Mead's social-psychological story of human origins was aimed specifically at accounting for the emergence of self-consciousness as a product of social and physical evolution, with particular emphasis on social factors. The following list highlights some of the main elements of that account:
This suggests a ratcheting process. This process is not simply a matter of individuals or species engaging in new activities but also of stabilizing the capacity to engage in those activities so as to establish new inheritable ready-to-hand abilities. It is not just a matter of composition and variation (including (pre)adaptation, exaptation, (re)appropriation, etc.) but also of selection and transmission. The combined effects of composition and variation plus selection and transmission constitute the innovation plus stabilization that characterize a ratcheting process.
Moreover, in the kind of story Mead offers, abilities, not genes, are the "units" of evolution, both biological and cultural. Many disagreements between biologists and sociologists concerning the role and relevance of evolutionary explanation can be avoided once we acknowledge that the notion of "abilities" cuts across both of these disciplines. The evolution of social life-forms need not be reduced to genes, or genotypes versus phenotypes. Genes are one kind of embodiment of stabilized achievements of new abilities, to whatever extent the latter are biologically transmittable from parent to child. But some abilitiese.g., the corpus of abilities that constitute languageare largely transmitted by other means, though the evolutionary stories we have to tell in such cases would still have to be in terms of evolutionary ratcheting processes.
Accepting Mead's account of human origins hinges particularly on being able to justify stages V through VIII, showing how these steps are evolutionarily enforceable without recourse to explanatory skyhooks. The story would have to be refined somewhat to accomplish this task. The particular claim to be addressed here is that momentous changes that ultimately gave rise to modern humanity are due to the emergence not just of language but of capacities to work with recursively enumerable grammars and thus languages in their most potent forms. Rudimentary proto-linguistic behaviors might be traced back to the earliest hominoids 25 million years ago, or maybe not until hominids appear some 6 million years ago. But fairly sophisticated symbolic languages may have been operative 100,000 or as much as 150,000 years ago. Hence the key to accounting for momentous advances occurring 50,000 years ago cannot just be language.
The story we want to tell requires a particular succession of biological ratcheting effects that developed in such a manner as to enhance human cultural ratcheting in unprecedented ways, attaining some kind of threshold momentum by 50,000 years ago. Such evolutionary advances would be the result not just of increasingly complex machinations of signs and the habits these signs may trigger but of the emergence of abilities to appreciate and freely manage (1) recursively enumerable grammars of symbols that (2) are suited for reference to abstract possibilities. In this sense, we have to account not just for the emergence of symbols or even languages, but of symbolic languages that allow open-ended freedom in the consideration and management of possibilities. This paper will address the second of these factors, namely, the gradual emergence and successive refinements of capabilities to refer to abstract possibilities.
Where Mead was more the sociologist and social-psychologist, Dewey tended to emphasize epistemological issues. Dewey developed a conception of experience that underlies most of his philosophical writingsa conception that is well-suited to ability-based explanations of human origins. This conception of experience is not just a notion of sensory excitations. Experience is instead a full-fledged interactive process in its own right. An experience is characterized by a pattern of "resolution" where disjointed, obstructed, or threatening organism-environment interactions are resolved, balanced, re-integrated. Experiences thus are episodes or moments of resolution, ongoing for as long as the experiencer is alive to protect and maintain its interests as a live creature. As such, experience in its primary modes is driven and mediated by instinct and habit, simple or complex, systematic or not.
Human experience, on the other hand, is a type of experience which, though driven by instinct and habit, is sometimes mediated by deliberate, purposeful reflection. One's habits will include and be refined and modified by such capabilities. More to the point, experiences would take on a new character that warrants a different label. An experience may become what is better termed an inquiry. Namely, an inquiry is an experience in which deliberate reflection has some effective mediating role to play (for better or worse) in the unfolding and completion of that experience. The question here with regard to human origins specifically concerns why and how such capacities to deliberate could have emerged.
