Feldman 1981; Boyd and Richerson 1985).
The current intellectual descendants of this unique Dunnellian form of evolutionary archaeology can now be grouped into two main strands of research, both united by a shared commitment to viewing the archaeological record in terms of the descent with modification of cultural traditions propagated by social learning (Lycett 2011). Some of the first attempts at using contemporary biological methods in archaeology stemmed from initial suggestions that the use of population genetic models of transmission and selection drawn from biology might be used to explain variability and change in artifacts making up the archaeological record (Boyd and Richerson, 1985; Cavalli Sforza and Feldman 1981). Here, it was suggested that these could provide useful means of testing hypotheses concerning patterns of cultural variation, because there has been increased recognition that at a statistical level, many of the factors known to structure variation in genetic data (e.g., population size, drift, dispersal, etc.) should also be considered when testing hypotheses about material culture variation (e.g., Bentley et al. 2004; Bettinger and Eerkens 1999; Eerkens 2000; Eerkens and Lipo 2005; Kohler et al. 2004; Lipo et al. 1997; Lycett 2008, 2011; Mesoudi and Lycett 2009; Neiman, 1995; O'Brien and Lyman 2000, 2003; O'Brien et al. 2008; Shennan 2000, 2001, 2006; Shennan and Wilkinson 2001).
A second stream of research has explored additional implications of a descent with- modification approach to artifact lineages but instead builds on earlier suggestions that modern phylogenetic concepts might have utility in the study of human cultural evolution (Foley 1987; Lycett 2011: 153). There is now a growing archaeological literature on the application of cladistic and related methods of phylogenetic reconstruction developed in biology to study the evolution of artifact lineages. Here, the general principles of applying these methods to archaeological data sets are now clear (e.g., O'Brien and Lyman 2000, 2003; O'Brien et al. 2001), and involves detailed characterization of artefact lineages linked by cultural transmission through time, with the appearance of new variants representing new branches on an evolutionary tree, whose history can be reconstructed through cladistic analysis (e.g., Buchanon and Collard 2007, 2008; Cochrane 2009; Cochrane and Lipo 2010; Lipo et al. 2006; Lycett 2007, 2009a, b, 2011; O'Brien et al. 2001; O'Brien et al. 2013; Prentiss et al. 2014). These methods have been argued to offer archaeologists a set of highly rigorous procedures for identifying critical variables and for measuring artifact variability and change in their evolutionary history (Prentiss 2011; and see Mesoudi et al. 2006:334; Mesoudi 2011).
In sum, archaeologists and anthropologists continue to study material culture in a range of very different ways and from contrasting theoretical perspectives. Application of an explicitly descent with modification approach to material culture has some deeper intellectual roots that stretch back into culture historical archaeology as practiced in North America and Europe, but its current and more concerted theoretical and methodological fluorescence is still a relatively recent development within archaeology (Prentiss 2011). It is clear, however, that this specialist and self-sustaining research literature is now growing apace.
More generally, however, phylogenetic studies of material culture evolution are still not that common (Gray et al. 2010:3929) One of the main challenges to all archaeological studies is having to infer social learning, cultural transmission, and other evolutionary processes such as drift directly from the documentation and analysis of continuity and change in artifact lineages. Although the studies cited previously illustrate that that this kind of approach is now logistically and empirically feasible, other, more detailed contextual insights into other aspects of human social behavior linked with the propagation of these cultural lineages tend to be lacking, or at very best, must be inferred from other archaeological evidence, which brings its own additional challenges. In contrast, studying material culture evolution in more recent, ethnographically documented or contemporary fieldwork settings can draw on independent information on many other related aspects of human behavior. Despite these obvious attractions, exploration of this more resolutely "ethnoarchaeological" branch of culture-evolutionary research has been rather limited to date.
Renewed Scope for Ethnoarchaeological Studies of Material Culture Evolution?
Ethnographic research generates rich opportunities for undertaking detailed recording and and analysis of variability in craft traditions in relation to social learning, pedagogic patterns and cultural practices, in combination with detailed contextual information pertaining to the the life histories, and social identities of the individuals or groups that produced the artifacts (Tehrani 2006). Also usually available is information on the social institutions and settlement systems within which of the craft traditions are being propagated. Examination of the former is important for understanding the intentionality and general agency of people involved in making the objects and sharing the traditions with others; insights into the latter can clarify how local factors such as kinship, interaction patterns, politics, and territoriality can lead to the formation of cultural boundaries. For example, increasing hostility between groups may reduce contacts and exchange, resulting in the formation of cultural boundaries and growing stylistic differences between local craft traditions.
