I

__________________________________

Some General Remarks on Linguistics and Language

_______________________________________

Inaugural lecture for the Chair of Comparative
Gramolar of the Indo-European Languages, given
December 17/29, 1870. al St. Petersburg University

GENTLEMEN: Inaugurating these lectures on the so-called com· parative grammar of the Indo-European languages. I consider it appropriate first to characterize, in general terms, the object of our studies and to define its relationship to other studies. The science to which the subject of my preseni talk belongs is linguistics; it follows that whatever is related to linguistics is also applicable to the comparative grammar of the Indo-European languages, although the latter has some features of its own which do not pertain to linguistics in general. This fact is inherent in the nature of the subject as well as in the methods which have been elaborated for it historically.

In the general characterization of our science, I shall try first of all to define its limits showing (1) what should not be expected of it, and (2) its essence. I shall then try to define its subject, i.e., the nature of language. My lecture will thus be divided into two parts: (1) linguistics ill general and the so-called comparative grammar of the Indo-European languages in particular, and (2) language in general, and especially the Indo-European languages. Both these parts are closely related to each other, overlap, and condition each other, making a precise division between them almost impossible.

As a starting point ill defining lhe science of linguistics. 1 shall take up the commonly held belief that grammar is the science which teaches us how to speak and write correctly in a certain language. This continues to be the view of many grammarians, who, in fact, define the subject of their science this way. Obviously, no one may impose on another a particular interpretation of a term, especially of a technical term. Thus we cannot demand that the general public or a certain school of grammarians discard their current definition of grammar. But we have a perfect right to tell them that such a definition and the attitude toward language which corresponds to it exclude grammar from the realm of science and class it with the arts, whose aim is the application of theory to practice. But can the student of language afford to limit his task in this way, and is this indeed the way in which the discipline of linguistics, as it has developed historically, presents itself?

The distinction between art in the broad sense (which includes more than the fine arts) and science in general corresponds to the dictinction between practice and theory, between invention and discovery. Art operates with technical rules and procedures, whereas science deals with generalization from facts, with their implications, and scientific laws.¹ Art has two aspects: (1) a continuous practice based on tradition, and (2) improvement in the means of realizing practical tasks. Analogously, the historical development of science presents: (1) the transmission of knowledge from one generation to another, and (2) the broadening and improvement of knowledge by diligent and talented scientists. Each step forward in art is made through invention; each step forward in science is made through discovery.

In the realm of language (as distinct from the science of language) one may also speak of art (or rather of arts) as applied to language in general and to individual languages. Among others these arts include:

A. Those which apply the results of science to everyday needs:

1) The first of these is the mastery of a language or languages, beginning at an early age and lasting throughout life. This process is partly a problem of didactics and partly a matter of independent work, the success of which depends on the ability and practical aptitude of the student. It may involve the mastery of (a) a native language, and (b) of foreign languages. Sources here depends to a large extent on the application of discoveries of linguistic science, which in the case of one’s native language provides an insight into (a) the acquisition of the language by the young child, and (b) its free and skillful use by the older child.² In the case of foreign languages it provides an informed and conscious approach that facilitates understanding, correct use, and fluency. The application of art in this area consists in the improvement of practice, exercises, and teaching methods.³ Midway between the study of a native language and the study of foreign languages is the study of the literary language, that language which unifies a whole people and facilitates the mutual understanding of its members, and which usually represents (with some modifications) the colloquial speech of the so-called educated class (in contradistinction to folk dialects). In countries where the literary language differs considerably from the local dialects (for example, in Germany), its study is for many people not much easier than that of a foreign language.⁴

2) The art of teaching language to deaf-mutes is characterized by methods all its own. Audible language is not accessible to deafmutes who possess only the language of pantomime. And although they may produce sounds which resemble language sounds, these sounds exist only for the hearer; the movements of the articulatory muscles are to the deaf-mutes of the same order as a grimace or gesture of the fingers. The training of deaf-mutes to pronounce words audibly takes into account the anatomical and physiological aspects of grammar.

Only a clear knowledge of the sounds of language (as opposed to their graphic representations) and of the origin and structure of words can provide

3) a good method of teaching children (and adults) to read and write a given language, and

4) an orthography and spelling which reflect the achievements of our science.

B. Everyday needs, the tendency toward convenience, simplicity, and facilitation of social intercourse, have brought forth in each nation possessing a literature of its own a single, more or less standardized, and educated language which serves to unify the members of the nation and which serves as a link between successive generations. Because of this very function, such a language unavoidably tends to be conservative and to resist the course of natural development. The role of human consciousness in this matter is of paramount importance.

The conscious and unconscious striving toward regularity, order, and correctness may, in turn, give rise to a linguistic purism which borders on pedantry⁵ and which leads to constant interference with the natural course of language, to the proscription of phenomena that appear to be irregular, and to prescriptions of how things should be. The grammarians concoct “inexorable” rules, together with exceptions to the rules,⁶ that are supposed to enable a foreigner to speak a given language “more correctly” (i.e., in accordance with the grammar) than the native. But grammarians, approaching language from such a viewpoint, do not understand its development: they do not know that seeming grammatical exceptions can be explained by the history of the language and are either remnants of past “rules” or harbingers of future ones.⁷ It is pleasant to legislate, if only in the field of language, and every (or almost every) practically oriented grammarian tries to arrogate to himself the right to pass verdicts in this area. It is very difficult to resist this temptation, and the disposition to organize and to improve one’s native tongue is shared even by enlightened and objective minds.⁸

If to the above-mentioned factors we add the effect of ever greater politeness and flattery, accuracy and logic, we obtain a more or less complete picture of the interference of man’s consciousness upon language, which at a certain point of development introduces true artificiality into language. Although the extent of this interference is limited, it leaves a definite mark on the structure and make-up of language.

C. In addition to the linguistic skills which form a bridge between science and life, between theory and practice, there is that skill which enables science to operate, namely its technical aspect. It includes, among others, the practice, continuity and dissemination of science, and allows for its growth, new discoveries, and methods; it speeds up the research of scholars and facilitates its mastery by beginners.

The main conditions of the pursuit of science are, in my opinion, the availability of adequate data and proper scientific method.

Adequate data can be gathered only through the study of pertinent phenomena, through the identification of the scientific facts which circumscribe the object of inquiry. In the case of language this means the practical study of languages, the categories of which we wish to comprehend scientifically and treat theoretically.⁹ Only by acquiring a practical knowledge of the languages one is dealing with can one avoid the kind of errors committed, for example, by Th. Benfey in his Griechisches Wurzellexicon, where the author interprets the Old Church Slavonic prazdǐnujǫ ([ferior) as “they beat me” instead of “I celebrate,” or ukradǫ (furor) as toben “to rave, make noise” instead of stehlen “to steal,” etc.¹⁰ For our purposes it is sufficient to understand the languages; to speak them fluently and to write them is desirable,¹¹ but not necessary.¹²

The gathering of data goes hand in hand with the elaboration of a scientific technique, or method , for (1) the analysis of facts and the inferences drawn from them, and (2) the presentation and transmission of scientific results. The latter includes all kinds of teaching aids and exercises, and such auxiliary means as translations from one language into another or from one form into another (e.g., morphological and phonetic translations), etc.

