Signs, Mind, and Reality is the late Sebastian Shaumyan's final version of his understanding of semiotics and linguistics.  He tries to lay the foundation of a semiotics/linguistics that works in the tradition of Saussure and early (French and Russian) structuralism.

The book is divided into ten chapters. Chapter 1 and 7 reflect on the methods of linguistics. Chapters 4 to 6 develop Shaumyan's main principles and theses. One of these is that the word is the more important linguistic unit than the sentence, which is elaborated in chapters 8 to 10. Chapter 11 introduces a small formalism to use with this semiotics. Chapters 2 and 3 place semiotics in a broader philosophical context discussing meaning and the construction of a shared (social) world.

It is difficult to write about Signs, Mind, and Reality in a non-sarcastic manner. The book is ripe with self-praise, self-admiration and hyperbole. Shaumyan praises his work as differing 'radically from all trends of contemporary linguistics'. He sees himself 'like Ariadne' advancing a new leading principle. His principle (of Difference) 'does for linguistics what the laws of Galileo and Newton have done for mechanics'.

In contrast to Shaumyan's praise for his own work he condemns the major approaches in current linguistics (especially cognitive linguistics in the form of generative/transformational grammar and formal linguistics/semantics) without detailed arguments or more convincing solutions to problems left open by these approaches. He just asserts that 'linguistics has found no clear answer' to questions of meaning (compositionality, it seems) and phonetic classes. He blames the whole idea of doing linguistics in a logic related way as misguided and as missing the proper field of study using notions 'alien to the nature of language'. He quotes Chomsky out of context to have a famous witness against formal approaches. He comes up with the ridiculous claim that generative grammar — which version? — claims that "the brown book" is generated by a transformation out of  "the book is brown" ; any undergraduate textbook on transformational grammar has something better to say about attributes and adjectival phrases than that. 'Syntax' is said to have forgotten the word by looking at sentences — and so forth. According to Shaumyan 'linguistics is going through a crisis of its foundations', only that nobody lives up to it, it seems.

At several points the examples and sub-theories do not live up to the pronouncements by Shaumyan. Often it is far from clear what is meant at all:

a)      Shaumyan's basic principle 'of Difference' works with an insufficiently clarified concept of sign (derived from 'the relation of being a sign of') that seems to contrast from both word and meaning. Claiming that meanings cannot differ without the sign differing has the prima facie incredible consequence that (river) 'bank' and the financial institute 'bank' are the same meaning. Shaumyan adds a clause to the principle to exclude signs in 'alternate distribution' thus making room for homonyms like 'bank'. The activity (to) 'pan' (with a camera in a movie) and the kitchen 'pan' are not in alternate distribution and nonetheless not the same meaning. The converse claim that 'only those signs are different which correspond to different meaning', also part of Shaumyan's principle would make, it seems, 'bachelor' and 'unmarried man' the same sign! Shaumyan's presents his main principle as if it spoke mostly for itself, which it does not.

b)      Meanings separated from signs are said to be mere 'concepts'. Then again 'Meanings are concepts.' Then again: 'The definition of the identity of meanings has nothing to do with whether they are mutually related or unrelated from the conceptual point of view.' Some clarifications seem needed here.

c)      Shaumyan sees each language as providing 'a folk model of the world', a familiar thesis. He adds, however, 'a collective philosophy unique to each language'. This may be read harmlessly, in as much as the world views overlap considerably. With Shaumyan it often sounds like the version of linguistic relativism that made it into the common mind as the so-called 'Sapir-Whorf-thesis' (claiming, according to folklore, Hopis having no concept of time, whereas Inuit had 13 words for snow etc.), which is refuted nevertheless.

d)     Phrases like 'the dialectics of language', 'the language-thought continuum' understood in analogy to 'the energy-matter continuum' may ring no bell in many readers. They seem to be gibberish: 'Semiotic identity is the repetition of differences.' Sometimes scientific phrases are misused: the 'language-thought duality' is said to be 'both continuous and discrete', which sounds remarkable and would be a plain contradiction, but it turns out that 'continuous' is not used in its mathematical sense here (being opposed to 'discrete') but in a more lax fashion of 'interwoven'.

e)      The whole methodological reflection consists in large parts of ill-digested discussions alive in the philosophy of science in the 50s and 60s (e.g. about building abstract models and confirmation). Apart from that it contains the methodological credo: 'Theoretical linguistics is not concerned with generalizations from the large amounts of data from a wide variety of languages. The proper business of theoretical linguistics is conceptual analysis of common, well-established facts.' This does not sound like empirical linguistics at all, rather like some dyed in the wool adherent of philosophical conceptual analysis.

f)       Taking the word as the basic unit, as Shaumyan does, makes semantically speaking no sense, since it is by the truth of sentences/statements that we connect to the world. That was one of Frege's main insights in semantics. Words are to be understood by their role in meaningful sentences/statements which may be evaluated for truth or falsity. Individual words say nothing, we understand them only because we have understood sentences in which they have occurred. Shaumyan adds the absurdly mistaken accusation that for Montague or Chomsky 'and most other linguists' words stand for objects (as their meaning)! Nonetheless he defines: 'What is called the meaning of a sign is the correspondence of the sign to the thing it represents'.

g)      The formalism provided is a cheap version of a categorical grammar (with functional application of types). By leaving out, for example, intensional types it is far from clear what structures should lay within its generative horizon.

Of course a lot is right in Signs, Mind, and Reality, but, for example, proposing 'sentence articulation laws' that deal with transitive and ditransitive predicates and attributes is way beyond the level of sophistication in current linguistics. Extracting such observations is not worth the effort. Signs, Mind, and Reality is a book that for some reason made it into a edited cognitive science series, but that no one needs to read.

© 2007 Manuel Bremer

Manuel Bremer, Heinrich-Heine-Universität Düsseldorf, Germany

Categories: Philosophical