SIL Electronic Book Reviews 2005-004

Introduction

When the study of knowledge and the study of language intersect, two distinct parties emerge. On the one hand are “ordinary language users” with their mysterious, abstract linguistic know-how. On the other hand there are linguists and philosophers—language theorists who approach the linguistic abilities of ordinary language users as a scientific discipline. This distinction represents the focus and locus of the “epistemology of language.”

Personally held convictions toward the following two related conundrums are prevalent and diverse: (1) How is it that we as humans are able to use language to meet the complex demands of reason, expression, and communication? (2) How should linguists best go about investigating and describing such intricate phenomena as these in ways that are actually fruitful in practice and faithful to the linguistic knowledge of ordinary language users? Whatever other questions may confront a linguist (or philosopher), these two must be among the most fundamental for extended reflection. Interestingly, personal convictions on these questions may not be consciously considered at all. Such issues as these are addressed by the authors brought together in Barber’s skilful selection and careful arrangement of The Epistemology of Language.

The book is an attractively crafted volume consisting of seventeen chapters from sixteen different authors. An introduction by Barber lays the groundwork for the discussion. The book is divided into four principal thematic parts (“Knowledge and Linguistics,” “Understanding,” “Linguistic Externalism,” and “Epistemology through Language”).

In this review, I will first attempt an overview of the book as a whole. I will then introduce each of these parts, seeking briefly to synthesize and interact with each of the chapters. Following these summaries, I will offer my own observations and reflections on the book’s positions and assumptions.

Overview

As an accessible entrance to the contents of an otherwise highly specialized field of study, Barber’s introduction is essential. His introductory chapter provides both the thematic bricks and the theoretical mortar necessary for structuring the book’s content and does so in such a way that is sufficiently approachable for a non-specialist. Readers will not be far into the introduction, however, before they realize that the book has been designed to bolster, advance, and repair a theory-internal point-of-view that can be traced back to Noam Chomsky and has come to be known as “conceptualism.” Barber offers the following summary of the theory (p. 3):

Detractors and critics have taken issue with both components of this model, for various reasons, and the authors of the book—explicitly in part one, but implicitly in the other four sections as well—seek to support the model both by answering specific challenges brought against it and by advancing our general understanding of it.

In other words, this book is not primarily intended to be a textbook on, or even a survey of, epistemological linguistics, although the book might well work as either. It is intended, rather, to serve as a twenty-first century launch pad for the Chomskyan-based theory of the language-mind interface known as “conceptualism.” Nevertheless, as would be expected, the issues dealt with are all highly relevant to contemporary debates occurring in both philosophy and linguistics related to the interaction between language and mind.

Part One: Knowledge and Linguistics

In her intriguingly entitled essay, “Rabbit Pots and Supernovas: On the Relevance of Psychological Data to Linguistic Theory,” Louise M. Antony sets out to deal with objections brought against component (ii) of conceptualism listed above. Are linguists actually able to describe the internalized grammars of native speakers in ways that are scientifically rigorous? Certain philosophers cast doubt on this possibility by claiming, for example, that linguistics cannot be psychologized because linguistics and psychology represent separate theoretical domains. Antony sides with Chomsky, nevertheless, in asserting that linguistics is in fact a valid branch of psychology. She makes her case by employing analogies from other sciences such as astronomy, palaeontology and archaeology, her prime analogy being an eleventh-century pot found in New Mexico which provided rare confirmation of an earlier-postulated date for the Crab Nebula supernova event. Her point is that that linguists can validly access other theoretical domains in order to make empirical claims. “Verification is holistic,” Antony (51) argues: “a suitable chain of inference could bring any fact to bear on any empirical issue.”

Also dealing with objections brought against the second component of conceptualism, Stephen Laurence follows Antony’s chapter by asking a closely related question: “Is Linguistics a Branch of Psychology?” His answer is yes; in fact, Chomsky himself has long held linguistics to be a branch of cognitive psychology. Taking a similar tack to Antony, Laurence uses analogies from biology and chemistry to argue against the inadequate criticisms of philosophers who have wished to substitute non-psychological alternatives for Chomsky’s conceptualist framework.

