Wolfgang Wildgen (University of Bremen, Germany) Linguistic functionalism in an evolutionary context

1. Wolfgang Wildgen (University of Bremen, Germany)Linguistic functionalism in an evolutionary context 40th Annual Meeting of the Societas Linguistica Europaea29 August - 1 September 2007University of Joensuu, Finland

2. Classicalfunctionalism • Functionalism in the tradition of Jakobson and Martinet is often understood in terms of an optimization of commu-nication (a minimum of phonological mergers, of ambiguity, etc.). • In Martinet’s concept of economy of linguistic change a level of economy is stabilized (after some kind of structural loss) or the system shifts from one economic maximum to a neighboring one. • One general tendency is that globally all current (full-fletched) languages (not jargons and pidgins) are at the same level of functionality; changes are only local shifts in a field of multiple (and grossly equivalent) alternatives.

3. Functionalism and evolution • If one considers long-ranging linguistic changes (millennia) or the develop-ment after some out of Africa move, i.e., since some proto-sapiens-language (100-200,000 y BP), it is rather obvious that the role of language in larger and highly organized societies must have affected the framework in which functions and degrees of optimization are defined. • An even bigger challenge to functionalism occurs in the context of some protolanguage, which describes the transition between the last common ancestor of humans and chimpanzees (LCA) in the period between australopithecines, Homo erectus and archaic Homo sapiens. Here functionalism has to be linked to the Darwinian notion of selection and the question arises what kind of selection was responsible for the emergence of language: overall selection by the environment, sexual selection, social (kin) selection or some combination with body-internal equilibriums between different selective pressures (a kind of self organization or morphogenetic process).

4. Emergence of functions The so-called functions of language (cf. Bühler and Jakobson) had to emerge from a prior configuration of communicative and behavioral functions, which were already present in mammals. Thus the theoretical foundation of functionalism in a larger context asks for: • The origin of specific functions of language in the field of more general communicative and social functions. How can new functions emerge? • A measure of the degree to which a function or a criterion for selection is fulfilled, i.e., a measure of optimization for functions. This questions points to processes of self-organization and levels of complexity.

5. Basic goals The goal of this contribution is to: • Ground the major concepts of linguistic functio-nalism in a Darwinian notion of selection. • Introduce the idea of self-organization in order to explain the emergence of new functions. • Specify measures of success, which allow for optimization and a dynamics of parallel optima (a landscape of optima between which a system can choose and which control its transformation).

6. Martinet’s functionalism and Darwinism • André Martinet turned radically against Saussure’s preference for synchro-ny (and a theoretical devaluation of diachrony) and against the neglect of language usage (parole) and thus of social variants which rival with a given linguistic norm. Already in his first treatise on French phonology (1933) he used the criterion “rendement fonctionnel”, i.e., functional load as measured by the frequency and relevance of a phonemic contrast. He developed this aspect further using the law of “least effort” applied to language by Zipf (1949). Finally, he proposed the principle of linguistic economy, which he exemplified in his treatise on diachronic phonology (first published in 1955: Économie des changements phonétiques). As in Trubetzkoy’s phonology, the function of information dominates although Martinet mentions Bühler's tripartition of communicative functions taking the perspective of the hearer. He says (Martinet, 1975: 35): • “L’auditeur, s’il connaît la langue employée, fait inconsciemment le tri entre ce qui le renseigne sur l’identité du locuteur, ce qui l’informe de son état d’esprit et de son humeur veut communiquer au moyen de la langue dont il se sert. “

7. Three forces of change Martinet (1975: 39) distinguished three types of (selective) forces: • Needs of a society in terms of communication (“nouveaux besoins d’une société en matière de communication”). • The internal dynamics of the linguistic systems governed by conflicting forces as the effort (for speakers/hearers) and the result (efficient transfer of information). • The system must achieve or retrieve a state of equilibrium between these forces. This trait is called dynamic. • As the learning of language is mainly an accumulative process of imitations, which result into traditions, the resulting language has a certain tendency to resist against language change.

8. “Darwin’s paradox” (Labov) • Labov (2001: 6-10) compares explicitly his observations on linguistic change with citations from Darwin’s book: The Descent of Man (1871) and he calls the underlying problem “Darwin’s paradox”. Here is his formulation of the paradox: • “The evolution of species and the evolution of language are identical in form, although their fundamental causes are completely different.” (ibidem: 14) • Labov points to the fact that the principle of Darwinian selection states the fundamental mechanism of biological evolution, whereas in languages rather the search for novelty or fashions governs the direction of change (ibidem: 14 f.). Thus the parallelism is only a superficial one, insofar as in the evolution of languages the fundamental (Darwinean) mechanisms responsible for the evolution of a species is absent.

