- Salikoko S. Mufwene (University of Chicago)
Title : What Makes Language Evolution So Complex and Difficult to Model Accurately?
Scheduled : Monday June 23, 2014, 16:00-17:30
Abstract: Like any other kind of evolutionary process, language evolution is unplanned. It is the outcome of how adaptively individuals interacting with each other both reproduce and innovate ways of communicating information successfully, a process marked by successive changes between time t1 and tn (t1, t2, t3
tn). While it is convenient and relatively easier to describe changing norms or communal patterns within a population, it is more difficult to account for the actuation of the changes. In the latter case, there are several ecological factors that bear on the evolutionary process, including: 1) the error-prone process of spreading innovations by learning and instantiating them (unlike in the inheritance of genetic traits); 2) variation in innovations, which generates competition between/among variants (during the convergence towards communal norms); 3) the periodization following from differences between the times of particular innovations (and their consequences); and 4) population structure (which constrains how innovations spread and how far). Perhaps more difficult is the fact that evolutionary theories have focused on documenting population-level outcomes of behaviors that are actually specific to and centered around individual agents, none of whom is aware of the full extant system at the time they innovate or borrow an innovation (and thereby contribute to spreading it). Often, they are not even aware of alternatives that already exist or are being innovated by others, though not exactly in the same way, at the time of their behavior. This is when the distinction between internally- and externally-motivated change (articulated inaccurately in traditional historical linguistics!) becomes significant to whoever in concerned with the actuation of change. Self-organization is of course part of the big picture, to the extent that it explains (or does it?) how everything comes together in the emergent system, but we do face the challenge of explaining the nature of the complexity inherent in language evolution. I submit an account that applies not only to structural changes but also to language vitality.
- Léo Léonard (IUF & Paris 3, UMR 7018) & Vittorio dell'Aquila (CELE, Milan & Vaasa)
Avec la collaboration da Antonella Gaillard-Corvaglia (Postdoc, Inalco)
Title : Algorithmic complexity applied to geolinguistic networks
Scheduled : Monday June 23, 2014, 17:30-18:00
Abstract: Dialectology has long been considered as a marginal field in linguistics, mainly concerned with the recollection of empirical facts, with low theoretical expectations. Nevertheless, thanks to quantification of dialectical data (dialectometry, see Goebl, 1981, 1982, 2002)), geolinguistics in particular turns out to be one of the most promising horizon for Complexity Theory (CT) - as much as CT opens new horizons for dialectology. We'll apply various methods to a Mazatec database from Paul L. Kirk (1966), providing 9000 tokens (750 cognates x 12 locolects): patristic distances (cladistics, see Hennig, 1950, 1966), Levenshtein algorithm (Beijering & al. 2008, Bolognesi & Heeringa, 2002) and dialect intelligibility tests (Kirk, 1970), in order to show multiplicity of prospects on a geolinguistic space. Mazatec as a case study for testing algorithmic complexity has been chosen on several grounds: i) it once provided the empirical base for a landmark study by Sarah Gudschinsky (1958) on the reconstruction of dialect diversification process (1958), ii) Kirk's data, with less than 10 000 tokens is easier to process than bigger data available on European languages, iii) we have thoroughly checked and revisited Kirk's data through fieldwork within the framework of an empirical research project (IUF, MAmP, 2009-14, see Léonard & al. 2012), iv) phonology and grammar of Mazatec dialects have been formalized within the same project (with declarative phonology and Paradigm Function Morphology). Conditions for a survey of algorithmic complexity are therefore met, allowing a multiplex modeling of Mazatec geolinguistics.
- Beijering K. Gooskens C. & Heeringa W. "Predicting intelligibility and perceived linguistic distance by means of the Levenshtein algorithm", Amsterdam, Linguistics in the Netherlands, 2008,), 2008, p. 13-24.
- Bolognesi, R. & W. Heeringa, "De invloed van dominante talen op het lexicon en de fonologie van Sardische dialecten", in D. Bakker, T. Sanders, R. Schoonen and Per van der Wijst (eds.). Gramma/TTT, Tijdschrift voor taalwetenschap,
Nijmegen University Press, Nijmegen, 9 (1), 2002, p. 45-84.
- Goebl Hans, "Eléments d'analyse dialectométrique (avec application à l'AIS)", Revue de Linguistique romane, 45, 1981, p. 349-420.
