Mental representation, communication and the transition from

Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Mental representation, communication and
the transition from animal to human
Marieke Schouwstra
UiL OTS, Utrecht University
July 21st, 2008
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Introduction
Introduction
It is sometimes said that animals do not talk, because they lack
the mental capacity. And this means: “they do not think, and
that is why they do not talk.” But—they simply do not talk.
Wittgenstein, Philosophical Investigations, 25.
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Introduction
Introduction
!
(Philosophical) accounts for describing human mental
representation and human communication generally use
concepts that only apply to humans (and not to animals)
!
These concepts are not useful in the debate about the
emergence and evolution of language and cognition
Recently: conceptual frameworks that do account for the
transition from animal to human
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Introduction
Two conceptual frameworks
!
Distinction between cued and detached representation
(Gärdenfors)
!
Pushmi/pullyu representations (Millikan)
Plan:
!
Analyze these concepts and formulate main differences
!
Evaluate them using empirical data
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Introduction
On the transition from animal to human
!
animals and humans
!
communication and mental representation
!
How do the two accounts apply their concepts to these
species?
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Introduction
On the transition from animal to human
!
Not closely related to humans
!
But: relatively complex
communicative behavior: honeybee
dance
!
“Most complex symbolic sytem
decoded in the animal world” (Crist
2004)
!
“One of the seven wonders of animal
behavior” (Gould and Gould 1995)
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Introduction
On the transition from animal to human
!
Not closely related to humans
!
But: relatively complex
communicative behavior: honeybee
dance
!
“Most complex symbolic sytem
decoded in the animal world” (Crist
2004)
!
“One of the seven wonders of animal
behavior” (Gould and Gould 1995)
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Introduction
On the transition from animal to human
!
Genetically closely related to
humans
!
But: quite different from humans in
linguistic capacity
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Gärdenfors
Cued and detached representation
[Gärdenfors(1995)]: two kinds of mental representation
Cued representation
Representation of something that is present in the current
external situation of the representing organism.
Detached representation
Representation of objects/events that are neither present in the
current situation, nor triggered by some recent situation.
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Gärdenfors
Mental representation from animal to human
animal → human
Transition from cued to detached representation
!
The distinction is not a sharp one: degrees of detachment
!
Representations with a detached flavor present in animals
!
Distinction is applicable to the mental realm; not to
communication
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Gärdenfors
Communication from animal to human
!
Signals
!
!
!
refer to cued representation
≈ refer to something in the outer environment
Symbols
!
!
refer to detached representation
≈ refer to something in the inner environment
→ language presumes something like an inner environment
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Gärdenfors
Honeybees in Gärdenfors’s framework
!
The honeybee’s communication is
compositional
!
But it is still in a cued manner
!
dance = signal
!
refers to cued representation of nectar location
!
bees don’t communicate about food without actually
coming from a food source
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Gärdenfors
Primates in Gärdenfors’s framework
!
The dimensions of speech and
thought are independent
!
One needs an inner environment in
order to have symbolic
communication
!
It is possible to have many cognitive
functions without having a language
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Millikan
Pushmi Pullyu Representations
!
“PPR’s tell in one
undifferentiated breath both
what is the case and what to
do about it” (Millikan 2004)
!
highly inarticulate; for very
specific tasks
!
communication and mental
representation
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Millikan
Pushmi Pullyu Representations
Examples:
The hen’s foodcall
The mental state of a bird by the sight of
the open beaks of the young birds
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Millikan
Pushmi Pullyu Representations
animal → human
Disassemblance of PPR’s into their directive and descriptive
parts (and recombination of these parts)
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Millikan
Honeybees in Millikan’s framework
!
The dance puts across two
messages at once:
1. There is nectar at X
2. Go and get the nectar
!
(but message 1 and 2 are not
articulated!)
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Millikan
Primates in Millikan’s framework
!
Language is constitutive of
developed human thought
!
“A very large portion of our
conceiving is done mainly or entirely
through the medium of language”
(Millikan 2005, p. 104)
!
→ Since primates do not have
language, their cognition is of a
different order than human cognition
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Millikan
Humans in Millikan’s framework
!
descriptive and directive information
separately
!
many representations that are not directly
useful
!