9. Major periods of radical change • Inside the well documented family of Romance languages (appr. 2,000-3,000 y). Martinet mentions the change of needs in highly organized empires like the Roman one and Bickakjian (2002) interprets changes in Romance and Germanic languages as having “advantages” for the speakers and hearers. • Inside the family of Indo-European languages (appr. 7,000 y) or inside some macro-phyla (14,000 y). • Language changes after the out of Africa migration of modern humans (100,000 to 70,000 y). • Changes in language capacity after some proto-language used by Homo erectus/Homo ergaster populations, from which the modern species Homo sapiens evolved (400,000 to 200,000 y). This would point to the before or during the first out of Africa migration (2,3 to 1,6 my).

10. Meta-representation (poetic function) representation Human language Semiotic sign Animal communication Non-semiotic sign expression appeal Sign-functions and their evolutionary significance For Bühler, functions (aims, intentions) are kinds of vital needs and thus presuppose the level of life (of animals). If such needs (or instincts in traditional terminology) are ge-neralized beyond animals and humans, a higher level of generalization can be reached.

11. The evolutionary interpretation of the triad of functions • The last two functions, expression and appeal are strongly linked, because the use and meaning of expressive acts asks implicitly for some receiver and appeal is without effect if no expressive content is transferred. We can use the label “social communication” (social calls, grooming, body postures, etc.) as a cover-term for both and distinguish it from functional referentiality (which first appears in the alarm-calls of e.g. velvet-monkeys). This simplifies Bühler’s triangle to a binary opposition between social communication (expression/appeal) and reference to the world common to all participants.

12. The transition to humans If representation is in its first stages already present in socially organized primates (or even in monkeys), the transition to humans concerns mainly: • The enrichment of representation, i.e., the lexicon and via self-organization the syntax; and • The emergence of meta-functions. • The most prominent case of meta-function concerns propositional attitudes and explicit performatives: • I believe that a snail is in the tree. • I tell you that a snail is in the tree.

13. If in a further step one assumes that representation emerges from ecological cognition (categorization of an ecology) and expression/appeal from some structure of the group (primitive, non-conscious social categorization of behavior), one obtains three inclusive levels, • where the inner circle is reached by all animals with a social organization and specific reactions (perception /motor control) to their environment, • the middle circle concerns animal communication with a minimal reference to the context and • the outer circle encompasses humans (and possibly some primates with self-awareness). The functions in Bühler’s triad emerge from ecological categories and from social categories already apparent in animal behavior.

14. Meta-representation Representation Ecological categories Social cognition Expression/appeal Meta-communication Functional hierarchy at three levels The inner circle is reached by all animals with a social organization and specific reactions to their environment, the middle circle concerns animal communication with a minimal reference to the context and the outer circle encompasses humans

15. The local differentiation of the referential function • Increase of social vocabulary referring to actions in the group (cf. the kinship terminology) and in relation to preys and predators (cf. the alarm calls as base line). • The increase in linguistically labeled distinctions in the ambient world; i.e., the differentiation of the lexicon of flora and fauna. • The complexity of utterance organization, i.e., the emergence of syntax. As in child development, the increase of the lexicon (1 and 2) asks for a proper phonological organization. Therefore, phonology (enabled be an efficient cognition/memory/motor planning of phonetic sequences) is a self-organized outcome of an increased lexicon. In a similar way syntax is a self-organized consequence of larger utterances, which are less context-dependent.

16. Self-organization and functionalism In relation to overall selective pressure this means that: • Adequate cognitive (perceptual, motor and memory) skills must be available. This increase is related to a bigger and more energy consuming brain, which in turn must be “paid” by the availability of high-energy food. • The power of the linguistic system can be decomposed in many different ways and distributed over the principal components: phonology, lexicon, syntax, discourse, i.e., many equally powerful forms of organization are possible. This is the basis for (de Saussure’s) arbitrariness in the lexicon and in many areas of morphology, syntax, and discourse. As a consequence, it becomes impossible to judge the functional power of a language in relation only to sub-components. Moreover, the context of usage becomes an important factor.