One may now ask whether the subject of our course, linguistics, falls within one of the above-mentioned practical pursuits. Will we attempt to provide instructions for the successful study of Ianguages by people with the gift of speech or by deaf mutes? Or will our course be a guide for the study of reading and writing, including orthographical prescriptions and rules? Or shall we be concerned with the purification and improvement of language and its application to everyday needs? Or will our lectures, finally, be an exposition of technique, methodology, and propaedeutics?

I must answer all these questions in the negative. The subject of our courses will not be the acquisition of a skill, nor its practice, nor a technique, but primarily the pursuit of science, theory, seientific knowledge, with science being interpreted as the exercise of the human mind on the sum total (or complex) of more or less homogeneous facts and concepts.

But the theoretical study of language can also be heterogeneous, depending on the interpretation of the tasks and methods of our science. If we ignore the purely practical aspect, whose aim is the mastery of foreign languages with a maximum of fluency and a minimum of reflection (which is in direct contradiction to the requirements of science, since it calls for a passive attitude toward foreign languages plus an ability to imitate, whereas science aims at a conscious mastery of facts by means of independent reflection—cf. the remarks above concerning linguistic skills¹³), linguistics or the historically developed, conscious, scientific investigation of languages and human speech in general can be discussed in terms of its three distinct approaches:¹⁴

1) A descriptive approach, an extreme empirical one, which sees its task in gathering facts and generalizing from them in a purely external manner, without attempting to explain their causes or to establish their affinity and genetic relationship. The adherents of this approach see all the wisdom of science in the preparation of descriptive grammars and vocabularies, in editing linguistic records, in preparing material without any conclusions, which for some reason or other are always considered overly bold or premature. This attitude stems in part from too strict a critical or skeptical turn of mind which flatly denies the possibility of any real science for fear of reaching the wrong conclusions or of stating hypotheses that might eventually be disproved; in part, however, this attitude must be ascribed to a kind of mental laziness and a desire to avoid the question of the usefulness and goal of gathering data, a desire which thus reduces science to a purely empirical endeavor, to some sort of meaningless game. These scholars defer the explanation of data ad acta,ad meliora tempora, losing sight of the important fact that the accumulation of details and their primitive, slavish, purely external explanation may eventually be useful not to themselves, nor to science directly, but to other investigators who will profit from these data if they are trustworthy and conscientiously prepared. It is obvious that if science is to avoid this situation of not seeing the forest for the trees, its practitioners must by all means abandon this approach.¹⁵ Nevertheless, as the first, preliminary step in science, pure description is indispensable. Accurate and faithful observation is, indeed, the first requirement of science, and it is given to few to do it to a high degree of perfection; many can look, but only few can see. Good descriptive grammars, text editions, and dictionaries will always remain a vital need of our science; without them theoretical conclusions of genius will lack a factual foundation.

2) The contrary of this modest and restrained approach is the speculative, philosophizing, aprioristic and childish approach, whose adherents recognize the need to explain phenomena but do not know how to go about it. They think up certain aprioristic (both general and particular) principles and force the facts without much ado, into the strait jacket of these principles. Herein lies the source of the most heterogeneous, biased grammatical theories concerning both the development of language and such language-related fields as history, the study of antiquities, mythology, ethnography, etc. These theories are the source of innumerable arbitrary and unfounded explanations which merely attest to the lack of common sense in their authors. Who has not heard the queer etymologies for which one would like to put Messrs. Etymologists in a lunatic asylum? As the alchemists tried to reduce all existence to primitive matter and a mysterious universal force, so some of the representatives of the aprioristic approach try to derive all the wealth of human speech from one or several consonantal groups.¹⁶ Nowadays alchemy is obsolete but linguistic alchemists are still with us; as a matter of fact there is little hope in the near future of driving fantasy and arbitrariness out of the field of linguistics.

The aprioristic trend in linguistics has in recent times given rise to the so-called philosophic school, which uses speculation and a limited knowledge of facts to construct grammatical systems that force linguistic phenomena into a logical strait jacket. Of course such systems may represent more or less sophisticated schemes of scholarly minds, products of the art of logic, impressive in their harmony and order; but because they violate and distort the facts for the sake of a narrow theory, they are no more than castles in the air and cannot satisfy the requirements of positive thinking.

If the descriptive, blunt, empirical approach merely holds back the development of science, the aprioristic, arbitrary, and childish approach leads it astray. Because of this, it is decidedly harmful.

3) The truly scientific, historical, genetic approach views Ianguage as the sum total of actual phenomena, of actual facts, and the discipline analyzing them as an inductive science. The task of an inductive science is (1) to explain phenomena by comparing them with each other, and (2) to establish forces and laws or the fundamental categories and concepts that connect the phenomena and present them as a chain of cause and effect. The first of these tasks is to present to the human mind a systematic knowledge of a certain sum of homogeneous facts or phenomena, whereas the second introduces into the inductive sciences an ever more prominent deductive element. In the same way, linguistics as an inductive science (1) generalizes the phenomena of language and (2) looks for the forces operating in language and for the laws that govern its development, its life. Of course, all facts have equal rights and can be viewed only as more or less significant; by no means can some be deliberately ignored, and it is simply ridiculous to sneer at facts. All that exists is reasonable, natural, and lawful; this is the watchword of any science.

Many see in “comparison” a special, distinctive mark of the modern science of language, and therefore quite readily and almost exclusively use such terms as “comparative grammar,” “the comparative study of languages” ( vergleichende Sprachforschung),“comparative linguistics” ( vergleichende Sprachwissenschaft),“comparative philology” (philologie comparée), etc. It seems to me that this view is motivated by a certain narrowness and clannishness; if we consider the intentions of the “comparative” grammarians and other “comparative” investigators of language, we can say that every science could be adorned with the epithet “comparative.” And thus we could speak of comparative mathematics, comparative astronomy, comparative physics, comparative psychology, comparative logic, comparative geography, comparative history, comparative political economy, etc. Indeed, comparison is one of the indispensable operations of all sciences, as it is the basis of thinking in general. The mathematician compares magnitudes, obtaining in this way data for his synthetic and deductive reasoning; the historian arrives at his conclusions by comparing different phases of a certain kind of human events. Comparison in Unguistics plays the same role as it does in the other inductive sciences: it enables us to generalize from facts and opens the road to the deductive method. From another viewpoint, however, “comparative grammar” has a historical meaning: it owes its origin to a new school, a new trend that has made significant discoveries. For this school comparison has meant the comparison of related languages and, more generally, of similarities and dissimilarities of Ianguages;¹⁷ but it does not mean comparison of linguistic facts in general, for the latter is a necessary prerequisite of all scientific analyses of language. A similar historical meaning attaches to the names “comparative anatomy,” “comparative mythology,” and so on. However, we must not forget that this is only a single moment in the history of a science, when comparison applied in a novel, hitherto unknown way, has yielded vast and completely new results. But if we are to name a science, not after its transient phases or particular scientific procedures but after its subject matter, then we should forgo such names as comparative grammar or comparative linguistics or explanatory grammar¹⁸ (erklärende Grammatik) or explanatory linguistics (erklärende Sprachwissenschaft) or comparative philology.¹⁹ The most appropriate name for the science of language (which is not unlike the natural sciences) would then be simply the study of languages and of human speech in general, the science of languages (jazykovedenie, jazykoznanie), or linguistics (lingvistika, glottika). The name should not commit us to anything but only indicate the subject of the field in which the scientific questions arise. At any rate, a science may be called by any name, and in particular it may be called “comparative,” so long as it is clear that comparison is not the end but only a means,²⁰ and that it is not the monopoly of linguistics but the common property of all sciences without exception.