Interestingly, one of the philosophers with whom Laurence takes issue is Michael Devitt—a scholar from the City University of New York who casts doubt on Laurence’s conclusions (if not Antony’s) through a rebuttal in Chapter Four, entitled, “Linguistics Is Not Psychology.” In this chapter Devitt objects that he has been misinterpreted by Laurence and explains that the Chomskyan account of systematic interaction between language and mind need not be undermined by claims that such interaction is not essentially psychological. In fact, Devitt argues, Chomsky’s claims are better undergirded and gain more predictive and explanatory power if we stop thinking of grammar as being true of psychological reality and start thinking of grammar as being true of linguistic reality. Drawing analogies for linguistic competence from such phenomena as bee dances, logic machines, chess players, and horseshoes, Devitt makes a crucial distinction between structure rules and processing rules: although the former actually have descriptive and explanatory potential, the latter are ultimately hidden and unknowable. He further argues that linguistic competence necessarily results in a product separate from its mental or psychological representations—a product that should be, prima facie, theoretically interesting to investigate in its own right.

In the next chapter, “Intentional Content and Chomskyan Linguistics,” Georges Rey deals with another tension in Chomskyan linguistic epistemology, a tension more closely related to the nature of the first component of conceptualism listed in the introduction—namely, the complex rules of language that an ordinary language user is theorized to possess. Simply put, Rey seeks to discover how Chomsky’s computational model can be made to account for and accommodate intentionality—a property of knowledge that is essentially representational and irreducible to the realm of the empirical. Chomsky himself denies that intentionalism has any place in his theory but, curiously enough (as Rey observes), Chomsky nevertheless abides by a “representational” model and frequently employs intentionalist expressions. Rey examines this contradiction in detail and concludes that, whether or not Chomsky will admit to it, intentionalism is both necessary and promising for advancing a Chomskyan computational model of language processing.

Robert J. Matthews’s question in the next chapter follows up on another significant issue in Chomskyan linguistic epistemology: “Does Linguistic Competence Require Knowledge of Language?” While others have argued that speakers know linguistic rules and principles subconsciously and merely “cognize” them tacitly through purely abstract syntactic computations, Matthews argues that in order to be truly linguistically competent, ordinary language users must in some sense employ—if not subconsciously compute—a semantic theory. Without semantics, ordinary language speakers would be unable to adequately map syntactic utterances onto their understanding computationally.

Part Two: Understanding

In chapter 7, “The Character of Natural Language Semantics,” Paul M. Pietroski further investigates Matthew’s semantic claims seeking to demonstrate, essentially, the following point: the semantic content of a given utterance in the mind of a natural language user cannot be equated with the “truth value” of that utterance in the mind of the same speaker. This issue has been the source of much debate and confusion at the intersection of philosophy and linguistics—especially since both sides of the distinction can be referred to as “meaning” in some sense. Pietroski argues that an equation of the two is a confusion of natural language with truth conditional propositions. Since natural language is necessarily linked to complex contextual factors, the meaning of such utterances cannot be treated as truth conditional propositions by philosophers. In the tradition of Plato’s Socrates Pietroski carries much of this argument (including its finer details and implications) through an engaging, whimsical dialogue between the two characters “Phil” (Philosophy) and “Ling” (Linguistics).

In the next chapter, “Grasping Objects and Contents,” Reinaldo Elugardo and Robert J. Stainton seek to explore the nature and implications of what they refer to as “subsentential speech.” How is it that we are able to understand utterances that involve conversational implicature, indirect speech acts, metaphor, and irony? When a speaker enters a room and waves a box of cigarettes, for example, saying, From France, how is it that the audience comes to understand the intended meaning? Much recent argument in the philosophy of language that links truth conditions to natural language is called into question by such phenomena. In fact, such subsentential speech strongly suggest that much more is required for understanding other speakers than knowledge of language—no matter how contextualized or disambiguated such knowledge may be. The authors argue in the end that although the mind does not translate the missing sentence parts into full natural-language sentences in order to understand them, understanding could not take place unless the mind at least converts both the uttered and non-sentential information into “mentalese,” the subconscious language of the mind.

