Clark, R., & Roberts, I. (1993). A Computational Model of Language Learnability and Language Change. Linguistic Inquiry, 24, 299-345.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....learning impossible in some cases (Niyogi Berwick, 1996) For further discussion of the limitations of the TLA, see (Fodor, 1998) A more complex form of random search is genetic algorithms. Such algorithms have been applied to many types of problems, including grammar learning (Clark, 1992; Clark Roberts, 1993; Pulleyblank Turkel, 1998) This type of learner maintains a population of several grammars, evaluating each on a collection of data. The grammars are then changed randomly, both via individual mutation and by combining elements of different grammars together, with preference given to grammars ....

Clark, R., & Roberts, I. (1993). A Computational Model of Language Learnability and Language Change. Linguistic Inquiry, 24, 299-345.

....community must also learn G. If not, how did the child s parents learn G, given that, by hypothesis, they were exposed to L In other words, there appears to be no room for endogenous language change, a point which has been recognized by generative theorists in recent years (Lightfoot 1991, 1999; Clark and Roberts 1993). Of course, if the conditions of linguistic transmission are altered, for example, by contact with another speech community, then change may well occur, since the linguistic experience of children of the community is likely to change. Since language change is ubiquitous, it might seem that the ....

....the learner, so that as we learn more about the latter, we have a hope of better understanding diachrony. In addition to these matters, however, there are issues concerning the robustness of 15 the evidence for linguistic structure being acquired that arise specifically in the context of change. Clark and Roberts (1993) make this point in discussing the loss of V2 in Middle French, an example that is worth citing at some length (see Roberts 1993 for a full discussion of the historical issues) Old French was a V2 language, like the Germanic languages; but C R argue that the evidence for the V2 property in Old ....

Clark, Robin and Ian Roberts. 1993. A computational model of language learnability and language change. Linguistic Inquiry 24:299--345.

....used to learn the characteristics of a wide variety of phenomena, both inside and outside linguistic and document retrieval domains. In addition to being applied in a variety of biological and industrial environments, they have been used to model and study 2 the historical changes in a language (Clark Roberts, 1993). The nature of linguistic phenomena may be learned through the application of other techniques, such as neural networks capable of learning to associate events with other events. Genetic algorithms were chosen for this work because at each stage in the evolutionary process, a full grammar and set ....

.... the same left hand side as the rule it replaces) A similar procedure may be used for the rules governing terms and their dominating non terminal symbols (grammatical tags) This would have the effect of weighting the offspring by those parts of a parent that have been most useful to the parent (Clark Roberts, 1993). To simplify the interpretation of genes in LUST, the results described below were derived from a system with a constant probability assigned to the possibility of each rule changing There are always three genes in use by the LUST system, numbered and . Genes and are always ....

Clark, R., & Roberts, I. (1993). A computational model of language learnability and language change. Linguistic Inquiry, 24, 299--345.

....example, Wanner and Gleitman (1982:12f) argue that children are predisposed to learn lexical compositional systems in which atomic elements of meaning, such as negation, are mapped to individual words. This leads to errors where languages, for example, mark negation morphologically. Similarly, Clark (1993) argues for a principle of contrast in lexical acquisition, suggesting that children hypothesize novel meanings for novel words, ignoring, at least initially, the hypothesis that a new word may be synonymous with a known one. This paper presents a model of the grammatical acquisition procedure in ....

.... generative linguistics is that language acquisition is the primary engine of language change (e.g. Lightfoot, 1979) In recent generative work on diachronic syntax, language change is primarily located in parameter resetting (reanalysis) during language acquisition (e.g. Lightfoot, 1992, 1999; Clark and Roberts, 1993; Kroch and Taylor, 1997) Differential learnability of grammatical systems, on the basis of learners exposure to triggering data from varying grammatical sources, causes language change. To a first approximation once the critical period for language acquisition is complete, the grammars ....