I have noted that linguistics investigates the life of language in all its manifestations, correlates the phenomena of language, generalizes from the facts, sets up laws of the development and existence of language, and looks for its operating forces.

Law here means a formulation, a generalization that states that under certain conditions, after a or b, there appears x and у _ג or that a and b in one domain of phenomena (for example, in one language or in one category of words or forms of a given language) corresponds to x and у in another domain.²¹ For example, one of the general laws of the development of language holds that a more difficult sound or group of sounds is replaced in the course of time by one more easily pronounced; another holds that abstract notions develop from more concrete ones, etc. There are seeming exceptions to these laws; nevertheless, under closer scrutiny the exceptions appear to be governed by certain causes, by forces that have prevented the causes or forces accounting for the general law from embracing the seeming exceptions. Once convinced of this, we must realize that our original generalization was inaccurate and incomplete and that the genus proximum of our law must be corrected by a limiting differentia specifica. Only then does it become clear that the seeming exception is, strictly speaking, only a corroboration of the general law.²²

It is appropriate to call the common causes, the common factors, which effect the development of language and determine its structure and composition, forces. Such forces are, for example:

1) habit, that is, unconscious memory;

2) a tendency toward convenience, which expresses itself (a) in the transition from more difficult sounds and groups of sounds to easier ones (in order to conserve muscular and neural activity), (b) in the simplification of forms (through the analogical influence of stronger forms on weaker), and (c) in the transition from the concrete to the abstract in order to facilitate abstract thinking;

3) unconscious forgetting and failure to understand (forgetting what was not known even consciously and failure to understand what could not even be understood consciously); but not fruitless, negative forgetting and lack of understanding (as is the case in conscious mental operations), but productive, positive forgetting and lack of understanding, which produce something new by encouraging unconscious new generalization;

4) unconscious generalization, apperception, i.e., the force by which a nation groups all the phenomena of mental life into certain general categories. This force is analogous to the force of gravity in planetary systems: just as the force of gravity produces certain systems of heavenly bodies, so the force of unconscious generalization accounts for certain systems and families of linguistic categories; and just as a heavenly body when it leaves the field of attraction of a given planet, moves in space alone until it is drawn into a new system, so certain words or forms whose connection with related forms is lost to the feeling of a nation (or, like loanwords that have at first no connection with the native words of a given language) stand by themselves until they are attracted into some family of words or category of forms through the speakers’ ability to create new words through analogy, etc.;

5) unconscious abstraction , the unconscious tendency toward separation and differentiation. If apperception is the centripetal force in language, this force (based on unconscious abstraction) can be compared with the second, or centrifugal force,²³ while both forces together make up the force of gravity in general.

The action of the above linguistic forces is deployed in two areas and presents, as it were, two aspects:

1) the purely physical aspect of language, its system of sounds and groups of sounds which is connected with the organic make-up of a given speech community and which is subject to the continuous influence of inertia (vis inertiae);

2) the feeling for the language of a given speech community. The speakers’ feeling for a language is not a fiction, not a subjective invention; it is a real, positive category (function) that can be defined in terms of its properties and effects, and which can be objectively confirmed and proven by facts.

The struggle of the above-mentioned forces determines the development of language. Of course, this struggle and the action of linguistic forces in general must not be personified; science does not deal with myths, but with representations and concepts. Laws and forces are not living beings; they are not even facts; they are products of man’s intellectual activity, whose purpose is to generalize, to correlate facts, and to find their common denominator, their common formula. Nor are they demonic ideas, as suggested by philosophers of a certain school; they are generic concepts (Artbegriffe), which are all the more powerful the wider the range of the phenomena which they encompass and explain. On the other hand, these laws and forces, as, in general, all concepts and conceptual categories, are not global in their kind, but sums of innumerable specific representations and concepts.

I shall refrain from a more detailed analysis of forces and laws, since (1) time does not permit, and (2) strictly speaking, it is the subject of logic, that science which deals with the foundation of knowledge and abstract thought in general. I shall only pose the question: can the general categories of linguistics²⁴ be treated as laws and forces in the same sense as the laws and forces of physics and other natural sciences? Indeed they can, since the forces and laws of the natural sciences are nothing but unifying formulas of thought, more or less successful generalizations. Their superiority lies only in the simplicity of the phenomena and facts which they examine and in the uninterrupted development of the sciences themselves, which have allowed them to apply mathematical computations and thereby to acquire a high degree of clarity and accuracy, while the highly complex processes of language and the brief history of the science of linguistics have kept its generalizations in a more or less precarious and unstable state. This, however, should not perturb us, considering that the general categories arrived at by the most recent schools in the biological sciences (zoology and botany) are hardly more accurate and clear; they too are only more or less successful generalizations and not forces and laws, if they are to be measured by the same exacting standards that we have become accustomed to apply in the fields of astronomy, physics, chemistry, etc.

It should be apparent from the foregoing remarks that two elements are inseparably linked in language: a physical and a psychological one (of course, these terms must not be interpreted as metaphysical but as generic concepts). The forces and laws and the life of a language in general are based on processes which are of concern to physiology (to anatomy and acoustics) and to psychology. But the same physiological and psychological categories make up a rigidly defined subject which is investigated by the historically developed science of linguistics. Most of the questions raised by the linguist are never broached by the physiologist or psychologist; consequently, linguistics must be regarded as an independent seience, not to be confused with either physiology or psychology. Similarly, physiology examines the same processes, laws, and forces of whole organisms whose abstract analysis is of concern to physics and chemistry; nevertheless, one cannot dismiss physiology in favor of the latter sciences.²⁵

Having thus defined, though only in an approximate and inexact manner, the subject of our discipline and the scientific approach that reflects best its present-day state, I shall try to draw an outline of its internal organization, that is, to present to you, gentlemen, the basic branches of linguistics as they have developed historically.²⁶

One must first distinguish pure linguistics, linguistics per se, whose subject is language itself as the sum total of more or less homogeneous facts that belong totally to the so-called manifestations of human life, and applied linguistics, whose subject is the application of the results of pure linguistics to questions pertaining to other sciences.

Pure linguistics presents two large areas of research:

A. the exhaustive study of empirically given, fully formed languages.

B. the study of the origin of human speech, the original formation of languages, and the investigation of the general psychophysiological conditions of their uninterrupted existence.²⁷

A. Empirical linguistics is divided into two parts: (1) in the first, language is examined as a composite of parts, that is, as the sum of heterogeneous categories which are organically (internally) interrelated; (2) in the second, languages as wholes are investigated according to their affinity and formal similarity. The first part is grammar, the analysis of the structure and composition of language; the second is systematics, or classification. The first can be compared to anatomy and physiology, the second to plant and animal morphology in a botanical-zoological sense.²⁸ It should be understood that, as everywhere in nature and in science, there are no sharp borderlines between them and that research in one field is based on and delimited by the results obtained in another field. In order to analyze the structure and composition of a certain language, it is very useful, perhaps necessary, to know to what formal type the language belongs; and in order to explain its phenomena by corresponding phenomena of related languages, one must define its place within a certain linguistic branch or family. Similarly, only an examination of the structure and composition of languages provides a solid foundation for their classification.

