Language, Philosophy of: Modern

views updated

Language, Philosophy of: Modern

Although discussions of language in seventeenth-and eighteenth-century philosophy foreshadowed many issues that came to full bloom in the twentieth century, before the twentieth century language was thought to have a secondary role in understanding the special place human beings have in the world. The fundamental concern was the problem of knowledge. How is it possible for human beings to have knowledge of the world? The solution for both rationalists and empiricists was to be found in the nature of mind since, as was universally held, all we can know directly are the ideas of our own minds. To solve this problem of knowledge, philosophers had to account for how we grasp generalities (and not just particulars) and how we determine which of our ideas represent the world truly.

Language was seen as the public conventional medium for communicating private thought. A simple denotational theory of meaning predominated. The meanings of words are their denotations, the objects that the words stand for or denote, or the ideas of particular objects. Proper names, like "Silver," denote particular objects. There was great controversy over what general terms, like "horse," denote. Three major positions emerged: (1) realism or Platonism, the view that general terms name real abstract objects (horsiness); (2) conceptualism, the view that terms stand for abstract ideas or concepts (the concept of horsiness); and (3) nominalism, the view that a term is general if it denotes more than one particular object. The important problem concerned the nature of abstract ideas. The solution to the problem of general words would follow from this.

Founders of the Twentieth-Century "Linguistic Turn"

In the twentieth century, Anglo-American philosophy took "the linguistic turn." Instead of seeking solutions to problems of knowledge and thought in an examination of the nature of our ideas, philosophers looked to the nature of language. This great shift began with Gottlob Frege's foundational work in mathematical logic.

Logical Syntax and Semantics

Well into the nineteenth century, Aristotelian logic dominated. Logic was seen as the study of thought itself. Gottlob Frege (1848–1925) revolutionized our conception of logic and its relation to thought and language. Frege's key insight was to see that formal arithmetic modeling can be used to display the structure of language. Just as "x y z" expresses the form of all instances of addition through the use of variables (x, y, z) and a special sign for the addition function (), so language can be modeled. "Horses are mammals" can be written in a special concept-script (Begriffsschrift ) that displays the different roles that the constituent words play in the sentence, "(x)(Hx → Mx)." This is done in a way that abstracts from the meaning of any particular sentence. The new logic, thus, distinguishes the formal features of language from meaning, laying the groundwork for the tripartite distinction between syntax, semantics, and pragmatics.

Syntax concerns the rules for combining expressions into well-formed sentences within the language while semantics gives us a theory of the meanings of words and sentences. On this view, language is a formal object that consists of a finite interpreted lexicon and a finite set of recursive rules for combining lexical items. Recursive rules are rules that can be applied repeatedly. This permits the construction of any of an infinite array of possible sentences. The lexicon consists of several kinds of terms—proper names, predicates, and relations, each distinguished by its distinctive role within a sentence. This conception of language, which continues to dominate discussion, is a representational theory of language, for it treats the essence of language as the representation of possible states of the world using finite resources. Pragmatics studies whatever practical and contextual aspects of language use are left.

Frege's semantic theory.

In trying to apply the traditional denotational theory of meaning to language, Frege identified a number of problems. One of the most important arises with "contingent identity" statements. The sentences "The Morning Star is the Morning Star" and "The Morning Star is the Evening Star" have different "cognitive values." If meaning is just a matter of denotation, then it is hard to see how this could be so. Frege's theory distinguishes two kinds of meaning: reference (what the term denotes) and sense (the mode of presenting the denoted object). Each meaningful expression must have both a sense and a referent. Names ("Walter Scott") and definite descriptions ("the author of Waverley ") denote particular objects by way of their mode of presenting the object. The sense of a predicate expression is the mode of presenting a concept, the referent of the expression. Questions about the nature of, and relation between, sense and reference have been at the center of philosophy of language ever since.

Logical analysis.

Frege came to logic and the philosophy of language through concerns with the foundations of mathematics. Bertrand Russell (1872–1970) and Ludwig Wittgenstein (1889–1951) shared this concern but saw the new logic as the key to philosophy in general. Both held that the new logic would enable philosophy to break with its metaphysical tradition, by showing that metaphysics had resulted from mistakes and confusions rooted in a failure to understand language properly. One must distinguish between the surface grammar of ordinary sentences and the underlying logical form of these sentences. The new logic is the means for characterizing this deep structure (or logical form), a structure hidden by surface grammatical form. Logical form is reached through a process of logical analysis.

