Wolfgang Wildgen The Evolution of Meaning and Discourse

1
Wolfgang WildgenThe Evolution of Meaning and
Discourse

• Cognitive Science
• Case Western Reserve University, 3rd of October
  2007

2
Contents

• 1. Introduction Catastrophic transitions in the
  evolution of life
• 2. From image to concept Thoms concepts of
  salency and pregnance
• 3. An illustrative fragment of catastrophe
  theoretic semantics
• 3.1 Changes in a quality space
• 3.2 The archetype of transfer and trivalent
  scenarios
• 4. From tool-manufacturing to propositional
  semantics
• 4.1 Instrumentality in higher mammals and man
• 4.2 Is tool-making a pragmatic source of
  propositional semantics?
• 5. The evolution of discourse
• 5.1 From ecological space to social pragmatics
• 5.2 From social pragmatics to discourse
• 5.3 A possible hierarchy of discourse functions
• 6. Consequences for an evolutionary grammar

3
Introduction

• Language, the exchange of meaningful messages,
  the systematic reference to a world beyond
  ourselves, the reflection on our use of language
  is a dramatic step beyond the behavior and the
  psychic states of other creatures and beyond the
  material world. It was therefore a major
  challenge for evolutionary thinking first during
  the controversies of the 18th century (Condillac,
  Rousseau, Diderot, Herder), later for Darwin and
  his followers. The central problem concerns the
  apparent perfection of human language and the
  difficulty to explain preparatory stages and
  their adaptive value.
• I will start with a parallel problem due to
  perfection and the lack of transition which
  concerns a much earlier step in evolution the
  evolution of the eye during the so-called
  Cambrian revolution some 500 my ago (cf. Park,
  2005).

4

• Charles Darwin in the Origin of Species (1859)
  treated the eye under the heading ORGANS OF
  EXTREME PERFECTION AND COMPLICATION. He admits
  that the idea that it has been created by natural
  selection is absurd to the highest degree.
• The Cambrian revolution affected only six out of
  38 phyla, but 95 of multicellular animals
  existing today have eyes. Therefore, it is
  probably the most decisive evolutionary step in
  the last billion years. Vision, i.e., the faculty
  to form images of selected aspects of the
  environment, triggered on arms-race which shaped
  bodies, behaviors, enhanced the control and
  perception of motion (a visually guided attack or
  escape) and created the world of colors we
  experience. In a sense, it created a world of
  meanings centered in the brains of animals (a
  kind of virtual reality)

5
Figure 1 The rough evolution of receptors for
different sense organs in geological time
(graphics from Parker, 2005)
6

• I will argue in the next section that the origin
  of language (roughly 2 my BP) produced a
  comparable catastrophic jump which could become
  the basis of evolutionary processes (losses and
  gains) in the future (the next millions of
  years).
• In the case of vision two subfields are
  coordinated and cannot be reduced to one another
• The physics of light/refraction/absorption, etc.
• The psychophysics of perception and the
  neurodynamics of image formation, storage and
  imagination.
• I will argue that with the emergence of language
  a third, non reducible domain is added, cultural
  significance and meaning. Therefore, any theory
  of language has to consider at least three levels
  and the context of their emergence.

7

• The physics of light emergence shortly after the
  Big Bang with the appearance of stars (other
  physical fields relevant for perception follow
  later, thus sound fields for hearing or chemical
  fields for smell and taste).Cf. Guths
  inflationary universe. The inflationary stage
  is followed by a dark age and later the creation
  of stars which emit light.
• The (neuro)psychology of image-formation
  emergence in the Cambrian revolution 500 my BP
  (later auditory gestalts rival with visual ones
  for dominance)
• The cultural significance of sign-behavior and
  meaningful social communication emergence 2 my
  BP (probably both in the auditory and the visual
  mode the auditory becomes dominant in humans).
• The architecture of a theory of language has to
  consider these levels.

8
From image to concept

• The Cambrian revolution created an image making
  machine which became the motor of evolutionary
  diversity and which controlled body shapes, color
  displays, camouflage, mimicry, pursuit and escape
  behaviors, social identification, and social
  cooperation. In a sense, the image-machine became
  the functional heart of higher organized animals
  and it is still at the heart of human behavior
  and culture.
• One may distinguish two factors in these
  dynamics, one is linked to the perceptual organ
  and neuronal centers (e.g., the eye, the visual
  pathway, area 7 and its projections), the other
  is linked to selectional forces, such as feeding,
  hunting prey, protection against predators, and
  sexual reproduction. René Thom (1991/2003) called
  these two forms salency (sensory apparatus)
  and pregnancy (biological needs). The major
  criteria of distinction are

