International audience; Three levels of intelligibility ought to be taken into account when one considers the origin and evolution of language: the level of the individual (as biological or psychological entity), the level of (proto-)linguistic forms and semiotic practices, and the level of social structures. In the present talk we shall be specifically concerned with the latter two dimensions of the evolution of language and symbolic activities. We contend that: (i) the psychological domain (beliefs, desires, abilities, dispositions) is inconceivable in the absence of a constitutive relation to social stakes and practices (especially various public semiotic practices); (ii) linguistic structures and functions cannot be understood in the absence of a broader perspective on the symbolic realm. Thus, in contrast to popular mentalistic wisdom, we put stress on public dimensions of symbolic activities, assuming that private dimensions can only be understood in regard to their constitutive relationship to the former. The precedence of the social and the semiotic over the psychological implies inter alia that there may not be such a thing as initial (or zero) cognitive state independent of our very social being. This applies to the individuation of psychological states (mental states, beliefs, desires) as well as to general dispositions and skills (e.g. postures, gestures, physiognomies, artefacts).The primary step in this study consisted in elaborating the framework of Symbolic Economies provisionally defined as complex systems in which transactions aim at assigning and transferring symbolic values (e.g. values of acts, roles, or even other signs) through a co-emerging semiotic medium. Our guiding intuition was that the functions of human languages should be viewed in the social context in which “exogamic” groups co-exist by means of ritualized “protocols” (gift, exchange, alliance, payment, sanction). All these protocols imply an exchange of symbols such as objects, gestures, or (proto)linguistic signs whose symbolic values arise from their involvement in ritualized actions which, in turn, could not be acknowledged or accomplished without those symbols. In this respect the analogy between money and language (which is recurrent in Western culture) constitutes a highly instructive field of inquiry (Lassègue et al., 2004a, b). Although money far postdates the emergence of language, uses of money unveil social, cognitive and semiotic constraints or resources which are presumably also involved in linguistic activity. Indeed, both money and language provide a “general equivalent” for an open series of occurrences by: (i) giving rise to abstract and fictive values in a sensible medium, and (ii) creating a universal reference system, which is fictive in nature, to regulate all emergent values, including the most basic and individualistic ones. The functional analogy is all the more valid if we depart from a strictly utilitarian point of view in economy, and a strictly conceptual / referential view in semantics. Our presentation will focus on the semiotic and (to some extent) linguistic conditions of social emergence of values, action types and categorical networks. Our proposal can be summarized as follows:• The evolution toward modern languages is intimately tied to the emergence of new social and symbolic systems, embodied in new practices. If modern language indeed emerged from a proto-language it was presumably not in order to better communicate an already established proto-knowledge but to sustain the development of new emerging needs and practices.• The origin and nature of primary values can neither be accounted for within a strictly utilitarian framework, nor in the framework of functional-adaptive theories of social evolution; although the adaptive dimension may obviously play a role (though often by hindsight), it can hardly account for the meaning and the effects of major creative inventions in technological, institutional or semiotic fields (coinage, print, use of grammar and of language dictionary, WWW). We are thus led to suggest that right from the beginning utility belonged to a network of values and stakes, which were non-utilitarian in nature. Moreover, as utility is constantly reassessed, a major shift of attention towards new semiotic media is accordingly required. New emerging values and transactions involve not only practical aspects in which a natural utility can be invoked, but also fictive entities, roles and institutions (e.g. spirits, gods, clans, totems, status etc.) mediating utilitarian stakes.• “Practical” and “mythical” dimensions are intertwined and hardly discernable in regard to the narrative-prescriptive constitution of the mankind, whether at the individual or social level. The great variety of socio-linguistic roles (and their corresponding abstract grammatical correlates – e.g. case roles) could never emerge and evolve in the absence of this intricate unity of the practical and the mythical.• The same holds with respect to the formation of social values, and to some extent, with respect to categories (whether concerning the social or the natural realm).• The “symbolic” is not what occurs once the logico-conceptual or causal thought proved unsuccessful. Rather, it is what in certain conditions brings the latter to existence. And, in any case, the essential stakes are always relative to the “symbolic”.• The “symbolic” is not a private competence generated by special brain powers and environmental conditions. The “symbolic” intrinsically relies on public semiotic activity, which itself becomes object of interest.• Human social phenomena do not emerge from the interaction between individuals granted with preprogrammed goals and ways of acting. In most cases, the occurrence of these phenomena depends on a social and semiotic “trigger” which is never under the control of a single individual.Two examples substantiate this proposal. First, we suggest that the grammaticalization of generic action structure may best be viewed as a part of a socio-semiotic process that involves ritualized scenarios and norms. Thus instead of focussing on syntax, that would be viewed as an autonomous and purely contingent generative device, we search for basic recurrent scenarios which instantiate the most generic dimensions of action and semiotically mediated transmission of values between agents and domains. In this context, the distinction between real and fictive agents appears to be useless, due in part to the intricate relationship between what pertains to mythical, practical, and ritual spheres, as many ethnologists described in regard to various so-called “archaic societies” (opening the way to the recognition of globally very same structures in any society). Studies in narratology (e.g. folktales) further illustrate this issue. Second, we argue that similar considerations apply with respect to the foundation of taxonomies (including natural kinds). In contrast to naively naturalistic and objectivist theses which assume that cultural differences are merely contingent on ecological variations, we view these variations as intrinsic to symbolic modes of thought that organize the categorical apparatus of various “primitive societies”. Even so-called natural kinds directly express the permanent interplay of perceptual, ritual, therapeutical, geographical, meteorological, cosmic, mythical and social dimensions of life. In a word, the emergence of modern languages may neither amount to an adaptive response nor be contingent on a singular autonomous event (be it biological or social), but should instead be viewed as a major part and outcome of a self-generating web of social needs and semiotic means. Lassègue, J., Rosenthal, V., Stewart, J., Visetti, Y.-M. (2004a). “Symbolic economies” and the origin of language. 5th International Conference on Evolution of Language, Leipzig, April 2004.Lassègue, J., Rosenthal, V., Stewart, J., Visetti, Y.-M. (2004b). A critical comparison between Money and Language. Language, Culture & Mind Conference, Portsmouth, July 2004.