Russell's classic example of an analysis that removed puzzlement was his treatment of definite descriptions. It was thought that definite descriptions ("the author of Waverley ") should be treated like proper names; to be meaningful they must refer to some particular. But this creates a problem for non-referring definite descriptions like "the present king of France." This is a meaningful phrase and yet there is no individual answering to this description. Russell's solution was to argue that the underlying logical form of the expression is quite different from what it appears to be on the surface. Analysis reveals that the sentence "the present king of France is bald" has the logical form of "There is one and only one person such that that person is now king of France and is bald." Russell's analysis eliminates the apparently referring expression "the present king of France." Correctly analyzed, this puzzling sentence turns out to be merely false.

Wittgenstein took this idea even further. In his first work, Tractatus Logico-Philosophicus (1922), Wittgenstein argued that all sentences (in any language whatsoever) have a determinate meaning. A sentence says that things are thus-and-so. Wittgenstein took this to imply that any sentence must be analyzable into a set of "elementary" sentences composed of constituent words that denote simple objects. Complex sentences are just functions of combinations of elementary sentences. Strings of words that look like sentences, but resist analysis, are not really sentences at all. They are nonsense. Only the sentences of ordinary factual talk and the natural sciences are meaningful. Ethical, aesthetic, and religious statements, though important to our lives, are strictly meaningless. Most (if not all) traditional philosophy is nonsensical on this view.

Logical Positivism and Its Challengers

The next important step was taken by a number of leading physicists, mathematicians, and philosophers who formed a group in Vienna known as the Vienna Circle. The Vienna Circle was committed to a repudiation of metaphysics in favor of science, and they saw in the ideas of Frege, Russell, and Wittgenstein the philosophical foundation for their movement. They embraced a philosophy of "logical positivism." At the heart of this movement was the thesis, known as the principle of verification, that only sentences making empirically confirmable claims are meaningful at all. Unverifiable sentences are nonsense. In the 1930s, many positivists had to flee Austria and Germany. As émigrés, they exercised a profound influence on philosophy in the English-speaking world.

Challenges to the positivist conception of language.

The 1950s and 1960s saw a reaction against the positivist theory of language. The challenge had two sources: the ordinary language movement of Oxford University and a powerful critique of the very idea of meaning by the American philosopher W. V. O. Quine of Harvard.

Ordinary language movement.

Wittgenstein's Philosophical Investigations (1953) and J. L. Austin's writings were the mainsprings of the ordinary language movement. These philosophers retained the positivist suspicion of metaphysics as confused theorizing, but they rejected the celebration of science at the expense of all other forms of understanding and its theory of language. Wittgenstein and Austin, in their different ways, argued that the theories of language represented by Frege, the Tractatus, and the positivist movement are themselves as confused as the metaphysical theories they deride. One of the striking consequences of the verificationist theory of meaning is that ordinary objects, from pigs to chairs to electrons, are logical constructions out of our sensory experiences, a view called "phenomenalism." Phenomenalism is the theory that all empirical sentences are analyzable into sets of sentences about actual and possible observations that would confirm (or disconfirm) the truth of the sentence. Meaning is ultimately rooted in sensory experience. "This is a penny" is thus analyzable into a set of experiential sentences: "this is copper-colored, round, hard and cool; or this is copper-colored, elliptically shaped, hard and cool; or …" The full analysis would consist of a potentially infinite disjunction of experiential sentences, one for each possible experience of a penny.

Critics argued that the positivist theory is belied by the realities of ordinary language. They used nuanced descriptions of our actual uses of language to challenge the claims both that language has an underlying logical form revealed through analysis and that referential theories of meaning, and phenomenalism in particular, are correct. Early referential theories state that the semantic relation between words and the world is fixed by ostensive definitions, that is, by naming an object. But ostensive definition cannot fix meaning since any object has many distinct properties, which the ostensive definition in itself does not distinguish. Defining "horse" by pointing to Silver does not show whether the expression is intended to be a proper name or to pick out Silver's horsiness or his being four-legged or white. So, it is concluded, reference cannot explain meaning, but presupposes it.

