Theories and Models of Language Change A Study of Self-Organization - PowerPoint PPT Presentation

Roland Mühlenbernd Introduction: The Evolut. Approach Universals of Sound Systems Theories and Models of Language Change A Study of Self-Organization Session 6: Models II - Emergence of Universals Agent Architecture Agent Interaction Simulation Results Conclusion Further Studies Roland Mühlenbernd Homeworks June 9, 2015

Review: Universal Darwinism Roland Mühlenbernd Introduction: The Evolut. Approach Universals of Sound Systems Mechanisms of universal evolution: A Study of Self-Organization 1. variation : continuing abundance of different elements Agent Architecture Agent Interaction 2. selection : number/probability of copies of elements - Simulation Results Conclusion depending on interaction between element features and Further Studies environmental features Homeworks 3. replication : reproduction/copying of elements What is the role of linguistic universals in an evolutionary model of language change?

Language Change - Broad and Narrow Sense Roland Mühlenbernd Introduction: The Evolut. Approach Universals of Sound Systems A Study of Self-Organization Agent Architecture Agent Interaction Simulation Results Conclusion Further Studies Homeworks

Linguistic Universals Roland Mühlenbernd Introduction: The Evolut. Approach ◮ are patterns that occur systematically across natural Universals of Sound languages Systems ◮ can be distinguished between A Study of Self-Organization ◮ absolute (e.g. all languages have nouns and verbs) Agent Architecture ◮ implicational (e.g. if a language is spoken, it has vowels Agent Interaction Simulation Results and consonants) Conclusion Further Studies ◮ are given on all linguistic levels: Homeworks ◮ phonology (e.g. symmetry of sound inventories) ◮ syntax (Greenberg universals, e.g. SOV → postpositional) ◮ semantics (Swadesh list, natural semantic metalanguage and its semantic primitives, basic color terms) ◮ pragmatics (e.g. generalized implicatures, speech acts) ◮ are given as innate cognitive structures or realized for functional reasons under communicative aspects?

Universals of Sound Systems (Exercise 1) Roland Mühlenbernd Introduction: The Evolut. Approach Universals of Sound Systems The UCLA Phonological Segment Inventory Database A Study of (UPSID) contains 921 different speech sounds with the Self-Organization following values (over all 451 languages of the database): Agent Architecture Agent Interaction Simulation Results ◮ average number of phonemes: 20 to 37 Conclusion Further Studies ◮ minimal number of phonemes: 11 (Rotokas, Pirahã) Homeworks ◮ maximal number of phonemes: 141 (!X˜ u) Example: Hawaiian phonemes: a, e, i, o, u, p, k, m, n, w, l, h, P

Universals of Sound Systems Roland Mühlenbernd Phonetic inventories of the world’s languages exhibit Introduction: The ◮ frequent and rare sounds: f ( [m] ) = 94 % , f ( [ ö ] ) = 1 % Evolut. Approach ◮ a tendency to symmetric inventories: Universals of Sound Systems ◮ f ( [ ď ] | [ O ] ) = 83 % ; f ( [ ď ] | ¬ [ O ] ) = 18 % A Study of ◮ f ( [t] ) = 40 % ; f ( [t] | [d] ) = 83 % Self-Organization Agent Architecture Agent Interaction Simulation Results Conclusion Sound sequences of the world’s languages exhibit Further Studies ◮ frequent and rare syllable structures: universal V , CV Homeworks ◮ follow a sonority hierarchy sonority class value plosives 1 fricatives 2 nasals 3 liquids 4 approximates 5 closed vowels 6 open vowels 7

Universals of Sound Systems (Exercise 2) Roland Mühlenbernd The repertoire and use of speech sounds in human languages Introduction: The are constrained. According to de Boer, explanations for those Evolut. Approach Universals of Sound regularities can by divided in the following two classes: Systems ◮ they are based on physiological features of the human A Study of Self-Organization vocal tract and articulators Agent Architecture Agent Interaction ◮ they are based on the human audible frequencies Simulation Results Conclusion ◮ they are based on innate human cognitive capabilities √ Further Studies ◮ they are based on functional constraints of a good Homeworks communication system √ ◮ enable learnable and robust communication ◮ redundancy and predictability ◮ easily to distinguish and to produce Note: It is hard to find out how specific capacities might have become innate