9

• Thom compares subjective pregnancies found in
  animals (and man) with objective pregnancies
  (forces) in nature. Linguistic meaning (concepts)
  which are necessary to form propositions by
  predication are at a point of convergence between
  biological pregnancies, natural forces, motion
  patterns and geometric forms

10
subjective
objective
free isotropic propagation diffusion
temperaturesoundchemical diffusion
biological pregnances
odor, taste, touch
physical fieldse.g. lightstate transitions
controlled propagation
color
valence-patterns
motion of solid bodies
constrained propagation
concepts words and syntax
phonetic gestalts
no propagation
geometrical forms
written words
11
The Lower Paleolithic revolution

• In humans and mainly in speaking humans this
  strict dependency on biological pregnancies
  disappears and the question is How did the
  propagation of meaning become to a large extent
  independent from basic survival mechanisms?
• As we assume that this transition was prepared in
  the stage of Homo erectus (2 my BP), biologically
  fixed with the speciation of Homo sapiens (ca.
  300.000 y BP) and fully unfolded since the
  populations of cave painters (ca. 40.000 y BP),
  we can call it the (Lower) Paleolithic
  revolution.
• This does not mean that the creation of meaning
  lost its biological significance but the Rubicon
  between a biologically controlled meaning
  propagation and a cultural control and social
  embedding of meaning has been crossed in this
  period (2 my-40.000 y BP).

12
The catastrophic transition to language
Non-language
Language
Communicative capacities
2 my BP to 40.000 y BP
Evolutionary time scale since the Cambrian
revolution
13
Thoms conjecture

• René Thom conjectured that the lexico-syntactic
  valences described by Tesnière (1959) or the
  case-frames enumerated by Fillmore (1968) are
  basically a reflection of restrictions imposed on
  natural processes. This hypothesis underlies
  catastrophe theoretic semantics (cf. Wildgen,
  1982). In semiotic terms, the relational
  architecture underlying language has a foundation
  in natural laws, or more provocatively, the
  archetypical architecture of linguistic
  utterances (sentences) is rooted in natural laws,
  it is an icon of the real world in which human
  beings live.
• As a corollary this explains why humans endowed
  with language are able to discover natural laws,
  use them for technology and control the ambient
  world which for all other beings, including
  non-human hominids, is opaque and just an
  all-mighty force which beings must endure
  passively.

14
An illustrative fragment of catastrophe theoretic
semantics

• The valence pattern is globally described as a
  conflict of pregances in Thom (1978c 76). If
  these conflicts are stripped off their specific
  intentional and real-life content, a formal
  topologico-dynamic pattern is left, which can be
  matched against the hierarchy of elementary
  catastrophes in Thom (1972). These archetypes
  (cf. for elaborations, Wildgen, 1982) are pure
  theoretical entities, which allow the formulation
  of a family of interesting hypotheses. Like the
  theoretical terms used in physics, they formulate
  a program of empirical research, such that some
  of the hypotheses formulated in these terms may
  be elaborated or falsified (cf. Wildgen, 1994,
  for relevant elaborations and corrections of the
  theoretical conjectures generated by Thom).

15
(a) to distort, to bend (German verbiegen) scales
qualitative scale _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - straight twisted, crooked
(b) to clean ( German reinigen)
qualitative scale - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ dirty clean, neat
If we assume a linear space with two poles we can
describe the process contained in the two verbs
above in the way shown in Figure 4. The curved
surface above describes the states of stability
and instability (the attractors and the repellors
of the system). The process makes a catastrophic
jump from one partial surface to another (e.g.
from 'dirty' to 'clean').
16
The archetype of transfer and trivalent scenarios
17
Evolutionary explanation of Thoms conjecture

• Thoms assumption is on one side tempting,
  because it could provide a much deeper foundation
  of linguistics than any current theory, but on
  the other side it cannot explain how human
  language could become a mirror of natural
  processes. In Wildgen (2004) I proposed a
  transition mediated by tool making and early
  technologies. In fact, language has not emerged
  in isolation, it rather came together with other
  symbolic forms (a term coined by Cassirer) like
  myth (ritual), art and technology. Lithic
  technologies used since more than 2 my could
  stand for a first stage (possibly in the context
  of rituals, an early proto-language and
  body-painting).
• Insofar as such technologies asked for a precise
  control of natural forces, human symbolic
  behavior was at the start parallel to a kind of
  scientific insight and corresponding conceptual
  elaborations.

18
From tool-manufacturing to propositional
semantics

• In the evolutionary line of primates, tool-use is
  reported both for new world apes and old world
  apes. The first show only the behavior of
  throwing objects (from above down to the bottom
  of trees) in attack and defense, whereas the
  second show a higher diversity of tool uses (cf.
  Becker, 1993 79-110). Rather sophisticated
  tool-use with beginning tool modifying is
  reported by Boesch (1993), who describes the
  nut-cracking behavior of wild chimpanzees of the
  Taï National Park (Côte dIvoire). The animals
  transport both nuts and hammers to roots, which
  are used as anvil. As stone hammers are rare and
  necessary to crack very hard nuts (Panda oleosa),
  they are transported and preserved. Wooden
  hammers may be shortened using fallen branches
  until they fit. Infants must learn the use of
  tools and different ways of passing on the proper
  method of use have been observed stimulation,
  facilitation, and active teaching.