This question is surely tied up with the emergence of capabilities to use symbols. For Dewey, this is one and the same with capabilities to manage and manipulate abstractions ("universals") as the contents and referents of deliberate reflection and thus as instruments in the regulation and control of experiences (also see Mead , 211). The important element that Dewey is introducing into the story here is a specific notion of what a symbol is, versus signs more generally. Symbols are conventional signs. But more importantly, according to Dewey, symbols are associated with abstractions corresponding to possible manners of action, ways of acting, or modes of being. These manners, ways, or modes easily encompass established abilities central to ability-based explanations of evolution, human or otherwise. Abilities are appropriately characterized as universals because they are in themselves intangible abstractions when serving as objects of reference, and because they subsist in such a way as to be ready to hand under any conditions, in any circumstances in which they may be relevant. We can introduce symbols into a story of human origins if we can provide an evolutionarily enforceable account not just of the conventional nature of symbols but also of their connection with universals.
Referential capabilities are key here insofar as they must eventually involve universals as objects of reference. Recall that Mead's story of human evolution can be cast as a development of new abilities by means of biological and cultural ratcheting processes: valuable new activities becoming engrained inheritable abilities. Any such ability is an evolutionary achievement if it has become a constitutive feature of a given creature's experience, regularly executable when called upon. Such an ability becomes in itself a thing and thus a possible referent insofar as it is engrained and thus is readily called up and implemented. The first question, then, is whether and how such things may become objects of reference.
The particular evolutionary developments we are trying to account for can be understood in part as an extension and refinement of our signifying behaviors, encompassing references not just to concrete tangible things but also to concrete abilities themselves as things in their own right. This is somewhat unusual insofar as abilities (e.g., your ability to walk) are intangible things, not objects of the senses. Initially, they are directly evidenced only as they are actually engaged and implemented. What we can see are instances of their execution, not the abilities themselves. Otherwise they are simply taken for granted (engrained, habituated), available for execution under suitable conditions. Nevertheless, for communication purposes, we can associate various sounds or marks or other signifiers with these intangibles just as we can associate different sounds with more tangible things.
Vervet monkey's are capable of signaling one another when a threat (e.g., an eagle) is detected, using specific alarm sounds signifying appropriate defensive behaviors (Cheney and Seyfarth ). Humans, on the other hand, can also use word sounds (like 'eagle') symbolizing the possibility of eagles and meaning associated ways of behaving toward them. This capability is a distinct achievement, but its evolutionary emergence does not require an explanatory stretch if possible ways of behaving can serve as things signified and thus as possible objects of reference. How did this come about? If in no other way, it would have come about in a social setting, in response to a persistent need to manage social relationships and coordinate shared activities.
Mead's build-up to the notion of a "generalized other" is pertinent here. This notion assumes a more-or-less stable capacity to objectify not just others' behaviours but others' possible ways of behaving. Stage V of Mead's story only involves capabilities to signify such possible ways of behaving, not yet a full-fledged sense of a generalized other. But having capabilities to signify possible activities is essential to being able to "take attitudes" of othersadopting a stance and a readiness to act in line with others' possible behaviors, to determine one's own behavior in light of anticipations of others' behaviors. We cannot and need not attribute a sense of self to a creature with such capabilities, nor does taking attitudes require reflective thought. But adopting attitudes in this sense promotes a form of reflexive determination of behaviors, given the capacity to align one's own actions with anticipated actions of other individuals. This early stage in Mead's story calls merely for stabilizing capacities to objectify and signify others' possible ways of behaving. Only then can those possibilities serve as prompts or guides or determinants for one's own behaviors.
Vervet monkeys may have such capabilities, in some rudimentary form, evidenced especially by cases of deception (Key ; Knight , ; Power ). Such capabilities would become an essential feature of managing social relationships and coordinating shared activities, allegedly having to be fairly well developed as stable clan sizes approach 150 or so individuals (Dunbar , , ; Aiello and Dunbar )something that only emerged roughly 150,000 to 130,000 years ago with anatomically modern Homo sapiens. We can reasonably surmise that, while rudimentary capabilities to objectify possibilities may go back much further in the evolution of Homo sapiens psychology, social pressures will have engendered sophisticated capacities to signify and refer to ways of behaving (in the absence of the actual behaviors themselves) in the psychology of archaic Homo sapiens and even more substantially in anatomically modern Homo sapiens.
