Approaches to the Implementation ofa correspondence theory of Meaning

Copyright Antar S. Abdellah 2003 1 Approaches to the Implementation of a correspondence theory of Meaning. By Antar S. Abdellah Semantics is the branch of Linguistics that has as its field the study of meaning and the correspondence between words – symbols of language- and their entities in the real world. Although Linguistics in general has long been the subject of exhaustive research, Semantics was the branch that has got the least attention in regard to forming a solid theoretical frame work for interpreting the association of words and sentences with their meanings. Syntax, Phonology and morphology on the other hand had rich and clear theories long before there was one for Semantics. One reason for this is perhaps that the subject of formal semantics is not just a lexical relation between a word and its denotation, but rather what constitutes meaning, the nature of this relation, and the necessary and sufficient components of meaning, among others. In this article, we will review the arguments that led to the implementation of a theory of meaning based on the correspondence between a symbol and its entity. Gamut (1991) explains that there are three main approaches to a linguistic theory of meaning; 1- the ‘meaning is use’ theory led by Wittgenstein that states that the meaning of a symbol resides in its use, 2- the ‘stimulus- response’ theory led by Skinner, Morris and Bloomfiled that states that the meaning of a symbol is identified with the set of stimuli which elicit the use of that symbol in response, and 3- the ‘partial correspondence’ theory led by Grice in which correspondence to entities is thought to account for one aspect of the total meaning of symbols. Gamut points out that in all three approaches, two main points are discussed; the nature of the relation between symbols and entities, and the nature of entities themselves. The nature of the relation between symbols and meaning has been the matter of dispute from the time of the Ancient Greek. One traditional view here is that of naturalism which views the relation between symbol and entity as extremely natural, a type of sound symbolism is inherent in the sounds themselves. Gamut concludes the viability of such approach by arguing that languages would have been naturally learnt without difficulties and onomatopoeia would be the same in all languages, and by referring to the inconsistency of using a certain sound to indicate certain emotion. Another approach is conventionalism which states that the relation between a word and a thing is not natural Copyright Antar S. Abdellah 2003 2 but conventional. These conventions have been discussed in many theories; so they are some times references and senses, and sometimes mental concepts, among other interpretations. De Swart (1998) tracks the historical order of the emergence of interest in Linguistics in general and Semantics in particular. He points out that a historical (diachronic)approach for studying languages was the prevailing attitude during the 19th century. With Saussure’s Course in General Linguistics (1916) the study of meaning was formulated in the shape of a signe which is a combination of a signifiant (form , sound pattern) and a signifiie (meaning or concept) with the relation between signifiant and signifie as arbitrary and based on conventions. De Saussure distinguished between ‘paradigmatic’ and ‘syntagmatic’ relations. The former focus on the relations of signs with other signs in a paradigm, the latter on the relation of signs with other signs in their structural environment. In the second quarter of the 20th century, Bloomfiled’s behaviorism dominated the linguistic scene. His stimulus-response model was the type followed in explaining animal as well as human behavior. He rejected the study of meaning on the grounds that it was not rigorous enough to allow scientific investigation. With Chomsky’s theory of generative grammar, and his latter discussion of the effect of transformations on the meaning of sentences, the interest in meaning was revived, and led consequently to the development of interpretative and afterwards generative semantics. Interpretative semantics proposes that we first develop syntactic structures that represent sentences and then turn these structures into semantic representations. Generative semantics on the other hand, develops devices that generates meanings and map them onto syntactic structures. Lexical decomposition was thus introduced as a device for identifying components of meaning. ‘Die’ therefore would be perceived as ‘ cease to be alive’ and ’kill’ as ‘cause to die’. Generative semantics, though a promising semantic theory, failed for a number of reasons; partly the theory was too vague to form an overall systematic theory of meaning, and partly there were serious problems with the basic ideas. The components that were proposed to form the meanings of words were not always having the same effects as the perceived meaning of the words themselves; an example is the difference between ‘kill’ and ‘cause to die’ in the following pair of sentences: 1- Andrew caused Alex to die on Sunday by stabbing him on Saturday. 2- Andrew killed Alex on Sunday by stabbing him on Saturday. Copyright Antar S. Abdellah 2003 3 where (1) is perfectly acceptable while (2) is semantically anomalous. The delay between the cause and effect is represented properly by ‘cause to die’ but it makes it impossible to use the single verb ‘kill’. Due to the shortcomings of generative semantics, syntax was considered the core of grammar and semantics was more marginal. Semantics was strictly taken as interpretative semantics; this means we first develop syntactic structures that represent sentences and then turn these structures into semantic representations. As a result the study of meaning was again marginalized in Linguistics. Eventually, the study of meaning has strongly been influenced by philosophical and logical traditions of semantics. With the formalization of propositional and predicate logic, logicians had powerful tools to offer for the study of meaning. Gamut (1991) refers to three variants of the correspondence theory of meaning. There are; 1-Conceptualism according to which meaning is a relation between symbols and the contents of consciousness, 2- Platonism according to which concepts are not mental entities but real things- only they do not belong to the world of observable phenomena but to the world of ideas-, and 3- Realism according to which linguistic symbols refer to concrete , observable real entities around us. In the remainder of this article we review some of the approaches that implement a correspondence theory of meaning. Logical semantics theory, Frege’s referential theory, context dependence theory, and the theory of concepts will be briefly discussed. In Logical