[Article contains additional citation context not shown here]

Clark, Robin and Ian Roberts (1993) `A computational model of language learnability and language change', Linguistic Inquiry, vol.24.2, 299--345.

....with local exploration of the search space and robust convergence to a target grammar given feasible amounts of potentially noisy or indeterminate input. Human language learners in special circumstances converge to grammars different from that of the preceding generation (e.g. Bickerton, 1984; Clark and Roberts, 1993). The model proposed has the same behaviour, though further work is needed to characterize the exact circumstances under which such behaviour will occur and whether this appears realistic with respect to attested cases of major and rapid grammatical change. Nevertheless, the need for such ....

Clark, R. and Roberts, I. (1993) `A computational model of language learnability and language change', Linguistic Inquiry, vol.24.2, 299--345.

....elements in the input as triggers to spur tree growth, let us first examine how children might do so. 2.1 Triggers in L1 acquisition Specific proposals concerning the nature of triggers for parameter setting in L1 acquisition have been put forth by e.g. Gibson Wexler (1994) J. Fodor (1992) and Clark Roberts (1993). Gibson Wexler consider a model whereby a single sentence type will enable the language learner to uniquely determine a set of parameter settings. Fodor, on the other hand, develops the notion of a designated trigger, according to which parameters designate what type of input will cause a ....

Clark, R. & I. Roberts (1993) 'A computational model of language learnability and language change', Linguistic Inquiry 24:299-345.

.... Brent, Computational Language Acquisition 2 knowledge of ever younger children (Golinkoff, Hirsh Pasek, Cauley, Gordon, 1987; Nelson, Jusczyk, Mandel, Myers, Turk, Gerken, 1995) At the same time, interest in the phonological, syntactic, and semantic aspects of the lexicon has grown (Clark, 1993; Gleitman Landau, 1994; Jusczyk, 1996) broadening a field in which morphological and syntactic rules had been the dominant focus. With the new methods and interests have come new theoretical approaches, including the class of theories known as bootstrapping hypotheses (Gleitman, 1990; Gleitman ....

....clauses in NPs, and so on. Originally, parameterization was thought of as resolving the mysteries of language acquisition after all, what can be so hard about setting ten or twenty binary parameters Recently, however, it has been pointed out that parameter setting is not necessarily so easy (Clark Roberts, 1993; Gibson Wexler, 1994) If cross linguistic variation is characterized by n different binary parameters, then there are 2 n possible combinations of settings that is, 2 n possible grammars. For example, if there are 20 parameters, then there are more than one million possible grammars. It ....

Clark, R., & Roberts, I. (1993). A Computational Model of Language Learnability and Language Change. Linguistic Inquiry, 24, 299--345.

....the particular grammar at work. Clearly, this problem is related to the first; the relationship between the grammars and the descriptions is dependent upon the hidden elements of the descriptions (here, the foot structure) Much recent work in language learnability (Dresher and Kaye, 1990) (Clark and Roberts, 1993) (Gibson and Wexler, 1994) has treated the indirectness of the relationship between overt forms and grammars in monolithic terms, without identifying the mediating role of full structural descriptions 1 . However, Tesar and Smolensky have suggested that decomposing the learning problem along ....

....paper is motivated by more than the articulation of a learning story for Optimality Theory. It is motivated by the goal of avoiding a pair of extremes in learnability work. One extreme is random search, including the triggering learning algorithm (Gibson and Wexler, 1994) and genetic algorithms (Clark and Roberts, 1993) (Pulleyblank and Turkel, to appear) Under this kind of approach, the linguistic theory specifies the space of possible grammars, but that is the extent of its contribution to learning. When confronted with evidence in conflict with the learner s current grammar, the learner randomly selects an ....

Clark, Robin and Ian Roberts. 1993. A computational model of language learnability and language change. Linguistic Inquiry, 24:299--345.