Note that this new sort of signification allows open and direct reference to possible activities, whether it be individual activities, activities of other individuals, or activities of whole groups. It allows reactions to possible activities or their results before such activities or results are actualized. It allows greater freedom of action guided by anticipation and "foresight" (in conjunction with memory and "hindsight") even if this may not yet involve deliberate thought. To some extent, reference becomes unhinged from actual objects. In time, foresight will be broadened insofar as creatures are able to manage wider swaths of options in given circumstances. It will be deepened insofar as they are able to manage more complex or lengthier sequences of options in those circumstances.
This raises questions concerning compositionality, even if the focus presently is on referential capabilities. In allowing for the deepening of foresight, we have to wonder how and to what extent referential capabilities may be composed to accommodate complex connections and associations among things. Tendencies to build compound actions from simpler ones go all the way back to the earliest life forms and thus predate the emergence of abstract signification. Complex sequencing of disjunctive habits is common in all animal species. Consider, for instance, the sophisticated migration behaviors of birds (Alerstam , Berthold , Berthold et al. ), or the complicated anti-snake behaviors of California ground squirrels (Coss , ; Coss and Owings ; Rowe and Owings ). Or simply using "hands" to mediate various actions (e.g., food-to-hand-then-hand-to-mouth) requires appropriate sequencing of activities (reaching, grasping, transferring, etc.).
The broadening and deepening of foresight is a natural variation on this pre-existing theme of compounding and complexifying abilities. Signs lend themselves especially easily to such compounding. It is easier to compound signs (e.g., sounds that signify eagle-avoidance behaviors) than it is to compound things signified (eagle-avoidance behaviors). At the same time, practice in compounding anything (sounds or behaviors) helps to hone the activity of compounding in its own right. Dennett's "crane" metaphor (, 7576) as an account of evolutionary exaptation is applicable here, specifically in that the development and establishment of compounding and complexifying techniques involving signs provides operational templates (cranes, as it were) that are applicable in domains of activity besides those on which they are "firmly based." In such ways, the compounding associated with exercising foresight could be transferred into refinements and elaborations of existing abilities to compound things of any sort whatsoever. We should not be surprised to see evidence of these successive refinements in the form of new and improved cultural artifacts and advances in the technologies by which they are produced.
We would further expect such foresight eventually to engender some appreciation of new or refined compound things that are foreseeable whether they come into existence or not. Hence the deepening and widening of foresight together with an extension or transfer of such capabilities to things in general, actual or not, can come together to generate not only activities of increasing complexity requiring substantial time and energy to unfold properly, but refined capacities to imagine complex activities without actually having to engage in them.
Such capabilities may be unique to humans. But there must be more to an account of human origins. So far the story interweaves elements of rudimentary "abstract" reference and compositionality, each mutually reinforcing the refinement and elaboration of the otherwith any new capabilities this engenders being transferrable back into material culture. The initial step leading to such developments in human evolution is a new sort of signification where signs are associated with abilities as things in their own right. The emergence of this latter capability will have taken several million years to stabilize into its more sophisticated forms, from the earliest hominids onward; but it is reasonable to assume that rudimentary forms of this capability would have become innate in the earliest hominids some five million years ago.
Nothing we have said so far requires or explains the introduction of symbols or symbolic reference, if by "symbols" we mean signs that are conventionally associated with their referents. Complex compositionality and reference to intangible abilities can characterize rudimentary sign systems designed to aid in exploiting connections and associations among things, without the need for symbols as such. We thus should distinguish three types of reference here: (a) The most basic kind of sign will be a thing that signifies some other thing, in the way that smoke signifies a fire. Our most basic instincts and habits will involve attunements to such involvements among things. These attunements will be operationally perspectival, depending on what abilities and sensitivities a given creature has, but they are not just conventional. (b) Wholly within such sign systems, things (e.g., gestures or other displays) may also signify abilities, in the way that a growl or a certain strut may signify possible impending combat or simply confirm social dominance, or in the way that menstruation may indicate imminent sexual receptivity. In turn, (c) such abilities may themselves signify other abilities, in the way that social dominance signifies access to mates, sexual receptivity signifies fertility, etc. Hence, even with sign systems devoid of symbols, the connections and involvements among things and abilities (1) may be quite complex in their compositionality and (2) may incorporate a rudimentary kind of abstractionwhere tangible things may signify intangible abilities, and intangible abilities may themselves signifying other intangible abilitiesassuming that intangible behavioral abilities (present in the form of "attitudes") can serve as signs at all.