Universals of Sound Systems Roland Mühlenbernd Sound systems of human languages are often optimized for Introduction: The criteria such as Evolut. Approach Universals of Sound ◮ acoustic distinctiveness (especially vowels) Systems A Study of ◮ articulatory ease (especially consonants) Self-Organization Agent Architecture → f ( [m] ) = 94 % , f ( [ ö ] ) = 1 % Agent Interaction Simulation Results Conclusion Further Studies Homeworks Fig: realization of German vowels Fig: vowel systems of human languages

Universals of Sound Systems Roland Mühlenbernd Introduction: The Evolut. Approach Universals of Sound Systems How do sound systems become optimized? A Study of Self-Organization Agent Architecture ◮ when children learn, they don’t explicitly optimize, but Agent Interaction imitate members of their community Simulation Results Conclusion ◮ their imitation is closer to the source material than Further Studies necessary for successful communication (see dialects) Homeworks ◮ suggestion: optimization is caused by self-organization in a population of language users

Self-Organization in a Language Community Roland Mühlenbernd Introduction: The Evolut. Approach Universals of Sound Systems A Study of Self-Organization Agent Architecture Hypothesis: “...the structure of human vowel systems Agent Interaction Simulation Results is determined by self-organization in a population Conclusion under constraints of perception and production.” Further Studies Homeworks Bart de Boer (2000), Self-organization in vowel systems

Self-Organization in a Language Community Roland Mühlenbernd Introduction: The Evolut. Approach Universals of Sound Systems Conditions for a model of self-organization A Study of 1. emergence of organization on a global scale : most Self-Organization Agent Architecture members have same inventories Agent Interaction Simulation Results 2. emergence due to interaction between members, not Conclusion optimization-actions of single members Further Studies Homeworks 3. non-local influence between members: no direct access to other members’ inventories 4. no pre-wired knowledge : members start with empty inventories; or as tabula rasa

Self-Organization in a Language Community Roland Mühlenbernd Introduction: The Evolut. Approach Universals of Sound Systems Why a computer simulation model for a system of A Study of Self-Organization self-organization? Agent Architecture Agent Interaction 1. phenomenon of self-organization is hard to predict from Simulation Results Conclusion just the description of the system Further Studies 2. computer simulations help to investigate life-like Homeworks phenomena 3. computer model formulates a hypothesis, its simulation synthesizes the phenomenon

de Boer’s Model Roland Mühlenbernd Introduction: The Evolut. Approach Universals of Sound Systems Properties of the model: A Study of Self-Organization 1. population of agents Agent Architecture Agent Interaction Simulation Results 2. agents can produce, perceive and remember speech sounds Conclusion in a human-like way Further Studies Homeworks 3. agents interact with others by imitating them 4. agents update their repertoires in dependence of interaction outcome (imitation success)

de Boer’s Model: Articulatory Space Roland Mühlenbernd Introduction: The Evolut. Approach Universals of Sound Systems A Study of Self-Organization Agent Architecture Agent Interaction Simulation Results Conclusion Further Studies Homeworks Fig: Articulatory space (from de Boer 1999, page 64)

de Boer’s Model: Acoustic Space Roland Mühlenbernd Introduction: The Evolut. Approach Universals of Sound Systems A Study of Self-Organization Agent Architecture Agent Interaction Simulation Results Conclusion Further Studies Homeworks Fig: Acoustic space (from de Boer 1999, page 42)

de Boer’s Model: Space Conversion Roland Mühlenbernd Introduction: The Evolut. Approach Universals of Sound Systems A Study of Self-Organization Agent Architecture Agent Interaction Simulation Results Conclusion Further Studies Homeworks Fig: Space Conversion (from de Boer 2000, page 7)

de Boer’s Model: Agent Architecture Roland Mühlenbernd Introduction: The Evolut. Approach Universals of Sound Systems A Study of Self-Organization Agent Architecture Agent Interaction Simulation Results Conclusion Further Studies Homeworks Fig: Agent architecture (from de Boer 2000, page 6)

de Boer’s Model: Agent Architecture (Exercise 3) Roland Mühlenbernd Introduction: The Evolut. Approach Universals of Sound Systems De Boer is using a prototype model for the agents to store A Study of vowels. In his model a prototype has the following features: Self-Organization Agent Architecture ◮ non-static √ Agent Interaction Simulation Results Conclusion ◮ space-segmenting Further Studies ◮ addable/removable √ Homeworks ◮ unerasable ◮ articulatory/acoustic aspect √ ◮ feature-based