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
Classical functionalism

• 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 Martinets 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-langua
  ge (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
Martinets functionalism and Darwinism

• André Martinet turned radically against
  Saussures 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 Trubetzkoys
  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)
• Lauditeur, sil connaît la langue employée,
  fait inconsciemment le tri entre ce qui le
  renseigne sur lidentité du locuteur, ce qui
  linforme de son état desprit 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 dune 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
Darwins paradox (Labov)

• Labov (2001 6-10) compares explicitly his
  observations on linguistic change with citations
  from Darwins book The Descent of Man (1871) and
  he calls the underlying problem Darwins
  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
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 Bühlers
  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 Bühlers triad emerge from ecological
  categories and from social categories already
  apparent in animal behavior.

14
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 Saussures) 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
  ones 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
Dunbars 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
Darwins 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 Martinets 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.
• Martinets 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.
  Darwins last principle.
• The third criterion mentioned by Martinet the
  accumulation and maintenance of traditions has
  an analogue in Darwins 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.