LINGUISTICS

It is clear that the term linguistics does not in itself define or delimit our area of study for us, least of all from its etymology. It does not carry any theoretical presuppositions that we are more concerned with the tongue than with the lips, the larynx, or the ears, or that we would necessarily rule out the study of gestural systems used by the deaf or by Australian aboriginal tribes. It is simply the conventional name in English for a discipline that would be the same by any other name.

The discipline of linguistics is understood instead in terms of a certain group of professionals, a cohesive literature, professional societies and meetings, organized faculties and graduate programs, and a set of interests and questions in relation to a currently agreed-upon range of observational phenomena. It reflects the existing division of labor in science.

Linguistics is not necessarily tied to a particular methodology or type of theory, for these are seen as means to the end of understanding the various phenomena that present themselves. Thus the named discipline is more subject to change than the name as new phenomena are explored, new methods utilized, new theories proposed, and new knowledge gained. Linguistics today is certainly much different from what it was when the name was first applied over a hundred years ago, and from what it was as recently as 1950, or even 1965.

The scope of data of interest to linguists includes ancient and modern written sources; recorded or transcribed speech; judgments by informants that utterances are the same or different, grammatical or ungrammatical; and subsidiary information such as definitions, translational equivalents, and typical situations of use. These are traditional sources of data for working out grammars or parts of grammars, comparative or historical studies, and the like.

Linguists have also turned increasingly to a broader range of data. There is interest in the influence of social variables in dialect studies and their role in historical change. There are studies of bilingualism, and of pidgins and creoles. Anthropological linguists study questions of kinship terminology, folk taxonomy, the ritual use of language, and the functions of language in society. There is research on the instrumental use of language and its relation to situations of use. Questions are raised concerning the psychological capacities and limitations of speakers and listeners, and the relation of speech to the psychological areas of perception, cognition, and concept formation. There is concern with how children learn to speak, and how they continue to learn as adults. In the areas of stylistics, poetics, and the artistic use of language, consideration is given to sentence structure, to the structure and coherence of texts, to devices of parallelism and contrast, to metaphor and other figures, and to questions of prosody. There are studies of discourse in natural conversations in small groups. There is growing interest in variables of pitch, loudness, and vocal quality; in facial expressions and body motion; and in variables of space and distance. Normal speech errors are studied, as are disorders of communication having a neurological, psychological, or developmental origin.

Linguistics, then, has for its de facto observational scope today a broad domain that covers the whole range of human communicative phenomena.

In this whole broad area it has been difficult to bring the observations under the tight control of theory, or sometimes even to accommodate them at all. Thus proposed definitions of the scope of linguistics in terms of the natural scope of any current formal theory are inevitably too limiting, leaving out of account the legitimate observational interests of substantial numbers of linguists. Furthermore, current theory seems to be suffering from other troubles of an internal nature as well.

SCIENCE

Linguists see their discipline as a science. This is a matter of choice, for some of the same data may also be approached philologically or through literary or other studies, which may not necessarily be scientific. Linguistics as a science is thus seen as a program, a way of approaching the data and of trying to gain understanding. The choice of a scientific approach was made at an early date. Rask, for example, in explaining the linguistic approach he was advocating, invoked the names of Linnaeus and Newton (Rask 1830). Schleicher placed linguistics among the natural sciences and carefully distinguished linguistics from philology (1850:1, 1860:119) and from the philosophy of language (1860:119). Bloomfield said that “it is only within the last century or so that language has been studied in a scientific way, by careful and comprehensive observation” (1933:3).

The scientific roots of linguistics stretch back to the ancient Greeks, as is the case for other sciences. A high regard for careful and comprehensive observation was shown by the Alexandrians in the third and second centuries B.C. as they searched for recurrent parallels in texts and worked out grammatical categories. In the same era the Stoics distinguished different levels in grammar by citing as evidence examples of well-formed nonsense and of ambiguity at one level that was resolved at a higher level. Varro, in about 45 B.C., pointed out that in the earlier disputes over analogy and anomaly, both parties deferred to observed usage (De Lingua Latina IX 1-3) (Kent 1938).

In modern science the data from observations by the senses are given first place over theory. Theories are inventions or creations of the scientist. They are designed to provide understanding, to interrelate the data in support of a larger picture, and thus to provide insight. Theories in this sense are not wild guesses. They are only proposed with data in view, and they meet their test against data. Theories always remain tentative: when firm observational data collide with the predictions of theory, it is the theory that must give way, no matter how deep our commitment to it may have been.

Theories can be flawed by the incorporation of special assumptions not supported by the data. Such assumptions might be introduced a priori, they might creep in from another discipline, or they might have been inherited through an unquestioned tradition.¹ One problem with unsupported assumptions is that they may mix with the data on which the theory is based, taint the data, and thus obscure our understanding and lead us into error.

Explicit special a priori assumptions can also lead to factionalism and strife. Unsupported by evidence, they can only be maintained by special appeals or the weight of personal authority.² This can generate puffery, empty polemics, intellectual bullying, and the great-man syndrome, all symptoms of an unscientific approach. They tend to make it even harder to keep an open mind, for they focus on arguing that someone is right rather than on trying to find out what is right.

Unsupported assumptions inherited from the tradition can be especially troublesome. They are often hidden and overlooked as part of the accepted background and way of thinking we have been taught. Thus they may continue to delude us all. They may lurk in our everyday way of speaking, or even in our technical terminology, and thus they may seem so natural as to be beyond question.

Linguistics has many times uncovered and rejected harmful special assumptions inherited from the tradition and otherwise unsupported. In an earlier era grammars of indigenous languages were often written with undue influence from Latin grammar. The assumption that all languages are structured on a general Indo-European plan, or that the categories of Latin grammar may be universally applicable, has been erased in the best work done today. In order to combat the bias from the school tradition, we take pains to instruct our students in the wide range of exotic features to be found in the languages of the world.

In the last century prescriptivism was thrown out in favor of scientific descriptivism. The scientific linguist was to study what people actually say, not what someone thinks they ought to say. The assumption that language is a prescriptive norm has to be combated, for example when obtaining data from an informant.

Finally recall that linguistics had to discard the Aristotelian assumption of the cross-cultural universality of meanings or mental experience (On Interpretation 1). This deeply entrenched assumption conforming to folk theory and common sense had to be challenged and rejected in the light of massive evidence from the comparison of unrelated languages and cultures.

As these examples illustrate, linguistics has acted in the same way as other sciences in trying to discover and remove assumptions and presuppositions unsupported by evidence. In modern science every assumption is open to skepticism and doubt. This is not Cartesian doubt which ends in the acceptance of intuitive truths. It is more like Baconian doubt directed against the idols of the den, of the marketplace, and the others. It is most like Galilean doubt, which meant the suspension of judgment on even traditionally held positions unless and until one has good evidence from the senses or can reason to the conclusion on the basis of good evidence from the senses.

Thus a science works with only a bare minimum of assumptions, building the rest on the solid support of observational evidence. In the more highly developed sciences, the assumptions implicit in the practice of working scientists are of the most general sort. To put it roughly, these are (1) that there is a world out there to be studied, (2) that it is coherent, so we have a chance of finding out something about it, (3) that we can reach valid conclusions by reasoning from valid premises, and (4) that observed effects flow from immediate causes (where there’s smoke there’s fire).³

The question remains, however, whether linguistics is or can become a science in this strong sense. It has been charged, for example, that an approach that advocates the study of linguistic competence is unscientific in that it removes its theories from the possibility of being tested against the observations of actual speech behavior. See Wilks (1972) and Derwing (1973) among others.

Doubts have been raised in some minds as to the legitimacy of linguistics aspiring at all to the status of a science (Robinson 1975:170, 182). One can grant to disciplines such as philosophy, philology, or the study of literature the possibility of a nonscientific approach to some of the same phenomena, if that is deemed appropriate by the individuals involved. But in linguistics there is a long-standing and widely accepted goal of approaching the phenomena of interest scientifically. This goal of disciplining the study by scientific methods and principles has mobilized not only the lip service but the deeds of virtually every major linguist since the founding of modern linguistics, and showing that a theory or piece of research is unscientific is still widely seen as the ultimate condemnation of it.