17. Selective value of communication and symbolic behavior • All functional models of communication should ask for the survival value of communication, because in a Darwinian framework only the fertility of the species (not its cleverness) counts. As communication is a type of information-sharing, a concept of (strong) reciprocity is needed. Under what circumstances did (reciprocal) sharing of information pay off? • The sharing of information on the ecology, on one’s own mind and on social relations (expression/appeal) follows from strong reciprocity. It also enforces a level of truthfulness of symbolic behavior. Cheating and lying by means of symbolic forms is, however, an alternative corresponding to the within-group selection of egotists. The equilibrium of both strategies and its stability is a phenomenon, which asks for further elucidation.

18. Dunbar’s hypothesis • Dunbar (1997) found that chimpanzees employ 20% of their time in grooming. These practices are necessary to uphold social solidarity, social roles (hierarchies), to control conflicts, etc.; i.e., grooming is a semiotic activity, a ritualized behavior abstracted from mutual hygiene. In bonobos, sexual activities are also ritualized for social purposes. • He argues that the percentage of time spent on grooming-activities depends on the size of the group. If the social organization of the group tends to larger communities, these techniques of solidarity and social peace become energetically too expensive. Vocal communi-cation, chatting, simply construing vocalized contexts of solidarity is an alternative. The most proficient actors in social communication get dominant roles in the tribe and reproduce at a higher rate. A run-away process makes this competence desirable and creates the necessary social power.

19. Levels of selection and language • In actual evolutionary biology a dispute arose about the proper units of selection: Are genes selected or organisms? Dawkins distingui-shes between replicators (genes) and vehicles (organisms). For selection by sexual partners, rivals, predators, ecological contexts only the vehicles, i.e., the entities which contain genes are visible, may die or survive (and procreate). • Other scientists proposed families, social groups or subspecies as entities of selection. This seems to be plausible, if a species shows a complex social organization, with a differentiation of roles and a high level of altruistic behavior. As this is certainly true for human societies and linguistic communities this debate is relevant for linguistic functionalism. Are populations, communities selected in relation to the level of linguistic competence or linguistic efficiency?

20. Selection at different levels beyond the genes • The genetic code (DNA) as a complex organization may select the activity and effect of genes which often rivals for resources in the translation process (RNA) and in epigenesis. • Specific cellular assemblies may be selected over others. Thus cancer cells can grow rapidly and even destroy the surrounding tissue, because under certain conditions they have a selective advantage. In the brain different cell-assemblies may rival for dominance and be selected or not. Thus brain activity can be understood as complex interaction between cellular agents under selection pressure. • The human organism as a whole is composed of many organs, which compete for resources.

21. Levels of selection beyond the gene • Body internal complexes under selective pressure. • Individual organisms under selective pressure (the classical case discussed by Darwin). • Kin-selection. Kins share genes, e.g., mother and daughter, or clans based on kinship. They form coalitions in search for resources or in defense against other individuals or groups (in a metaphorical sense kin-groups act in favor of the permanence and expansion of their genes; cf. Dawkins, 1994). • Larger social groups sharing genetic features, or, in the case of humans, culture and language, may be the relevant units of selection. At the stage of civilizations cultural traditions and communalities may act like genetic proximity. If this case, biological selection is replaced by cultural selection. As Dunbar showed, cultural selection is, however, grounded in biological selection, which is only effective indirectly.

22. Darwin’s principles • The principle of serviceable associated Habits. — Certain complex actions are of direct or indirect service under certain states of the mind, in order to relieve or gratify certain sensations, desires &c.; and whenever the same state of mind is induced, however feebly, there is a tendency through the force of habit and association for the same movement to be performed, though they may not then be of the least use. (Darwin, 1872/1969: 28 f.) • The principle of Antithesis. — ... when a directly opposite state of mind is induced, there is a strong and involuntary tendency to the performance of movements of a directly opposite nature, though these are of no use; and such movements are in some cases highly expressive (ibidem). • The principles of actions due to the constitution of the Nervous System, independently from the first of the Will, and independently to a certain extent of Habit (ibidem).

23. Comparison with Martinet’s levels • The first criterion of Martinet: needs driving the evolution of language, opens a large field. Certainly the ecology and the social environment in the different stages of humanization were radically different. • Martinet’s second criterion, i.e. conflicting forces and the search for an equilibrium of forces in the language system (at the levels of: phonology, morphology, syntax, lexicon) can be related to selective forces in the brain; cf. Darwin’s last principle. • The third criterion mentioned by Martinet; the accumulation and maintenance of traditions has an analogue in Darwin’s first principle which mentions “serviceable habits”.

