Transcript Reflections on the shape of a “protolanguage” and the

From animal communication to
human language.
A comment to the papers of
Jacques Vauclair and
Tecumseh Fitch
Wolfgang Wildgen
Contribution to the Conference: Nature,
Culture and Language: Learning from
Animals?
Essen, December 5, 2005
1
The common base line of the two
authors
 They
are both firmly based on a Darwinian
ground.
 Both refer to new results in brain imaging
(with human and apes) and molecular
genetics (more pronounced in Fitch when
he discusses FoxP2).
 The major concern is: a comparative
behavioral analysis of humans, apes,
mammals, birds, insects.
2
Differences between the two contributions
(from a linguistic perspective)
 Vauclair
rather advocates the hypothesis
of a gestural origin of language. He more
specifically points to the intentional use of
gestures.
 Fitch enumerates five phylogenetic paths
to language. He remains rather skeptical
about the gestural hypothesis and favors:


The hypothesis of a prosodic protolanguage,
A dual stage synthetic theory.
3
Selection scenarios

Vauclair distinguishes two kinds of reinforcement:


Mands > realization to the benefit of the speaker,
Tacts > agreement, sympathy, surprise, laughter.

The latter is more important in the case of
language, the first dominates in other primates.
 Fitch distinguishes: natural selection, sexual
selection and kín-selection. He argues in favor of
the last as a selection factor for language because
low-cost and honest signals are made possible by
“less conflict of interest” in kin-communication.
4
What seems to be rather neglected
(from the perspective of a linguist)



The evolution of man since the australopithecines; mainly
the capacities of homo erectus, homo neander-thaliensis,
and homo sapiens (cf. however the evolution of speech
control in Fitch 2000).
The cultures of primates are discussed, but the cultural
evolution documented by stone industries (since 2 my BP),
“art” since 700.000 BP and Paleolithic cave paintings seem
to be rather irrelevant (or nonconclusive) for the
understanding of language origin.
Only a few linguistic aspects are discussed: protolanguage
(by Fitch), functions of language (injunctive, exclamatory,
informative) and speech acts (mands versus tacts) by
Vauclair and analytic versus synthetic sentence
understanding, functions of formants (in Fitch)
5
Suggestions for Vauclair (from a
linguist’s view point)
Vauclair’s position hand
language could
consider artifacts and art as hand-made semiotic
objects and thus as intermediary between gesture
and language.
 The crucial question: When did phonic language
take over? must be answered. Was it before a
protolanguage with a large vocabulary and a
simple syntax were developed or after that date?
 The fact that language-like gestured systems
depend heavily on the cultural availability of
spoken languages should be considered.

6
Suggestions for Fitch (from a
linguist’s view point)
Fitch’s favored scenarios could lead to an anthropology
of music and a research into holistic patterns in modern
languages. The basic question is: How similar/different
are the semiotic systems : music versus (spoken)
language.
 The complexity measure based on the Chomsky
hierarchy which does only concern the signal-structure
(its syntax) must be replaced by a complexity measure of
semantic schemata (e.g. one based on dynamic systems
theory)
 The possibility that phonetic self-organization is
responsible for signal complexity in syntax must be
further discussed.

7
Remarks on the specification of the
goal: grammar of human languages



Current linguistic theories are rather based on a
longstanding tradition of normative grammar and the
analysis of written language (grammaticality and
competence in Chomsky’s terms).
They have a historical bias towards logical (analytic)
descriptions and lack dynamic or self-organization
models.
As a consequence they may define a mistaken goal of
evolutionary explanation, insofar as the intrinsic relation
to holistic action patterns (frame analysis goes into this
direction) or to multi-channel cognition (visual
imagination, musical structure) are misrepresented in the
standard models.
8
First sketch of an evolutionary
grammar as a possible goal

A system of rules and even one with basic
categories, modules and principles is not able to
map the inherent (and not just parasitic)
developmental, historical and evolutionary
processes.
 The grammatical tradition of normative
grammars, school grammars, competence
grammars falls short of these demands.
 In a more radical move one may even reject the
algebraic, logical models and rather use
formalisms stemming from dynamic system
theory, as the have a genuine dynamic
dimension.
9

A protolanguage must categorize events and
actions (by proto-verbs) and must discriminate
stable entities (by proto-nouns). The question
arises, if temporal, dynamic, quantitative,
qualitative relations between them can be
mastered and to what degree.
 Two basic delimitations of a protolanguage
(already discussed by Bickerton) are:



phrase structure (X-bar)-structures)
government (case-frames).
In my opinion there are intrinsic complexity
barriers which have blocked the elaboration of a
protolanguage for a long (evolutionary) timespan, which separates the proto-species of
Homo erectus and the species of Homo sapiens.
10
Representation of actions and
events




The action-concept GRASP involves two stable entities: Kanzi
(his hand) and the object.
Every point on the lines in Fig. 2 correspond to the perception/
control of a stable entity.
The bifurcation and its environment (the singularity) correspond
to the dynamic event/the relation
The whole schema could ft the sentence: Kanzi took the banana
the banana
Kanzi took the banana
Kanzi
took

Dynamic schema of GRASP.
11
Elaboration
The GRASP-schema can be elaborated in
two basic directions:
 A simpler schema with one entity involved
in the event,
 More complex schemata with three of four
entities involved.
 A hierarchy of ontological levels, which
enable a multiple interpretation of the
schemata (cf. Wildgen, 1994 for details).
12
Basic questions of comparative
biolinguistics
 Are
the levels of grammatical analysis and
the functional distinctions in grammar
adequate for a comparative biolinguistics?
 How can the semantics and pragmatics of
language (context, situation, usage) be
defined in a biolinguistic perspective?
 What were the selection parameters on
language and where do we find their
modern correlates in linguistic behavior?
13
Conclusions




The two papers give a excellent overview and critical
discussion of some of the major fields of a theory of
language evolution. The differences between the authors
point to major not yet resolved controversial issues.
Other interdisciplinary fields more concerned with human
cultures and languages and their results should added.
The proper goal : to understand human language from
an evolutionary perspective, should not just start from
standard proposals in linguistics (mostly using
Chomsky’s authority).
This goal: human grammar must be redefined based on
insights in the debate on language origin, such that other
aspects of human language are put to the frontline and
systematically assessed using the data of living
languages (i.e. linguistic methodology must be adapted).
14
Further references
 Papers
on my homepage:
 http://www.fb10.unibremen.de/homepages/wildgen.htm
 My monograph:
 The Evolution of Human Language.
Scenarios, Principles and Cultural
Dynamics, Benjamins, Amsterdam , 2004
15