|
|
|
|
|
The
seminar is held in hybrid
format, in person (Múzeum
krt. 4/i Room 224) and
online.
________________________________________________________
|
| June 2 (Friday) 4:15 PM Room
224 + ONLINE |
|
Gergely Kertész* and Dániel Kodaj**
|
|
*Institute of Philosophy, Research Centre for the Humanities, Budapest
**Department of General Philosophy, Institute of Philosophy, Eötvös University Budapest
|
| |
|
In defense of teleological intuitions |
|
According to recent work in experimental philosophy, folk intuitions
about metaphysical issues like parthood, persistence, and biological
essences, are heavily teleological. Some authors think that these
intuitions belong to a “benighted view of nature” and therefore we
should mistrust folk intuitions in metaphysics. We argue that
teleological folk intuitions could be veridical. Our argument is based
on an emerging naturalistic theory of biological functions, the “the
Organismic View” championed by Alvaro Moreno, Matteo Mossio and others.
The gist of this view is that biological systems are characterized by a
special circular causal regime where each part of the system contributes
to the boundary conditions of other parts and of the whole.
|
|
| June 16 (Friday) 4:15 PM Room
224 + ONLINE |
Mátyás Lagos
|
|
Department of Logic, Institute of Philosophy,
Eötvös University Budapest
Hungarian Research Centre for Linguistics, Budapest
|
| |
|
Analogical generalisation in formal languages |
Analogical
generalisation is widely believed to play a key role in the acquisition
and processing of human languages: if we have learned that “zorg” is a
verb, then when we encounter the word “zorged” for the first time, we
can guess that it most likely means something like “to zorg in the
past”, by analogy to the relatedness of other pairs of words like
“walk” and “walked”. The two key questions in this talk are: (1) What
might be a suitable explicit, formal definition of the process of
analogical generalisation? And, (2) Given such a definition, is it
possible to actually prove that being able to rely on this process
makes it easier for us to acquire a language?
The language acquisition process can be mathematically modelled using
the notion of “language identification in the limit” (first outlined by
Gold (1967)), where the task of the idealised learner is to infer a
correct grammar for a given formal language after being shown a finite
number of well-formed expressions in that language. This task is
impossible for formal languages in general, but Angluin (1982) has
shown that it is solvable for a certain subclass of languages called
“reversible regular languages”.
Drawing on Angluin’s paper and my own research, in this talk I will
propose a formal definition of analogical generalisation in the setting
of language identification in the limit, and I will investigate whether
being able to use analogical generalisation could be proven to make the
grammar inference task easier.
References:
- Angluin, D. (1982): Inference of reversible languages
- Gold, M. (1967): Language identification in the limit
|
|
|
|
|