Chapters 9 and 10 are entitled, respectively, “Knowledge of Meaning” and “Understanding and Knowledge of What Is Said.” In the former, Stephen Schiffer takes on the tedious task of describing what it is to know what a given expression means. In the latter, Elizabeth Fricker seeks to describe how it is that a hearer of natural language gains knowledge from belief about a given utterance. Schiffer holds that knowing what something means does not simply consist of knowing propositions about it; but, consists, rather, in occupying a specific processor role with prescribed conditions in relation to what is being said. In her chapter, furthermore, Fricker holds that we do not understand what has been said simply by knowing what has been said—the two are distinct, though related, processes.

Alex Barber himself writes chapter 11, “Truth Conditions and Their Recognition.” In this chapter he proposes that in order to properly describe and explicate the relationship between meaning and truth conditions, we must introduce intentionality and belief. Simply put, Barber’s reanalysis holds that an ordinary language user’s semantic beliefs form the basis of his or her linguistic intentions. Through this unique combination, truth conditions are granted the possibility of a relationship with semantic knowledge without necessitating an exact correspondence between the two. This turns out to be a key distinction—one that enables a break with many assumptions held by the analytic philosophical tradition spanning most of the twentieth century.

Part Three: Linguistic Externalism

In “Externalism, Logical Form, and Linguistic Intentions,” Peter Ludlow sets the stage for the penultimate section of the book by wrestling with an interesting problem: since there is very good evidence to suggest that our internal mental states are at least partially determined by a speaker-external environment, why shouldn’t we suppose that the logical forms of our utterances are also dependent on external properties? The logical forms of our utterances become an issue of debate in such cases as the content of names that refer to individuals who are fictitious such as Santa Claus or are possibly fictitious such as Plato’s Socrates. If Socrates was an actual person, our reference to him as a literal, living individual in history past has a different truth value than if he never existed as a literal, living individual. Thus assuming truth value is contingent on speaker-external factors would seem to be quite valid. The theory runs into problems elsewhere in the use of language, however, as Ludlow himself describes.

In chapter 13, Gabriel Segal follows up with a related essay entitled “Ignorance of Meaning.” Segal argues for a “holistic” account of knowledge in which differing cognitive content is necessitated by differing beliefs and assumptions surrounding given utterances in a given language user’s idiolect. On this account, the word tiger, for example, would mean something quite different between two speakers, one of whom believes that tigers originated in Africa and the other of whom believes that they originated in India. Segal admits however, that either would be afraid to make the tiger angry if he or she were placed in a cage with one—thus conceding that although meaning may differ radically from person to person it is still in some sense simultaneously influenced by social normative content. He ends his article by asserting that theorists of cognitive content are still very ignorant of the nature of meaning themselves.

In the next chapter, “Externalism and the Fregian Tradition,” Jessica Brown takes issue with some of Segal’s arguments set forth in the previous chapter. She holds that internalist notions of cognitive content such as Segal argues for are largely incompatible with what she calls “socially individuated” cognitive content—otherwise known as “social externalism.” Instead of the internal, organic approach to knowledge that Segal argues for, she presents as an alternative that, “a subject may have two thoughts with the same content at a single time” (453).

Part Four: Epistemology through Language

In chapter 15, Alexander Miller asks, “What Is the Acquisition Argument?” After explaining the argument, he seeks to counteract it in the interest of semantic realism. Semantic realism holds that in order to understand certain sentences that we evidently understand, we must be able to grasp truth conditions that transcend the abilities of our conscious recognition. But the acquisition argument holds such a scenario to be impossible since all of our training in the use of sentences teaches us to recognize only those truth conditions that we are able to be consciously aware of. Miller’s critique of this argument hinges largely on a critique of the nature of language acquisition itself, which takes place under circumstances that are not highly rational to start with.

