|
Center for Cognitive Science The Puzzle of the Mind |
||||||||||||||||||||||||||||||||
| Fall
2004 Colloquium
|
||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||
|
8 Business Meeting
OCTOBER
17 Jay
Atlas |
Regular colloquia are Wednesdays, 2:00pm - 4:00pm, at 280 Park Hall, North Campus and are open to the public. Refreshments are served. (Calender of Events: Fall 2004) For related CogSci events please go to the Department of Computer Science and Engineering and the Department of Philosophy. If you are interested in receiving email announcements of each event, please subscribe to one of our email mailing lists.
Abstracts
Wednesday,
September 15, 2004 280 Park Hall, North Campus Ingvar
Johansson. "Concepts and Classifications in the Gene Ontology" Since a couple
of years there exists on the web the Gene Ontology (http://www.geneontology.org/index.shtml).
The Consortium behind it says that its goal is "to produce
a structured, precisely defined, common, controlled vocabulary for
describing the roles of genes and gene products in any organism."
In my talk, I will show that this classificatory effort displays
two features of general importance for cognitive science. First,
it shows implicitly how hard it is to keep - in practice - a concept
and its extension distinct. Second, and more importantly, it deviates
from traditional Aristotelian-Linnaean non-evolutionary taxonomies
of animals, plants and dead matter. In the latter, it is required
that subordinate concepts are subsumed under only one superordinate
concept on the overlying level, but the Gene Ontology allows explicitly
a concept to be multiply subsumed; it allows so-called "multiple
inheritances." I will argue that this fact shows the need to
analyze in more detail a seldom noted distinction between "subsumption"
and "specialization." From a purely linguistic-sentential
point of view, this distinction corresponds to the fact that sentences
of the form "P is f?ing" (example: "Paul is running")
can have to two different kinds of relations to sentences that are
more precise and specific. When "P is f?ing" has been
specified by means of an adverb, as in "P is f?ing fast,"
the relation between the sentences corresponds to that of subsumption,
whereas the relation between "P is f?ing" and specifications
such as "P is f-ing on Y" ("Paul is running on the
road") and "P is f-ing at midnight" are claimed to
correspond to that of specialization.
For a printable version of this file click here
Wednesday,
September 22, 2004 280 Park Hall, North Campus Greg
Carlson . "Meaning and Concepts" Meanings of
lexical items are often identified with concepts by many psychologists
and linguists alike. I will agree with this. I will also agree that
meanings of entire utterances are not concepts (at least, not of
the same type). This appears to create a fundamental conflict about
the nature of meanings expressed by language. I am going to argue
that the conflict is apparent. Using evidence from linguistic theory,
this talk is aimed at squaring the two conceptions of meaning by
proposing an integrated system of semantic interpretation which
takes meanings of lexical items to be concepts exactly of the sort
(some) psychologists and linguists say they are and mapping them
into meanings of the type (some) linguists, chiefly formal semanticists,
say they are. The linguistic structures examined most closely are
"weak indefinites" found, to my knowledge, in all languages,
and object-incorporation structures found in many languages. For a printable version of this file click here
Wednesday,
September 29, 2004 280 Park Hall, North Campus Michelle
Gregory "Speech
production in humans and machines:
For a printable version of this file click here Wednesday,
October 6, 2004 280 Park Hall, North Campus Morten
Christiansen "The
Role of Phonology in the
When learning their language children face a difficult ?chicken-and-egg? problem. Discovering the syntactic constraints governing their native language requires being able to assign individual words to lexical categories, such as nouns and verbs. Lexical categories, on the other hand, are only useful for acquisition insofar as they support syntactic constraints. In this talk, I consider how phonological cues in combination with distributional information may be used for solving this "bootstrapping" problem in language acquisition, and the possible consequences that such multiple-cue integration has for adult processing. I report on computational analyses of child-directed speech and connectionist simulations, quantifying the usefulness of phonological and distributional cues, and showing that there are learning mechanisms that can integrate them efficiently. On a theoretical level, these results suggest that multiple-cue integration becomes a crucial part of the child's emerging language system, and thus should also affect adult processing as well. I present results from on-line sentence processing experiments to corroborate this prediction by demonstrating the impact of phonological cues on adult language processing. I conclude that the integration of phonological cues with other types of information is integral to the computational architecture of our language system both in acquisition and adult processing.
For a printable version of this file click here Wednesday,
October 13, 2004 280 Park Hall, North Campus Michael
Webster
"Adaptation and the phenomenology of perception" To
what extent do we have shared or unique perceptual experiences?
I will discuss how the answer to this question is constrained by
known processes of sensory adaptation. Adaptation continuously renormalizes
visual coding according to the stimuli currently before us. These
adjustments have a large influence on how the world looks, and thus
should influence whether it looks the same or different to others.
