Transcript On the Origin of Language

Origin of Language: the Hardest
Problem in Science?
Eörs Szathmáry
Collegium Budapest
Eötvös University
The major transitions
(JMS & ES, 1995)
*
*
*
*
* These transitions are regarded to be ‘difficult’
Why is language so interesting?
• Because everybody knows that only we talk
• …although other animals may understand a
number of words
• Language makes long-term cumulative
cultural evolution possible
• A novel type of inheritance system with
showing “unlimited hereditary” potential
What is so special about human
language?
• Basically, it is the fact that we make
sentences using grammar
• Languages are translatable into one another
with good efficiency
• Some capacity for language acquisition
seems to be innate
• THE HOLY GRAIL IS THE
EMERGENCE OF SYNTAX
Understanding language
evolution is difficult
Three interwoven processes
• Note the different time-scales involved
• Cultural transmission: language transmits itself as
well as other things
• A novel inheritance system
Language is not Weismannian
soma
germ
soma
germ
protein
DNA
germ
sentence
Neural
germ
representation
protein
DNA
sentence
Neural
representation
Design features of language
• Compositionality (meaning dependent on how
parts are combined)
• Recursion (phrases within phrases)
• Symbolicism (versus icons and indices)
• Cultural transmission (rather than genetic)
• SYMBOLIC REFERENCE and SYNTAX
A simple experiment (Hauser &
Fitch)
• Finite state grammar
(AB)n is recognizable
by tamarins
• Phrase structure
grammar AnBn is
NOT.
• Humans recognize
both
Our evolutionary relatives
What has happened on our linage in the past few
million years so that our genes allow for the
development of a brain that can sustain syntax?
Words are symbols, Saussurean
signs
Object
Concept
TREE
Symbol
Word representation is
distributed…
…and is related to the somatosensory handling of the
designated object
Principles and parameters
• Principle: a universal property of human
language, assumed to be innate.
• Parameter: a two (or more) valued choice
determining a general property
distinguishing one type of language from
another.
Syntactic processes and
information
• Colourless green ideas sleep furiously
• Structure building (phrases, etc.)
• Checking agreement (e.g. in German noun
phrases must be marked for case)
• Mapping thematic roles (John loves Mary,
Mary loves John)
• Complexity (the dog was chased by the cat)
• SYNTAX IS NOT WORD ORDER!!!
The D- and S-structures
• The sentence is: Mary
was chosen
The traditional view
• Broca’s area: the “seat
of syntax”
• Wenicke’s area: the
seat of semantics
(fluent aphasia)
• Double dissociation
• Unfortunately (?) not
quite true
New data on Broca
• One can have syntactic deficit with intact Broca
• Affected Broca does not always produce problems
in morphosyntax
• Some Broca aphasics have problems with
semantics as well
• Broca lesion neither necessary nor sufficient for
syntactic deficit
• BUT may be essential for COMPLEX sentences
(a problem with working memory?)
Neuroimaging studies of syntactic
processing
• By comparing syntactically complex to
simple sentences
• By comparing sentences to lists of unrelated
words
• By comparing sentences containing nonreal words to normal ones
• Comparing sentences with syntactic
violation to those without
Semantic and syntactic violations
Syntactic violation versus
• Correct sentences
• Semantic violation
• Other violation
• Semantic violation versus
• Correct sentences
• Syntactic violation
Where is syntax in the brain?
• In many areas
• These include some parts of the RIGHT
hemisphere
• None of these areas is exclusively dedicated
to syntax
• Broca: semantics phonology, memory,
music perception
• INCONSISTENT WITH A STRICTLY
ANATOMICAL MODULAR VIEW
Resolution (Kaan & Swaab, 2002)?
• Maybe there is a dissociation at the cellular
level between these functions, below
resolution
• Maybe the combination of these areas forms
a unique network
• Different parts of the network are recruited
to different syntactical tasks
• MAYBE, BUT WHY NOT IN APES?
An even more radical resolution:
The Language AmoeBa (LAB)
hypothesis
• Szathmáry, E. (2001) Origin of the human
language faculty: the language amoeba
hypothesis. In (J. Trabant & S. Ward, Eds.):
New Essays on the Origin of Language.
