RT list: Oslo: Workshop on metarepresentation, communication & culture

As many of you will already know, there is a workshop on
metarepresentation, communication & culture later this month at the
Centre for the Study of Mind in Nature, at the University of Oslo.
Some details of the programme are below.

For online registration and links to accommodation in Oslo, see the
conference website: http://folk.uio.no/nicholea/metarepcc/

This conference is on 17th and 18th June, right after another
workshop (on metarepresentation and non-literal language use) which
is on 15th and 16th June. That workshop has its own website: http://
folk.uio.no/nicholea/metarep/

I hope to see you in Oslo.

Best,
Nick

Nicholas Allott

Postdoctoral research fellow
CSMN
University of Oslo

www.csmn.uio.no/homepages/nick/

--------------------

METAREPRESENTATION, COMMUNICATION & CULTURE

Programme

Time and place: Jun 17, 2009 09:00 AM to Jun 18, 2008 01:00 PM,
Seminarrom 454, P.A. Munchs hus, Blindern Campus, Oslo, University of
Oslo

Dan Sperber (Institut Nicod, Paris): Overview

Olivier Mascaro (Institut Nicod, Paris): Metarepresentations and
children’s epistemic understanding of deception.
High-level abilities related to the understanding of deception will
be evidenced in three-year-old children: intentional manipulation of
ignorance and beliefs, and the capacity to treat communicated
information as false. I will argue that these capacities involve two
types of metarepresentational ability: the capacity to represent
representations (often seen as developing around four years of age,
here evidenced at three years of age) and the epistemic assessment of
representations (which has not been investigated in young children
before). I will show that despite these capacities, three-year olds
encounter consistent difficulties in handling deception, and will
develop two hypotheses to explain this discrepancy. These analyses
will open up perspectives for the study of mindreading. In
particular, two reinterpretations of standard false belief tasks
(sometimes seen as paradigmatic tests for metarepresentational
abilities) will be presented and will receive original empirical
support.

Fabrice Clément (Département de sociologie, Université de Genève):
What is the minimal cognitive equipment for real trust? The 3-year-
olds' problem

Hugo Mercier (PPE Program, University of Pennsylvania): Reasoning as
a metarepresentational device
Dual process theories are now ubiquitous in psychology. In the field
of reasoning, this dichotomy has been framed as that between
analytic / rule-based mechanisms and heuristic / associative
mechanisms, but no principled explanation has been offered for this
very dichotomy. A possible solution is to divide inferential
mechanisms into intuitive and reflective mechanisms. Reflective
mechanisms are a distinct type of inferences because they pay
attention to reasons (they are reasoning proper). Since reasons are
representations, reflective mechanisms must be a type of
metarepresentational mechanisms. In this presentation, I will develop
a working model of reasoning based on this assumption and try to
explain how reasoning can lead to epistemic improvement.

Christophe Heintz (Konrad Lorenz Institute, Vienna): Inference to
convincing explanation
I will use the argumentative theory of reasoning to clarify why and
how some inferences are guided by the goal of explaining. Lipton
(2005) nicely shows that some important aspects of scientific
reasoning can be accounted for in terms of inference to the best
explanation. However, he characterises inferences to the best
explanation through their semantic and epistemological properties
rather than by specifying their psychological underpinning. I will
show that the argumentative theory of reasoning provides a means to
understand which psychological processes philosophers of science have
referred to when talking about inference to the best explanation.
This will lead me to claim that inferences to the best explanation
are in fact inferences to convincing explanation. With this
psychological characterisation of inferences to convincing
explanation, one is better able to specify the semantic and epistemic
properties of a key inference at work in scientific reasoning.

