Logicalization

R.E. Jennings

LLEP, Department of Philosophy,
Simon Fraser University, Burnaby, British Columbia,
Canada, V5A 1S6
(jennings@sfu.ca)


Prerequisites: There are no formal prerequisites for this course except sufficient logic for familiarity with truth-tables and connectives.

Summary: For the dispassionate observer, if one could be found, it would be a curious matter that in the study of language, so much attention should be paid to semantic questions. The reason is that to observation, language does not present itself as a semantic phenomenon. Indeed, what would a semantic phenomenon be? Evidently, language, more obviously spoken language but written language as well, is a physical, and more particularly a biological phenomenon, one human organism producing sound streams and inscriptions that have neural and other physical effects on other human organisms. Why should it then be studied under any other heading than that of biology? Why should we not engage ourselves in constructing explanatory theories using methods similar to those of the empirical sciences? One reason that has suggested itself to philosophers is that the functional significance of language is conventional. By this is not meant that there has been any convention at which such matters have been settled, but rather that matters have settled themselves in ways which are in important respects arbitrary, a fact attested to by the variety of human languages, which for all practical purposes work equally well, though using different sound and inscriptional schemes for similar functions. But these facts alone do not justify an abandonment of ordinary physical explanatory methods; they merely shift the subject matter of the investigation. The causal significance of any particular linguistic activity itself has a causal history; the aim of the biology of language is to seek detailed particular causal explanations for present causal roles and to enunciate the laws governing such second-order causal histories as these laws emerge. There is no reason to suppose that in principle this cannot be done at whatever level of detail interests us; there is no reason to suppose that in principle a useful detailed semantic account of language can be given.

There are some very general constraints upon such a theory. First, it must be capable of integration with organic biology and so with the rest of physical theory. Second, the methods that it adopts, and the types that it recognizes must be such as to make it capable of integration with an eventual evolutionary account of the emergence of language itself. Thus, if it uses the language of meaning, then it will understand meanings as types of physical effect rather than as, say, intentions. The evolutionary question becomes: how did the genus of specifically linguistic effects evolve from the genera of non-linguistic ones, and from which? The more local question within a specific language is: how did the effects of particular linguistic items evolve from the effects of ancestral linguistic items? In some instances we can give satisfying answers to such questions for linguistic items for which we can find no satisfying semantic account.

The aim of this course of lectures is to bring this theoretical approach to the understanding of the so-called logical vocabulary of natural language, with main examples drawn from English connective vocabulary. The broad facts of the matter are well understood by historical linguists. All of the functional, and therefore all of the logical connective vocabulary of every natural language has evolved from lexical ancestral vocabulary, principally from adjectives, adverbs, and verbs of spatio-temporal physical relationship. T be usefully informative, an account must o do more than gesture vaguely toward etymologies.

The process by which connectives evolve is called logicalization. In the English language it has yielded, for example, or from other (meaning second as in Ďevery other dayí), and from the language of spatial boundary, and if from the ancestor of the verb give. Since logicalization continues, lexical vocabulary of current English may be expected to yield some of the connective vocabulary of successor languages. Although there are residual semantic questions about the connectives that have actually evolved, the process itself is best described in biological rather than semantic terms, and the semantic questions that remain are best answered in the light of that biological account.

Course Outline: Lecture 1: The Semantic Illusion
We donít need a semantic understanding of the connective vocabulary of a language in order to use it in speech. That we do not understand the semantics of connectives of English becomes evident in our flawed attempts at semantic accounts. A quick survey of introductory logic texts shows us where we go wrong and at the same time shows us the sort of thing that needs to be explained.

Lecture 2: The Case of Or
Language as a biological phenomenon. A quick look at the biological approach and a sampling of the method. Species defined. Meanings as species of effect.

Lecture 3: Where Connectives Come From
Composition and compositionality. It is tempting to suppose that meaning is in some way compositional. In fact composition is the instrument by which meanings evolve. This lecture includes a survey of the evolutionary processes at the early stages of logicalization.

Lecture 4: Mutations
It is a tempting assumption that the semantics of logical vocabulary is a reliable bedrock upon which a larger semantic theory of language can be constructed using broadly similar methods. In fact, from a biological perspective, connective vocabulary is among the most fragile of the constituents of language, and composition is the instrument by which it is changed. The apprehension of speech involves syntactic replication. Undetected faulty replication produces mutations. The formal notion of duality (the relationship between, for example, and and or, and all and some) generalized to accommodate the historical developments by which dualized uses evolve through mutations. The stages of mutation are explained. Mutation and idiom: how lexical vocabulary resists mutation.

Lecture 5: What Logicalization Can Teach Us About Language Origins
More general issues: where language comes from, where syntax comes from, the connection between language and technology. Ignorance explained: the consequences of a biological understanding of language for approaches to traditional philosophical questions.

The most common reaction to this approach used to be "Why canít we just have Grice? Itís so much simpler." None of the lectures will explicitly mention Grice on connectives, but some may find it useful to have seen a contrasting approach. They should read some of Griceís essays, including 'Logic and Conversation'. Others may wish just to reflect on these questions: "Why do we say such things as 'My but thatís a lovely X'?" "Why did Richard Nixon say `I have no doubt but that the Viet Cong will be defeated'?".

Course Notes:
How Connectives Evolve in Natural Language
What Logicalization Can Teach Us About the Origins of Language
Proof and Conversation
The Semantic Illusion
The Meanings of Connectives
Course Slides (PowerPoint format)

Back to course listing.