- Goebl Hans, Dialektometrie. Prinzipien und Methoden des Einsatzes der numerischen Taxonomie im Bereich der Dialektgeographie, Vienne, Verlag der Öst. Akademie der Wissenschaften, 1982.
- Goebl Hans, "Analyse dialectométrique des structures de profondeur de l'ALF", Revue de Linguistique Romane, 66, 2002, p. 5-63.
- Gudschinsky Sarah, "Mazatec dialect history", Language 34, 1958, p. 469-481.
- Gribbin, John, Deep Simplicity, Chaos, Complexity and the Emergence of Life, London, Penguin, 2004.
- Hennig Willi, Grundzüge einer theorie der Phylogenetischen systematik, Berlin, Deutscher Zentralverlag, 1950.
- Hennig Willi, Phylogenetic Systematics. Urbana, University of Illinois Press, 1966.
- Kirk Paul L., Proto-Mazatec phonology. PhD dissertation, University of Washington, 1966.
- Kirk Paul L., "Dialect Intelligibility Testing: The Mazatec Study", International Journal of American Linguistics, 36-3, 1970, p. 205-211.
- Léonard Jean Léo ; dell'Aquila, Vittorio & Gaillard-Corvaglia, Antonella "The ALMaz (Atlas Linguistico Mazateco): from geolinguistic data processing to typological traits", STUF, Akademie Verlag, 6, 2012, p. 78 -94.
- Stefan Balev (ISCN, Le Havre), Gérard H.E. Duchamp (LIPN, Paris 13) & Jean Léo Léonard (IUF & Paris 3)
Title : Visualizing and Revisiting Dialect Intelligibility Networks: Mazatec as a case study
Scheduled : Monday June 23, 2014, 18:00-18:30
Dialect Intelligibility Testing (DIT) has long been the focus of attention in Native American linguistics (Hockett, 1958: 321-30), as an alternative standpoint to taxonomies based on comparative or quantitative linguistics. It has contributed to enhance the heuristic value of continuous chain models, over discontinuous tree-like models in dialectology. DIT also highlights epigenetic trends over phylogenetic and ontogenetic assessments on dialect variation. In this talk, we'll revisit Kirk's and Casad's data on Mazatec mutual intelligibility patterns (Kirk, 1970; Casad, 1974: 46-51, 167-79) with visualizing tools such as GraphStream (http://graphstream-project.org/), and we'll question relevant thresholds in Mutual Intelligibility Networks (MIN). We'll suggest a finer-grained grid than the one initially used by Kirk & Casad, in the seventies, and we'll show how previously unobserved communal clusters may emerge from algorithmic complexity, out of raw MIN data, through visualization devices now currently used in processing complex systems.
- Àngels Massip i Bonet (Departament de Filologia Catalana, Universitat de Barcelona)
Title : Complexity as a framework for understanding transdisciplinarity
Scheduled : Monday June 23, 2014, 18:30-19:00
Abstract: The goal of this talk is, on the one hand, to introduce the results of the work we have undertaken in recent years in the group Complexity, Sociolinguistics and Communication (http://www.sociocomplexitat.ub.edu/grup-de-complexitat-comunicacio-i-sociolingueistica) at the University of Barcelona and, on the other hand, to provide some prospects on how the modeling of sociocomplexity can be improved and further developed.
We will first highlight the importance of current epistemological agreement, in order to work from a transdisciplinary standpoint. We also wish to review which languages and which tools are appropriate for modeling sociocomplexity.
In order to ground the theory on relevant linguistic and sociolinguistic topics, we will point at a complexic vision of linguistic change, exemplified with data from Catalan and French. We will also provide insight on language shift, comparable to other ones about the issue at stake in the STDL (Space, Topology, Diffusion & Language) workshop.
- Bastardas-Boada, Albert (2013), Language policy and planning as an interdisciplinary field: towards a complexity approach, Current Issues in Language Planning, DOI: 10.1080/14664208.2013.829276.
- Bastardas i Boada, A., & Àngels Massip i Bonet (2013), Llengua i complexitat. Presentació, LSC Llengua, Societat i Comunicació 11, pp. 1-4. (Monogràfic dedicat a Llengua i Complexitat).
