2.–4. 4.
|
Remi van Trijp
|
 |
Sony Computer Science Laboratory Paříž
Odborné zaměření: |
| 2. 4. / 17:20 / č. 300 |
1. Evolutionary Linguistics and Fluid Construction Grammar |
| How do languages emerge? How do they evolve? These questions are among the great unsolved mysteries of mankind, and they form a fundamental scientific challenge.
This first lecture introduces Evolutionary Linguistics, a young but thriving research field that investigates the cognitive mechanisms and cultural processes that underlie language evolution. A theory of cultural language evolution rests on two biologically inspired concepts that are applied to the domain of language: (a) linguistic selectionism, which involves processes that create variation in a population (e.g. reanalysis, schematization) and processes that select some variants to become dominant (e.g. communicative success, cognitive effort, social conformity); and (b) self-organization, which occurs when speakers and listeners align their communication systems during their linguistic interactions, and which explains how an entire speech community may converge on novel linguistic conventions without central control. This lecture also introduces Fluid Construction Grammar (FCG), which embodies the grammatical component of a theory of cultural language evolution. FCG has been explicitly designed to allow linguists to write a computationally precise account of linguistic knowledge, processing, acquisition and evolution. It is available as an open source software tool at www.fcg-net.org. Background reading: – Steels, L. (2011). “Introducing Fluid Construction Grammar.” In: Steels, L., editor, Design Patterns in Fluid Construction Grammar. Constructional Approaches to Language, 11. Amsterdam: John Benjamins.
– Steels, L. (2011). “Modeling the Cultural Evolution of Language.” Physics of Life Reviews 8, 339-356.
– Steels, L. (2012). “Self-Organization and Selection in Cultural Language Evolution.” In: Steels, L., editor, Experiments in Cultural Language Evolution. Amsterdam: John Benjamins.
– Steels, L. and Loetzsch, M. (2012). “The Grounded Naming Game.” In: Steels, L., editor, Experiments in Cultural Language Evolution. Amsterdam: John Benjamins.
Available at www.emergent-languages.org together with the other chapters of the book. |
|
| 3. 4. / 15:50 / č. 300 |
2. Language Games as an Integrating Experimental Paradigm |
| Evolutionary linguistics hypothesizes that the key to explaining “language systems” (i.e. grammatical paradigms) is to understand their function in communication. Taking a cognitive-functional view, a language system can be considered as a particular solution to a complex problem, so explanations need to take into account the nature of the problem, the biological and cognitive endowment of the language user, the social relations between language users, and the context in which communication takes place. This second lecture focuses on an experimental paradigm called “Language Games” that integrates all these aspects of language usage.
Language games involve embodied language users that communicate with each other about real-world actions and objects. In order to be successful at the game, a complete processing model needs to be implemented, including sensorimotor processing, conceptualization and interpretation, and production and parsing. The first part of this lecture illustrates the paradigm through a computational reconstruction of German for the domains of space and argument structure. The second part of the lecture shows how these processing models can then be exploited for explaining attested cases of language change. Background reading: – van Trijp, R. (2011). “A Design Pattern for Argument Structure Constructions.” In: Steels, L., editor, Design Patterns in Fluid Construction Grammar. Constructional Approaches to Language, 11. Amsterdam: John Benjamins.
– Spranger, M. and Loetzsch, M. (2011). “Syntactic Indeterminacy and Semantic Ambiguity: A Case Study for German Spatial Phrases.” In: Steels, L., editor, Design Patterns in Fluid Construction Grammar. Constructional Approaches to Language, 11. Amsterdam: John Benjamins.
– Steels, L. (2012). “Grounding Language through Evolutionary Language Games.” In: Steels, L. and Hild, M., editors, Language Grounding in Robots. Heidelberg/Berlin: Springer.
– van Trijp, R. (in press). “Linguistic Assessment Criteria for Explaining Language Change: A Case Study on German Definite Articles.” To appear in Language Dynamics and Change.
|
|
| 4. 4. / 15:50 / č. 18 |
3. Language Strategies and Language Evolution |
| How can new language systems develop from scratch? While the second lecture discussed experiments in which a population of artificial language users starts with a reconstructed language, this final lecture tackles experiments in which these agents start without a conceptual and linguistic inventory. Instead, they are endowed with “language strategies”, including operators for learning and invention, and they have to self-organize a shared communication system without central control. The lecture discusses three experiments in the domain of color, space and grammatical agreement.
Background reading: – Bleys, J. (2012). “Language Strategies for Color” In: Steels, L., editor, Experiments in Cultural Language Evolution. Amsterdam: John Benjamins.
– Spranger, M. (2012). “The Co-Evolution of Basic Spatial Terms and Categories.” In: Steels, L., editor, Experiments in Cultural Language Evolution. Amsterdam: John Benjamins. – Beuls, Katrien and Steels, Luc (2013). Agent-Based Models of Strategies for the Emergence and Evolution of Grammatical Agreement. PLoS ONE 8(3): e58960. Available as open access at http://dx.plos.org/10.1371/journal.pone.0058960
Available at www.fcg-net.org/design-patterns/ |