EXPLANATIONS IN TERMS OF PEOPLE

An adequate understanding of linguistic phenomena has long been seen as requiring an analysis ultimately in terms of psychological and social realities. Thus in addition to the incompatible goals of linguistics that it should study language and that it should be a science, there is a third goal, sometimes acknowledged, not always accepted, that linguistics should seek explanations in terms of people. There are a number of examples in the history of linguistics where explanations in terms of people have been proposed. Let us review a few.

In classical antiquity Plato had Socrates say that the original words have been completely buried by those who wished to dress them up, and that the inserting of extra letters in words is the work of people who care nothing for the truth, but only for the shape of their mouths (Cratylus 414 B-E). Varro explained the function of grammatical processes such as derivation and inflection as saving people from having to memorize a large vocabulary in toto (De Lingua Latina VIII 4–7)(Kent 1938). Thus he said that new slaves introduced into a large household could quickly inflect the names of all their fellow slaves in the oblique cases, provided only that they had heard the nominative.

In the earlier period of modern linguistics the relations between language families worked out by the comparative method were interpreted in terms of the wanderings of peoples from one place to another, carrying their language with them. For example, Fick (1871:1045ff) tried to work out their itinerary in geographic and temporal detail. Or the data were interpreted in terms of an original gradation or continuum of dialects in which regional languages grew up by neighboring people adopting the way of speaking of prestige centers (Schmidt 1872). Changes were seen in terms of people making analogies, a point emphasized by the neogrammarians. Or observed changes were seen as due to problems in communication caused by homonyms resulting from sound change (Gilliéron and Roques 1912). Or the rate of change was seen as related to the actual density of communication among speakers (Bloomfield 1933:326, 340).

In more recent years there have been investigations of the social motivation of sound change (Labov 1963), and of the social factors responsible for speakers alternating between different styles or different languages (Gumperz 1962). It has been hypothesized that the properties of human temporary memory are related causally to language structure and language change (Yngve 1960). Hall (1962) has seen a creole as developing when a pidgin acquires native speakers. Naro (1978) has traced the origin of pidgin Portuguese partly to native speakers of West African languages being captured, taken to Portugal, and there taught Portuguese so they could be used as translators. Silverstein (1972) has offered an interpretation of Chinook Jargon as a drastically reduced form of each speaker’s primary language so that each speaker in speaking it may retain his own syntactic base. These are just a few of the many attempts to understand the observed phenomena in terms of the realities of people and what they do, that is, in terms of psychological and social realities.

There have been continued calls for linguistics to turn away from the domain of language and concern itself more with the domain of people. When modern scientific linguistics started, it was seen as part of the natural history of man (Rask, Bopp, Schleicher, and others). The neogrammarians in their manifesto pointed out that language is not a thing which leads a life of its own outside of and above human beings, but that it has its true existence only in the individual, and hence all changes in the life of language can only proceed from the individual speaker. It was not the Greek language which dropped final t, but those among the Greeks with whom the sound change started (Osthoff and Brugman 1878:xii)(Lehmann 1967:204, 209). Hermann Paul stated the program of linguistics as

showing how the single individual is related to the community; receiving and giving; defined by the community and defining it in turn; and how the younger generation enters on the heritage of the elder. [1889:xxx]

But Saussure said that linguistics has accorded too large a place to history and must turn back to grammar. Nevertheless he placed langue in the minds of a collectivity as had linguists before him.

As linguistics turned back to grammar, Sapir pointed out that:

it is peculiarly important that linguists, who are often accused, and accused justly, of failure to look beyond the pretty patterns of their subject matter, should become aware of what their science may mean for the interpretation of human conduct in general. Whether they like it or not, they must become increasingly concerned with the many anthropological, sociological, and psychological problems which invade the field of language. [1929:214]

Gardiner (1932:6) spoke of the crisis of grammar, and proposed to try to find out how speech works by putting back single acts of speech into their original setting of real life. Malinowski saw that linguists were faced by a dilemma.

Once we recognize . . . that “language is a form of activity, a mode of human behavior, perhaps the most important,” the question arises: Can we treat language as an independent subject of study? Is there a legitimate science of words alone, of phonetics, grammar, and lexicography? Or must all study of speaking lead to sociological investigation, to the treatment of linguistics as a branch of the general science of culture? [1937:172]

But Bloomfield abandoned his earlier reliance on psychology and criticized Hermann Paul for accompanying

his statements about language with a paraphrase in terms of mental processes which the speakers are supposed to have undergone. The only evidence for these mental processes is the linguistic process; they add nothing to the discussion, but only obscure it. [1933:17]

Bloomfield’s influence was on the side of moving linguistics back to the study of speech, language, and grammar, and away from the study of people.