In contrast to the substantial archaeological literature on cultural transmission cited previously, a much smaller body of empirical research has emerged to address this potentially rich vein of inquiry. Generally, the main goal is to work with cultural trait lists to document geographic variability in material culture, and then use a range of cladistic and related methods of phylogenetic reconstruction to reconstruct the evolutionary histories of material culture traditions within relatively well-known ethnographic and historical settings (see chapter 2). Recent studies have examined Turkmen textiles (Tehrani and Collard 2002; Collard and Tehrani 2005, 2009a, b , 2013; Tehrani et al. 2010), indigenous clothing traditions in Northwest Siberia (Jordan 2009), various craft and architectural traditions in California (Jordan 2007; Jordan and Shennan 2003, 2005, 2009), and on the Pacific Northwest Coast (Jordan and Mace 2006, 2008; Jordan and O'Neill 2010); Polynesian bark-cloth traditions (Larsson 2011), European table cutlery (Riede 2009), Polynesian canoe designs (Rogers and Ehrlich 2008; Rogers et al. 2009; and see Gray et al. 2010:29-31), and musical instruments (Tëmkin and Eldridge 2007).
While all these studies have generated important contextual and historical insights into the contingent ways in which material culture traditions can potentially evolve, their focus has remained on understanding the evolutionary dynamics of a single craft tradition within one specific cultural or geographic setting. Some have attempted to compare evolution of craft traditions to language, but only a handful have attempted to examine more complex scenarios in which multiple material culture traditions potentially coevolve and the extent to which they exhibit deeper historical congruence (see Jordan and Mace 2006, 2008; Jordan and Shennan 2009; Riede 2009). This leaves an important gap in the current research literature, because the capacity to maintain multiple traditions is one of the defining features of human culture (table 1.1). Investigating how a range of different cultural traditions evolve is also central to the evaluation of all four of Boyd et al.'s (1997) models (figure 1.3). Likewise, the general focus on single-craft traditions within each ethnographic case studies also leaves it unclear as to the kinds of cross-cultural commonalities and divergences that might characterize overall material culture evolution at a more general comparative level (but see Collard et al. 2006a, b). Clearly, substantial progress has clearly been made on many fronts, but much more empirical research remains to be done: the current research literature remains rather fragmented, with somewhat disconnected insights into the evolution of material culture traditions dispersed across multiple case studies; no studies to date have attempted to undertake a strategic and more explicitly comparative and contextual analysis of how multiple material culture traditions might evolve across a more targeted range of ethnographic settings.
To conclude, it is clear that there is now scope for making substantive progress on some of the most important themes, topics, and debates surrounding the broader evolutionary analysis of material culture traditions. In particular, there are exciting opportunities for a more explicitly coordinated cross-cultural study of how multiple material culture traditions evolve across a range of analogous-or at least broadly comparable-culture-historical settings. Conducting this research project within an ethnoarchaeological framework, linking social action to histories of artifact production, would also mean that the main insights would have a much deeper relevance to general debates spanning anthropology and archaeology about the main factors that generate variability and change in material culture traditions. In fact, undertaking exactly this kind of integrated and interdisciplinary research project now emerges as the central goal of this book.
Designing an Empirical Study: How Do Multiple Material Culture Traditions Evolve?
With so little empirical investigation into how multiple material culture traditions evolve, it is important to embed the case studies within a well-structured framework that enables sets of mutually reinforcing insights and understandings to feed into one another.
At a starting level, the book develops an explicit focus on evolution of hunter-gatherer technological traditions, which provides a basic sense of ethnographic unity to what otherwise might be a disparate collection of case studies. First, the technology of hunter-gatherers tends to consist of relatively simpler data sets, and their relationships with immediate environments tend to be more localized and direct, reducing the number and complexity of potentially confounding variables. While a descent-with-modification approach could, of course, be applied to cultural traditions within a wide range of different settings, such as modern nation-states and complex urban societies, progress on the basic arguments and issues is probably best achieved via a sustained focus on understanding technological evolution in small-scale societies.
Second, the central focus on foragers also provides a useful framework for comparative integration of the results and insights. This is because hunter-gather populations have been the focus of systematic cross-cultural documentation and analysis (Jordan and Cummings 2014), generating a common vocabulary and shared terminology for describing different aspects of behavioral variability and major social institutions such as mobility regimes, kinship, and postmarital residence rules, settlement systems, and land-holding regimes (Kelly1995, 2013). Moreover, all three case studies are also set in higher-latitude environments, where hunter-gatherer populations face broadly similar ecological challenges consisting primarily of the uneven distributions of resources over the landscape at different times of the year (figure 1.4). Populations in all three areas have responded to these opportunities and challenges via similar strategies of seasonal storage, complex mobility regimes, and distinctive social institutions that allocate people in highly structured ways to specific social groups, resources, and territories. These general contextual behaviors are important in studies of cultural transmission because these kinds of common institutions may also serve as TRIMS that structure how particular traits evolve across populations and cultures.
