Example: confidence

Synchronic vs. Diachronic Explanation and the Nature of ...

And the Natureof the Language FacultyStephen R. Anderson11 Department of Linguistics, Yale University, New Haven, CT,USA, 06520 8366 Xxxx. Xxx. Xxx. Xxx. YYYY. AA:1 22 This article s ((please add article doi))Copyright YYYY by Annual rights reservedKeywordsSynchrony, Diachronic change, Explanation , Language Faculty,learning algorithm, cross-linguistic generalizationAbstractThe nineteenth century conception that linguistic structure was to beexplained by recourse to the histories of languages was largely aban-doned with the rise of Synchronic theories in the twentieth century, buthas recently returned to prominence. While traditional generative the-ories of language have tended to attribute cross-linguistic regularitiesto constraints imposed on the class of possible grammars by the hu-man Language Faculty, some scholars have argued that this isoften amistake: that there are no (or at least very few) real substantive uni-versals of language, and that the regularities in question arise on thebasis of common paths of Diachronic change having their basis in fac-tors outside of th

understanding of the ways sound systems can be altered in transmission across generations, a point that will be important in discussion below. interpretations and connotations in the literature which might give rise to confusion. www.annualreviews.org • Synchronic vs. Diachronic Explanation and the Language Faculty 3

Tags:

  Transmissions, Sound, Synchronic, Diachronic

Information

Domain:

Source:

Link to this page:

Please notify us if you found a problem with this document:

Other abuse

Transcription of Synchronic vs. Diachronic Explanation and the Nature of ...

1 And the Natureof the Language FacultyStephen R. Anderson11 Department of Linguistics, Yale University, New Haven, CT,USA, 06520 8366 Xxxx. Xxx. Xxx. Xxx. YYYY. AA:1 22 This article s ((please add article doi))Copyright YYYY by Annual rights reservedKeywordsSynchrony, Diachronic change, Explanation , Language Faculty,learning algorithm, cross-linguistic generalizationAbstractThe nineteenth century conception that linguistic structure was to beexplained by recourse to the histories of languages was largely aban-doned with the rise of Synchronic theories in the twentieth century, buthas recently returned to prominence. While traditional generative the-ories of language have tended to attribute cross-linguistic regularitiesto constraints imposed on the class of possible grammars by the hu-man Language Faculty, some scholars have argued that this isoften amistake: that there are no (or at least very few) real substantive uni-versals of language, and that the regularities in question arise on thebasis of common paths of Diachronic change having their basis in fac-tors outside of the defining properties of the set of cognitively accessiblegrammars.

2 This review surveys evidence for that position, primarily inthe domain of phonology but also including morphology and is argued that in phonology, there are at present no convincinglydemonstrated substantive universals governing the set of possible reg-ularities, and that the generalizations we find should be attributed toa combination of contingent historical developments and biases in thelearning algorithm that relates available data to the grammars learnersacquire. In morphology and syntax, it is argued that some apparentgeneralizations are indeed the product of Diachronic change rather thansynchronic constraint, though no broader conclusion is INTRODUCTION .. 22. Sources of the properties of linguistic systems.

3 33. Explanation of properties of sound systems .. Explanatory bases for sound Patterns .. The Bases of Distinctive Features .. 114. Explanation Beyond Phonology .. Diachronic Explanation in Morphology .. Diachronic Explanation in Syntax .. 175. Conclusion .. 191. INTRODUCTIONThe Nature of the scientific study of language have been subject to a variety of interpreta-tions over time, and the places linguists have looked for explanatory principles have variedaccordingly. For example, the Neogrammarian revolution inthe 1870s brought a focus onthe systematic description of historical change, especially sound change, and with that camethe claim that the only genuinely scientific study of language is historical (Paul 1880, ), and a concomitant search for Explanation in the regularities of the shift of attention to Synchronic systems in the early years of the twentiethcentury that is generally associated with de Saussure (1916[1974]), the locus of potentialexplanation shifted from the ways linguistic patterns arise over time to the properties ofthose patterns themselves.

4 For much of the first half of that century, however, the linguist stask was taken to be the careful recording and analysis of theexternal manifestations oflanguage: sets of sounds, words, sentences, and texts in as wide a variety of individuallanguages as possible. Categories developed in the course of this endeavor, such as thoseofphonemes, morphemes, immediate constituent analyses,etc. were taken to be those ofthe linguist s analysis, validated to the extent they helped to elucidate the structure of thetexts under consideration. Calling the linguistics of the period descriptive should be takenquite literally: the aim of the field was to develop complete and accurate descriptions ofthe observable facts of the world s languages rather than explanations of those the Cognitive Revolution of the latter half of the century came a major shiftof the object of inquiry from the external manifestations oflanguage to the systems ofknowledge and the cognitive capacity that underlies the ability of someone who knows alanguage to produce and understand linguistic objects.