- Durand, Daniel (1979), La Systémique. PUF
- Marco Patriarca (National Institute of Chemical Physics and Biophysics, Tallinn, Estonia)
Title : Models of language competition
Scheduled : Tuesday June 24, 2014, 14:00-15:00
Abstract: This contribution presents an overview of language competition models that have been introduced and studied for many years by now in different fields such as mathematics, physics, and linguistics. The overview covers some relevant models which resemble the classical mathematical models of competition between biological species. A first aspect that we try to clarify is the statistical interpretation of the models and their corresponding meaning, in particular concerning the role of bilinguals in the competition between two languages. We begin with the Baggs-Freedman models, proceed through the Abrams-Strogatz and Minett-Wang models, and then turn to more recent models.
Also the spatial side of language spreading is considered, by illustrating various ways in which the spreading of language (features) through space and time has been modeled so far in order to take into account geographical, cultural, and social factors.
- Adam Lipowski (Faculty of Physics, Adam Mickiewicz University, Poznan, Poland)
Title : Dynamics of Naming Games: Why is it slow and how to make it faster?
Scheduled : Tuesday June 24, 2014, 15:00-15:30
Abstract: Naming Game is an important model of agreement dynamics. It might be used to describe emergence of a common vocabulary but it was also used to describe an opinion formation in a large scale sensor network, simple grammar or leader formation mechanism. Since the time to reach consensus is an important characteristics of Naming Game, its dynamics has been intensively studied. We show that some previous results on the Naming Game needs to be modified. It turns out that due to the formation of stripe structures dynamics of this model is much slower than expected. We also show that one can construct a Naming Game that does not suffer from the formation of such structures and has a much faster dynamics. We suggest the relation of the dynamics of this model with some percolation problems as well as with some other statistical mechanics models.
Lipowska Dorota & Adam Lipowski, 2013. "Phase transition and fast agreement in Naming Game with preference for multi-word agents", eprint: http://arxiv.org/abs/1308.3781
- Didier Demolin (Gipsa-lab, Université de Grenoble, Alpes, Fr)
Title : Evaluating the complexity of phonological systems
Scheduled : Tuesday June 24, 2014, 15:30-16:00
Abstract: Phonological complexity has been the subject of many discussions for a century, e.g. Zipf and more recently Pellegrino et al. (2009). However, even though there is an agreement in seeing phonological systems as complex adaptive systems, we are still far from able to measure phonological complexity. Factors typically considered are the inherent phonetic complexity of elements in a phonological inventory, the role of combinatorial possibilities and the combination of frequency of occurrences of different elements (Maddieson 2009). Coupé¡¥t al. (2009) emphasize that a system is said to be complex if it is structured in different levels; if the properties of the global level differs from those of the elements of the basic level and if the systemic properties cannot be derived linearly from the basic ones. Phonetic and phonological systems reflect various types of constraints (biological, cognitive, linguistic and social) but the understanding of their interactions and integration is still quite limited. Taking into account the various levels of phonological systems together, gestures, features, vowels, consonants, syllable, suprasegments) is a possible way to start evaluating their complexity.
Complexity in phonology also focuses on the relation between the number of phonemes in a language and the number of allophones. Do we find more allophones where there are fewer phonemes? Do we have fewer allophones when there are more phonemes? Even if phonological descriptions are not always comparable and have been made at different times, we believe that this is an important study to be undertaken and this will be discussed in this presentation.
- Coupé C., Marsico, E. & Pellegrino, F. (2009). Structural complexity in phonological systems. In Pellegrino, F. Marsico, E., Chitoran, I. Coupé (Eds.). Approaches to phonological complexity. Berlin, Mouton de Gruyter. 141-169.
- Maddieson, I. (2009). Calculating phonological complexity. In Pellegrino, F. Marsico, E., Chitoran, I. Coupé (Eds.). Approaches to phonological complexity. Berlin, Mouton de Gruyter. 85-109.
- Pellegrino, F. Marsico, E., Chitoran, I. Coupé (Eds.). (2009). Approaches to phonological copmplexity. Berlin, Mouton de Gruyter.
- Zipf, G.K. (1949). Human behavior and the principle of least effort. Cambridge (MA), Addison-Wesley.
- SW4: Round table/discussion - Space, Topology, Diffusion & Language : prospects for future research
Scheduled : Tuesday June 24, 2014, 16:00-16:30