Jackendoff, Ray (2002) Foundations of Language: Brain, Meaning, Grammar, Evolution. Oxford University Press, xx+477pp, hardback ISBN 0-19-827012-7, GBP 25.00
Pritha Chandra, Centre of Linguistics and English, Jawaharlal Nehru University, New Delhi (India).
'Foundations of Language' stands out as a unique endeavour to bring together issues that have been the concerns of many contemporary researches. The dilemma of most present day linguistic researches is that they miss the essential linkages and interfaces between various aspects of language, thus failing to formulate a holistic approach indispensable for a complete and comprehensive understanding of the architecture of the language faculty. The book successfully indulges in the arduous task of bringing together various theoretical assumptions about language and filling up the missing links in the study of the mind and language. Committed to this task, it targets an interdisciplinary audience with many of the linguistic jargons being successfully moulded to the taste of non-linguists.
The book is divided into three parts - each with four chapters, with one additional chapter linking up the manifold ideas distributed throughout the book. The first part lays down the 'psychological and biological foundations' of language, primarily dealing with issues that have shaped generative linguistics for the last thirty years or so - mentalism, combinatoriality and nativism. The next part carries forward the doubts and questions posed in the previous chapters regarding the conjectures of the existing generative paradigms, and lays out alternative 'architectural foundations'. The final part -'semantic and conceptual foundations' focuses essentially on framing a more sophisticated theory of semantics, necessary for forging a meaningful and dialectical interconnection between the human mind and the objective reality.
The uniqueness of human language is the complex and rich structure inherent in all possible linguistic constructions; even a very simple sentence like 'The little star's beside a big star' can be conceptualised at four different levels - phonological, syntactic, semantic/conceptual and spatial. Structural specificity has therefore rightly occupied the centre-stage of linguistic research, rendering it the status of a highly specialised field. In the first chapter, the author reflects on these issues and constructs the two-fold agenda for linguists working within the generative paradigm: recognising the four levels and their tiers of structure, and secondly, postulating possible ways of correspondence between them.
In the second chapter, terms like f(unctional) mind and f- mental are introduced in a bid to reduce the tensions generating from otherwise commonly used terms like `representation', `symbol' in the generative tradition. Rejecting theories that vouch for a complete analogy between linguistic structures and neural instantiations, the notion of f-mind insinuates that diverse linguistic notations encode particular dimensions of the state-space, characterised by the combination of the states of neurons in relevant parts of the brain.
In the next chapter, Jackendoff deals with the various rules used to beget legible sentences in the grammar. These rules are of three types - formation rules, derivational rules and constraints, along with other specific rules for the construction of lexical items from smaller units. Extending on the language-neural instantiation relationship developed in the previous chapter, these rules are too formal constructs that could be conceptualised either as data structures stored in long-term memory or as partial descriptions of the operations of the processor itself. The last chapter in the first part is used to clarify some of the common misconceptions usually associated with the term Universal Grammar. Jackendoff considers that mentalism and connectionism are two extremes - one focuses entirely on pre-specification of linguistic universals and structures in the mind, while the other takes more of an empiricist viewpoint to the manner in which children acquire language. The right path, the author claims "lies undoubtedly between these two extremes".
The fifth chapter provides an alternative to the existing generative models for the language faculty. Jackendoff's alternative model perceives more than one combinatorial system interacting with each other through interface systems. In other words, each of the four language specific components in the present model is equally generative.
The alternative model also calls for a distinction between elements stored in the long- term memory (lexical items) and those constructed online from working memory (grammatical words), and between productive morphology (containing mostly regular forms) and semi-productive morphology (with partial regularities). All these different processes are dealt with extensively in chapter six, that contemplates a dynamic version of the lexicon, with the Universal Grammar providing design space for words and stable 'unmarked' pre-specified lexical rules as well as the 'marked' ones.
The seventh chapter sets a deeper probe into the concept of 'modularity', with the focus now primarily placed on the interactions of the various components of grammar and the implications this new architecture has for processing. Jackendoff's 'structure constrained modularity' states that levels of structure communicate with each other through interface modules where correlation is carried out through specific computations. There are in fact degrees of modularity rather than absolute modularity, with the interactions enhancing as the interfaces between two components enrich. These stipulations flag another deviation from standard generative linguistic assumptions about the evolution of the language faculty. Chapter eight is entirely devoted to this theme, where Jackendoff contends that language is by large a consequence of natural selection. His claim is that the new grammatical machine allows for a well- articulated account of the way the human language kept enriching and adding new structures to accommodate to the increasing pressures of adequately conveying complex thoughts and concepts. The evolution of the language faculty is regarded in the author's own words, as "the successive addition of more and more 'tricks' to the toolkit".
Generative grammar, according to the author has never had a clear understanding of semantics; likewise, semantic theories have also not been very explicit in their definition of linguistic semantics and its relation to human cognition and conceptualisation. Extending on the new grammatical model, the ninth chapter argues against the reduction of semantics to syntax, instead postulating semantics as an independent combinatorial structure. Linguistic semantics is conceptualised as arising from the interface between conceptualisation and linguistic structures corresponding to syntax and phonology.
The concept of semantics is further clarified in the tenth chapter, which focuses on notions of truth and reference and the way language relates to our perception of the world and thought processes. A detailed sketch of the way, language is used to convey names, abstract objects and kinds, imparts vital inputs about the intricacies of human language.