According to this goal, linguistics should adhere to the program and approach of modern science, including particularly that it should build on the accepted foundational assumptions of science referred to previously, that theories must be answerable to observational data, and that any other assumptions unsupported by data, whenever they are discovered, should be rejected.

LANGUAGE

Besides the goal of disciplining its studies by scientific methods and principles, linguistics continues to embrace another long-standing goal. It should study language. Indeed, the discipline is often characterized simply as the scientific study of language.

But what is language? The term is used in a number of senses in everyday language, with which we need not be concerned here. In linguistics there are two major senses. There is the quasi-technical sense where the term is used to refer to the phenomena that are studied in linguistics, or more narrowly, verbal phenomena, texts, and the like. In this sense saying that linguistics is the scientific study of language would say nothing more than that linguistics is the study of certain phenomena of interest, a matter we have already discussed under linguistics. To avoid confusion and to highlight an important distinction, we will continue to use paraphrases for this sense such as linguistic phenomena or communicative phenomena. We will confine our use of the term to its central technical sense whereby language is seen as that which lies behind the phenomena, not the phenomena or data themselves. The goal then simply says that linguistics should study the nature of language as the object lying behind our observations.

What, then, is language? A number of conceptions and definitions have been proposed. But in spite of their differences a common core of agreement can be discerned.

Schleicher, who brought the comparative method to a high degree of development, and was noted for drawing family trees of languages, recognized four parts of grammar: the study of sounds, morphology, the study of function, and syntax (1859). He further distinguished between general grammar and the grammar of particular languages, and between historical linguistics and descriptive linguistics (1860:124-5).⁴ Schleicher saw linguistics as a science on a par with other comparative sciences in biology and geology. For him language was an organism like the other organisms of nature—animal, vegetable, and mineral—a kind of fourth kingdom. The term organism, which appears to reflect an older animistic view of nature, followed the usage of earlier linguists such as Herder, Wilhelm von Humboldt, and Bopp, and the philosopher Hegel. Schleicher later emphasized that the material basis of language is in the brain and the organs of speech with their nerves, muscles, etc., and that at least at present it is known only through its audible effect in speech (1865).

Whitney (1867:48–51) said that language is an institution—the work of those whose wants it subserves, and in their sole keeping and control. For him it was the product of a series of changes effected by the will and consent of men, but developing in the absence of reflection and conscious intent. He called it a grand system with a highly complicated structure comparable with an organized body. It had its existence only in the minds and mouths of those who use it (1867:35), and it consisted of the associations of ideas and their signs (1867:11).

Saussure, in speaking of language, still used the term organism from time to time, but more often he called it a system. According to Saussure, language

is both a social product of the faculty of speech and a collection of necessary conventions that have been adopted by a social body to permit individuals to exercise that faculty. [1916:2;1959:9]

It is a storehouse filled by the members of a given community through their active use of speaking, a grammatical system that has a potential existence in each brain, or, more specifically, in the brains of a group of individuals. For language is not complete in any speaker; it exists perfectly only within a collectivity. [1916:30;1959:13]

It is a system of signs in which the only essential thing is the union of meanings and sound-images, and in which both parts of the sign are psychological. [1916:32;1959:15]

Saussure’s view of language as a system of signs has been very influential.

Sapir defined language as

a purely human and noninstinctive method of communicating ideas, emotions, and desires by means of a system of voluntarily produced symbols. [1921:8]

This view stresses the idea of communication and appears to focus more on the external symbols produced than on a sign relation existing in the brains of a collectivity. But Sapir goes on to say that

language . . . consists of a peculiar symbolic relation—physiologically an arbitrary one—between all possible elements of consciousness on the one hand and certain selected elements localized in the auditory, motor, and other cerebral and nervous tracts on the other. . . . Hence, we have no recourse but to accept language as a fully formed functional system within man’s psychic or “spiritual” constitution. [1921:10]

He adds that the study of language is

to be an inquiry into the function and form of the arbitrary systems of symbolism that we term languages. [1921:11]

Here and elsewhere Sapir championed the patterning of language as a system of symbols and its psychological reality.