Meaning seems better explained in terms of how words are used in connection with our actions and interactions with the world and each other. Meaning is not the denoted object, but the use to which the word is put in practice. This idea of meaning as use brings with it a holistic conception of meaning rather than the atomistic conception of referential theories of meaning. A number of different positions developed within this movement: (1) Some advocate a use theory of meaning, which replaces the idea of language as a single systematic totality with that of an array of overlapping ways of using language. It replaces the search for necessary and sufficient experiential conditions for the applicability of an expression in favor of criterial grounds, where criteria constitute necessarily good evidence for the presence of some object while nonetheless falling short of entailment of the presence of that object. (2) J. L. Austin and others introduced the speech act theory of language, according to which the act of utterance and the context within which it occurs is the starting point for a theory of meaning. This too leads to a holistic conception of meaning and blurs the distinction between what belongs to semantics (the meanings of our words) and what belongs to pragmatics (the background and contextual considerations that inform actual speech). (3) Lastly, there are those who see in the ordinary language critique of the representational picture a rejection of the possibility of theorizing about language at all. There can only be the diagnosis of error in any such attempt.

Quine's philosophy of language.

The second great challenge to the dominant picture of language comes from W. V. O. Quine (1908–2000) with his critique of the distinction between analytic and synthetic truth ("Two Dogmas of Empiricism," 1951) and his more general attack on the very idea of meaning (Word and Object, 1960). Sentences that are analytically true are true solely in virtue of the meanings of the constituent words. The truth of synthetic sentences, on the other hand, is a function both of the meanings of the words and of the nature of the world. It is the difference between "Bachelors are unmarried" and "Sam is a bachelor." Quine argues that the sentences we accept as true hang together holistically in a "web of belief" that can be adjusted at any point. He concludes that there is no point in classifying some sentences as true by meaning and some as empirical.

In Word and Object, Quine introduces what proves to be the greatest challenge to the dominant view. With "the museum myth of meanings" (words as labels for objects) fully discredited, we must look at language in a new way. The way to understand meaning is to ask what translation preserves. Much translation is customary, so we should look at radical translation. The situation of radical translation is one in which the linguist seeks to translate a language that is wholly alien to him. We find that there is no uniquely correct way to specify the translations of the unknown language into the known language ("the indeterminacy of translation") and no way to specify determinately what objects or properties the terms refer to ("the inscrutability of reference"). Further Quine attacks the idea that language has a logical form that is captured by formal logic. He argues that such an ideal is at best the regimentation of a part of our language. Quine's challenge set the agenda for the 1960s and 1970s.

Philosophy of Language since Quine

Donald Davidson (1926–2003) builds on Quine's project, developing his own interpretationist theory of language. Davidson argues that truth is the fundamental semantic notion and that meaning is given by specifying the truth conditions for a sentence. This specification can only be achieved by a method of radical interpretation, which requires treating the speaker as holding mostly true beliefs. This approach differs from the representationalist picture in the following ways: Reference is not the fundamental semantic relation between language and the world, but a derivative relation; meaning and belief are interdependent—there is no way to separate them in a principled fashion; the method of radical interpretation requires, as a normative constraint, that the speaker be rational.

The 1990s brought a defense of inferentialist theories of meaning against representationalist theories. On an inferentialist view, the meaning of a sentence is a matter of the inferential relations that acceptance incurs. Instead of thinking of a sentence as representing a state of affairs, we should think of the sentence as entitling and committing the utterer to say and believe other things. To assert "Horses are mammals" is thereby to be committed to holding true "Horses are warm-blooded" and to rejecting as false "Horses lay eggs."

New theories of reference.

The second reaction to Quine was the construction of new theories of reference that are held to avoid the problems of the simple denotational theory. These are causal theories of reference and "direct reference" theories. Causal theories of reference identify the reference relation between words and properties with causal relations. Direct reference theories state that the object or property referred to just is the meaning of the expression; the reference relation is not explained by way of ostensive definition or some special causal chain. Both approaches free the notion of reference from necessarily involving some kind of mental act. Saul Kripke introduced key elements of both theories in his highly influential book Naming and Necessity (1972). There he maintains that referring terms "rigidly designate" their objects. Once an object has been baptized, that name denotes its object in any possible world in which the object (is specified to) occur(s). What secures the link for those who have no direct relation to the object is the socially transmitted causal chain that extends from the baptism to subsequent uses of the name. Kripke extended his theory of rigid designation and socially sustained causal chains to natural kinds as well as individuals.