This latter type of involvements among intangibles will of course be more directly and easily managed and manipulated with the emergence of tangible symbols to track and refer to these intangibles. Precedents for this kind of thing are already present in vervet monkey's alarm callssounds that may at least be regarded as proto-symbols. There is an element of conventionality at work there, in that just which alarm sound signifies which kind of threat is more or less arbitrary and yet is uniformly recognizeda matter of common attunements to invariant sound-to-thing associations in the case of vervet monkeys. As stable clan sizes increase and as management of social relationships and coordination of shared activities become more complex in hominids, we would expect an increase in these "vocabularies" of proto-symbols, pertaining not only to common interests in various natural resources (food and water sources) and different manners of social interaction (mating with, fighting, giving to, taking from) but growing to encompass references to different kinds of individual identification and relative social status (sex, mood, group membership and rank, alliances and common kinship, dominance relations) as well as to fairly complicated if not subtle internal threats (cheaters and freeriders). It is easy to see how social signaling and particularly "gossip" (allegedly having a role similar to grooming in the maintenance of social relationships, but not limited to one-on-one interactions) would be an important step toward the emergence of full-fledged language in human beings (Dunbar , [1998a], [1998b]). It initially produces relatively rich vocabularies of proto-symbolsconventionally determined things (noises, gestures, postures, facial displays, and sequential arrays thereof) linked more or less unambiguously to specific referents. All that is involved here is an elaboration of something that vervet monkeys can already do, but now aimed as well at managing and exploiting complex social environments.
Coupled with the need to detect and identify freeriders and freeriding, deceivers and deceptions, among other things, the capacity to allow conventionally determined proto-symbols to refer to abilitiesnot just to cheaters or cheatings but to dispositions to cheat, etc.would emerge naturally as a variation on capacities that already exist in the function of social signals like growls, struts, smiles, or other significant gestures. At some point we should begin to see the emergence of more refined capabilities to acknowledge and manipulate relationships among abilities themselves, insofar as relationships of significance among tangible proto-symbols could be utilized to reflect involvements among the abilities they refer to. Insofar as this makes possible a flexible quasi-syntax of composable proto-symbols, we would expect to see the emergence of proto-language, moving beyond mere vervet-like holophrastic signaling to include simple compositionality and complex reference (Bickerton , , , , ; Calvin and Bickerton ; see also Botha ). In this context, proto-symbols would begin to function more like true symbols, or symbolic signs (to turn Mead's phrase around).
Thus we come to stage VI in Mead's story. As the capacity to adopt attitudes becomes a reliable capability, one moves into a position to draw on the attitudes of others in one's ongoing conductto repeatedly and reliably adopt and thus "consult" others' possible attitudes in the course of pursuing one's own agendas. As this capability becomes common within whole groups, facilitated by common use of symbolic signs, it serves to establish and stabilize various social "institutions," reflected in attitudes about attitudes not only of particular individuals but of whole groups. Groups may develop and share common attitudes concerning ways of behaving that are in a sense more than the sum of the behaviors of their individual members. These attitudes may be "taken" and "drawn on" just like any other attitude, and thus we would begin to see the institutionalization and ritualization of certain ways of behaving. We can imagine an example of this in the cosmetic use of red ochre to symbolize menstrual blood, facilitating the stabilization of female coalitions as institutions designed to distribute assets of "immanent fertility" of individual menstruants across whole coalitions (Power and Watts ; Power and Aiello ; Power , ; Watts ). There is evidence that late archaic Homo sapiens and early anatomically modern Homo sapiens engaged in such ritualization, though the general practice of forming attitudes about group attitudes (not to mention iconic-signaling through the use of earth pigments) may of course go back further than that.
Pursuing things a bit further, what we have by now are improved conditions for an individual's being able to concretize a generalized sense of what others of one's own "kind" (one's "kin") are like, that is, for developing a sense of a "generalized other" introduced in Stage VII of Mead's story. This is more likely to emerge as one becomes attuned to social institutions (rituals, symbolic behaviors, etc.). One's social environment begins to take on more tangible forms with the emergence of ritual or other kinds of symbolic behavior, allowing individuals to solidify a generalized sense of that social environment. Ritual enhances existing regularities and introduces a degree of regularity and constancy into at least some social practices where it may not exist otherwise. One may of course become attuned to these social regularities just as one becomes attuned to any regularities anywhere else in nature. The coalescence of these social attunements into engrained abilities thus gives rise to possible reference to a generalized other.