Third, to strengthen the framework for cross-cultural analysis of results, broadly similar sets of hunter-gatherer container and transport technologies, clothing, and built structures are examined across all the case studies, adding further conceptual linkages. Many of these craft traditions have an explicit gender association, enabling some interesting contrasts to be traced out between the histories of "male" and "female" technologies within and between the different case studies, as well as contrasting evolutionary patterns in more collective social traditions like house building versus more individualized crafts such as basketry.
Fourth, the entire phenomenon of cumulative culture emerged in first in prehistoric forager societies, so studying propagation of cultural traditions in these kinds of modern yet broadly analogous societies can also shed light on some of the basic mechanisms and factors that structure social learning and general cultural propagation within mobile small-scale communities who derive much of their primary subsistence from hunting, fishing, and gathering (Whiten 2011). Moreover, many early technological traditions also emerged first in forager settings, ranging from lithic tools through to the world's first pottery containers, and the surviving material residues of these crafts now form an extensive component of the earlier archaeological record for many world regions. Of course, by drawing such parallels between ethnography and archaeology-basically adopting an ethnoarchaeological approach to analysis of material culture evolution-is not to argue that modern or ethnographically documented hunter-gatherers are somehow devoid of historical dynamism and are trapped in an earlier social evolutionary stage or that they have been unaffected by the effects of culture contact and colonialism. For example, the Siberian case study (chapter 3) examines cotemporary communities of "commercialised hunter-fisher-gatherers" with long histories of colonial contact (Jordan 2003, 2011a, b), while the Western North American case studies (chapters 4 and 5) examine the relatively high-density populations who occupied rich ecological niches, and whose societies and technological traditions were the outcome of many millennia of localized cultural developments. Ethnoarchaeologists have long been aware of the limitations to their approach but believe that important insights into the complex links between material culture and social action can still be produced (Lane 2014).
The interlocking themes of propagation, coherence, and historical congruence unite all the case studies, but these topics are are deliberately examined across a range of intersecting social scales. The case studies employ linguistic diversity as a general measure of social scale, and in many cases, also use language or dialect speech communities to define specific ethnolinguistic populations. These scales range from single-dialect communities, a chain of related dialect communities, through to regions with high levels of linguistic diversity, where populations speak languages with very different histories. Organized in such a way, the insights generated at one scale of analysis can also feed into interpretations and understandings at other scales, generating better comparative understandings of how different kinds of material culture tradition tend to evolve.
Organization of the Book
This opening chapter has outlined the main research themes and identified the key questions to addressed (see table 1.4), starting with a clear definition of the problem to be solved (Boyd and Richerson 2005: 65).
Chapter 2 aims to provide a bridge between current debates about best to understand and explain the evolution of craft traditions and the conduct of empirical research. It examines how comprehensive material culture data sets can be collected in ethnographic settings, and explains how the main research questions can be addressed by applying a single methodological framework. This approach is based on the use of a range of relatively straightforward techniques drawn from evolutionary biology. Choice to use these specific methods was guided by the book's more general anthropological and archaeological readership and represents a relatively simple and easily accessible approach to the kinds of material culture and linguistic data sets that other regional specialists should also be able to reproduce and start to analyze. Emphasis is also placed on the importance of integrating wider ethnographic insights so that observed patterns of cultural evolution can be contextualized within specific historical settings.
Chapter 3 is set in Northwest Siberia and focuses primarily on understanding the microscale propagation of a range of material culture traditions, examining the different scales of coherence that tend to emerge. Here, the emphasis is very much on the agency of individuals as they practice different craft and architectural traditions, but these propagation processes are also examined from the perspective of the artifacts that are produced. Data are derived from the author's own ethnoarchaeological surveys (1997-2005) illustrating how active field research is an ideal method for generating new data and studying cultural propagation within complex living settings. Here, the methods are used as part of a "pattern recognition" approach to understanding the extent to which more coherent material culture traditions are likely to emerge.
Chapter 4 picks up on these Siberian insights but expands the social scale to examine the inheritance of three material culture traditions across communities of Pacific Northwest Coast hunter-fisher-gatherers who speak a chain of related Coast Salishan dialects. The main concern here is exploring coherence at the macro-scale of cultural evolution; in particular, it examines whether local community traditions have evolved via branching or blending processes, and the degree to which language and craft traditions exhibit historical congruence. As such, the chapter looks first at the general ethnographic context and then examines the process of evolution from the perspective of the artifacts being reproduced, attempting at the end to link these patterns to the agency of individuals responsible for the propagation of the traditions. This approach highlights the important role played by local social institutions, which appear, in some cases, to play the role of TRIMS, playing a major structuring role in the propagation of these different cultural traditions and producing distinctive macroscale patterns of material culture diversity.