19
Is tool-making a source of propositional
semantics?

• Basic script of tool manufacturing
• Seeking for materials.
• Using both hands, such that one hand fixes the
  material, which has to be shaped, and the other
  controls a tool used for shaping.
• The tool is adapted to specific contexts it
  becomes the blade of a knife, the point of an
  arrow, the body of an ax, etc., or it is used to
  perform one phase of a process, e.g., cleaning
  the fur of an animal the fur is already the
  result of a longer goal-oriented process
  beginning with the hunting of the animal. The
  mastering of tool-production allows the
  production of cultural objects and art these may
  become objects of value. Elaborated tools and
  objects of art show geometrical abstraction and
  iconicity.
• A further stage produces pictures (signs) of the
  hand, the instrument which shapes tools.

20
Simple pebble tools from the Olduwan-gorge
(around 2 my BP)
Refined and small tools from the
Magdalenian-culture (around 16.000 y. BP)
21
The evolution of discourse

• The pragmatics of action with hands establishes a
  micro-level of emerging functions which elaborate
  the relation between cause and effect.
• At the macro-level human housing and
  house-building is a domain where structures
  emerge, which can be reorganized in the shape of
  space-oriented communication, linguistic
  orientation in space and memory of narrative
  contents related to space.
• The background of these processes is given by the
  ecological/situational context. Some objects or
  context features become culturally significant.
  These are mainly
• places (of living, of chase, etc.),
• tools and the techniques of their use,
• motion patterns, gestures, gestured signs, dance,
  
• The relevance of places (in space and time), of
  spatial orientation and categorization are of
  primordial importance for the semantics of
  natural languages as the tradition of localistic
  theories shows.

22

• Already in the last common ancestor of humans and
  chimpanzees (LCA), contextual space acts as an
  external memory of affordances, which is
  indexically given by paths (of social locomotion
  and predator/prey-locomotion), harvesting
  locations (and times), dangerous locations,
  places for sleep, courtship, housing, frontiers
  of territories, etc. These indexically loaded
  areas and places function like a catalyst of
  social action, insofar as they can coordinate
  social perception and action.
• As soon as space is more specifically organized
  in relation to cognition and social use, it
  unfolds in a cycle of social investment.
  Architecture and the spatial organization of a
  village (or later a town) are clear examples.
  This level is autocatalytic insofar as the
  spatial organization becomes itself a cyclic
  structure in which different functions cooperate.

23
Semiotically invested subspaces
housing
fire place
myth. space
public space
ritual
tool making
outside
chase, harvest
24
From social pragmatics to discourse

• Tool manufacturing, body art and rituals may be
  either preadaptations enabling the emergence of
  language or already be parallel and fostered by a
  protolanguage. Due to the non-permanent nature of
  spoken language, we have no chance to check which
  of the alternatives is valid. The fact that
  language usage is primarily a social
  communicative phenomenon encourages the search
  for a cultural/social origin of language and
  discourse. In this perspective it is not
  predication or propositional structure, but
  discursive processing in social contexts which
  must be foregrounded.
• Therefore, one must ask, if discourse functions
  like narrative, descriptive, argumentative or
  ritual discourse had a survival value in early
  human populations (before hominisation), which
  differs sharply from the survival patterns in
  chimpanzees and other primates which did not
  evolve a linguistic capacity.

25
A possible hierarchy of discourse functions

• Classical speech-act theories placed the
  proposition (and its elocution) at the center and
  added illocution and perlocution. The perlocution
  (the impact on the audience) and its social
  effects are neglected. The relation of language
  use to its contextual evaluation and thus to its
  selective relevance is excluded. An evolutionary
  account must start from perlocutionary effects,
  like A persuades / convinces B (via an
  utterance), A evokes positive feelings / gets
  help /in/by B (via an utterance) A contributes
  verbally to the solution of a problem / teaches /
  helps to find a solution (via an utterance).
• If the perlocutionary effect is increasing the
  fitness of the group, such a feature (and the
  underlying faculty) can be selected. As no other
  human species with lesser communicative faculties
  exists, it is impossible to test the selective
  advantage our species got and why. The only
  approach which is feasible concerns the analysis
  of actual discursive effects.