....with local exploration of the search space and robust convergence to a target grammar given feasible amounts of potentially noisy or in20 determinate input. Human language learners in special circumstances converge to grammars different from that of the preceding generation (e.g. Bickerton, 1984; Clark and Roberts, 1993). The model proposed has the same behaviour, though further work is needed to characterize the exact circumstances under which such behaviour will occur and whether this appears realistic with repsect to attested cases of major and rapid grammatical change. Nevertheless, the need for such ....

Clark, R. and Roberts, I. (1993) `A computational model of language learnability and language change', Linguistic Inquiry, vol.24.2, 299--345.

....and seem to. For the motion verbs, I sampled walk, run, and move. 3 Examples include the probabilistic context free phrase structure grammars used by computational linguists (see Charniak, 1993) the Competing Grammars model of Kroch (1989) the probabilistic Principles and Parameters model of Clark and Roberts (1991). Table 1: Quantitative data from the history of be going to 3 dimensions. Year Source 1. Place 2. Agt 3. Nonagt Tokens 1590 Shakespeare 64 35 0 31 1695 Defoe 47 45 8 62 1796 Austen 48 43 10 150 1841 Dickens 22 64 15 151 1884 Hardy 18 60 22 149 1907 Lawrence 24 47 31 142 1911 ....

Clark, R. and Roberts, I. (1991). A computational model of language learnability and language change. Linguistic Inquiry, 24(2):299--345.

....a useful purpose, in that it allows us to se the evolutionary effects of other factors, such as bottlenecks (see below) all the more clearly. Perfect access to primary linguistic data is a basic assumption of classic work in language learnability theory and related theory of language change (e.g. Clark Roberts (1993); Gibson Wexler (1994) Niyogi Berwick (1997) It is not a problematic assumption, because it is clear that it could be relaxed to partial access all of the tike, or perfect access some of the time (or both) so long as such access is sufficient. Pre defined meanings: The extant models all ....

Clark, R., and Roberts, I., (1993) "A computational model of language learnability and language change" Linguistic Inquiry, 24,2:299-345.

....in turn influence subsequent development of language via linguistic selection. 1. 2 Linguistic Selection In recent generative linguistic work on diachronic syntax, language change is primarily located in parameter resetting (reanalysis) during language acquisition (e.g. Lightfoot, 1992, 1997; Clark and Roberts, 1993; Kroch and Taylor, 1997) Differential learnability of grammatical variants, on the basis of learners exposure to triggering data from varying grammatical sources, causes change. Language can be viewed as a dynamic system which adapts to its niche of human language learners and users (e.g. ....

Clark, R. and Roberts, I. (1993) `A computational model of language learnability and language change', Linguistic Inquiry, vol.24.2, 299--345.

....human languages. For example, Wanner and Gleitman (1982:12f) argue that children are predisposed to learn lexical compositional systems in which atomic elements of meaning are mapped to individual words. This leads to errors where languages, for example, mark negation morphologically. Similarly, Clark (1993) argues for a principle of contrast in lexical acquisition, suggesting that children hypothesize novel meanings for novel words, ignoring, at least initially, the hypothesis that a new word may be synonymous with a known one. How do sometimes inaccurate biases of this kind arise and how pervasive ....

.... Berwick, 1997a,b) or account for typological, statistical and implicational universals (e.g. Kirby, 1996, 1997, 1998) In generative work on diachronic syntax, language change is primarily located in parameter resetting (reanalysis) during language acquisition (e.g. Lightfoot, 1979, 1992, 1997; Clark and Roberts, 1993; Kroch and Taylor, 1997) Differential learnability of grammatical systems, on the basis of learners exposure to triggering data from varying grammatical sources, causes change. This can be modelled as an evolutionary process in which variant source grammars provide competing constructions which ....

[Article contains additional citation context not shown here]

Clark, R. and Roberts, I. (1993) `A computational model of language learnability and language change', Linguistic Inquiry, vol.24.2, 299--345.

