On the Origin of Language

In the Western world the study of language began as a philosophical inquiry into origins.¹ The Greeks (Third and Fourth Century B.C.) initiated the study of language essentially to explain its origin. The Conventionalists hypothesized that the relationship between the form of language (i.e., primarily the sounds and words) and meaning was essentially arbitrary, a convention of society. The naturalists hypothesized that the form of a word (i.e., its sounds) had a natural association with its referent in the real world. Only certain sound combinations (words or parts of words), however, were directly associated as an imitation of an object, its sound or an idea directly associated as an imitation of an object (e.g., kookaburra).

In an effort to explain how most of language, which is not so directly relatable to meaning, derived from an onomatopoeic beginning, the discipline of etymology began. Through studying the derivational history of words (etymology) the naturalists intended to demonstrate that the origin of all of language was ultimately relatable to words which directly reflected the meanings of their referents.

The first philosophical forum on language eventually developed into a discussion on the regularity of language patterns. Two basic theoretical positions emerged as explanatory frameworks for language, that which opted for irregularity and that which insisted that language was essentially regular. From the pre-eminence of the latter position it became popular to explain the irregularities of language on the basis that language somehow became corrupted with improper usage through time; this theoretical position regarded the older forms of language to be the purer forms.

By the Nineteenth Century there was a severe reaction to the highly speculative nature of the philosophizing about the original language of man which had characterized much of the study of language up until then. The interest was still historical, but the goal was not so idealistic. It was a romantic era of a rediscovery of the national past; the mother tongues of nations and families of nations rather than the mother tongue of the whole human race became the focus of attention. The romantic nationalism was a definite influence, but perhaps a more basic cause of the more realistic goal was the reaction to previous unscientific speculations. The felt need was to take a more scientific approach by analyzing empirical data. Thus was ushered in the period of systematic comparison of languages for the purpose of reconstructing the historical past.

During the Nineteenth Century largely under the leadership of German scholars an impressive amount of detailed scholarly work was done. Building on Sir William Jones' discovery that Sanskrit was genetically related to Latin and Greek and other European languages as well, these early historical linguists began to develop principles of language comparison. The availability of historical data not only made possible advances in the reconstruction of the original Indo-European language² (proto Indo-European), it also enabled linguists to describe the processes of change by which the proto-language developed into the diversity of the many Indo-European languages.

The German based 'neo-grammarian' school is known for its contribution to the study of sound change in the last quarter of the Nineteenth Century. The neo-grammarians, through meticulous analysis of historical text material, demonstrated the striking regularity of sound change. Hermann Paul (1846 -- 1921), the foremost theoretician of the neo-grammarians, identified convenience as the central mechanism of sound change; within the framework of convenience he categorized three types of sound changes under the mode of 'mispronunciation.'³ Leonard Bloomfield (1887-- 1949) was an early American structuralist who extended the neo-grammarian position with greater detail. He catalogued the mechanisms of sound change as two types: stabilizing vs. deteriorating or simplifying mechanisms. He documented at least three stabilizing changes characterized as reformation and compensatory processes. In the simplifying category Bloomfield documented no less than eleven processes of sound change.

Detailed documentation by the neo-grammarians of various processes of language change, especially those of sound change, contributed greatly to the statement of two basic principles of language change, 1) the process of streamlining and 2) the process of restructuring. Martinet, one of the most eminent historical linguists in the Twentieth Century, is credited with formulating these two principles of language change. He refers to the restructuring process, which maintains adequate communication, as being in conflict with the streamlining process which manifests (in language) the human tendency toward reducing effort to a minimum.

Linguistic evolution may be regarded as governed by the permanent conflict between man's communicative needs and his tendency to reduce to a minimum his mental and physical activity. Here as elsewhere, human behavior is subject to the law of least effort. (Martinet 1964:167).

The law of least effort effects a relentless streamlining of the status quo, reducing complexity and redundancy, which in turn eventually leads to restructuring adjustments in the various systems of language to help maintain an acceptable level of communication. The restructuring principle could be termed the law of conservation of communication.

It would be misleading to imply that Hermann Paul was not interested in the origin of language; the question of origin certainly interested Paul as it still does linguists today. The essential difference between modern linguistics (the past 175 years) and that of the two previous millennia is that linguistics has moved from the purely philosophical realm to that of the empirical sciences. Linguists are still intrigued with the question of origins but their speculations on the origin of language must be based on observable facts about language.

Two important basic principles of language must be mentioned, the streamlining effect of least effort and the compensatory maintenance of communication, or restructuring; two observations that relate to these are worth noting.

a) Primitive languages: No group of human beings today, even those living in a stone-age culture, speak what could be conceived of as a primitive language. Furthermore, no known language in all of history was in any sense primitive. Elgin remarks, 'The most ancient languages for which we have written texts -- Sanskrit, for example -- are often far more intricate and complicated in their grammatical forms than many contemporary languages.' (Elgin 1973: 44) This, of course, is no surprise to us if the inevitable processes of simplification observable today have consistently been operating for all or most of human history (this is in itself of course indeterminate, but we can at least conclude that simple material cultures do not imply simple languages).

b) Creativity of language: The vocabulary may be considered to be the most creative area of language and even here, 'For the most part, people tend to re-adapt existing lexical material rather than create entirely new material.' (Langacker 1967:186). Apart from re-adapting and extending existing vocabulary items from within a particular language, words or parts of words are commonly borrowed from other languages. A language seldom exhibits creativity in the sense of inventing new and unique forms.

The English pronoun system illustrates the two basic principles in action today. The oldest English pronoun system distinguished three numbers (singular, dual, plural) for each of 1st, 2nd, and 3rd persons. Today standard English distinguishes only singular and plural. The previously 'extravagant' system has been streamlined by neutralizing the difference between duality and plurality. In addition, with the 2nd person 'you' the singular-plural distinction has been lost, resulting in unacceptable ambiguity (ineffective communication) at times. (One of the first times I asked the girl who is now my wife for a date I ended up taking a whole carload of people on an outing because she wasn't sure whether I meant 'you-singular' or ‘you-plural’; I was too embarrassed to expressly exclude everyone else who was there at the time. As far as I was concerned, that was an unacceptable ambiguity!) The restructuring presently going on in English to remedy this situation involves adding particles from elsewhere in the grammar into the pronoun system. A new 2nd-person-plural pronoun is being formed in the northeastern United States by adding the normal noun pluralizer "s" to the pronoun "you" resulting in the plural "yous" (pronounced the same as "ewes"). The much-publicized "Southern dialect" has restructured the system in another way. A phrase-level quantifier "all" has been added in a contacted form of "you-all" resulting in "y'all." In both cases the restructuring process is clearly adaptive rather than innovative.

Many linguists, apparently including Martinet, believe that the two opposing principles equalize each other. Langacker states, 'Just as there are no primitive languages, there are no 'corrupt' languages. Languages change, but they do not decay.' (Langacker 1973:17) This is a difficult point to verify. Support for this claim seems to have been well documented in a historical review of the sound system of Spanish. While sounds have changed, the number of distinctive features in the system have remained fairly constant; from the viewpoint of information processing the overall process of change has not altered the potential for communication. Many of Bloomfield's examples of simplification due to sound change appear to affect syntactic categories and vocabulary. The loss of noun case endings in English is a case in point. The relative complexity of communication potential of the resulting system is difficult to evaluate. While the case ending functions of identifying subjects, objects, etc. have been shifted to another level in grammar (word order on the clause level now usually identifies the subject and object) it is difficult to judge how the two systems compare in efficiency of communication. Some questions to be asked would be, 'What happened to previous functions of word order in clauses?', 'Is word order now overloaded with the jobs of case-role encoding and the indicating of old and new information in a discourse?', 'Is focus involved?'

Language is fantastically complex. Its built-in means of combining and recombining (nesting) of its various levels has suggested to many leading linguists that language is theoretically infinite though not practically so in everyday usage. It almost sounds too complex to be able to detect any significant leveling out of language any more than one could detect by observation that the sun is burning itself out.

As far as I am aware no linguist seriously purports that the restructuring process of language overrides the streamlining process resulting in a qualitative positive development of language. If we decide that language did originally develop, possibly evolving from animal communication, we can only do so assuming evolution to be a universally valid principle. This type of a priori reasoning was the basic fallacy of pre-Nineteenth Century 'speculative grammar' which was pre-scientific in the modern sense of the word.

Furthermore, the observable data neither indicate that such a period of pre-historic development even existed, nor do they suggest a cause of the subsequent state of equilibrium or process of simplification that would have to have come into operation at some time after such a pre-historic development. Noam Chomsky, one of the most prominent linguists of this century, has indicated that human language and animal communication are not even comparable entities, they are so different.

Either the streamlining and restructuring processes balance each other or the streamlining process is gradually reducing language to a limited system of over-generalities. Either human languages have always existed with essentially the potential they exhibit now or they once exhibited greater potential for precise communication than they now do.

Labov, a prominent contemporary sociolinguist, comments on these two processes in his effort to understand the place of language in an evolutionary framework:

It is plain to most linguists that the 'destroy and rebuild' theory of linguistic evolution is equivalent to claiming that the whole process is dysfunctional. For the systematic part is the destructive one, and the analogical re-shaping seems to be making the best of a bad job. And if the principle of least effort is the evil genius behind the destruction, we can only look at language change as a kind of massive testimony to original sin. (Labov 1973: 245)

In the remainder of his thesis Labov does not provide any relevant alternative to the dysfunctional role of language change. He does not deny that language change results in diversification of languages and not in overall complexity or adaptive radiation. He rather looks for functional evolutionary result in the development of human society. Thus he suggests that language diversity provides relative cultural isolation, maintaining cultural pluralism which presumably promotes the evolution of human society. But he cannot provide a functional role for language change purely within the development of language. Language development seems to be dominated by a dysfunctional (non-evolutionary) process. Conclusion:

Regardless of how we might attempt to fit language into the broader picture, looking at language by itself there is no evidence that language is the product of any positive developmental process. Language is in a state of consistent change which at best seems to maintain a state of equilibrium.

__________________________________
¹ The Indian tradition antedates the work of the Greeks. Panini (Fourth Century B.C.) culminated the work of his predecessors with a grammatical description of Sanskrit which has been acclaimed as being the most detailed and comprehensive grammar ever done. The Indians were more interested in an accurate accounting of Sanskrit rather than answering philosophical questions.
² Work in the Nineteenth Century was based essentially on the classical languages (Latin and Greek) and the then-oldest-attested Indo-European language, Sanskrit. Hittite was identified as an older Indo-European language subsequent to the uncovering of Hittite inscriptions in the 1870's about 150 kilometers east of Ankara, Turkey. Hittite became extinct by 1200 B.C.
³ Mispronunciation or changed pronunciations were caused by convenience not to be confused with laziness, neglect, or some natural ease of pronunciation of individual sounds. Instead, mispronunciation is an effect involving the assimilation of sounds and the influence of the symmetry (pattern pressure) of the entire sound system. These changes were not considered to be deteriorative in the sense of 'corrupting' language as linguists of earlier periods had suggested.