26

• If the scout can describe the place and number of
  a herd of bison accurately, the group will follow
  him and bring food to the clan, which will not
  starve and thus survive.
• If the experienced warrior can give a good story
  for his undertakings others will join a new
  enterprise and learn from his experience how to
  overcome the enemy.
• If the perpetrator can defend his cause
  effectively he will not be expelled or killed.
• These examples show three different discourse
  functions descriptive, narrative and
  argumentative.

27
Discourse as basic achievement

• My hypothesis is that rhetoric (and possibly
  poetic) functions stood at the beginning when
  discourse emerged. I will only discuss he
  narrative function in the folowing.
• The central concern in the narrative is
• How can a sequence of events/actions be
  segmented/compressed into sentences?
• How are these arranged such that not only the
  temporal sequence can be derived but also spatial
  itineraries and causal effects can be imagined or
  reproduced? The problem concerns a mapping of
  time, space and cause/force in a text such that
  an easy and reliable understanding by the
  audience is made possible.

28
Evaluative and relevance functions

• This is, however, not sufficient. In each
  narrative text, evaluative and relevance
  assigning processes have to be controlled. As
  Labov (1972) has shown, the Abstract/Title must
  sketch the relevance of the story which will be
  told, the Climax separates the Complication and
  the Result and spans an arch of interest for the
  audience. In many cases, self-evaluative
  information is distributed over the story, etc.
  Thus even simple stories contain two components
• Time/space/force mappings
• A socio-evaluative profile or a relevance
  component in which social values are exchanged
  (respecting the audience and self-advertisement)

29

• These factors point into the direction of a
  twofold functionality of (narrative) discourse.
  It has a referential function (mapping a sequence
  of events/actions) and a socio-evaluative
  function. In the emergence of (narrative)
  discourse, two different selective processes must
  have cooperated. If in small talk the
  socio-evaluative component dominates, this does
  not mean (as Dunbar suggests) that discourse
  emerged from social contact (grooming).
• The two factors have probably different
  evolutionary histories. The referential function
  elaborates cognitive functions already developed
  since the Cambrian revolution (helped by bigger
  brain, which was made possible by high energy
  food and allowed for the construction of
  sophisticated tools). A further function is based
  on the evolution of social groups and their
  organization and more specifically of
  cooperative/competitive processes in dense social
  networks.
• The key to the solution, the social organization
  of human populations 300.000 y ago and that of
  neighboring human species in competition with
  them is not accessible empirically.

30
Consequences for an evolutionary grammar

• A grammar is called evolutionary, if its
  architecture reflects the order in which
  important linguistic features emerged and
  respects the natural (causal) relations between
  components which were selected at different
  stages (e.g., 2 my, 500.000 y, 100.000 y,
  50.000 y, 5.000 y BP).
• In conclusion of the facts and hypotheses exposed
  in this lecture, a grammar should first consider
  the cognitive basics, i.e., the mapping of space,
  time, force (cause) into a language.
• Secondly, it should pay attention to discourse
  organization in relation to social functions of
  language.

31
Self-organization and the arbitrariness of
languages

• The basic factors which shaped human language led
  to numberless but functionally equivalent
  individual languages/dialects/jargons/repertoires,
  etc. This feature was called the arbitrariness
  of the linguistic sign by Ferdinand de Saussure.
• In reality it is only the effect of multiple
  processes of self-organization which fulfill the
  basic cognitive and social functions. As the set
  of concepts grows, and at the same pace the
  length of utterances, the fine-grained structure
  of languages is only grossly constrained by the
  basic functions. Internal measures of economy and
  optimality select one or several solutions and by
  a law of conservation the system stops the search
  for other solutions.
• The differences between languages are the
  out-come of a process weakly constrained by the
  basic functions and selected by mechanisms of
  self-organization, which allow for many
  equivalent solutions.

32
Some bibliographical hints

• Labov, William, 1972. The Transformation of
  Experience in Narrative Syntax, chapter 9 of
  Labov, W., Language in the Inner City. Studies in
  the Black English Vernacular. Philadelphia
  University of Pennsylvania Press, 354-396.
• Parker, Andrew, 2005. Seven Deadly Colours. The
  Genius of Natures Palette and How it Eluded
  Darwin, Free Press, London.
• Thom, René, 1990. Semiophysics a sketch,
  Addison-Wesley, Redwood City, Calif
• Wildgen, Wolfgang, 1982. Catastrophe Theoretic
  Semantics. An Elaboration and Application of René
  Thoms Theory, Benjamins, Amsterdam/Philadelphia.
• Wildgen, Wolfgang, 1994. Process, Image, and
  Meaning. A Realistic Model of the Meanings of
  Sentences and Narrative Texts, Benjamins,
  Amsterdam/Philadelphia.
• Wildgen, Wolfgang, 2004a. The Evolution of Human
  Languages. Scenarios, Principles, and Cultural
  Dynamics, Benjamins, Amsterdam/Philadelphia.