....language is fully innate. The least constrained LAD is one with solely genes all parameters. Such an LAD could in principle learn any one of the 256 logically possible languages. In this extreme, there are no innate constraints on variation. 4. 3 Utterances as triggers As Clark and Roberts [8] do, we will treat each utterance in the simulation as a trigger for a particular subset of the set of possible grammars. To take a concrete example, the first Hungarian sentence below could potentially provide the learner with evidence that she is hearing a sentence produced from a language with ....

Robin Clark and Ian Roberts. A computational model of language learnability and language change. Linguistic Inquiry, 24:299--345, 1993.

....with local exploration of the search space and robust convergence to a target grammar given feasible amounts of potentially noisy or indeterminate input. Human language learners in special circumstances converge to grammars different from that of the preceding generation (e.g. Bickerton, 1984; Clark and Roberts, 1993). The model proposed has the same behaviour, though further work is needed to characterize the exact circumstances under which such behaviour will occur and whether this appears realistic with respect to attested cases of major and rapid grammatical change. Nevertheless, the need for such ....

Clark, R. and Roberts, I. (1993) `A computational model of language learnability and language change', Linguistic Inquiry, vol.24.2, 299--345.

....project concerns language and glossogenetic language change. That is the diachronic change that a particular language goes through over a number of successive generations. Many examples of such changes can be seen historically. For example, one of the syntactic changes that has occurred in French (Clark Roberts 1993) is that null subjects have been lost. Compare: 1.1) Ainsi thus (they) s amusaient bien had fun cette that nuit. night. Modern French) 1.2) Si thus (they) firent pro made grant great joie joy la the nuit. night. Old French) In modern French the subject must be explicitly ....

....framework for the work done here, along with the work on triggers by Gibson Wexler (1994) which Niyogi Berwick s work in turn is based upon. The parsing complexity issues are looked at through the work of Kirby (1996) and the work of Hawkins (1994) on which some of Kirby s work is based. Clark Roberts s (1993) work is of interest because it also tries to incorporate ideas about acceptability within a given parameterisation along with issues of fitness and complexity. 2.1 Triggers In many respects the work of Niyogi Berwick is an extension and formalisation of work done by Gibson Wexler (1994) ....

[Article contains additional citation context not shown here]

Clark, R. & Roberts, I. (1993), `A computational model of language learnability and language change', Linguistic Inquiry 242, 299--345.

....all of the above data structures using phrase markers and to encode the phrase markers as bit strings. I will argue for this move in section 4 below. For the moment, I will simply develop a general encoding scheme. The encoding method that I will discuss here is based on the method developed in Clark (1993) and can be thought of as a programming language for syntactic representations. It should be noted, however, that the method described here is not intended to be optimal, nor is it intended to be completely general. In particular, the scheme given here is not a prefix code for tree structures. ....

....the definition in either (18) or (66) requires that parameter expression be unambiguous. The computational problem of sorting through the interactions of the principles and parameters to arrive at the correct set of parameter settings for the target language is discussed in Clark (1990;1992) and Clark Roberts (1993); I will leave that problem aside and focus on the proper metric for the information content of parameters. Let us take the information content of a parameter value to be the Kolmogorov complexity of the least structure that expresses that parameter value: 70) Parameter Complexity The ....

Clark, R. & I. Roberts (1993). "A Computational Model of Language Learnability and Language Change", Linguistic Inquiry, 24.2, pp. 299-345.

No context found.

R. Clark and I. Roberts. A computational model of language learnability and language change. Linguistic Inquiry, 24(2):299--345, 1993.

No context found.

Clark, Robin and Ian Roberts. 1993. A computational model of language learnability and language change. Linguistic Inquiry 24(2):299-345.

No context found.

R. Clark and I. Roberts. A computational model of language learnability and language change. Linguistic Inquiry, 24(2):299--345, 1993.

No context found.

R. Clark and I. Roberts. A computational model of language learnability and language change. Linguistic Inquiry, 24(2):299--345, 1993.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC