by Noam Chomsky
What is more, there is little recognition in philosophical grammar of the intricacy of the mechanisms that relate deep to surface structure, and, beyond the general outlines sketched above, there is no detailed investigation of the character of the rules that appear in grammars or the formal conditions that they satisfy. Concomitantly, no clear distinction is made between the abstract structure underlying a sentence and the sentence itself. It is, by and large, assumed that the deep structure consists of actual sentences in a simpler or more natural organization and that the rules of inversion, ellipsis, and so on, that form the full range of actual sentences simply operate on these already formed simple sentences. This point of view is explicit, for example, in Du Marsais’s theory of construction and syntax, and it is undoubtedly the general view throughout.106 The totally unwarranted assumption that a deep structure is nothing other than an arrangement of simple sentences can be traced to the Cartesian postulate that, quite generally, the principles that determine the nature of thought and perception must be accessible to introspection and can be brought to consciousness, with care and attention.
Despite these shortcomings, the insights into the organization of grammar that were achieved in Cartesian linguistics remain quite impressive, and a careful study of this work can hardly fail to prove rewarding to a linguist who approaches it without prejudice or preconceptions as to the a priori limitations on permitted linguistic investigation. Beyond these achievements, the universal grammarians of the seventeenth and eighteenth centuries have made a contribution of lasting value by the very fact that they posed so clearly the problem of changing the orientation of linguistics from “natural history” to “natural philosophy” and by stressing the importance of the search for universal principles and for rational explanation of linguistic fact, if progress is to be made toward this goal.
Acquisition and use of language
We have so far extracted from “Cartesian linguistics” certain characteristic and quite important doctrines regarding the nature of language and have, quite sketchily, traced their development during the period from Descartes to Humboldt. As a by-product of this study of langue, and against the background of rationalist theory of mind, certain views emerged as to how language is acquired and used. After a long interlude, these views are once again beginning to receive the attention that they deserve, although their appearance (like the reappearance of the central ideas of transformational grammar) was, in fact, a largely independent development.
The central doctrine of Cartesian linguistics is that the general features of grammatical structure are common to all languages and reflect certain fundamental properties of the mind. It is this assumption which led the philosophical grammarians to concentrate on “grammaire générale” rather than “grammaire particulière” and which expresses itself in Humboldt’s belief that deep analysis will show a common “form of language” underlying national and individual variety.107 There are, then, certain language universals that set limits to the variety of human language.108 The study of the universal conditions that prescribe the form of any human language is “grammaire générale.” Such universal conditions are not learned; rather, they provide the organizing principles that make language learning possible, that must exist if data are to lead to knowledge. By attributing such principles to the mind, as an innate property, it becomes possible to account for the quite obvious fact that the speaker of a language knows a great deal that he has not learned.
In approaching the question of language acquisition and linguistic universals in this way, Cartesian linguistics reflects the concern of seventeenth-century rationalistic psychology with the contribution of the mind to human knowledge. Perhaps the earliest exposition of what was to become a major theme, throughout most of this century, is Herbert of Cherburys De Veritate (1624),109 in which he develops the view that there are certain “principles or notions implanted in the mind” that “we bring to objects from ourselves . . . [as] . . . a direct gift of Nature, a precept of natural instinct” (p. 133). Although these Common Notions “are stimulated by objects,” nevertheless, “no one, however wild his views, imagines that they are conveyed by objects themselves” (p. 126). Rather, they are essential to the identification of objects and the understanding of their properties and relations. Although the “intellectual truths” comprised among the Common Notions “seem to vanish in the absence of objects, yet they cannot be wholly passive and idle seeing that they are essential to objects and objects to them . . . It is only with their aid that the intellect, whether in familiar or new types of things, can be led to decide whether our subjective faculties have accurate knowledge of the facts” (p. 105). By application of these intellectual truths, which are “imprinted on the soul by the dictates of Nature itself,” we can compare and combine individual sensations and interpret experience in terms of objects, their properties, and the events in which they participate Evidently, these interpretive principles cannot be learned from experience in their entirety, and they may be independent of experience altogether. According to Herbert:
[They] are so far from being drawn from experience or observation that, without several of them, or at least one of them, we could have no experience at all nor be capable of observations. For if it had not been written in our soul that we should examine into the nature of things (and we do not derive this command from objects), and if we had not been endowed with Common Notions, to that end, we should never come to distinguish between things, or to grasp any general nature. Vacant forms, prodigies, and fearful images would pass meaninglessly and even dangerously before our minds, unless there existed within us, in the shape of notions imprinted in the mind, that analogous faculty by which we distinguish good from evil. From where else could we have received knowledge? In consequence, anyone who considers to what extent objects in their external relationship contribute to their correct perception; who seeks to estimate what is contributed by us, or to discover what is due to alien or accidental sources, or again to innate influences, or to factors arising from nature, will be led to refer to these principles. We listen to the voice of nature not only in our choice between what is good and evil, beneficial and harmful, but also in that external correspondence by which we distinguish truth from falsehood, we possess hidden faculties which when stimulated by objects quickly respond to them.