More recently, there have been calls for developing an ethnography of speaking (Hymes 1962), and for linguistics to concentrate on investigating how people use language to communicate (Yngve 1969). Sociolinguists have increasingly tried to focus on people interacting in social groups, but Chomsky has reiterated his position of idealizing away from people and social groups. His concept of competence represents a move back to the domain of language and grammar, although competence is supposed to exist in the head of the native speaker.

Not only have explanations been given in terms of people, but actually all the data for linguistics, without exception, also stem from people. Recall that the observational domain of linguistics today covers the whole broad range of human communicative phenomena.

At least since the beginning of modern scientific linguistics, the issue of the two domains has been with us, the one concerned with speech, language, and grammar, the other concerned with people and the psychological and social reality. The interrelationships of these domains, their clashes, and our confusions continue to be debated widely. The question of which should be dominant continues as an underlying theme in the linguistics literature, where it is conceived in terms of current conceptions of grammar. For example, there is the current issue of whether or not people actually carry out transformations when they speak. Or, more generally, do they in speaking execute grammar rules of any sort, or are these to be seen as part of some different domain? We tend to forget that this is just the current version of a very old question.

THE FALLACIES OF THE PSYCHOLOGICAL
AND SOCIAL REALITY OF GRAMMAR

In spite of the many calls to focus on people instead of on language, and in spite of the widely acknowledged need for explanations in terms of people, we find that essentially all linguistic thought directed toward the domain of people has actually been put forth in terms of language and grammar.

Plato, when he talked about people caring more for the shapes of their mouths than for the truth, was talking in terms of changes of letters (sounds) and names using the current concepts in the domain of speech and language. Varro, in talking about slaves not having to learn separately all the inflected forms of names, spoke in the grammatical terms of nominative and oblique cases. The neogrammarians in their manifesto spoke about some among the Greeks dropping final t when the sound change started. Gardiner wanted to put back the single acts of speech into their original setting of real life, but ended up talking about words, meanings, sentences, and the like. Yngve discussed a model hypothesized to represent a speaker in some respects, including a mechanism and permanent and temporary memory organs, but the permanent memory was populated with a grammar, and the mechanism moved symbols representing grammatical entities into and out of the temporary memory as it produced grammatical sentences. And Chomsky talked about the competence of the native speaker, but ended up discussing his current version of transformational-generative grammar, which was then supposed to underlie performance in some way, although its relation to performance, he now says, is not just a problem but a real mystery.

We have attempted to focus on people, on a psychological and social reality, but the phenomena have been interpreted in terms of language and grammar. We have looked at people but have seen speech-sounds, forms, and meanings. Instead of people we have seen parts of speech, sentences, paradigms, rules, and the rest. This has required the introduction of unsupported assumptions to create the objects of language and grammar, the assumptions constituting Bloomfield’s fundamental assumption of linguistics or its equivalent. Then, since we are interested in explanations in terms of people, the grammatical results have been reinterpreted in terms of people. This has required the introduction of further assumptions, that grammar has a psychological reality and a social reality, that is, that grammar has an existence in the physical domain. Such assumptions are fallacies.

Let us look more closely at what has happened.

First we have looked at people in terms of language and grammar: the observations in linguistics, every one of them, are actually observations of people, but the usually tacit assumptions creating the traditional objects of language and grammar have interposed themselves between our eyes and the phenomena. Thus the linguistic results flow in fact not only from the data but also from the assumptions. Since the assumptions are unjustified, the results are thereby insecure. It is difficult to know to what extent they flow from the observations and to what extent they flow from the assumptions. The results may seem coherent and may even be able to predict other observations. They may even appear to be adequate, say, 80 percent or more of the time, thus increasing our confidence in them and reinforcing an idea that linguistics is an empirical science, and that we are on the right track. But the widely held perceptions of adequate empirical support for linguistic results are a mistake if unsupported assumptions have also been accepted. We know from our experience with Latin grammar applied to indigenous languages that a theory based on inappropriate and unsupported assumptions can appear to be adequate in many respects, yet be unacceptable in the end. To the extent that the unsupported assumptions creating the objects of language and grammar adversely affect the results, we will experience serious problems of adequacy. And to the extent that different linguists proceed on the basis of somewhat different assumptions, they may achieve different areas of apparent adequacy, and there may be areas where their claims stand in conflict. Such unsupported assumptions represent an a priori, arbitrary, and nonempirical or untestable component at the very heart of the discipline, leading to conflicting views of language, endless argumentation, and lack of convergence on scientifically adequate results.

