According to this generally accepted point of view, the natural world around us cannot be an expression of language. Language is only that which represents it. However, this perspective does not hold up so well to deeper philosophical analysis. We see, for example, that even animals have rudimentary language. Since animals are part of nature, are we then to say that their language is not language? Are we humans above nature so that our use of language represents nature while not being part of nature? Who are "we" that we are distinguished from "things and relations in the world around us? How would we describe such a transcendent we or us? Why do linguistics texts rarely address this metaphysical characteristic of language or the human qualities which are presumed to be responsible for it?
Also, from an Absolutist metaphysical perspective, nature (the world of things and relations) may be an expression of Deity or less abstractly, the expression of an advanced race of beings. Homo sapiens has been on this planet for only a small fraction of the time for even one 200,000,000 year tour of the galactic centre. What kind of achievements might we be capable of after a few such round trips? Might we not be capable of building entire planets and genetically engineering new species? We may well create what we now call nature or at least major portions of it. If asked why we made a particular new planet that way and its new species the way we made them, what would be the answer? Perhaps just: It was a poetic expression, a language which represented how we thought a planet should be. Such advanced intelligences may make their mark upon the world as we make our marks on paper, calling it written language.
These comments in speculative philosophy serve to show that we should not accept the conclusions of present linguistics texts as the final word on this subject, nor to they pretend to be so. Trask and Maybin write in "Introducing Linguistics" that "All languages exhibit stimulus-freedom, the ability to say anything at all, including nothing, in any circumstances" (p. 67). Thus they open the door to a metaphysical examination of language. After all, where does "free will" or the stimulus freedom of language fit into conventional deterministic science? When deterministic science seeks to dominate linguistics such important issues may be lost. Weisler and Milekic refer to another freewill-related aspect of language when they write in their Glossary that linguistic determinism means "Linguistic structure determines the nature of thought". And that should cause some thought on the part of computing science people when they turn their attention to NLP (natural language programming). When machines can speak as well as humans, will we recognize them as having our powers of thought? If we then say "I think, therefore I am " as Descartes did, are we forced to recognize machines as sentient?
THE COMPONENTS OF A LANGUAGE
Fromkin, Rodman and Hultin (2001, page 16) give us the following statement on the components of language:
"In our sense, the GRAMMAR includes everything speakers know about their language - the sound system, called PHONOLOGY; the system of meaning, called SEMANTICS; the rules of word formation, called MORPHOLOGY; and the rules of sentence formation called SYNTAX. It also, of course, includes the vocabulary of words - the dictionary or LEXICON." (Block letters are mine).
This does the job of macro-analysis of language well but I must issue a word of warning that on occasion some authors stray a little from the definitions given. Lyons (1991, page 4) uses the expression LANGUAGE-SYSTEM in lieu of GRAMMAR. Fromkin et al are using 'grammar' in a comprehensive way. Lyons narrows its usage. He says, "...a language-system comprises both a grammar and a lexicon." (pp. 3-4). Under grammar he includes phonology, syntax and semantics but not lexicon. In any case, the reader can usually determine the definitions from context. Sometimes the word GRAMMAR is also used in an even narrow way to refer to SYNTAX. Thus we may say that a grammatical error has been made, meaning that the sentence structure is faulty.
Trask and Maybin credit Dionysius Thrax (2nd to 1st C BC) with the eight-fold taxonomy of grammar in use today: nouns, verbs, articles, pronouns, prepositions, conjunctions, adverbs, participles (p. 3)
THE LANGUAGE OF MACHINES
Wyatt asks in "Writing Your First Computer Program" (2000, page 3), "Why use languages?" (ie computer languages). The answer he gives is similar to the answer we might give concerning why we learn natural human languages. "Computer languages provide a way for humans to communicate with a computer" he says and "Computer languages allow interaction with the computer". (page 4). On page three he expresses this as "communication and guidance". Using the language of the machine, we can then write a program which exercises some capability of the machine from simply putting text on a screen to approximations to human natural language in complex dialogues.
Perry writes in his "Absolute Beginners Guide to Programming" that computer language and program can be used to teach the machine. "After it is taught to add, subtract, multiply, and divide, the computer can then perform any math necessary as long as a program is supplied to direct it" (page 45). Perry lists a number of common computer languages on pp. 28-29 including C and Machine Language.
MACHINE LANGUAGE
The expression "machine language" is used to describe a language in binary form. The charged and uncharged microscopic, transistor-like units in a computer store the values which we are manipulating as either data or programs. We can call them charged and uncharged units or more conventionally, 1's and 0's. Thus the system of language is entirely binary at its foundation. That is, it is based on two values, 1 or 0, and all "higher" languages in the machine including NLP at human equivalency when it arrives, are binary at their machine foundation. "Higher level" languages like C, Visual Basic and so on, are so termed because they are higher or closer to, natural human language.
This is an awesome statement which too easily gets glossed over: A machine which is expected to some day attain human equivalency in language, is, at its foundation, only a collection of bits or binary digits.
Weisler and Milekic differentiate between natural language and a "formal language" like C or the language of logic. They say, "formal language ... includes invented languages such as logical languages and computer languages" (page 319). Could machines some day turn the tables on us and invent a language which is even "higher" than the natural human language we now seek to simulate with a machine?
references for above texts available upon request