I. To give a step-by-step analysis of a language, one may subdivide grammar into three basic parts: (1) phonology (phonetics), or the study of sounds, (2) word-formation, in the broadest sense of the word, and (3) syntax.

1. The first condition for a sucessful study of sounds must be a strict and clear distinction between sounds and their graphic representations. Since no orthography is completely consistent or accurate in rendering sounds and sound combinations, and since inadequate methods of teaching and practice have contributed, or, to be more fair, have not removed the misconceptions stemming from a primitive concrete conception of reality, the observance of the above-stated condition requires that one approach the analysis of sounds in parallel with writingw; each sound or group of sounds must be compared with the corresponding graphic sign or letter or group of letters. The subject of phonetics includes:

a) the analysis of sounds from a purely physiological point of view, the natural conditions of their formation, their development, their classification, their division (even here we cannot approach language apart from man, but must regard sounds as acoustic products of the human organism);²⁹

b) the role of sounds in the mechanism of language, their value for the feeling of a speech community, is determined not so much by their physical properties as by their physiological nature and their origin and history, which involves the analysis of sounds from the viewpoint of morphology and word formation;

c) the analysis of sounds from a historical viewpoint: the genetic development of sounds, their history, and their etymological and morphological affinities and correspondences.

The first, physiological (a), and the second, morphological (b) parts of phonetics analyze the laws and conditions of the existence of sounds at a given stage or moment of the language (the statics of sounds). The third, or historical part studies the laws and conditions of the development of sounds in time (the dynamics of sounds).

2. Word formation, or morphology, traces the gradual development of language: it reconstructs the three periods of this development (monosyllabism, agglutination or free juxtaposition, and inflection). The parts of morphology are:

a) etymology, or the study of roots;

b) stem formation, or the study of derivational suffixes and of stems or bases;

c) inflection, or the study of desinences and complete words that are found in highly developed, inflected languages.

As generally in nature and science, here too it is difficult to draw clear-cut boundaries and to decide whether a given question falls into one or another category. The transition from a lower to a higher stage of development (or from an earlier to a later stage) is, after all, accomplished, not by leaps and bounds, but slowly, gradually, and imperceptibly.

3. Syntax, or word combination, investigates words as parts of sentences and defines them by their function in connected discourse in the sentence (a function which determines their division into parts of speech); it studies the meaning of words and forms in their interrelation. It, furthermore, analyzes whole sentences as parts of larger units, and the combination and relationship of these units.

Just as anatomy is not applicable to the study of all organisms (for example, osteology applies only in the study of vertebrates), not all of the above-mentioned parts of grammar apply to all languages. Thus, for example, the study of monosyllabic languages (whose main representative is Chinese) involves only phonetics and syntax; of word formation there remain only the problems of etymology, that is, the analysis of particular kinds of roots.

The grammatical study of a language requires adherence to the chronological principle, that is, the principle of genetic objectivity with regard to the development of language in time. This principle is founded on the three following propositions:

1) No language is born suddenly; it is the result of a gradual and unique development through different periods spanning many centuries. The periods of development do not replace one another as one sentry replaces another, but each period creates something new which, in the imperceptible transition from one stage to another,³⁰ prepares the basis for the subsequent development. The cumulative effects of changes occurring over a period of time can be observed only at discrete stages; in the natural sciences they are called strata. Similarly, one may speak of strata with regard to language, and the separation of strata is one of the principal tasks of linguistics.³¹

2) The mechanism of a language (its structure and composition) at any given time is the result of all its preceding history and development, and each synchronic state determines in turn its further development.

3) It is inappropriate to apply to the structure of a language at a given period the categories of a preceding or following period. It is the task of the investigator to describe a language structure precisely as it is at each period and only then to demonstrate how the structure and composition of that period could give rise to the structure and composition of the following period. The requirement of genetic objectivity also applies to the investigation of different languages; it is unscientific to ascribe arbitrarily the categories of one language to another language; science must not impose alien categories on its object of study, but must seek in it only what really exists, what defines its structure and composition.³²

Grammatical problems can be tackled in two ways: they can be discussed in terms of the general categories of ouroscience, in terms of the homogeneity of the particular facts, or they can be treated in terms of their genetic development.³³ The first approach identifies similar phenomena in various areas of human speech, or in all Ianguages accessible to the investigator, or within a strictly defined group of languages (or even in one language), with the ultimate goal of formulating general categories, the laws and forces which explain linguistic phenomena. The second approach describes the natural course of a language, abstracting and systematizing only as much as is indispensable for any science; otherwise it traces the internal development of its subject matter (either from the oldest to the most recent period or only of one particular period). This internal history of a language (or languages) must be clearly distinguished from its external history,³⁴ which approaches Ianguage ethnologically, from the point of view of the fate of its speakers, and which is consequently a part of applied linguistics, inasmuch as it applies its systematics to ethnography and ethnology (it is in this sense that we speak of applied linguistics as bearing on other sciences). Ordinary grammars select and describe only a certain period in the history of a language. But in order to be truly scientific, one ought to compare the state of the language at this period with the entire development of the language.

The degree of scientific perfection reached by contemporary linguistics is such that, with the amount and precision of available empirical data, and with the knowledge of the history and current tendencies of a language, and drawing upon comparable findings in other languages, we can generally predict the future internal development of that language or construct its past in the absence of written records.³⁵

For lack of time I shall not give examples, the more so because in this course many examples will be brought to your attention. Of course, with regard to the future, these scientific (but not prophetic) linguistic predictions are by no means as exact as the predictions of, let us say, astronomy. They are only capable of pointing out in general terms a future phenomenon, but are incapable of determining the exact moment of occurrence of such a phenomenon. But even this is quite encouraging and proves the validity of current research methods, and brings linguistics closer to the goal of all inductive sciences, that is, the broadest possible use of the deductive method.

II. Systematics, classification of languages, must not be undertaken with a view to superficially facilitating their study by ordering them on the basis of randomly selected or preconceived characteristics. A truly scientific and modern classification of languages must take into account their natural development and be founded on their genuinely distinctive properties.

In the field of cognate languages,³⁶ that is, of languages that have developed from a common language and have merely modified the same original material (as a result of the diverse conditions that affected the speakers of these languages), classification itself is only a different aspect of the history of a language. We need only regard these languages as entities, or better, as complexes of meaningful sounds and sound groups presenting a unified whole in the mind of a given people and, on the other hand, to single out the properties of individual languages which set them apart from others or which they share with a larger group in order to realize that the history of language yields eo ipso a genetic classification of Ianguages. At the same time one must keep in mind that the principle of the genetic differentiation of related languages must be sought not, as is usually done, in phonetic, lexical, and formal differences, but in the general tendencies that determine the unique development of the entire mechanism of a language, since only these tendencies constitute characteristic and invariant properties which enable us to identify individual languages in a family of more or less closely related languages.

Along with genetic classification, there is morphological classification, which distinguishes languages in terms of their structural properties, i.e., of those properties which make up the second part of grammar: morphology, or word formation.³⁷ The morphological difference between various groups of languages is the result of an original and fundamentally different world view of the speakers of these languages, which must have preceded the actual formation of these languages and contributed, in fact, to their formation.³⁸ Therefore, morphologically different languages cannot be genetically related.³⁹ Conversely, genetically different languages may belong to the same morphological type, exhibiting an identical, or at least similar structure.⁴⁰

Cutting across the genetic and morphological classification there is the division of inflected languages into primary and secondary, synthetic and analytic. In the primary languages the composition of words is still keenly felt and flexion is still expressed by means of endings, etc. In the secondary languages, words are made up of mere sequences of sounds, and inflectional relationships are expressed by means of independent words. The basic character of the two types is, nevertheless, the same: they only employ different means to render similar tendencies and functions. Here too, one can hardly draw sharp boundaries, in view of the great variety of transitional states. Moreover, even the most evolved flexional languages exhibit rudiments of the secondary type, while the Ianguages that have moved furthest in the direction of the secondary type retain traces of their original primary structure.

Of the seven known, scientifically described, and genetically ordered groups of languages,⁴¹ two have attracted the greatest attention, for they are most highly advanced in their morphological, flexional structure and their speakers have formed the most civilïzed and influential nations in world history. These are the Semitic and the Indo-European groups (branches) of languages.⁴² The latter has been subjected to a particularly close scientific scrutiny, and the method of studying it has been applied to the study of other groups as well.

The method of our science can best be understood when it is applied to a strictly defined family or branch of languages. We shall study the Indo-European languages exclusively, investigating their structure and composition during their historical development and their mutual relationships. The Indo-European (Aryo-European) branch is divided into eight separate families: (1) Indic, (2) Iranian, (3) Greek, (4) Romance, (5) Celtic, (6) Germanic, (7) Lithuanian, and (8) Slavic. These families separated at different times from a common original language, but some of them remained together longer than others and present more resemblances. On the last matter scholarly opinion is divided, and one must confess that the methods used for resolving it are far from satisfactory (cf. the discussion above on the principle of the genetic classification of languages). We may regard as settled the close kinship of Indie and Iranian, of Slavic and Lithuanian, and probably, of Greek and Romance.⁴³

The oldest written records of the Indo-European branch are represented by the Vedas, the sacred literature of the Indians. Their language, Sanskrit, later became, in a somewhat modified form, a common literary vehicle. Thus Sanskrit is not a popular language, but rather a sacred literary language of the same order as Old Church Slavonic (for the ancient Slavs) and contemporary High German (for the Germans). Sanskrit is very important for the study of the Indo-European branch in general both because of the detailed and sucessful analysis which it received at the hands of native grammarians and because of its transparent structure. However, Sanskrit is only one of the members of the IndoEuropean branch of languages (it is not the proto-language, being but one of the derived languages with the oldest records), and it does not exhaust the wealth of questions raised in the study of this branch; strictly speaking, similar questions arise in a detailed historical study of any other language or family (for example of Slavic, Germanic, or Greek). But a thorough knowledge of the grammatical structure of Sanskrit will do no harm, as long as one avoids the widespread fetishization of Sanskrit and as long as one does not force its categories on the study of other languages.

B. Grammar and systematics complete the scientific investigation of historically existing languages. The second part of pure linguistics deals with questions that lie outside the domain of historically attested facts, such as the beginnings of human speech, its original formation, the general psychological-physiological prerequisites of its existence, the influence of a people’s world view on the development of their language and, conversely, the influence of language on the world view of its speakers, on the psychological development of a people, etc. Many investigators of Ianguage relate these questions to anthropology and psychology, but it seems to me that since they refer to language, they should also be studied from a linguistic viewpoint, the more so as their solution depends on data provided by the empirical part of our science, by historical linguistics.

Applied linguistics includes:

1) the application of grammatical data to quesions of mythology (etymological myths),⁴⁴ antiquities, and the history of culture (by comparing cuturally and historically important words, which are especially pertinent to the reconstruction of a prehistoric period, the study of which constitutes the field of so-called linguistic paleontology). Grammatical investigations are also relevant to the study of mutual relations between various peoples;

2) the application of systematics to ethnography and ethnology and to the history of peoples in general (the classification of languages in relation to the classification of peoples); and, finally,

3) the application of the results of the second part of linguistics (the beginning of language, etc.) to questions of anthropology, zoology, etc. (the contribution of the linguist to these fields is, however, secondary).

In the preceding exposition, I have tried to define linguistics, to point out its basic questions, and to present its internal organization and historical development. But so far I have not posed the question of what is language, though a clear, even a negative, definition would be quite instructive.⁴⁵

Before answering this question, I consider it necessary to refute most emphatically the prejudice held by some scholars that language is an organism. This view was formed on the basis of analogy, though it must be apparent that analogy does not constitute a proof. The recourse to analogy is, in effect, a way of avoiding genuine, serious analysis. It gives rise to empty talk, to scholarly phrase-mongering, which misleads superficial minds and serious people alike. Without entering into a more detailed analysis and critique of the thesis that language is an organism, and without bothering to define the nature of an organism, I shall only remark that an organism, like an inorganic substance, is something tangible that fills a certain space and which, furthermore, eats, drinks, breeds, etc.,⁴⁶ whereas man’s speech (and the existence of language depends on that, of course) is connected with the movements of his speech organs and the resulting vibrations in the air that produce the sense impressions which are associated with corresponding representations in the mind of the hearer and the speaker.⁴⁷ To treat language as an organism is to personify it, to detach it completely from the speakers, from man, and to lend credence to the story of a certain Frenchman who was supposed to have said during the winter campaign of 1812 that his words did not reach his listener’s ear but froze in the air halfway. If language is an organism, it must indeed be a very delicate one, and the parts of this organism, words, could hardly be expected to brave the severe Russian cold.

I shall forgo analyzing all the errors and fallacies that ensue directly or indirectly from the misconception that language is an organism.⁴⁸ But before I submit the final definition of language, I shall draw your attention, first, to the distinction between human speech in general (as a sum of all present and past languages) and between separate languages, dialects (or the languages of each individual).⁴⁹ Second, I would like to point out the distinction between language as a complex of constituent parts and categories that exist only in potentia, as a sum of all possible individual variations,⁵⁰ and language as a continually recurrent process based on the social character of man and his need to give concrete expression to his thoughts and to communicate them to other human beings (language—speech—the human word).

In the light of all that I have said, as well as what has not been fully said, and even of what has been left unsaid, I propose the following definition of language: Language is the audible result of the normal activity of muscles and nerves.⁵¹ Or: language is a complex of separate and meaningful sounds and groups of sounds which are unified into a whole by the feeling of a certain people (as a collection of perceiving and unconsciously generalizing individuals) who form, in turn, one category, one intellectual species, owing to the language which they all share . . ....

NOTES

_________________________________

1. In the historical development of any art (not the fine arts only), one must distinguish between unconscious art, that is, ordinary practice preceding theory (although even here unconscious invention is possible), and conscious art guided by theory and knowledge. In the same way, in science one can distinguish between the accumulation of fragmentary knowledge by the savage or uneducated man and the critical analysis of facts and the conscious generalizing of the educated and capable man.

2. See “On the Study of the Native Language in General, and in Particular during Childhood. From the Discussions of I. I. Sreznevskij” (Izvestija imp. Akademii nauk po otdeleniju russkogo jazyka i slovesnosti, vol. IX, 1860, pp. 1-51, 273-332; special reprint; fasc. 2, SPb, 1861). Here we also include all textbook-grammars and other books whose aim is to facilitate the study of their native language and the mastery of foreign languages by children.

3. Here we include the art of translating well from foreign languages into one’s own language and vice versa.

4. The study of rhetoric pertains to linguistics only externally, i.e., as socalled orthoepy; internally, i.e., in terms of the choice and arrangement of thoughts, rhetoric pertains to dialectics or logic.

5. Cf. Jakob Grimm, “Über das Pedantische in der deutschen Sprache,” Kleinere Schriften, I (Berlin, 1864), pp. 372-73.

6. Such rules and exceptions can only evoke an aversion for grammar in an independent, positive, and objective mind. Goethe said of grammatical rules and exceptions: “Die Grammatik missfiel mir, weil ich sie nur als ein willkürliches Gesetz ansah; die Regeln schienen mir lächerlich, weil sie durch so viele Ausnahmen aufgehoben wurden, die ich alle wieder besonders lernen sollte” (Aus meinem Leben. Dichtung und Wahrheit, I).

7. “Alle grammatischen Ausnahmen schienen mir Nachzügler alter Regeln, die noch hier und da zucken, oder Vorboten neuer Regeln, die über kurz oder lang einbrechen werden” (J. Grimm, Über das Pedantische in der deutschen Sprache,” p. 329).

8. Thus, for example, even Schleicher, who considered himself strictly an observer of the natural development of language and made no allowance for the interference of free will upon the natural development of human speech, was concerned with the purity of his native language; he frequently spoke out against various “nonorganic,” as he called them, phenomena in the German language that had been introduced by ignorant schoolmasters and called upon his fellow countrymen to avoid such transgressions. In particular, the second part of his Deutsche Sprache is full of passages remarkable for their sermonizing vein and patriotic zeal. Here the practical approach was carried to an extreme. This book had the incidental purpose of awakening the “Nationalgefühl” of the Germans; in my opinion this is almost the same as trying to stimulate an appetite by an article on cooking.

9. “As an ‘empirical’ linguist,” says Schleicher, “I am firmly convinced that only a working knowledge of languages can be the basis for linguistic studies and that one must first strive, as far as possbile, to become acquainted with the languages selected as the subject of research. Only on the basis of solid, positive knowledge can one do anything worthwhile in our science. Didicisse juvat. Thus, one who wants to devote himself to Indo-Germanic linguistics must first thoroughly study all the older Indo-Germanic languages, read texts, etc. Anyone who neglects some languages as less important will surely regret this later.” (Schleicher, in Die Wurzel AK im Indogermanischen von Dr. Johannes Schmidt. Mit einem Vorworte von August Schleicher [Weimar, 1865], p. iv).

10. Cf. Aug. Schleicher, Die Formenlehre der kirchenslawischen Sprache . . . (Bonn, 1852), p. xi.

11. It is desirable to develop as fine a feeling for the languages studied as the general education of former times (the 16th and 17th centuries in Western Europe) enabled one to acquire for the so-called classical languages (Latin and Greek, but mainly Latin).

12. Knowledge and understanding of languages differ from command more or less as knowledge of physiological processes differs from their performance (of course, the disparity between the respective subjects makes for the inaccuracy of the comparison).

13. And for one who studies the theoretical side of linguistics, it is quite useful to acquire the widest possible knowledge of various languages, as I mentioned above.

14. Cf. System der Sprachwissenschaft von К. W. L. Heyse, . . . (Berlin, 1856), §§ 5-9, pp. 6-21; Geschichte der Sprachwissenschaft und orientalischen Philologie in Deutschland . . . von Theodor Benfey (München, 1869), pp. 1-12 ff.

15. The natural consequence of this approach is a narrow particularism which denies the propriety of comparing similar phenomena in different languages and restricts itself to the confines of a single language.

16. Thus, for example, all the words of all languages are regarded as arising from consonant groups signifying “rooster” in Dr. Max Müller’s Bau-WauTheorie und der Ursprung der menschlichen Sprache . . . von Dr. Christoph Gottlieb Voigtmann,… (Leipzig,1865); cf. Johannes Schmidt, in Zeitschrift für vergleichende Sprachforschung, XV, pp. 235-37. Voss concludes that all the words are from the original groups phy ō , feo, and geo; cf. J. Grimm, “Über Etymologie und Sprachvergleichung,” Kleinere Schriften, I, p. 307.

17. Recently there has been a tendency to compare human language with that of animals; this kind of comparison may be expected to yield completely new results.

18. As is known, the explanation of phenomena is the basic goal of all science; explanation, therefore, cannot be considered the monopoly of one or a few of them.

19. To identify philology with linguistics means, on the one hand, to narrow the scope of their respective subjects (since philology deals with all phenomena of the mental life of a people, not only with language), and on the other hand, to define them too broadly (since philology has so far confined itself to a certain people or group of peoples, while general linguistics investigates the languages of all peoples). Philology, however, as it has developed historically, is not a homogeneous science but a conglomerate of parts of different sciences (linguistics, mythology, history of literature, cultural history, etc.) which find their common denominator in the fact that the carriers of these heterogeneous phenomena, which make up the subject and scientific problems of philology, are the same. Hence we can speak of classical (GraecoLatin) philology, Sanskrit, Germanic, Slavic, Romance philology, and many others.

20. Some scholars, in fact, see all the wisdom of linguistics in comparison for comparison’s sake (ars gratia artis), and they forget that there are many other interesting aspects of scholarly pursuit.

21. This is the basis for distinguishing laws of development in time from laws that define the synchronic state of a subject in space (or at any given moment of its existence), that is, for distinguishing that which undergoes change from what is essential and fundamental. The laws of one type pass into the laws of the other type; they are mutually related.

22. The necessary conditions for a scientific law are: (a) with regard to the subject: identification, clarity, and accuracy; (b) with regard to the object: generality.

23. Besides these and similar forces that affect the whole life of language, we must assume the existence at a certain stage of the development of mankind of another force (albeit a comparatively weak one): the influence of man’s consciousness on language. This influence unifies the forms of language and in its own way improves it, since it is a consequence of the striving toward the ideal, as discussed above (in the analysis of linguistic skills). Although the influence of consciousness on language is manifested fully only in some individuals, its results are, nevertheless, transmitted to the whole speech community, slowing down the development of language, counteracting the influence of unconscious forces, which generally tend to accelerate development, all this for the sake of making language an agent of unification and the mutual understanding of all the present, past, and future members of a speech community. This (force) produces a certain stagnation, whereas languages which are not subject to the influence of human consciousness develop rapidly and without inhibitions. The role of consciousness also expresses itself in the (conscious and unconscious) influence of books (and literature in general) on the language of a literate people (e.g., the influence of Church Slavonic books on the pronunciation of the clergy in the Orthodox Slavic countries), the influence of literary on the popular language (e.g., the influence of Church Slavonic not only on the inventory of words but on the structure of spoken Russian) or the acceptance of certain bookish and newspaper expressions as stereotyped phrases and clichés in everyday language; cf. Zeitschrift für Völkerpsychologie und Sprachwissenschaft . . . , ed. by Lazarus and Steinthal, V, pp. 106-9.

Sometimes, despite all the efforts of investigators, it is impossible to discover what forces or causes bring about particular phenomena. In such a case, the question of cause must remain unanswered until more favorable circumstances allow the explanation of the phenomenon in terms of cause and effect. For a methodical mind, it is impossible to accept the existence of phenomena without cause and at the same time to pursue science seriously. Nevertheless, many scholars, in analyzing the various manifestations of the so-called inner life of man, prefer mystical explanations to natural ones and introduce into science completely unscientific categories such as purposefulness, chance, interference of demonic forces, etc., not only in cases where definite causes have not yet been discovered but even in cases where the phenomenon in question is explainable by forces and laws already known to science. To seek an objective purpose in phenomena in order to explain them is inadmissible. To say, for example, that “every historically formed nation lives in order to manifest and fully develop the capacities and traits bestowed upon it by Nature (!), in order to create a special culture, to contribute its share to human civilization” amounts to imposing upon science heartfelt and even noble intentions, letting loose one’s fantasy, forgetting that the development of science (the preachings and the daydreams of idealists notwithstanding) is predicated on the question “why?” (and not “for what purpose?”) and on the answer “because” (and not “in order to”). Scholars of this nonscientific stamp explain the general character of the manifestations of a given people, which is determined by its disposition and external conditions (or what we usually call “culture” and “civilization”) as something bestowed upon them from above. These preachers of supernatural forces readily speak of the “spirit of a people,” the “spirit of language,” the “spirit of an age” (for example, in explaining particular phenomena by the spirit of the age), and the like, forgetting Goethe’s cogent remarks:

Was ihr den Geist der Zeiten heisst,

Das ist im Grund der Herren eigner Geist.

A positive-thinking man first asks: aut . . . aut, i.e., whether purposefulness, destiny, free will, chance,* dogma, etc., and similar beautiful ideas can serve as instruments of explanation or not. If we begin to explain the most trivial phenomenon by purposefulness, destiny, free will, and the like, we must henceforth always admit similar explanations. To treat reality as a heap of incoherent and disconnected phenomena is to dismiss all causality, all science. I repeat: a methodical mind cannot at one and the same time admit the existence of phenomena without cause and not reject science. Science does not allow for compromises: it demands cool, unprejudiced, and abstract thinking.

24. We must distinguish between the categories of linguistics and the categories of language: the former are pure abstractions; the latter, however, are the living part of language: sound, syllable, root, base (theme), ending, word, sentence, different categories of words, and the like. The categories of Ianguage are also categories of linguistics, but they are based on the speakers’ feeling for the language and on the objective and unconscious conditions of the human organism, while the categories of linguistics in the strict sense are predominantly abstractions.

25. Cf. Theodor Benfey, Geschichte der Sprachwissenschaft, pp. 8-9. On the whole, however, all sciences constitute but one science, whose subject is reality. Individual sciences are the result of the attempt to divide the labor; yet this division is based on objective data, that is, on the greater or lesser similarity and affinity of phenomena, facts, and scientific questions.

26. In addition to linguistics proper as the study of language, linguistics includes two kinds of scientific pursuits which have been left aside here: (1) the history of linguistics (the investigation of the development of linguistic concepts and their realization in literature and pedagogy) which forms a part of the general history of the sciences, but which must be practiced by students of language, since they alone have a special interest in it and are sufficiently equipped to deal properly with the history of their science; and (2) linguistic propaedeutics, methodology, the theory of scientific techniques, whose task it is to work out the best methods of studying and furthering a science in all respects (to work out rules for study, research, and presentation).

27. Here it is apropos to mention a question of linguistic methodology which may contribute to the more accurate definition of some problems that arise in the various branches of pure linguistics. Such is the question of gathering material and the preliminary operations performed on it.

The material utilized in the first, positive section of pure linguistics falls into three categories:

1) The directly given material, or the multifarious variety of living languages accessible to the investigator. This also includes the national language in all its variety, the spoken language of all classes of society, be they rich or poor, the language of the peasantry as well as of the educated class . . . This material should include the language of all layers without exception: the speech of street urchins, tradesmen, hunters, workmen, fishermen, etc.; the language of different age groups (children, adults, old people) and of people under special conditions (for example, the language of pregnant women), the Ianguage of personalities, individual language, the language of families, etc. Of further interest are place names, proper names, and traces of foreign influence (something akin to linguistic fossils).

2) Records of language (in chronological order), literature not in the sense of esthetic or cultural products, but as a record of language. The contemporary literature of modern languages is only a document; it is not the language itself. For the reading of ancient monuments, paleography is a necessary linguistic tool. But documents are never a sufficient record of a given language, and data derived from their investigation must be supplemented by the study of the structure and composition of the contemporary language, if one exists, and by deductive reasoning and comparison with other languages, if it is extinct. Usually the documents of a less civilized people transmit information about their language more faithfully than do those of peoples which have created artificial literary languages and writing systems. In the same way, present-day material that is of interest to linguistics includes, for example, letters of uneducated and semiliterate people. Linguistic records include not only entire works, but also single words and phrases occurring in a foreign language (cf. my O drevnepol’skom jazyke do XIV stoletija, sections 1-3). While the study of a living language can be justly compared with zoology and botany, it would be inaccurate to compare the analysis of documents with paleontology, inasmuch as language is not an organism and words are not parts of an organism; consequently, words cannot leave visible traces, real imprints (fossils) of their former existence, as do organisms or parts of organisms of animals and plants. Written records present only arbitrary, visual signs (graphs) of the audible sounds of language and allow conclusions about language only by analogy. The linguist cannot even see the structure of living languages (although he may hear their sounds), and he can judge the language of written records only by comparison and other scientific means. At the same time, the naturalist can sometimes re-create the structure of extinct organisms on the basis of fossils.

The preliminary operations based on the material of living languages and documents consist of the presentation of all the richness of languages in a published form, through the preparation of descriptive grammars and dictionaries.

3) Indirect material for inferences and conclusions about language is provided by: (a) children’s language (which throws light on the formation of sounds, their alternation, the feeling for the root, the tendency to differentiate, etc.; (b) natural flaws in the pronunciation of individuals; (c) the pronunciation of deaf mutes; (d) the pronunciation of foreigners and their general treatment of a foreign language (this throws light on the difference between languages and on the nature of the languages compared).