‘inner articulation’ of mental representations
but PPR’s still present too:
!
!
!
Eye blink
‘We don’t eat peas with our fingers!’
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Analysis
Common emphasis
detachment from here and now (and directly necessary)
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Analysis
Differences: honeybee dance
Millikan
One primitive PPR that directs and describes
Gärdenfors
Signal that refers to cued representation
!
In Gärdenfors, the directional part of the bee’s message is
‘inferred’ by the receiving bee
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Analysis
Differences
!
Millikan: mental capacities at least partly dependent on
linguistic capacities
!
Gärdenfors: complex mental capacities possible without
linguistic capacities
In an evolutionary setting:
!
Gärdenfors: development of cognitive capacities first, then
language
!
Millikan: development of certain cognitive capacities only
after development of language
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Analysis
Generalizing. . .
!
What was the order of things in the course of evolution?
!
Private domain: cognitive functions, mental representation
!
Public domain: communication and language
!
→ private-initial
!
→ public-initial
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Analysis
Private-initial and Public-initial
private-initial
much of cognition was in place before language emerged
!
Also defended by Jim Hurford (e.g. in Hurford 2007)
public-initial
the emergence of more advanced communication forms has
boosted the development of thought
!
See also W.F. Harms (Harms 2004)
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Homesign data
Empirical evidence
!
The two conceptual frameworks correspond to different
hypotheses about the order of emergence of sophisticated
cognition and sophisticated communication
!
which framework is more useful → which prediction is
more credible
!
In order to see that, we can test the predictions against
empirical data
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Homesign data
Homesign as empirical evidence
!
Deaf children that grow up in hearing families
!
No use of conventional sign system
!
These children develop improvised sign systems
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Homesign data
Displaced reference in homesign
Displaced reference
!
The property of language that it can be about things that
are not here and now
!
‘Core feature’ of language
Observation
All children that develop a homesign system display displaced
reference
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Homesign data
Displaced reference in homesign
Examples
!
Requesting a bottle
!
Remarking that the baseball game wouldn’t have been
cancelled if it hadn’t rained yesterday
Conclusion (Morford et al. 1997)
“A conventional language model is not necessary for children to
be able to extend their communication beyond the here and
now”
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Homesign data
Conclusion
!
At first sight: support for private-initial
But. . .
!
What exactly is ‘a conventional language model’?
!
Is there really no input for the homesigning child?
!
How exactly do we proceed from homesign data to general
conclusions?
!
What are the underlying assumptions?
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Homesign data
General conclusion
!
New theoretical frameworks are necessary to provide
better perspectives on the similarities and differences
between humans and other animals
!
The frameworks I have presented have different underlying
assumptions about the order of things in the course of
evolution
!
‘Choosing sides’ should happen on the basis of empirical
data
!
Going from the data to the frameworks is not
straightforward
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Homesign data
A view on conceptual frameworks
Necessary in the future:
!
Continuous interaction between theoretical and empirical
approach
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht
Introduction
Gärdenfors
Millikan
Analysis
Homesign data
Homesign data
References
Gärdenfors, P. (1995).
Language and the evolution of cognition.
In Lund University Cognitive Studies 41.
Gärdenfors, P. (1996).
Cued and detached representations in animal cognition.
Behavioral Processes, 36, 263–273.
Gärdenfors, P. (2004).
How Homo Became Sapiens: On the evolution of thinking.
Oxford University Press.
Hurford, J. R. (2007).
The Origins of Meaning.
Oxford University Press.
Millikan, R. G. (2004a).
On Reading Signs: Some Differences between Us and the Others, pages 15–30.
MIT Press, Cambridge, MA.
Millikan, R. G. (2004b).
Varieties of Meaning.
MIT press.
Millikan, R. G. (2005).
Language: a Biological Model.
Oxford University Press.
Morford, J. and Goldin-Meadow, S. (1997).
From here and now to there and then: The development of displaced reference in homesign and english.
Child development, 68(3), 420–435.
Mental representation, communication and the transition from animal to human
UiL OTS, Universiteit Utrecht