Animal communication and pre-linguistic childhood communication (roughly in the first year of life) share two fundamental characteristics above and beyond the absence of coded vocabulary and grammar: (i) Non-displaced reference; that is communication only about here-and-now, you-and-I, this-and-that-visibles. The past and non-immediate future may or may not be contemplated, but they are never communicated about. And (ii) the total absence of informative–declarative and interrogative–speech-acts; so that communication is confined to manipulative gestures, be they requests, commands, warnings or solicitations. What Bickerton asked at the end of his note, perhaps somewhat facetiously and implicitly, is how would Theory of Mind explain the evolution of displaced reference; that is, of communication about remembered past events or planned future actions. The immediate answer is that there is no direct causal connection. But this is only part of a complex story.

In his recent book “Adam’s Tongue” (2009), Bickerton outlines one possible selectional pressure that could have led to the rise of displaced-reference communication in the hominid line–planning for big-game scavenging. The crucial ingredients are indeed there–the existence of adaptive-vital information that is not about here-and-now but about there-and-then, and further is not shared equally among group members. In all fairness, however, other proposed selectional pressures would qualify just as well by the same cluster of criteria: Planning for a group big-game hunt, planning for inter-group warfare, reporting on big-game presence elsewhere, reporting on enemy presence elsewhere, scouting reports on desirable faraway new territory, and more. (For a thoughtful review of such a list, see Számadó and Szathmáry 2006).

The second issue, raised obliquely in Bickerton’s contribution to this forum but more extensively in subsequent communications, is the relevance of mind-reading to human communication in general, and in particular to pre-grammatical pidgin, a communicative mode that Bickerton (1981) had re-christened as “proto-language”. Since I have written extensively on this issue (e.g. Givón 2005), the following is but a tip-of-the-iceberg reminder. The best-known early arguments for the relevance of Theory of Mind to human communication were made by two philosophers, J. Austin (1962) and H.P. Grice (1968/1975). They involve the difference between the use of declarative, interrogative or manipulative speech-acts. Both Austin and Grice point out that the implicit assumptions a speaker entertain about the hearer’s current belief and intention states are radically different in the use of these three speech-acts. What is more, such knowledge is embedded in a network of socially-shared conventions of usage and propriety, whose violation by the speaker are immediately noted by the hearer. In pre-grammatical pidgin, a single noun, say “apple”, may stand for either “This is an apple” (declar.), “Where is the apple?” (interrog.), or “Give me the apple?” (imper.). But while the speech-act distinctions may not be coded by grammar, both speaker and hearer observe their usage conventions (?felicity conditions’) just as stringently. The same Theory-of-Mind dependence of communicative acts, be they grammatical or pre-grammatical, has been shown in multiple areas of human discourse (Givón 2005). As two great primatologists, D. Cheney and R. Seyfarth (2007) have sagely observed, “mind reading pervades language”. So while Sarah Hrdy may not have set out to talk about language evolution, by elucidating a plausible adaptive pressure for the rise of Theory of Mind, she inadvertently also elucidated one of the core adaptive pre-requisites for the genesis of symbolically-coded human communication.

References

Austin, J. (1962) How to Do Things with Words, Cambridge, MA: Harvaerd University Press
Bickerton, D. (1981) Roots of Language, Ann Arbor: Karoma
Bickerton, D. (2009) Adams’ Tongue, NY: Farrar, Strauss & Giroud
Cheney, D. and R. Seyfarth (2007) Baboon Metaphysics, Chicago: University of Chicago Press
Givón, T. (2005) Context asa Other Minds: The Pragmatics of Sociality, Cognition and Communication, Amsterdam: J. Benjamins
Grice, H.P. (1968/1975) “Logic and conversation”, in P. Cole and J. Morgan (eds 1975) Speech Acts, Syntax and Semantics, 3, NY: Academic Press
Számadó, S. and E. Szathmáry (2006) “Competing selective scenarios for the emergence of natural language”, Trends in Ecology and Evolution, 2

We thank our featured essayist Sarah Blaffer Hrdy and all those who have sent in their responses, offering such diverse perspectives.