Nicolas Baumard (Institut Nicod, Paris): What is intuitive in
intuitive theories?
The term “intuitive theory” has been widely used in the past two
decades of cognitive science: intuitive psychology, intuitive
physics, intuitive biology, etc. However, its use has led to
confusion between different kinds of phenomena. Here, I propose to
distinguish between intuitive and metarepresentational or reflective
theories. Intuitive theories are produced by adaptive dispositions,
which appear early in infancy, are universal and may be shared with
non-human animals. For instance, humans (and other animals) have an
intuitive physics, selected by evolution to help them catch and
launch objects. But humans also reflect on questions such as “What is
the trajectory of a ball?” or “Why do objects fall?” These questions
have led humans to construct metarepresentational theories, either by
reflecting by themselves (naïve theories) or by learning cultural
reflections (folk theories). The theory of the four elements or the
Newtonian theory are examples of such folk reflective theories aiming
to answer reflective questions. In contrast, intuitive physics is
only a specialized device, selected to solve adaptive problems, such
as predicting trajectories, without trying to understand the actual
laws of the universe. I argue that many metarepresentational theories
such as impetus, vitalism, teleologism, dualism, or some religious
beliefs have wrongly been labelled “intuitive” by psychologists. I
suggest that there is good evidence that these theories are not
produced by an adaptive disposition and are rather the product of
metarepresentational abilities.

Dan Sperber (Institut Nicod, Paris): Demonstrations
Imitation and communication are the two main mechanisms of cultural
transmission in humans. Imitation, however, is more likely to occur
and to be effective when the behaviour to be imitated is accompanied
by ostensive cues intentionally provided by the model (see the work
of Csibra and Gergely on pedagogy for arguments and evidence). In
such cases, the behaviour to be imitated is not just performed, it is
demonstrated. Much of cultural knowledge is transmitted through
demonstrations. In demonstrations, ostensive stimuli such as
pointing, exaggeration or slowing-down of particular features of the
behaviour to be imitated and explicit instructions indicate and
highlight those aspects of the behaviour that are to be imitated.
Interestingly, these ostensive stimuli themselves are not to be
imitated, and imitators are rarely if ever confused in this respect.
I argue that metarepresentational abilities are at play when
imitation is used intentionally to transmit cultural knowledge.

Christophe Heintz (Konrad Lorenz Institute, Vienna): Is this
cognitive device reliable enough for scientific cognition?
Scientific thinking includes assessing the reliability of cognitive
devices. It requires decisions about when and how a cognitive device
can be used. I will start with the analysis of the historical and
psychological processes through which cognitive functions are
ascribed to *external* cognitive devices. This process will be
illustrated by the history of the four-colour problem. I will then
argue that a similar process is at work in determining which *mental*
cognitive devices can be used for solving specific scientific
problems. Although scientists do not normally think of mental devices
as such, they nonetheless -- as part of their scientific practice --
assess the usefulness of the intuitions that are the output of mental
cognitive devices. During the talk, I will emphasise the role of meta-
representations in the cognitive and historical processes that I will
analyse.

Olivier Morin (Institut Nicod, Paris): Doubting epidemiological
accounts of institutions
Durkheim said that the ideas that are shared within a society are not
social because they are shared: rather, they are shared because they
are social. Epidemiological approaches to institutions (Heintz 2007;
Sperber 2007, 1984 - see also Millikan, 1998) try to reverse this
perspective. Social norms and institutions, it is argued, are nothing
more than distributions of sets of shared representations; what sets
them apart from other shared representations is the fact that their
diffusion is caused by a representation that they contain, a
representation that both represents the distribution and causes it.
For example, a chain letter is a public representation containing
instructions urging me to copy the chain letter and send it to my
friends. These instructions form the metarepresentational and
regulative part of the letter. This regulative representation is
causally responsible for the spread of the letter as a whole.

Chain letters are ideal examples for the epidemiological theory, for
two reasons. First, the causal effects of the regulative
representation (which prompts many people to send the letter to their
friends) tend to match its content (the regulative representation
says that the letter should be sent to friends). Second, these causal
effects of the regulative representation tend to stabilize the chain
letter: to make it more widespread and to preserve its content. In
other words, the normative content of the regulative representation
is fully reducible to its causal effects on the diffusion of the
letter - its epidemiological effect. I will argue that most
institutions do not resemble chain letters: the representations that
govern their distribution cannot be reduced to their epidemiological
effects. The epidemiological approach does not do away with the
normative aspect of regulative representations by stressing their
effects on cultural diffusion. We should not be too quick to decide
whether institutions are social because they are shared, or shared
because they are social.

General discussion

---------------------
Received on Wed Jun 3 21:03:40 2009