The eleventh and twelfth chapters deal with two very important aspects of meaning, lexical semantics and phrasal semantics respectively. The author deals extensively with issues like lexical decomposition, its drawbacks and prospects, polysemy, and other subtle similarities and differences in meaning arising from levels of cognition. Building up on notions like state and event functions, the author postulates structures (conceptual structures and spatial structures) within lexical meaning. The twelfth chapter draws parallels between lexical semantics and phrasal semantics to the extent that both use certain basic conceptual combinations and other principles like variable satisfaction, modification and lambda extraction. However, phrasal semantics also reflects the extremely rich and complex coordination between grammar, independent well- formedness conditions on conceptual structure and the construal of context. The final chapter presents the concluding remarks.
COMMENTS It is indeed a formidable task to embark upon a naturalistic study of an abstract phenomenon as language, and to construct a mathematical/logical model for it, suited to explain its immense complexity. In this approach, linguistics is elevated to the standards and sophistication expected from any genuine natural science using higher levels of abstraction to generate principles governing the phenomena under study, against the crudity of mechanistic "physicalism". As Chomsky (2000) cogitates, "A naturalistic approach to linguistic and mental aspects of the world seeks to construct intelligible explanatory theories, taking as "real" what we are led to posit in this quest, and hoping for eventual unification with the "core" natural sciences, not necessarily reduction." In this regard, Jackendoff's effort is definitely commendable, in that he not only relates the results of linguistic researches with other disciplines, a beneficial enterprise for both sides, but also that he examines the existing generative paradigms' attempts with the required scepticism, essential for further opening up of the field.
The underpinning of the minimalist programme has been that a speaker's internal or individual language (I-language) consists of a computational procedure and a lexicon, where "language variation appears to reside". Drawing in lexical items from the lexicon, syntax generates a more complex array of features, with the generative engine bifurcating into the Phonetic Form and the Logical Form. This paradigm has come under severe criticisms from various quarters, on various issues. However, one of the major loopholes pointed out has been its undermining of the concept of semantics. Chomsky explicitly states, "As for semantics, insofar as we understand language use, the argument for a reference-based semantics (apart from an internalist syntactic version) seems to be weak. It is possible that natural language has only syntax and pragmatics; it has a "semantics" only in the sense of "the study of how this instrument, whose formal structure and potentialities of expression are the subject of syntactic investigation, is actually put to use in a speech community". What could be concluded is that semantics is not necessarily reduced to syntax, as more often claimed, but that it does not form part of the naturalistic inquiries into the language faculty. Linguistic structures need not necessarily bear any relation to the semantics finally attributed in actual use. This view only adds to the status of semantics as an important, independent (but related) field of study, rightly exemplified by its immense complexity and variation in different social environment and belief systems.
Further, when one talks about the 'centricity' of syntax in the organised structure of the language faculty, it is not at all equivalent to building up a deterministic framework, rather it draws from the task of dealing with the organisation of any complex, probabilistic system. Like any self-regulating and self-organising complex system, the language faculty has to satisfy the basic cybernetic laws, which require its factorisation into various sub-systems, along with the satisfaction of Godel's 'incompleteness theorem', which forces a hierarchic structuring. A hierarchy does not signify that of privilege, status or seniority, rather it simply means that the tension unresolved within a system requires a `meta'-system to resolve it. The functional 'centricity' of syntax derives from this inescapable 'natural' law.
Another point of contention among present day linguists revolves around the evolution of the language faculty in hominids. While some like Pinker and Bloom (1990), and the author himself argue for an incremental development of language as a response to natural selection, others like Gould and Lewontin (1979) contest for the position that the emergence of the language faculty was a consequence of natural selection-driven heterogeneous developments converging on a brain structure. "Objects designed for definite purposes can, as a result of their structural complexity, perform many other tasks as well. ...Our large brains may have originated "for" some set of necessary skills in gathering food, socializing, or whatever; but these do not exhaust the limits of what such a complex machine can do." (Gould, 1980) Language faculty in this framework emerges as an un-purposed component of the brain, a necessary corollary of macroevolution. 'Hyper- selectionism', on the other hand, tries to explain every macroevolution as slowly accumulated microevolution, and is unable to explain the historicity of evidenced long-term stasis. The author tends to adhere to this position, which is already under revision in molecular biology, in direct study of fossil sequences, etc.
But then science develops through its tensions, by evidencing and counter-evidencing principles and facts. The greatness of this 'bound to be a classic' lies in its uncompromising posing of questions and possible answers. It definitely establishes a landmark where the linguistic research must take a radical leap, towards the definite establishment of linguistics as a science.
REFERENCES Chomsky, N. (2000) New Horizons in the Study of Language and Mind. Cambridge University Press.
Gould, S.J. and R.C. Lewontin (1979) The spandrels of San Marco and the Panglossian paradigm: a critique of the adaptationist programme. Proceedings of the Royal Society, London B205:581-98.
Gould, S.J. (1980) The Panda's Thumb. Penguin.
Pinker, S. & P. Bloom (1990) Natural language and natural selection. In Behavioural and Brain Sciences 13: 707-784.
ABOUT THE REVIEWER:
ABOUT THE REVIEWER Pritha Chandra is associated with the Centre of Linguistics and English, Jawaharlal Nehru University. She has been teaching linguistics in a college in the University of Delhi. Her specialisation is generative syntax with a specific focus on argument structure, Case and EPP.