I love Jose Angel’s image of upgraded honeybees, and agree with him that multiple feedback processes must have been involved in the evolution of Homo sapiens and agree also with Derek Bickerton that language and the processes shaping it were critically important in terms of what makes us the intellectually curious and argumentative species we are today. About all I have to say about language though can be found on pp. 37-38 of Mothers and Others in a section labeled “Logically, Language Comes Later”. It reads:

Unquestionably, the uniquely human capacity for language enhances our ability to connect with others and exponentially increases the complexity of the information we can convey. But language is not just about conveying information, as in warning others to “Look out!” An animal alarm call does that. Even vervets (which are Old World monkeys after all, not even apes) have specific calls that alert conspecifics to danger and also inform them whether the threat is from the air and likely to be a predatory bird, as opposed to something scary on the ground, like a snake. Honeybees convey surprisingly precise information about the location of food (how far away and in what direction) by the type and duration of their ritualized “dance” movements. Animals have all kinds of ways of communicating information about their environment or state of arousal… (note 14).
The open-ended qualities of language go beyond signaling. The impetus for language has to do with wanting to “tell” someone else what is on our minds and learn what is on theirs. The desire to psychologically connect with others had to evolve before language. Only subsequently do the two sets of attributes co-evolve. As (the psychiatrist Peter) Hobson puts it, “Before language, there was something else—more basic . . . and with unequalled power in its formative potential.” (note 15). If we are looking for sources of human empathy, these emotion-laden quirks of mind had to evolve before language. Even before humans began actually speaking to one another in a behaviorally modern way, their immediate hominin ancestors already differed from other apes in their eagerness to share one another’s mental states and inner feelings. In this sense, these creatures were already “emotionally” modern long before they became “anatomically” or “behaviorally” modern” and were routinely using speech to converse with one another (more on this chronology in chapter 9). The ancestors of people who acquired language were already far more interested in others’ intentions and needs than say chimpanzees are. What we need to explain is why.

In Mothers and Others that is as far as I sought to go. The rest of the fascinating and very complex story about how human language evolves I leave to Tom Givon, Derek Bickerton, Jose Angel and others to tell us.

Hrdy suggests that a mother’s empathic responses are contingent, not only on her relation with an infant, but on her network of social relations. This includes a multiplicity of allomothers, fathers, sisters, mother, aunts and cousins with whom she lives, and on whom she may be mutually dependent. (It is important that Hrdy prefers to consider fathers “allomothers”, sidestepping the zoological term “alloparent”.) A mother benefits from the support of these allomothers, not only through their attention to her infant, but in the development of her own empathic relationship with the baby.

For fathers, this idea raises interesting suggestions about the importance of men in an infant’s life, not simply in paternity, but indirectly through relation with mothers. Early research on men and fatherhood tended to consider men’s involvement as mating behavior, energy expended prior to conception. More recent research of fatherhood in small scale societies suggests that paternal attention has a direct impact on an infant’s chance of survival after birth , and that multiple fathers increase that chance even more . What studies of paternity often fail to attend to, however, is the importance of the father in the “birth triad,” that is in relation to both the mother and child. Hrdy’s attention to empathic relations between mothers and allomothers embeds an infant in a community of caregivers, including mothers and fathers and a diversity of female kin. This becomes a mutually supporting group, which like the strings of a hammock support each other as the entire cloth holds the infant.

As postindustrial pressures force human into smaller and smaller nuclear units, fathers assume ever more-important roles in the diminishing circle of allomothers. The effects are felt in every hospital birthing room I’ve observed, where fathers (not mothers or aunts or sisters) are asked to provide the primary emotional support for birthing mothers. And I have no doubt that many a mother has had an easier time relating to her new baby because of the support of an empathic father!

REFERENCES
Hill, Kim and Magdalena Hurtado. (1995) Ache Life History. New Jersey: Aldine de Gruyter.
Beckerman, S., Valentine, P., (eds) (2002) The Theory and Practice of Partible Paternity in South America, Miami: University Press of Florida.

REFERENCE:
Bickerton, Derek. 2009. Adam’s Tongue: How Humans Made Language, How Language Made Humans. New York: Hill & Wang.