(pp. 105–106)
It is only by the use of these “inborn capacities or Common Notions” that the intellect can determine “whether our subjective faculties have exercised their perceptions well or ill” (p. 87). This “natural instinct” thus instructs us in the nature, manner, and scope of what is to be heard, hoped for, or desired” (p. 132).
Care must be taken in determining what are the Common Notions, the innate organizing principles and concepts that make experience possible. For Herbert, the “chief criterion of Natural Instinct” is “universal consent” (p. 139). But two qualifications are necessary. First, what is referred to is universal consent among “normal men” (p. 105). That is, we must put aside “persons who are out of their minds or mentally incapable” (p. 139) and those who are “headstrong, foolish, weak-minded and imprudent” (p. 125). And although these faculties “may not ever be entirely absent,” and “even in madmen, drunkards, and infants extraordinary internal powers may be detected which minister to their safety” (p. 125), still we can expect to find universal consent to Common Notions only among the normal, rational, and clearheaded. Second, appropriate experience is necessary to elicit or activate these innate principles; “it is the law or destiny of Common Notions and indeed of the other forms of knowledge to be inactive unless objects stimulate them” (p. 120). In this respect, the common notions are like the faculties of seeing, hearing, loving, hoping, etc., with which we are born and which “remain latent when their corresponding objects are not present, and even disappear and give no sign of their existence” (p. 132). But this fact must not blind us to the realization that “the Common Notions must be deemed not so much the ou
tcome of experience as principles without which we should have no experience at all” and to the absurdity of the theory that “our mind is a clean sheet, as though we obtained our capacity for dealing with objects from objects themselves” (p. 132).
The common notions are “all intimately connected” and can be arranged into a system (p. 120); and although “an infinite number of faculties may be awakened in response to an infinite number of new objects, all the Common Notions which embrace this order of facts may be comprehended in a few propositions” (p. 106). This system of common notions is not to be identified with “reason.” It simply forms “that part of knowledge with which we were endowed in the primeval plan of Nature,” and it is important to bear in mind that “it is the nature of natural instinct to fulfil itself irrationally, that is to say, without foresight.” On the other hand, “reason is the process of applying Common Notions as far as it can” (pp. 120–121).
In focusing attention on the innate interpretive principles that are a precondition for experience and knowledge and in emphasizing that these are implicit and may require external stimulation in order to become active or available to introspection, Herbert expressed much of the psychological theory that underlies Cartesian linguistics, just as he emphasized those aspects of cognition that were developed by Descartes and, later, by the English Platonists, Leibniz, and Kant.110
The psychology that develops in this way is a kind of Platonism without preexistence. Leibniz makes this explicit in many places. Thus he holds that “nothing can be taught us of which we have not already in our minds the idea,” and he recalls Plato’s “experiment” with the slave boy in the Meno as proving that “the soul virtually knows those things [i.e., truths of geometry, in this case], and needs only to be reminded (animadverted) to recognize the truths. Consequently, it possesses at least the idea upon which these truths depend. We may say even that it already possesses those truths, if we consider them as the relations of the ideas” (§26).111
Of course, what is latent in the mind in this sense may often require appropriate external stimulation before it becomes active, and many of the innate principles that determine the nature of thought and experience may well be applied quite unconsciously. This Leibniz emphasizes, in particular, throughout his Nouveaux Essais.
That the principles of language and natural logic are known unconsciously112 and that they are in large measure a precondition for language acquisition rather than a matter of “institution” or “training” is the general presupposition of Cartesian linguistics.113 When Cordemoy, for example, considers language acquisition (op. cit., pp. 40ff.), he discusses the role of instruction and conditioning of a sort, but he also notices that much of what children know is acquired quite apart from any explicit instruction,114 and he concludes that language learning presupposes possession of “wholly developed reason [la raison toute entière] for indeed this way of learning to speak is the result of discernment so great and reason so perfect that it is impossible to conceive of any more marvelous” (p. 59).
Rationalist conclusions reappear with some of the romantics as well. Thus A. W. Schlegel writes that “human reason may be compared to a substance which is infinitely combustible but does not burst into flame on its own: a spark must be thrown into the soul” (“De l’étymologie en général,” p. 127). Communication with an already formed intellect is necessary for reason to awaken. But external stimulation is only required to set innate mechanisms to work; it does not determine the form of what is acquired. In fact, it is clear “that this acquisition [of language] through communication already presupposes the ability to invent language” (Kunstlehre, p. 234). In a certain sense, language is innate to man; namely, “in the truer philosophical sense in which everything that, according to the usual view, is innate to man, can only be brought forth through his own activity” (ibid., p. 235). While Schlegel’s precise intentions, with many such remarks, might be debated, in Humboldt the Platonism with respect to language acquisition is quite clear. For Humboldt, “to learn is . . .always merely to regenerate” (op. cit., p. 126). Despite superficial appearances, a language “cannot properly be taught but only awakened in the mind; it can only be given the threads by which it develops on its own account”; thus languages are, in a sense, “self-creations” [Selbstschöpfungen] of individuals (p. 50; Humboldt 1999: 43–4):
Language-learning of children is not an assignment of words, to be deposited in memory and rebabbled by rote through the lips, but a growth in linguistic capacity with age and practice.
(p. 71)
That in children there is not a mechanical learning of language, but a development of linguistic power, is also proven by the fact that since the major abilities of humans are allotted a certain period of life for their development, all children, under the most diverse conditions, speak and understand at about the same age, varying only within a brief time-span.
(p. 72; Humboldt 1999: 58)
In short, language acquisition is a matter of growth and maturation of relatively fixed capacities, under appropriate external conditions. The form of the language that is acquired is largely determined by internal factors; it is because of the fundamental correspondence of all human languages, because of the fact that “human beings are the same, wherever they may be” [der Mensch überall Eins mit dem Menschen ist], that a child can learn any language (p. 73).115 The functioning of the language capacity is, furthermore, optimal at a certain “critical period” of intellectual development.
It is important to emphasize that seventeenth-century rationalism approaches the problem of learning – in particular, language learning – in a fundamentally nondogmatic fashion. It notes that knowledge arises on the basis of very scattered and inadequate data and that there are uniformities in what is learned that are in no way uniquely determined by the data itself (see note 114). Consequently, these properties are attributed to the mind, as preconditions for experience. This is essentially the line of reasoning that would be taken, today, by a scientist interested in the structure of some device for which he has only input–output data. In contrast, empiricist speculation, particularly in its modern versions, has characteristically adopted certain a priori assumptions regarding the nature of learning (that it must be based on association or reinforcement, or on inductive procedures of an elementary sort – e.g., the taxonomic procedures of modern linguistics, etc.) and has not considered the necessity for checking these assumptions against the observed uniformities of “output” – against what is known or believed after “learning” has taken place. Hence the charge of a priorism or dogmatism often leveled against rationalistic psychology and philosophy of mind seems clearly to be misdirected. (For further discussion, see the references in note 110.)
The strong assumptions about innate mental structure made by rationalistic psychology and philosophy of mind eliminated the necessity for any sharp distinction between a theory of perception and a theory of learning. In both cases, essentially the same processes are at work; a store of latent principles is brought to the interpretation of the data of sense. There is, to be sure, a difference between the initial “activation” of latent structure and the use of it once it has become readily available for the interpretation (more accurately, the determination) of experience. The confused ideas that are always latent in the mind may, in other words, become distinct (see note 111), and at this point they can heighten and enhance perception. Thus, for example, a
skilful and expert limner will observe many elegancies and curiosities of art, and be highly pleased with several strokes and shadows in a picture, where a common eye can discern nothing at all; and a musical artist hearing a consort of exact musicians playing some excellent composure of many parts, will be exceedingly ravished with many harmonical airs and touches, that a vulgar ear will be utterly insensible of.
(Cudworth, op. cit., p. 446; Cudworth 1996: 109)
It is the “acquired skill” that makes the difference; “the artists of either kind have many inward anticip
ations of skill and art in their minds” that enable them to interpret the data of sense in a way that goes beyond the “mere noise and sound and clatter” provided by passive sense, just as the informed mind can interpret the “vital machine of the universe” in terms of “interior symmetry and harmony in the relations, proportions, aptitudes and correspondence of things to one another in the great mundane system” (ibid.). Similarly, in looking at and “judging of” a picture of a friend, one makes use of a “foreign and adventitious” but preexistent idea (pp. 456–457; Cudworth 1996: 109). Once this distinction between learning and perception has been noted, however, the essential parallel between the cognitive processes that are involved outweighs the relatively superficial differences, from the point of view of this rationalist doctrine. For this reason, it is often unclear whether what is being discussed is the activity of the mind in perception or in acquisition – that is, in selecting an already distinct idea on the occasion of sense, or in making distinct what was before only confused and implicit.