Note how this may lead to activities appropriately termed "thinking." By engaging in ritual activities, one is interacting with one's society, not just with other individuals. This generic activity may become an engrained ability; and this intangible ability may in turn become an object of reference, allowing one to anticipate generic events in one's social environment (depending on its degree of constancy) and to determine one's own behaviors in light of such anticipations. Mead thus introduces "reflexive discourse" at this point, at least in a rudimentary form, as a process of anticipating responses of the social environment as a whole to one's own actions. This process, while it need be neither silent nor covert, may take place in imaginationusing imagery (MacWhinney ) in the absence of actual behaviorsprecisely because of one's capability to anticipate and survey possible behaviors and their possible effects. It becomes reflexive discourse in the sense that (1) it is a "reflexive" interaction among attitudes, "coming back" on oneself, where (2) the individual alternately takes on the attitudes of a generalized other as well his/her own attitudes as an individualeach successively modified in response to the other, perhaps in sequences or patterns typical of or derived from actual ritual activitiesthus being a "conversation" of imaginable activities.
We may call this activity "thinking" insofar as it is to some degree disengaged from actual or potential circumstances that it is about and which somehow compel one's attention. It is important to recall Dewey's claim that this thinking activity will take place in a context of organism/environment interactions, brought to bear under conditions of and in response to felt needs. Thinking will thus be about troublesome circumstances requiring some kind of resolution, taking place as a means for accomplishing that resolution, supplementing the instincts and habits that might otherwise function adequately in much of one's ordinary experiences. The "trouble" may be large or small, momentous or trivial; but some such stimulus is needed to prompt thinking activities and to impel them in some direction. The key point here, generally speaking, is that this reflexive conversational activity allows one to disengage from given circumstances precisely in order to survey possibilities and to better determine ways of resolving problems or satisfying felt needs in those circumstancesnot to disengage entirely (whatever that might entail) but to guide one's engagements in ways that improve on what basic instinct and habit would otherwise provide. Thinking is then an activity in its own right, no doubt acquiring its own habits, instincts, and refinements thereof. It would involve a kind of temporary and mediating disengagement of reflexive-conversational habits from habits at largestimulated by felt needs but also designed to actively bear on actual responses to such stimuli.
Earlier instances of this capability in human evolution could not properly be called "talking to oneself" if there is no such thing as "talking" to be reflexivized. But conversation, at least in the form of conversations of gestures and signs (vocally, by hand, or by countenance), has a longer ancestry and will have been ensconced in the repertoire of hominid abilities well before archaic and anatomically modern Homo sapiens appeared on the scene. The reflexivization of such capabilities would emerge then as a succession of attitudes, taken by one and the same individual alternating various roles (in the place of multiple interlocutors). The alternation of roles and corresponding succession of attitudes would be constrained by actual social institutional norms and corresponding involvements among attitudes and related modes of behavior. Imagine, for example, a succession of attitudes of guilt, shame, fear, disapproval, disgust, defiance, angerin various ordersthat might be initiated in an individual targeted by community gossip. Couple that with the individual's sense of a generalized other and a capacity to assume alternately the role of the community and the role of the individual in that community. Associate some of those attitudes with the generalized other and some with the individual, and we have the makings of a reflexive attitudinal conversation. Languages of thought could thus have been born out of reflexive attitudinal proto-symbolic conversations.
Elaborations and refinements of such reflexivized conversation will of course be facilitated by the emergence of symbolic behaviors and ritual activities. Regularities in the social environment will have been enhanced, and one's sense of a generalized other will be correspondingly more robust. Symbolic vocabularies will facilitate and expand the range of this reflexive conversation in ways that standardized gestures and proto-symbolic vocabularies would not. The eventual refinement of vocal speech would of course enhance reflexivized conversation in such ways as to be, in part, like an "inner" flow of speechtalking to oneself, reflexivized discoursethough it is important to note that this reflexive speech would be just a refined type of attitudinal conversation, and it need not be heavily constrained by the same rules and tempos of overt conversational speech from which it is nevertheless derived.
The fossil record indicates that vocal tracts (and thus vocalization capabilities?) of late archaic Homo sapiens (perhaps as early as 200,000 years ago) were virtually the same as modern Homo sapiens. Rudimentary vocalization capabilities no doubt emerged much earlier (with Homo ergaster/erectus?). Evidence points to possible rudimentary control of vocalization in Homo habilis 2 million years ago, and more so 500,000 years ago in earlier archaic Homo sapiens (the common ancestor of Homo sapiens neanderthalensis and Homo sapiens sapiens). Certainly by 125,000 if not 150,000 years ago, anatomically modern Homo sapiens had the anatomy needed for vocal speech. Given the likely emergence of symbolic ritual in this latter time frame, we should not be surprised if refined forms of thinking, in the sense just described, were emerging at least by 150,000 years ago and were well-established by 125,000 years ago. Rudimentary forms of thinking may have been developing for many hundreds of thousands of years, depending of course not just on refinements of vocal or other communicational means (driven by increasing clan sizes?) but also on increasing stabilization of societal forms and conditions needed for individuals developing a sense of a generalized other (or minimally a capacity to engage in the alternate role-taking characteristic of reflexivized attitudinal conversation).
Still, we do not have a full slate of conditions in place yet for a satisfactory account of human origins. To complete that story, we would have to back up and consider factors that have been only briefly touched on in earlier sections. The focus so far has been on the development of a sense of a generalized other as an object of reference and thus as a crucial element in the emergence of capacities to think. Along the way, we have considered ways in which the emergence of symbols and symbolic behavior would enhance these capacities. Other factors we must consider concern developments of compositional complexity of sign systems and symbol systems in particular. In this context we may give an account of evolutionary changes that ushered in new ways in which symbols were usedchanges which may have been crucial to the emergence of capacities for deliberate innovation that characterize behaviorally modern Homo sapiens. We will not pursue this line of investigation here, but we will conclude by considering the final stage of Mead's account of human evolution simply in terms of referential capabilities.
As of stage VII in Mead's evolutionary story, we have accounted for human capacities to think. Stage VIII further accommodates the emergence of self-consciousness and individual minds, given that the activity of thinking by itself is not sufficient to engender a sense of self.
This final stage of development would best be facilitated by a robust sense of a generalized other insofar as the reflexive conversation of attitudes that constitutes thinking takes on a new complexion where one's sense of a generalized other becomes identified as "me." A generalized other in essence provides one with individual or personal character. Thinking is still a reflexive conversation where the individual takes alternate roles, adopting a succession of attitudes of those respective roles. But the role of the generalized other now becomes more distinct as one's own role ("me"), having more or less definite features and characteristics, while the other role ("I") is just the live individual (as a member of the community represented as the generalized other), remaining indistinct but actively engaged in this reflexive conversation. What is new here is the development of individual character or "personality" as something that can operate independently, freed (to some degree) from cultural constraints dictated by the social environment. The new development here is one in which the thinker becomes in a sense an internal society-of-two, capable of generating its own reflexive culture: "the self, as that which can be an object to itself, is essentially a social structure, ... in a certain sense provid[ing] for itself its social experiences, and so we can conceive of an absolutely solitary self" (Mead , 204). Mind, as an arena of reflexive conversation, is not recognized as private until a "solitary self" arises.
Note the possibility of yet further developments, with or without a strong sense of self. Activities of using symbols may solidify as stable abilities to use symbols, in which case we may have symbols associated with abilities to use symbols. Obviously this generates a potential ratcheting regress (not the only one we are interested in here). Evidence of advanced developments along these lines comes with thinking about thinking, writing about writing, etc., which is clearly well-ensconced in human development by at least 2500 years ago if not sooner (5000 years ago?). Prior to the emergence of robust self-consciousness in particular individuals, such a regress could have yielded ethnic art and symbolic culture. With the emergence of individual selves we would begin to see individual authorship and innovation.
It is remarkable that Mead could take an account of human origins as far as the emergence of self-consciousness, from simplest life forms, wholly governed by rules and constraints of naturalistic, evolutionary explanation. In tracing this story carefully from beginning to end, one finds only one small step after another, each being evolutionarily feasible on the basis of what has been previously accomplished.