Berlin/New York: Mouton/de Gruyter, pp.
41-51.
Recuerdos de mi vida (Cajal, 1917,
pp. 345–350)
“At that time, the generally accepted idea that the
differences between the brain of [non-human]
mammals (cat, dog, monkey, etc.) and that of
man are only quantitative, seemed to me unlikely
and even a little offensive to human dignity. . .
but do not articulate language, the capability of
abstraction, the ability to create concepts, and,
finally, the art of inventing ingenious
instruments. . .
seem to indicate (even admitting fundamental
structural correspondences with the animals) the
existence of original resources, of something
qualitatively new which justifies the
psychological nobility of Homo sapiens?. . . ’’.
Species-specific differences in
cortical microstructures do exist
Differences in the primary visual
cortex among primates (Preuss et al)
In monkeys: the
honeycomb
Modifications in
evolution
The difference in gene expression
patterns
• Despite our close
genetic relationship to
chimps
• The epigenetic
difference in the
brains seems
enormous
The evolutionary approach
genes
selection
development
learning
behaviour
environment
Impact of evolution on the developmental
genetics of the brain!
Crucial facts for LAB
•Localisation of language is not fully genetically
determined: even large injuries can be tolerated before a
critical period.
•Language localisation to certain brain areas is a highly
plastic process, both in its development and its end
result.
•It does seem that a surprisingly large part of the brain
can sustain language: there are (traditionally
recognised) areas that seem to be most commonly
associated with language, but by no means are they
exclusive, either at the individual or the population
level, during either normal or impaired ontogenesis.
•Whereas a large part of the human brain can sustain
language, no such region exists in apes.
Crucial theses of LAB
•The language amoeba is the neuronal activity
pattern that essentially contributes to processing of
linguistic information, especially syntax. It is a
dynamical manifestation of Chomsky’s language
organ, as it were
•An appropriate and rather widespread connectivity
pattern of the immature human brain renders it a
potential habitat for the emerging language amoeba.
•This condition does not require too many altered
(probably regulatory) genes, but there are great risks
involved, which make this “major transition”
difficult indeed.
Variation and selection in neural
development
• Changeux’s version
• There is vast
overproduction of
synapses
• Transient redundancy is
selectively eliminated
according to functional
needs
• The statistics and the
pruning rules for the
network architecture are
under genetic control
The structure of the visual system
Partial crossing at the chiasm allows for
stereoscopic vision
Development of the columns of
ocular dominance
• The initial overlap
decreases with time
• Visual input is
NECESSARY for
columnar development
Genes and visual input make up
for normal vision
• Synapses are pruned
during development
• A blindfolded eye
does not send sensory
information to the
cortex
• It’s column shrinks to
negligible size
• Reversible within the
CRITICAL PERIOD
The FOXP2 gene is mutant in a
family with SLI
• SLI: specific language impairment
• In the KE family the mutation is a single
autosomal dominant allele
• Another individual has one copy deleted
• TWO intact copies must be there in
humans!
• The mutation affects morphosyntax:
Yesterday I went to the church
Possible regulatory modes of the
FOXP2 gene
Interpreting the nature of SLIrelated conditions
• Sometimes SLI affects
specifically grammar
• Sometimes if affects other
linguistic functions
• Sometimes several other
functions are affected
• Outcome must depend on
the region of expression of
the (genetic) disturbance
in the developing brain
Nucleotide substitutions in the
FOXP2 gene
• Bars are nucleotide substitutions
• Grey bars indicate amino acid changes
• Likely to have been recent target of selection
Coevolution of the language and
the brain
• An old idea (Wilson):
increased brain size leads
to more complex
behaviour
• Which in turn, due to
increased environmental
complexity, selects for
increased brain size
• Another crucial
component: genetic
assimilation
Rapoport’ scheme applied to
language
One method of finding out (within
ECAgents)
•
•
•
•
Simulated dynamics of interacting agents
Agents have a “nervous system”
It is under partial genetic control
Selection is based on learning performance
for symbolic and syntactical tasks
• If successful, look and reverse engineer the
emerging architectures
Between linguistic input and
output…
Transmission dynamics in
simulated agents