semantics theory, meaning is a relation between the symbol of a language and certain entities which are independent of this language. We choose a set of entities as a domain and specify the relation between the predicate-logical language in question and this domain. By means of an interpretation function, the constant symbols are assigned individual domain elements, and the predicate symbols are assigned a set of domain elements as their reference. This means that the meanings of the symbols are always their references. Within this theory there are some principles of which the principle of semantic compositionality and the principle of extensionality are the most celebrated. The former states that ‘every sentence ..is the result of a syntactic construction process Copyright Antar S. Abdellah 2003 4 which builds it up step by step, and in which every step can receive a semantic interpretation’. The latter states that ‘if X and X’ have got the same truth value, then X’ may be substituted for X in ? without a change of truth value’ One weakness of a logical theory of semantics is that there are natural languages for which the semantic methods of standard logical systems are not adequate. Another criticism is that if the theory proposes that the meaning is the same as reference, then how can it account for proper names that have no reference in real life like Odysseus or that once had a reference but no more like Socrates. It seems that there must be a strong link between meaning and reality, a point that led Frege to make up his theory of sense and reference. Frege’s theory proposes that there are three important concepts; sign , sense and reference. By sign he means a name, a combination of words or even a letter, by reference he means “the object that which the sign refers to, by sense he means the thought where a “mode of presentation is contained”. Proper nouns have a reference and a sense. The difference between sign and reference is that to a single sign there is a definite sense and a definite reference, while to a definite reference (an object) there does not belong only a single sign. Likewise one sense has different expressions in different languages and even in one language; for example “John” has the reference of “a man”, and the sense “the thoughts associated with this person”, while ‘man’ as a reference does not signify ‘John’ only. Frege discusses in details the case of the following two sentences: 3- the morning star is the morning star. 4- the morning star is the evening star. Although the morning star is the same thing as the evening star, the two sentences are quite different; statement (3) is a tautology, an analytical a priori truth, but (4) expresses a significant astronomical discovery and as such is a synthetic a posteriori statement. Whereas (3) must always have been accepted by everyone, (4) was considered untrue for a long time. Frege concludes that reference and sense are not the same but are to some extent independent from each other; it is quite possible to know what an expression means without being familiar with its reference, and vice versa. Because standard extensional logical is restricted to propositions which are not context- dependent, the intensional logical theory emerged as a context dependent approach to meaning. In this theory the interpretation of meaning is done against the background of the context in which a Copyright Antar S. Abdellah 2003 5 word is used. This context provides ‘the here and the now’ on which the truth of a situational sentence depends, so a sentence like ‘It is raining’ will be true in a given situational context if it happens to be raining in that context. A sentence in the past, however, is more complicated because it introduces an affair which is not present in the given context. That’s why it is often necessary to take other contexts in consideration. The intension of a predicate then is ‘a function which assigns to any given context the set of individuals forming the reference of that predicate in that particular context’. Another important approach to the correspondence between symbol and meaning is the theory of mental representations. This theory places a new level between words and the world: a level of mental representation where a noun is said to gain its ability to denote because it is associated with some mental image , or concept, in the speaker/hearer’s mind, or as Sapir puts it meaning is “a convenient capsule of thought”. Thus in this theory the meaning of a word is a combination of its denotation and a conceptual element. Concepts may be associated with single words; i.e., lexicalized concepts, and may be associated with the meanings of sentences. But one serious problem for this approach is that people have different mental images, or concepts, for the same thing; democracy, freedom, or love. Therefore, it was important to have the development of determining a set of necessary and sufficient conditions by which a concept will be agreed upon. In this sense, concepts are lists of bits of knowledge that form the necessary and sufficient conditions for some thing to be an example of that concept. Saeed (1999) points out that the necessary and sufficient conditions may be required for determining the definition of a concept rather than its meaning as ‘we can use a name for a person or a place knowing little or nothing about the referent’. In order to solve this problem, some semanticists(Rosch and Mervis among others) proposed the notion of prototypes where ‘a model of structured concepts around a central or typical members of a category with a shading off into less typical or peripheral members’. Other approaches include Fillmore and Lakoff frames and idealized cognitive models where they claim that speakers have folk theories about the world based on their experiences and rooted in their culture. Accordingly they differentiate between the dictionary-type definition of a word and its encyclopedia-type entry of cultural knowledge about the word. In this view using a word involves both combining semantic knowledge and encyclopedic knowledge. Copyright Antar S. Abdellah 2003 6 Finally It seems that a theory of meaning has to take in consideration as much complexities of the relation between symbols and their meanings as there are complexities of the human nature itself. In this article light was shed on some remarkable theories that tried to interpret the nature of this relation. _____________________________________ Bibliography De Swart. H.(1998) Introduction to Natural Language Semantics. CSLI. Stanford. Frege. G. (1975) in Donald Davidson and Gilbert Hartman (eds) The Logic of Grammar. Dickenson Publishing, Encino, CA. Gamut, LTF.(1991) Logic, Language and Meaning, vol II: Intensional Logic and logical grammar. University of Chicago Press. Saeed. J (1999). Semantics .London: Blackwell publishing. *Word count: 2111words.