As I read Jacqueline Solway’s reminder about Lewis Henry Morgan’s long-ago views on collective parenting my first thought was to tell her about recent books and articles by Camilla Power and Chris Knight. Minutes later, Camilla Power’s cogent summary of her views arrived in my in-box, and is now posted at OTH. I can not improve on her concise summary of the relevant literature. Solway is correct of course, that there were serious problems with Morgan’s and Engel’s ideas — particularly Morgan’s notions about “stages” in human evolution. But as both Power and Solway are aware, this does not mean these early theorists were wrong about everything. I agree with Power that what we need now are tests of archeological and paleontological models which start out by reassessing and reconsidering long dismissed views about collective child-rearing.
In addition to the oft-noted androcentric biases, much of the problem here has to do with disciplinary compartmentalization. On this point, I completely agree with Catherine Caldwell-Harris. Thus in addition to archeological, paleontological, demographic, genetic and ethnographic evidence, folklore and emerging information from neuroscience and the comparative study of both infant development and parental and alloparental responses, this revised model-building needs to be informed also by sociocultural studies of caretaking such as Jacqueline Solway is now doing among former foragers. Bill Benzon suggests we consider taking into account the source of the appeal behind anime characters as well, and he may be right. I don’t know enough about anime to comment, but surely casting wide the net to encompass diverse perspectives is the point of this Forum.

As a social-cultural anthropologist I lack the expertise to address the evolutionary questions raised by Hrdy and her interlocutors; therefore I will confine my comments to resonances between Hrdy’s account of human or proto human origins and key debates within the anthropological study of kinship, debates that have broader epistemological implications for the study of human society. I then extend the discussion to consider Hydy’s analysis in light of observations based upon three decades of fieldwork amongst Kalahari residents, including former foraging peoples.

In reading Hrdy’s account I am struck by how her hypothesis on the role and significance of alloparenting and intersubjectivity in our ancestors’ child rearing practices appears to support, as much as an hypothesis regarding our unobservable evolutionary heritage can, the theories put forward by early social evolutionists such as Lewis Henry Morgan (1870) regarding the origins of human society and human social organization.

Morgan’s specific evolutionary outline has been rightly criticized for being overly speculative and empirically wrong in many instances. No proof exists for the postulated pre-historic stage of primitive promiscuity and group marriage that he predicated upon an existing system of classificatory kinship terms in which biological parents and children were not distinguished from their siblings or other collateral relatives. The evolutionists’ subsequent stages of social organization, based largely on available empirical data at the time, reveal a ‘progressive’ shrinking of kinship groupings with the eventual emergence of the nuclear family, and not coincidentally, less alloparenting.

The details of Morgan’s and others’ analysis are not apposite here but what is important to note is that the starting point is the collective and not the nuclear family or even the irreducible biological mother-child relationship that has been central to alternative kinship theories; in particular to Malinowski who argued for the primacy and universality of the nuclear family and universal significance of the biological mother-child bond. This bond provided the fundamental building block of society and kinship terminology extended out according to a ‘kinship algebra’ (Malinowski 1930).

For Morgan, Engels (1972) and others, kin relations, including those entailed in parenting, were to be read ‘down’ or ‘inwards’ from the collective and not out from a dyadic relationship. Even after the supposed stage of group marriage when biological mother-child relations attained recognition, it was assumed that the role of mothering/parenting was still carried out by numerous, possibly interchangeable, adults who bore the same designated kinship relationship to the child. If Hrdy is right, then her argument lends support to the general approach of the evolutionists who, in their quest to understand the human social condition, started with a focus on the group and not on a specific dyadic relationship.

In sum, if Hrdy is right, that alloparenting preceded human language, and the evolutionists are right in interpreting classificatory kinship systems as implying cooperative child rearing, then the two hypotheses reinforce one another. Further, if the evolutionists were correct, then the erosion of alloparental care as the default human condition is a more recent phenomenon than some have thought it to be. Alloparenting is not simply a convenient support for otherwise isolated mothers.

In my own research in the Kalahari, Botswana among both San former foragers and Bantu-speaking Kgalagadi, I observed alloparenting to be the norm. However, unlike the scenario identified by Givón above, biological mothers do not necessarily voluntarily read the intentions and emotional dispositions of others and then entrust their newborns to them — they often have little or no choice. Alloparents abound and in many cases if the alloparents are in a senior relationship to the mother, they claim their right to care for the child and in some instances, to take the child and assume primary parenting rights, rarely, if ever, over the objections of biological mothers.

As the Kalahari debate of the 1990’s highlighted, it is dangerous to project the contemporary behaviour of recent foragers to a Pleistocene past; all extant peoples have a history and multiple events and influences impact upon the present. However, with the greatest of prudence, if we can infer anything about our ancestors from the behaviour of existing recent hunter-gatherers, we can adduce widespread alloparenting, a great deal of cooperative child rearing, much of it embedded in kinship systems bearing resemblance to those depicted by Morgan and other evolutionists in which the responsibilities for raising the next generation and provisioning for their material welfare do not fall only to one or even a few people but are, in a general sense, the responsibility of a broader collective.