If two individuals are exposed to and thus adapted by different
environments, then their perception will be normalized in different
ways and their subjective experiences will differ. This will be
illustrated through examples of the effects of adaptation on color
perception and face recognition. When intrinsically different individuals
are exposed to a common environment, their perception will instead
be normalized in common ways and their subjective experiences will
be similar. This will be illustrated through examples of the influence
of adaptation on the perception of image blur. Adaptation may partly
serve to highlight how the present scene differs from the history
of scene properties we have adapted to, and thus much of what we
notice about the world may be a visual aftereffect. For a printable version of this file click here Wednesday,
October 20, 2004 280 Park Hall, North Campus Constance
Clarke "Adapting
to foreign-accented speech: Implications for In
contrast to written language, speech contains a great deal of variability.
The acoustic signal corresponding to a particular word or phoneme
can vary tremendously when different people say it. The sources
of this variability include, but are not limited to, the size and
shape of the speakers vocal tract, the rate of speech, and the speaker's
dialect or accent. Because there are few, if any, consistent acoustic
markers that reliably signal a particular phoneme such as /d/ or
/a/, it is not clear how listeners so easily recover the intended
linguistic information. One way listeners may deal with variability
in speech is by learning about the speech characteristics of the
people they encounter. There is now good evidence that listeners
are better able to understand the speech of someone whose voice
they are familiar with. Listeners also seem to tune their perceptual
criteria to the current talkers characteristics within just seconds
of speech. Unfortunately, theoretical models of spoken word recognition
are lagging behind the experimental findings. Current popular models
assume constant, abstract representations of phonemes and words,
and are unable to explain rapid perceptual learning. In this talk
I will present evidence of rapid adaptation to speech characteristics,
explore some fundamental problems with traditional models of spoken
word recognition, and propose some new directions that may be fruitful
in accommodating both the stable and the flexible aspects of human
speech perception. For a printable version of this file click here Wednesday,
October 27, 2004 280 Park Hall, North Campus David
Mark, Ph.D.
"Ethnophysiography: Recently,
ethnophysiography has been defined as an ethnoscience of landscape.
Ethnophysiography explores the meanings of terms used in various
languages and cultures to refer to the landscape and its components.
Ethnophysiography has objectives similar to ethnobiology, which
studies folk names and categories for plants and animals, but differs
in important ways. Ethnobiology often uses scientific taxonomy as
a baseline for assessing folk categories for plants and animals;
however, variation in landforms and waterbodies is not constrained
by mind-independent natural kinds in any obvious way. Thus ethnophysiography
could contribute to the understanding of categorization in general
by examining categorization of an inorganic natural domain. Ethnophysiography
also can provide a valuable basis for multilingual access to geographic
databases. Examples of differing language-specific conceptualizations
will be presented. These are mainly drawn from a comparison of arid-landscape
terms in an Australian aboriginal language (Yindjibarndi) with terms
and their definitions in English. We will conclude with a description
of plans for extending the work to Native American languages in
Arizona and New Mexico, and with a listing of open problems for
future research. For a printable version of this file click here
Wednesday,
November 17, 2004 280 Park Hall, North Campus Jay
Atlas, Ph.D. "What
Do Reflexive Pronouns Tell Us about Belief? -- Commonsense
intuitions about folk-psychological concepts, linguistic intuitions
about the semantics and pragmatics of 'believes' sentences, and
philosophical views about mental states have resulted in strong
philosophical theses about the nature of belief, conscious belief,
and a Cartesian "transparency" of the mind. I shall look
at logico-linguistic evidence for 'believes' sentences that refutes
a basic principle of self-awareness of one's belief-states that
is defended by philosopher of mind Sydney Shoemaker and many others.
Then I give a philosophical critique that shows why Shoemaker's
arguments for that view should never have been expected to succeed
in the first place. And I suggest some implications for Richard
Moran's use of Moore's Paradox sentences to defend the metaphysical
"inner"/"outer" and the epistemological "objective"/"subjective"
distinctions and to support the transcendence of the "inner."
For a printable version of this file click here Wednesday,
December 8, 2004 280 Park Hall, North Campus William
Badecker ,
Ph.D. "Do
we process grammatical agreement by
For a printable version of this file click here
|
|||||||||||||||||||||||||||||||
You are
visitor
to this
site.
Last updated on Sunday, February 19, 2006 9:37 AM by H. Jones
Contact:
ccs-cogsci-contact@buffalo.edu
The Center for Cognitive Science, University at Buffalo, State University
of New York, 201 Bell Hall, Buffalo, NY 14260
Phone: (716) 645-3180 ext. 125, Fax: (716) 645-3464, Stuart Shapiro, Ph.D.,
Professor and Director.
© Copyright 2004, Center for Cognitive Science, University at Buffalo, All Rights Reserved.