Then, second, in order to achieve understanding in terms of people, the resulting grammar is thrust into the head of native speakers or predicated of the community in some way. This requires a second set of assumptions: that grammar has a psychological and a social reality. On these assumptions, objects in the domain of language, which have no external existence independent of theory, but are based instead on a priori assumptions, are reified—given a reality in the real world—a psychological or social reality. There is a psychological reality, and a social reality, but to identify either or both of these with language in the technical sense of a sign relation between sound and meaning given by grammar and lexicon is a fallacy. It is a category mistake of the most serious kind: they exist in different domains altogether. Furthermore, the results in the domain of language and grammar are already unsteady, since they rest in part on the unsupported assumptions creating the objects of language and grammar.

Thus the goal that linguistics should seek explanations in terms of people is incompatible with the goal that linguistics should study language.

THE GRIP OF THE TRADITION

In spite of the many plausible explanations in terms of people and the extreme scarcity of reasonable explanations in terms of language alone, theories of language and grammar still hold sway. As an interesting side issue one may speculate as to the reasons for this continued preoccupation with the objects of speech, language, and grammar. First, the extensive teaching of traditional grammar in the schools for more than two millennia has assured these ancient ideas a continuing place on the intellectual landscape instead of their being discarded.¹ Second, the continued interest of philosophers in logic and epistemology has often turned their attention to concepts of language in a theory of signs, thus keeping this aspect of the tradition alive. Third, the theory of language and grammar has been the most highly developed body of theory in the whole area, the most successful in dealing with complexity and detail, and might be expected to preempt the field for this reason alone. Fourth, no good alternative has been available even for those who have been dissatisfied with current theory. Fifth, most of the technical linguistic concepts available, such as morpheme, meaning, vowel, sentence, and the like are defined in terms of language and grammar, so any careful thinking in the area tends to bring in the preconceptions of this theoretical structure. And sixth, the related familiar concepts available in everyday language similarly have their meaning in the realm of words, language, meaning, and the like. Consequently it has been very difficult to think clearly in the domain centered on people without being drawn into the ancient conceptual structure and terminology of speech, language, and grammar—the Sapir-Whorf hypothesis in action.

The psychological and social reality of grammar fallacies are inherent in present conceptions of linguistic study. They follow nearly automatically from the course of development of scientific linguistics in the last century and a half. First there was the demand of science for careful observation and the basing of etymologies and statements of language change on solid observational evidence rather than on free speculation or philosophical preconceptions. Then there was the scientific demand for the accountability of theory to the evidence, which required increased explicitness and formalization in the theory so as to relate to the details of observation. This naturally led to further developing and improving the already highly developed traditional grammatical theory, an effort of approximately the last hundred years. Finally there was the demand of science for understanding and explanation, and the realization that explanations must ultimately relate to people. When grammatical theory is linked to efforts at understanding in terms of people, the fallacies arise. This is undoubtedly the reason that Bloomfield a half century ago tried to draw away from explanations in terms of people. But the demands for explanations in terms of people remain and even seem to have been increasing, bringing the fallacies ever more into play and exposing to view many confusing and difficult problems in the discipline.

But whether these speculations are correct or not, it is evident, in conclusion, that trying to interpret observations of people in terms of language and grammar and then trying to understand people indirectly in terms of language and grammar is injurious to our attempts to achieve a scientific understanding. It involves an unjustified assumption of the objects of language and grammar. It thus interposes an inappropriate theoretical structure between our observations of people and our understanding of people. It leads to a reliance on arbitrary or a priori criteria for judging statements that cannot be tested against evidence from the senses, and thus it tends to lead us into empty polemics and endless argumentation. And it leads us into the psychological and social reality of grammar fallacies. In these ways it frustrates our efforts to erect an adequate linguistic science.

Indeed, many of the current difficulties in the discipline mentioned earlier can be seen as expected symptoms of this underlying conceptual problem in the foundations of linguistics—the fact that the goal of linguistics to study language, which is incompatible with its goal to be a science, is also incompatible with its goal to seek explanations in terms of people.