Chapter 5 casts the widest net and focuses on macroscale cultural inheritance among populations inhabiting Northern California, an area of very high linguistic diversity. The main goal is to examine how variability in social institutions affects the coherence and historical congruence of different craft traditions. This approach also examines the general ethnographic context and focuses on reconstructing evolutionary patterns from the perspective of the artifact lineages, linking these patterns back to the role of the agents who propagate these traditions. Again, certain social institutions can often serve as TRIMS, but only for certain traditions, and only in certain geographic areas.
Chapter 6 concludes the book with a cross-cultural analysis of results derived from each case study, and is structured by detailed discussions of the three main themes of propagation, coherence, and historical congruence. Finally, the chapter aims to explore the general relevance of a descent-with-modification approach to the broader analysis of cultural traditions and highlights themes and questions requiring further research.
This book argues that the vast range of material culture and technology made and used by humans forms a unique form of social tradition. In fact, all cultural habits can be understood as collections of information that are passed on to others by social learning. It is argued that this process of cultural transmission has some general analogies with the ways in which genetic information is inherited and cumulatively adjusted within and between populations: in short, cultural traditions can also be argued to evolve via a broadly similar process of descent with modification.
Drawing on a wide range of methodological and disciplinary sources, this book aims to apply such a descent-with-modification perspective to the study of variability and change in material culture traditions. Recent work using a similar framework has generally been limited to broader theoretical syntheses and position statements, single empirical case studies, edited collections, or general summary overviews of the evolved human capacities for cumulative culture and social learning. In particular, ethnographic research into the evolution of material culture traditions has also been rather limited and reflects a general challenge generated by a shortage of suitable high-resolution data sets suitable for this kind of approach. In contrast, Technology as Human Social Tradition undertakes the first book-length empirical analysis of cultural transmission and the evolution of complex technological traditions, presenting some new data sets and focusing on detailed case studies, so that the propagation, coherence, and congruence of specific material culture traditions can be linked to other behavioral factors, such as the role of social institutions, all within an explicitly comparative framework.
This latter goal is supported by a sustained focus on northern hunter-gatherers, their material culture and social institutions, and by the application of the same methods and approaches across these different historical and cultural settings. This central focus on material culture also means that the results and insights from these essentially ethnographic case studies have central relevance to some central archaeological debates. As such, the book can also be defined as a comparative, cross-cultural exercise in hunter-gatherer ethnoarchaeology, and, with a central focus on the role of social traditions and material culture-and not adaptation and ecology-also aims to take this specialist field into some productive new research directions (David and Kramer 2001; Lane 2014).
Therefore, the book is primarily aimed also at a more general anthropological and archaeological readership as much as at existing culture-evolutionary researchers, and the methods deployed also represent a range of widely available and also readily accessible approaches to researching material culture traditions for those without an advanced training in computational biology. In fact, it is hoped that by reading this book, which consists of a series of worked empirical examples of how a descent-with-modification approach can be applied to the study of different material culture data sets at an interlocking range of social and spatial scales, other regional specialists will start to explore some these ideas and approaches, and by drawing on their in-depth regional knowledge and period-specific expertise, will generate and analyze a new generation of archaeological and ethnographic data sets that are, frustratingly to date, such a rarity (Whiten 2011: 945).
Finally, and more generally, the three recurring themes of themes of propagation, coherence, and historical congruence in material culture traditions that stretch through this book clearly emerge from a dual inheritance perspective on culture that has its roots in (modern) evolutionary biology. Other intellectual lineages also play a role: highlighting human material culture and technology as an embodied and embedded social tradition means that human decision-making processes, cumulative history, and local contingency all become central to the research, intersecting in a novel and productive way with long-running archaeological and anthropological debates about enculturation and enskillment, the operation of human agency, the performance of culture, and, at larger and more enduring scales, with long-standing questions about how best to understand the origins and social significance of specific distributions of material styles over time and space, still generally described as "archaeological cultures."
Therefore, rather than an abstract position statement or theoretical overview, the book aims to empirically demonstrate several points of intellectual convergence between some of the new and emerging strands of culture-evolutionary research and a range of more established interpretive, historical, and particularistic approaches, the most important being that a shared focus on the study of inheritance-both cultural and biological-will serve to unite future studies of human diversity, both in the more distant past and also in the present.