Leonard Bloomfield (1926) spoke of vocal features or sounds, and stimulus-reaction features of speech. Then by definition an act of speech was an utterance, and the totality of utterances that could be made in a speech-community was the language of that speech-community. A speech-community was defined as a community in which successive utterances are alike or partly alike. The vocal features common to same or partly same utterances were forms, and the corresponding stimulus-reaction features were meanings. The study of sames and differents led in Bloomfield’s hands to an inductive method for working out the structure of a language from observational data. He summarized:

To put it briefly, in human speech, different sounds have different meanings. To study this co-ordination of certain sounds with certain meanings is to study language. [1933:27]

A brief account cannot do full justice to the views of these linguists, but we can elaborate further should it prove necessary. In spite of the obvious differences in these positions, we can see a thread of agreement in what has become the canonical view: language is conceived at once as a relation between sound and meaning, the system or code behind particular utterances or texts, postulated on the basis of evidence and having predictive power, described in terms of grammar and lexicon, and incorporating a series of levels variously understood but typically involving phonology, morphology, syntax, and semantics. This view, in one form or another, is widely, almost universally, held in the discipline today:

A grammar of a language, in the sense in which I will use this term, can be loosely described as a system of rules that expresses the correspondence between sound and meaning in this language. [Chomsky 1970:52]

A language, by its nature, relates sounds (or graphs, i.e. marks on paper or the like) to meanings. [Lamb 1966:1]

Similar statements from many others could be quoted.

THE DISTINCTION BETWEEN LANGUAGE
AND PEOPLE

Contrasting with this canonical view of language and grammar that focuses on a sign relation between sound and meaning and on language as a system of signs, there are other persistent threads where there may be less agreement among various linguists. These are focused not on the system of signs but in some way on the people involved. Since we want to investigate the conceptual structure of linguistics, and we will be discussing such questions as the psychological reality of grammar, the interests of clarity dictate that we carefully distinguish between the conceptual domain of language and the conceptual domain of people, sometimes distinguished in the literature but often curiously confused.

Thus for Schleicher we want to distinguish his view of grammar and the grammars of particular languages and of the parts of grammar, and perhaps his concept of language as an organism or system, from his view of the material basis of language in the brain and the organs of speech.

For Whitney we would distinguish language as a grand system with a highly complicated structure consisting of the association of ideas and their signs, and perhaps his concept of language as an institution, from his concept of its existence in the minds and mouths of those who use it.

For Saussure we would distinguish language as a system of signs uniting meanings and sound-images from its psychological manifestation; the social product of the faculty of speech from the faculty of speech; the collection of conventions from the social body adopting them; and the grammatical system from its existence in the brains of a group of individuals.

For Sapir we would distinguish the system of symbols from its purely human and noninstinctive realization as a method of communicating; language as a peculiar symbolic relation from the elements of consciousness and motor and other cerebrally localized elements; language as a functional system from its existence within man’s constitution; and the function and form of arbitrary systems of symbolism from their psychological reality.

For Bloomfield we would distinguish vocal and stimulus-reaction features of speech from the psychological aspects of speakers (which he did not propose to study); the totality of utterances that could be made in a speech-community from the speech-community; and language as the co-ordination of certain sounds with certain meanings, and human speech, from the humans and what they do when they speak and understand.

Thus we will try to maintain a conceptual and terminological separation of language, speech, and grammar from the relevant psychological and social aspects of the real world, whatever they may turn out to be. We will try to eliminate any terminological ambiguity.⁵

It should be noted that our distinction is not between the formal and the informal, or between the abstract and the concrete, or between a theory of something and the thing it is a theory of. Nor is it Saussure’s distinction between langue and parole, or the distinction between a grammar and a mechanism for producing and understanding sentences according to the grammar, or the distinction between competence and performance.

Instead we want to distinguish language from the communicative aspects of people; speech or utterances as a product from what people do; and theories in the domain of speech, language, and grammar from theories in the domain of people (Yngve 1981). This distinction has roots in antiquity, as we shall see.