24. Levels of social evolution and its consequences for language • With the rise of hunter gatherer populations (probably since some 100.000 y BP) complicated rules of kinship classification were developed. • The stage of large civilizations with long ranging exchange and global cultural patterns may have begun with the cave painters in Europe and Northern Africa (Sahara). With the Neolithic revolution concentrated highly dense areas of habitation arose. This led to the first well documented high civilizations with writing system in Egypt, Mesopotamia, India (perhaps China) with central power, social hierarchies, towns, armies, fixed religions etc. • The functional differences we may observe between the modes of communication of hunters/gatherers, agricultural populations and modern industrial societies can help us to imagine the tremendous differences in the functional profile of languages (the pragmatics) which existed 1 million → 500,000 → 100,000 y → 40.000 y → 5.000 y ago.

25. Functional differences and social codes • Functional differences are not necessarily linked to differences of linguistic complexity. Since all human populations living today have the same cognitive level (and the same statistical range of variation in the cognitive equipment) their languages are cognitively at the same level. • In spite of this, the language developed and used can respond to different needs. Once an adult learner has adapted his linguistic competence to a specific profile (of needs), he may have difficulties to readapt to a different profile (characteristic of other societies, cultures, social groups). Normally, the child in the process of socialization and language acquisition can overcome this barrier (if she is exposed in her behavioral training to other communicative profiles). Therefore, one has to differentiate between communi-cative profile and the complexity of grammars. Only the first ones react to cultural and ecological differences.

26. Conclusion • Linguistic functionalism can in principle be grounded by Darwinian principles of selection, although differences in the functional profile do not directly affect the genetic basis of language. As evolution does not stop, all forces, which are operative in the speciation of humans and in the emergence of human language capacity are still there and relevant. The effect of these forces becomes only visible after a very long (evolutionary) time, but many microprocesses of social adaptation and language change operate in the framework defined by evolutionary biology (and with reference to Darwinian selection). • This framing can be neglected as long as the biological/neural architecture and dynamics of language have not been discovered or are not a major concern of linguistic research. With the rise of comparative ethology and neurolinguistics, however, the grounding of linguistic functionalism in the natural sciences and mainly in evolutionary biology became an unavoidable necessity.

27. Some bibliographical suggestions • Bichakjian, Bernard H. (2002). Language in a Darwinian Perspective.Bern: Lang. • Darwin, Charles (1872/1969). The Expression of the Emotions in Man and Animals. [Reprint, 1969]. Culture et Civilisation, Brussels [1st edition London 1872]. • Dawkins, Richard (1994). Das egoistische Gen., Heidelberg: Spektrum Verlag. (English original title: The Selfish Gene.) • Dunbar, Robin (1997). Groups, Gossip, and the Evolution of Language. In: Schmitt et al. (Eds.), New Aspects of Humans Ethology. New York: Plenum Press. • Keller, Laurent (ed), 1999. Levels of Selection in Evolution. Princeton: Princeton U.P. • Labov, William, 2001. Principles of Linguistic Change, vol. 2: Social factors. Oxford: Blackwell. • Martinet, André, 1975. Studies in Functional Syntax. München: Fink.

28. Schleicher, 1863, Die Darwinsche Theorie und die Sprachwissenschaft, Bohlau, Weimar. • Tomasello, M. (1999). The Cultural Origins of Human Cognition. Harvard University Press. • Wildgen, Wolfgang 1977a. Differentielle Linguistik, Entwurf eines Modells zur Beschreibung und Messung semantischer und pragmatischer Variation. Tübingen: Niemeyer. • Wildgen, Wolfgang, 1977b. Kommunikativer Stil und Sozialisation. Eine empirische Untersuchung. Tübingen: Niemeyer. • Wildgen, Wolfgang 1994. Process, Image, and Meaning. A Realistic Model of the Meanings of Sentences and Narrative Texts, Amsterdam: Benjamins. • Wildgen, Wolfgang,, 2003a. Die Sprache – Cassirers Auseinandersetzung mit der zeitgenössischen Sprachwissenschaft und Sprachtheorie, in: Sandkühler, Hans Jörg and Detlev Pätzold (eds.), 2003. Kultur und Symbol. Ein Handbuch zur Philosophie Ernst Cassirers, Kap. 6, 148-174. • Wildgen, Wolfgang, 2004. The Evolution of Human Languages. Scenarios, Principles, and Cultural Evolution, Amsterdam: Benjamins. • Wildgen, Wolfgang, forthcoming 2008. Kognitive Grammatik. Klassische Paradigmen und neue Perspektiven, Berlin: De Gruyter.