REFERENCES
Engels, Frederich (1972 [1884]) The Origin of the Family, Private Property and the State. New York: International Publishers.

Malinowski, Bronislaw (1930) Kinship. Man 30:19-29.

Morgan, Lewis Henry (1870) Systems of Consanguinity and Affinity of the Human Family. Washington: Smithsonian Institution Press.

I’d like to pick up on Hrdy’s discussion of alloparents and the following remark by Michele Pridmore-Brown:

“As for children themselves, it is no accident that, as I have discussed elsewhere (Times Literary Supplement, May 22, 2009), the winsomeness of famous characters like Pip in Great Expectations or Heidi in the children’s story of that name or the orphan girl in Silas Mariner are crucial to the unfolding of their creators’ plots. That winsomeness is in fact the narrative motor of countless tales—whether it’s the winsomeness inscribed in physical features or of the mind-reading connection-seeking variety or, most often, a combination of the two.”

The winsome characters that interest me are all those large-headed and large-eyed characters in manga and anime, Japanese comics and cartoons, respectively. It would be one thing if those and associated traits (think of Konrad Lorenz’s infant schema) were used only in depicting infants and toddlers, where they are representationally appropriate. But that’s not at all the case. Those traits are also quite common in the depiction of older children, teens, and adults. Why?

For those who aren’t familiar with manga and anime, let me say that I broach this issue because these particular visual characteristics are among the first things people notice upon becoming acquainted with these works. Though not a universal practice, their use is quite common and quite striking. It’s a question that always comes up in introductory commentary, whether in hardcopy or online media. Now …

Given that the characters in these stories act in age-appropriate ways (according to the conventions of the particular imagined universe) it doesn’t make much sense to think of this as an infantilization of the characters – though that has been suggested by Takishi Murakami, a Japanese artist and culture critic. I suggest, instead, that the pervasive use of these traits is directed at the audience, inviting us to consider these characters, of whatever age, as though we were their alloparents. Thus we have a parental concern for their welfare even though they may be a bit wayward; we’re willing to indulge them, e.g., as they experiment with gender roles (as they often do).

Alas, I’m not in a position to cite a strong argument on this point and so I offer it only in the spirit of speculation. Continuing in that spirit, I have some other observations.

In the first place, if the speculation is to have any force at all, it implies that we do not routinely assume the role of alloparents to fictional characters. What role(s) do we routinely assume? As far as I know, we don’t know. Empirical research on audience response is sparse and qualitative research is only less-than-sparse. In either case, I do not know that this issue having been addressed. The default assumption seems to be that we respond in our own person, whatever that may be.

Still, why are these traits so prevalent in manga and anime? The use of these traits is not confined to manga and anime; they certainly appear in other cartoons and comics. However, while manga and anime have existed in Japan since the early 20th century, they have flourished after WWII to an astonishing extent. Manga and anime have been particularly inventive in exploring gender roles and they have been open to a wide variety of cultural influences. Does encouraging alloparental reading and viewing somehow facilitate cultural transition and transformation? The question is not exclusively a Japanese question as manga and anime have migrated out of Japan over the last two or three decades and have (arguably) become pervasive in international youth culture.

And just who is the audience? In Japan the audience of manga and anime certainly includes teens and adults as well as children. The situation in the United States is a bit different. Since WWII cartoons and comic books have been regarded as media for children, which was not the case pre-war. But this is only one aspect of the question.

Regardless of whatever assumptions obtain in a given locale, what does it mean for a six-year-old to play at being an alloparent to fictional characters and what does it mean for a twenty-six-year-old to do the same? I haven’t the foggiest idea.

At this point we’ve got two sets of issues occasioned by certain representational traits in manga and anime. One set of issues centers on how individuals enact the role of reader (of manga) or viewer (or anime). One isn’t simply being oneself. Something else is going on, something having to do with how you regard the characters in the story. The old standby of “identification” won’t do. The other set of issues has to with social and cultural practices concerning the reader’s or viewer’s role. Under what circumstances is that role framed as being an alloparent to the characters in the story? Why Japan after WWII?

Needless to say, sorting through these matters is not going to be easy.