Realism vs. Anti-Realism.

This debate concerning the nature of reference and whether it can explain the meaningfulness of language is at the heart of a philosophical debate that dominated the 1980s and continues to be felt. In the 1980s, the great metaphysical debate between realism and idealism was recast in terms of the relation between language and the world. This debate concerns both how reference and truth should be understood and whether the correct theory of meaning for language is a truth-conditional theory or an assertability conditional theory. A truth-conditional theory holds that the meaning of a sentence just is the truth condition for that sentence. An assertability conditional theory of meaning states that the meaning of a sentence is given by its assertability conditions, namely, the conditions by which one can recognize the statement is true. These are distinct versions of the representationalist picture of language. This debate became focused on vying theories of truth. Michael Dummett, the leading defender of the assertability conditional account, characterized the difference between realist and antirealist positions in terms of whether a statement could be meaningful and yet transcend any possible evidence for its truth or falsity.

This interest in truth as a semantic value has generated a lively debate. Both the realist and the antirealist hold substantive theories of truth. The realist identifies the truth of a sentence with its corresponding to the facts. The antirealist assesses truth in terms of coherence with other sentences and our recognitional capacities. Challenging both of these positions are "deflationary" theories of truth. Such theories deny that truth has a substantive nature (whether correspondence or coherence). The truth predicate is a device for semantic ascent, for moving from the sentence "snow is white" to the sentence "the sentence 'snow is white' is true." There are things that one can say by talking about sentences that are not easily said in other ways.

Other issues.

The 1990s saw a number of developments. Most important was the debate between inferentialist theories of language and representationalist theories. Other important issues concerned the nature of "vague" predicates like "bald" that have fuzzy borders (is the "fuzziness" a feature of the property itself or our ignorance?), the compositionality of meaning (is it essential to the explanation of meaning?), innate rules of grammar, and the infinite reach of language. Many of these issues have become highly technical and specialized, but they are all in the service of the great debates about the general character of language and meaning.

See also Language and Linguistics ; Language, Philosophy of: Ancient and Medieval ; Linguistic Turn .

bibliography

Austin, J. L. How to Do Things With Words. Edited by J. O. Urmson. Oxford, U.K.: Clarendon Press, 1962.

——. Philosophical Papers. Edited by J. O. Urmson and G. J. Warnock. Oxford, U.K.: Clarendon Press, 1961.

Ayer, A. J. Language, Truth, and Logic. 2nd ed. New York: Dover Publications, 1946.

Brandom, Robert. Making It Explicit: Reasoning, Representing, and Discursive Commitment. Cambridge, Mass.: Harvard University Press, 1994.

Davidson, Donald. Inquiries into Truth and Interpretation. Oxford, U.K.: Clarendon Press, 2001.

Dummett, Michael. "What Is a Theory of Meaning? I and II." In his The Seas of Language. Oxford, U.K.: Clarendon Press, 1993.

Frege, Gottlob. "Begriffsschrift" and "On Sense and Reference." In Philosophical Writings of Gottlob Frege, edited by Peter Geach and Max Black. Oxford, U.K.: Blackwell, 1969.

Kripke, Saul A. Naming and Necessity. Cambridge, Mass.: Harvard University Press, 1972.

Quine, W.V.O. "Two Dogmas of Empiricism." In his From a Logical Point of View: 9 Logico-Philosophical Essays. New York: Harper & Row, 1961.

——. Word and Object. Cambridge, Mass.: MIT Press, 1960.

Russell, Bertrand. "Lectures on Logical Atomism." In Logic and Knowledge, edited by Robert Charles Marsh. London: George Allen and Unwin, 1956.

——. "On Denoting." In Logic and Knowledge: Essays, 1901–1950, edited by Robert Charles Marsh. London: George Allen and Unwin, 1956.