5 With this, in turn, came a shift fromthe desire to provide maximally accurate descriptions of the observable data in particularlanguages to the effort to understand the fundamental natureof the cognitive faculty oflanguage, and to explain why the systems we find are as they areand not the centrality of the cognitive underpinnings of thisemerging conception of lan-guage, it made sense to think of the search for explanations of linguistic structures andregularities in terms of the study of the Language Faculty itself. This was to be developedthrough a precise characterization of Universal Grammar1, the cognitive endowment ofHomo sapiensthat supports our capacity to acquire and use particular systems of natural1I avoid use of this term in the present review, because it has taken on a wide range of diverse2 Stephen R.

6 AndersonFigure 1 Grammars and their sources: (a) Only languages consistent with the evidence of the primary datacan be acquired; (b) Only languages accessible via the available learning procedure can beacquired; and (c) Only cognitively possible grammars can beacquired (Anderson 2008). in the 1980s, however, and acquiring momentum in the early years of the newmillenium, proposals were made that in a way mark a return to the Neogrammarian view:the suggestion that in fact, much of what we find in particularlanguages is the productnot of necessary constraints imposed by the Language Faculty but rather of the contingentoutcome of the historical paths by which these languages have developed over tension between the search for explanations of the regularities we find in systemsof language across the world, on the one hand in the Nature of human cognitive organi-zation, and on the other in paths of historical development,is the subject of the presentreview.

7 Because much of the discussion in the literature hasfocused on properties of thesound systems of languages, this will be reflected below, butit is important to note thatthe basic issue of Explanation is in principle just as relevant toother domains of linguistic structure, and some attempt will be made to illustrate this formorphology and for Sources of the properties of linguistic systemsOur evidence for the Nature of language, of course, comes from the observed propertiesof particular languages, and from the inferences we can makeabout the grammars (inthe sense of systems of knowledge) underlying these. When weask for the foundations ofthe particular properties we observe, however, there are several distinct potential loci ofexplanation to factors determining the content of particular grammarsare illustrated schemati-cally inFigure 1.

8 Grammars arise in the individual on the basis of the learner s experiencewith utterances in the surrounding community. As input to the process of acquisition, it isimportant to be clear that the Primary Linguistic Data should not be identified with thebrute physical facts of these utterances, but must also takeinto account the filtering role ofthe perceptual systems through which these are presented tothe mind for work of John Ohala in particular ( Ohala 1981, 1993,inter alia) has stressed the ex-tent to which properties of the perceptual system operatingon speech data are crucial to anunderstanding of the ways sound systems can be altered in transmission across generations,a point that will be important in discussion and connotations in the literature which might give rise to Explanation and the Language Faculty 3 The Primary Linguistic Data are mapped onto a specific grammar by some learningalgorithm characteristic of human cognition.

9 Since the grammar that results is not simplya registration of the perceptually processed utterances that gave rise to it, some non-trivialprinciples of inference must be involved. The output of thisprocess must fall within thespace of grammar systems that are cognitively possible for humans. The two aspects, thecharacter of the space that is the range of the learning algorithm and the Nature of thatalgorithm itself, are frequently conflated in the notion of the human Language Faculty, butthey are distinct: it is logically possible that there are some outputs of the learning algorithmthat would lie outside the space of possible grammars (and which must thereby be rejectedor adjusted), and also that there are some grammars which, while cognitively possible, arenot accessible from any data on the basis of the learning algorithm.

10 The distinctive role ofthe learning algorithm will take on special significance below, but in general we will referto the Language Faculty as a unified question of the existence of a Language Faculty in this sense is often conflated withthat of the domain-specificity of its components, but this isnot logically necessary. Thereis no question that the ability to acquire and use natural languages is a species-specificproperty ofHomo sapiens, grounded in the biology of our species. Whether or not some or even all of that ability derives from broader aspects of human cognition, applicablein other domains beyond language, is strictly speaking irrelevant to the point that such acapacity exists and that its properties can be the object of scientific can view the Learning Algorithm as a system of inference that maps a particularcollectionDof Primary Linguistic Data onto some specific (cognitively possible) grammarG.


Related search queries