I’m an evolutionary anthropologist particularly interested in Darwinian modelling for the emergence of symbolism, whether that’s linguistic, ritual, religious, artistic. All of human life, in all cultures, organises their most important affairs – allocation of food, allocation of sex partners – through systems of collectively shared fictions. That’s a hell of a difference from any other extant ape. What intrigues me most in relation to Hrdy’s argument – which I strongly support – is how much difference between us and Homo erectus? If H. erectus was, H. heidelbergensis and Neanderthals besides evolving H. sapiens were all presumably cooperatively breeding apes too. It would be hard to argue otherwise, given their expanded brain sizes, which place even more stress on the energetic requirements for mothers. If we accept Hrdy’s basic hypothesis, what were and why were there differences among all these species?

Hrdy states very clearly that she’s dealing with the prequel, but she signposts down a specific evolutionary road that should give us more power to make predictions against the fossil and archaeological records. Although she stresses demographic flexibility in her cooperative breeding model, it is fundamentally a sewing together of Kristen Hawkes and colleagues’ ‘grandmother’ hypothesis with the Tomasello school of developmental psychology, which highlights intersubjectivity, that is, a willingness to share what I am thinking with you, and seek to know what you are thinking of my thoughts, apparent even in very young children. It’s important to stress that ‘intersubjectivity’ is something over and above ‘theory of mind’ because it denotes a two-way traffic in mindreading and necessitates allowing one’s own mind to be read. Chimpanzees don’t do that, unless acculturated to a human world. Given their politics of Machiavellian power struggle, natural selection is not likely to favour a chimp who gives away its thoughts to others.

In this view then, H erectus world was very different from chimp world. Early H erectus, from about 1.8-1.5 million years ago, remains an enigmatic creature, suspended somewhere halfway between apelike and humanlike lifeways. Not really like chimps, but not really like us either. Archaeologically there is no secure evidence of life by a hearth, and no suggestion of symbols, art, ritual or language. Evidence from the key specimen, the sub-adult ‘Nariokotome’ boy of 1.5 mya, is somewhat contradictory about how fast he would have grown up and matured; aspects of dentition remain apelike, suggesting a fast-track to maturity closer to chimps than to us (Dean et al. 2001). Yet, with the brain twice as big as a chimp’s, and larger body size, this hominin must have been living comparatively longer, and becoming sexually mature later.

James O’Connell, Kristen Hawkes and Nicholas Blurton Jones in their original ‘Grandmothering and the evolution of Homo erectus’ article (1999) located precisely the palaeoclimate, archaeological and fossil contexts for the onset of these behavioural changes, producing a parsimonious model for evolution in the direction of all the special characteristics of human life history: not only longer lifespans – notably women’s long post-reproductive lifespan – and delayed sexual maturity, but also the puzzling combination of early weaning with short interbirth intervals going along with longterm childhood dependency. With accessible resources for vulnerable weanlings failing in the drying climate, those juveniles who had mother’s mothers alive to help them find and process food would survive better, improving the mortality rate, enabling selection for longer lifespan, and particularly selection on females living past reproduction (incipient menopause). Those mothers who had helper mothers available would be able to wean earlier and reproduce faster. This cooperation in production and reproduction between grandmother, mother and offspring not only accounts for evolving human life history but also the shared cooperative goals that Tomasello, and Hrdy, argue are the basis for intersubjectivity.

Peter Ellison’s suggestion that we should also examine cooperation in production is well taken, since we need to know as much about H. erectus economics and distribution of resources as possible. However, the point about the grandmother model is that cooperation in both production and reproduction evolve hand in hand. And once this shift in behaviour occurs, a whole cascade of consequences, some of which are brought alive by Hrdy, some of which we don’t fully understand yet, would be in train.

The number one point to make about the grandmother model and Hrdy’s extension of it to cooperative breeding is that they rest on female kin sticking together as a default in Homo. Hawkes, Hrdy and several other contributors to volumes on grandmothering strategies and early human kinship (Voland et al 2005; Allen et al 2008) are now establishing this as a revolutionary new orthodoxy, opposed to the old-fashioned and ideologically coloured notion of ‘patrilocal bands’ (see Knight 2008). We now need more testing of palaeontological and archaeological models which adopt this starting point and explore its ramifications.

The second major point to make about the Hawkes plus Hrdy trajectory is that both leave males somewhat in the margins, and the issue of how males became involved in investment in increasingly large-brained and demanding offspring is left cloudy. Hrdy does provide rich evidence on the endocrine systems of sensory entrapment by which babies can get dads, or brothers or uncles to fall in love with them. But I’m still thinking of Homo erectus as this transitional halfway house. Hrdy is perfectly right to argue that people have done an awful lot of thinking, or fantasising, about male-female relations and sexual selection in hominin evolution, and comparatively little about the all important childhood development issues. For both Hawkes and Hrdy, female kin get on with the job, since they can’t rely on males being reliable. The demographically flexible cooperative breeding networks Hrdy envisages act as child welfare safety nets compensating for the extreme variability of male commitment to investment.

Because female fertility is altered by the grandmother strategy, since mothers with allocare support would tend to have shorter interbirth intervals and be fertile more often, this must affect male behaviour. Potential male strategies in response could vary. More dominant males might attempt to target fertile females opportunistically, roving from one to another, while less dominant males could pursue a strategy of hanging around more regularly, offering provisioning and protective support to a particular female and her kin. As interbirth intervals shortened, investor males who waited around, rather than competed for other mates, should get more reproductive payoffs. Such a picture of variability in male commitment fits Hrdy’s observations of stark differences among even modern human fathers.

Hawkes and colleagues of course have a further hypothesis about male investment for a subsequent stage, drawing on their observations of Hadza hunters who vastly prefer to hunt big game, even though this means they may be successful less often. During the Middle to Late Pleistocene (associating to Homo heidelbergensis), hunting strategies become more effective and reliable. Males were motivated to hunt big game as ‘show offs’. Rather than hunt small to medium game for their own offspring alone, they demonstrated quality by generously providing big game to the whole camp (Hawkes and Bliege Bird 2002). So females gained male investment via mating effort rather than specifically paternal strategies. However, there is no strong argument from Hawkes as to what causes the shift in male behaviour and productivity between Homo erectus and later encephalized humans.

My view is that we should follow through the logic of cooperative breeding strategies and female coalitions as the basis of human social organisation into the last half million years of the Mid to Late Pleistocene (c. 500,000 to 130,000 years ago), when H. heidelbergensis and its descendant sister species, ourselves in Africa and the Neanderthals in Eurasia, rapidly expanded brain sizes to over triple that of chimpanzees. If males got their act together to start supporting the cooperative breeding enterprise in this period, which appears in archaeological evidence, then it very likely was female kin coalitions who got them organised. That they succeeded in gaining the extra energetic investment is proved by the extraordinary expansion of brain size. The precise sexual selection models that emerge give us not only a parsimonious account of human life history, and the cooperative basis for intersubjectivity, but also yield testable hypotheses for the emergence of symbolic strategies as the matrix for ritual, language and art (see Power 2009). The Female Cosmetic Coalitions model is currently the only behavioural ecological account which explains the archaeological record of red ochre usage, ochre ‘crayons’ and engraved pieces of haematite, alongside shell beads, now revealed as the earliest evidence of symbolism in the African Middle Stone Age, arising with our species (Watts 2009).

Allen, N. J., H. Callan, R. Dunbar and W. James (eds) 2008 Early Human Kinship. From sex to social reproduction. Oxford: RAI/Blackwell Publishing.
Dean, M. C., M. G. Leakey, D. Reid, F. Schrenk, G. T. Schwartz, C. Stringer and A. Walker 2001. Growth processes in teeth distinguish modern humans from Homo erectus and earlier hominins. Nature 414, 628-631.
Hawkes, K. and R. Bliege Bird 2002. Showing off, handicap signalling, and the evolution of men’s work. Evolutionary Anthropology 11, 58-67.
O’Connell, J. G., K. Hawkes and N. G. Blurton Jones 1999. Grandmothering and the evolution of Homo erectus. Journal of Human Evolution 36, 461-485.
Knight, C. 2008. Early human kinship was matrilineal. In N. J. Allen, H. Callan, R. Dunbar and W. James (eds) Early Human Kinship. From sex to social reproduction. Oxford: RAI/Blackwell Publishing, pp.61-82.
Power, C. 2009 Sexual selection models for the emergence of symbolic communication: why they need to be reversed. In R. Botha and C. Knight (eds) The Cradle of Language. Oxford: Oxford University Press, pp. 257-280.
Voland, E., A. Chasiotis and W. Schiefenhövel (eds) 2005. Grandmotherhood. The evolutionary significance of the second half of life. New Brunswick, NJ: Rutgers University Press.
Watts, I. 2009. Red ochre, body painting and language: interpreting the Blombos ochre. In R. Botha and C. Knight (eds) The Cradle of Language. Oxford: Oxford University Press, pp. 62-92.