James Higginbotham wraps up the book with chapter 16, entitled “Remembering, Imagining, and the First Person.” In this final chapter Higginbotham develops our understanding of “de se knowledge” (or “knowledge of the self”) through reflections on such linguistic phenomena as pronominal and reflexive anaphora in English. The sentences I remember falling downstairs and I remember myself falling downstairs, for example, reflect what may be fundamentally different memories through which we may gain distinct perspectives on de se knowledge. In the former example, which uses a zero anaphor, the speaker is remembering the experience in a much more immediate, first-personal way. In this case the speaker asserts that he or she is immune from error. The second utterance, on the other hand, leaves open the possibility that the speaker never actually fell at all. Higginbotham examines many more such variations and points out ways in which such distinctions made by the use of anaphoric reference in English have implications for various issues in the applications of logical form.

Observations and Conclusions

Overall, I appreciate the flair and expressiveness with which these authors are able to handle highly technical topics. The general content of the articles is also instructive, relevant, and informative. I would recommend the book to anyone who wishes to be up-dated on current debates in the philosophy of language, or to anyone interested more generally in the interaction between language, mind, knowledge, thought, meaning, intuition, and logic.

I should point out, however, that the authors appear to be primarily from philosophy departments rather than linguistics departments—a fact that is relevant in itself for field linguists who might approach the book as a potential read. Given this background one would not expect to find extensive linguistic data from a typologically broad cross-section of world-languages to back up the linguistic claims made in these articles—a feature that may undermine its integrity in certain regards. In the words of James A. Matisoff (1978:230), “If a linguist has really steeped himself in an alien language or language family, he may well be in a much stronger position to advance valid theoretical claims . . .”

In fact, my enthusiasm toward an otherwise engaging subject matter is somewhat muted by my own developing assumptions regarding linguistic knowledge and ordinary language users—a set of related observations that seem to be neglected by Chomskyan conceptualism. I owe most of these assumptions to a slowly-growing acquaintance with the functionalist traditions, a collection of linguistic theories that seem to have as much to say about mind and knowledge as they have to say about actual linguistic data and ordinary language users. Contrast, for example, the following two functionalist perspectives with those advocated in Barber’s volume. First, suppose I approach the mind-language situation from the proverbial top down; but instead of assuming that language is a “complex system of rules or principles,” I assume that language is “a system of communicative social action” (Van Valin 1993:2). In fact, by opting for the latter (functionalist) perspective, my model of linguistic epistemology will be affected as much as my data analysis. On the other hand, suppose I approach the situation from the “bottom up;” but instead of assuming that a linguist’s job is to codify a given speaker’s abstruse linguistic competence computationally in a generative grammar, I assume that a linguist’s job is to codify the ways in which ordinary language users employ speech pragmatically to conventionalize constraints on inference (LaPolla 2003). Once again, by opting for the functionalist perspective, both my data analysis and my understanding of the nature of linguistic knowledge in the minds of the “ordinary language users” who provide me with language data will be influenced toward markedly different conclusions.

The contributions of certain authors (such as Laurence, pp. 69–106) might especially benefit from such functional reanalyses. Whether functional reanalyses of linguistic epistemology would be derivative from—or opposed to—the general Chomskyan framework of competence and conceptualism remains to be adequately demonstrated. Devitt (pp. 107–139) comes tantalizingly close to providing an epistemological rationale for functionalism in his reply to Laurence, but he ultimately remains verbally loyal to the generative program (p. 124).

This caveat, significant though it may be, should by no means be taken as a wholesale rejection of Barber’s brilliant collection. In fact, numerous statements and claims in the book (such as the following by Pietroski (p. 222): “People refer to things, words do not”) give good reason for thinking that Chomskyan thought itself is yielding more and more ground to the functionalist program. Also of notable significance, the theoretical tone of the book tends to affirm the demise of Wittgensteinian logical positivism—a philosophy of language that, as Barber himself notes (pp. 30–31), resulted in unfortunate complications for philosophy in general and for the philosophy of language in particular. In Barber’s words: “bad linguistic solutions to genuine philosophical concerns often depend on questionable claims about language” (p. 31). For the sake of progress in linguistics and philosophy alike, we may hope that the near future will enjoy a continued decrease in dubious claims regarding the fundamental nature of language in particular.