The second part of pure linguistics, which deals with the beginning of the human word, with the primordial formation of sounds, etc., provides us no direct, only indirect, material, from which we can draw analogical inferences and conclusions:

1) The linguistic development of an individual illuminates the primordial formation of language, since we are taught by the natural sciences that the individual recapitulates, on a smaller scale, the changes of the species. Of primary interest here is the observation of the infant, of the young child, beginning to babble (revealing from the earliest age the rudiments of his future language). Observations thus made can be projected mutatis mutandis into the epoch of primordial speech. Nevertheless, analogical statements in this area should be made with great caution, since our baby differs from the primordial man who is about to begin or has just begun to speak: (1) zoologically: (a) in a collective sense, in that it represents another degree of development of the human species, a different brain structure and nervous system; (b) in an individual sense, in that it represents another degree, another stage of individual development; (2) the baby finds himself among people speaking a ready-made language, which has been formed over many generations, each of which added something to the language of its predecessors; the baby faces from the start ready-made cultural relationships, while primordial man lived in a close relation with nature and submitted to its influence passively.

2) The comparison of various degrees of cultural and intellectual development of various peoples leads us to the conclusion that the contemporary state of humanity simultaneously presents different stages of its development ranging from half-wild savages to the highly advanced Caucasian tribe (race) (cf. the simultaneous existence in a society of children, adolescents, adults, elders, etc.). To gain an approximate picture of the primordial state of language, it is most instructive to investigate the languages of savages. When direct observation is impossible, one may draw this information from works of other scholars and from dependable descriptions by travelers.

3) The study of the general trends and direction in the development of languages enables one to retrace this development backward and to arrive at more or less tenable conclusions concerning the formation of primitive Ianguages, even though these appear to us in advanced stage of development. It is clear that some modern languages retain, if only in a rudimentary state, much that constituted the essence of primitive languages.

The indirect observations concerning the original formation of languages should be corroborated by anatomical and physiological studies of man’s nervous system, which should, perhaps, serve as a point of departure.

28. This comparison must not be taken literally for, as we shall see below, language is not an organism, whereas anatomy and physiology, just as the morphology of organism, deal with actual organisms. The correctness of the comparison is based on the identity and similarity of the mental processes that occur in both areas.

29. The investigation of the sounds of language from a physical point of view must take into account the findings of physiology and acoustics. Some investigators of language show no interest in acoustics and physiology, depending in this matter on their own resources. I think that the scientific study of a subject can profit from various kinds of investigation and must not ignore the results of related sciences. Otherwise one is always performing the work of a Sisyphus < . . .>

30. The imperceptible transition from one state to another, the imperceptible effect of the slowly but thoroughly operating forces in language (as of other phenomena in life) can be expressed by the algebraic formula O X  = m, which means that when an infinitely small change occurring at a given moment is repeated an infinite numbers of times, the final result is a noticeable, definite change. In the passage of time, extension of space, the action of a drop of water on a stone, etc., there is always an elusive, critical point when something vanishes without even leaving a tangible trace.

31. The first attempt to bring together and to synthesize this kind of investigation and to establish the separate layers in the formation of IndoEuropean is Georg Curtius’ treatise, “Zur Chronologie der indo-germanischen Sprachforschung,” Abhandlungen der philologisch-historischen Classe der Königl. Sächsischen Gesellschaft der Wissenschaft, 111, Leipzig, 1867. Curtius distinguishes seven main periods in the formation (Organisation) of the Indo-European languages: (1) the period of roots ( Wurzelperiode); (2) the period of root determinants (. Determinativperiode); (3) the period of primary verbs ( primäre Verbalperiode); (4) the period of stem (base) formation ( Periode der Themenbildung)՛, (5) the period of complex verbal forms ( Periode der zusammengesetzten Verbalformen)՛, (6) the period of the formation of cases (Periode der Casusbildung)՛, and (7) the period of adverbs (Adverbialperiode). One of the main conclusions of his work is that language employed the same means at different times in completely different ways ( dass die Sprache dieselben Mittel zu verschiedenen Zeiten in ganz verschiedener Weise verwendete [p. 193]). The different changes of identical sounds under identical conditions can likewise be explained only by assuming a different chronology of these changes.

32. The error of many scholars is that in a genetic classification they carelessly compare languages of different stages of development, such as Sanskrit and Slavic, Sanskrit and English (the first is a very ancient, and the second a highly advanced Indo-European language).

33. As for the two main approaches of the inductive sciences (see above), it should be observed that the second approach chiefly generalizes and explains phenomena in terms of their genetic and synchronic relationship, whereas the first approach seeks to establish general laws and forces.

34. The external history of a language is closely connected with the fate of its speakers, the fate of its people. Its study comprises the geographic and ethnographic diffusion of a given language and the influence upon it of foreign languages (and vice versa). It deals, further, with the literary or spoken use of a given language, with the social status of its speakers, with its expansion in space (e.g., the spread of French, German, English, and the so-called universal languages) or in time (e.g., Latin, Greek, Church Slavonic), with its use by and function for other peoples, and other such questions. The internal history of a language, on the other hand, deals with its natural development apart from the fate of its speakers but of course not apart from the physical and psychological make-up of its speakers. The internal history can ignore the fate of the language, focusing only on the changes which occur within the Ianguage. Internal history investigates how a people speaks at a certain time or over the course of many centuries, and why it speaks in this way, while external history asks how many people speak the given language and where. The first approach concentrates more or less on the question of quality, the second on the question of quantity < . . .>The external and internal histories of a language (which are the object of a science, not the sciences) are related. The effect of the former on th latter seems to be stronger than the converse. The influence of foreign languages, literary treatment, native command, geographic conditions of the country determine the more or less rapid development of a given language and its specific character. The internal history of a language, the degree of its flexibility (though by itself an insignificant factor) determine the faster or slower development of its literature and are decisive factors in determining the question of the change of language, that is, when a given Ianguage is to be viewed, because of its deep changes, as a new language and its dialects as new and autonomous entities. The subject matter of the external history of a language is largely the same as that of history and history of literature. The historian often touches on the external history of the language when he discusses the expansion of a people, its education, and the flowering of its literature.

35. One of the important tasks of scholarship is to re-create the so-called original and basic languages ( Ursprachen and Grundsprachen ), i.e., the Ianguages represented in modified form by actually existing languages. But one must keep in mind that the reconstruction of the original and basic languages represents not real phenomena but deductively obtained scientific facts.

36. The external test of the genetic relationship of languages is the possibility of phonetic translation from one language into another, that is, the possibility of rendering each word (except borrowings) of one language in the form of another language in accordance with specific sound laws and sound correspondences.

37. From this point of view, languages are most generally divided into monosyllabic, agglutinative, and flexional.

38. For an example, let us take the two extremes, the monosyllabic and the inflected languages. Inasmuch as language determines the national frame of mind, people speaking monosyllabic languages have never had (as a people or nation) and never will have a need to express conceptual relationships by means of material forms, sounds, and sound groups; we may even assume that speakers of such languages lack the conception of formal relationships altogether, which is not the case with speakers of flexional languages. The effort of abstraction which has led to the distinction of noun, verb, and other parts of speech is in the monosyllabic languages rendered through a strictly defined word order (e.g., the predicate follows the subject, the attribute precedes the subject, etc.). In the flexional languages, this task is achieved by means of purely formal elements.

39. This, of course, does not prevent a people from substituting a language of one structure with a language of another structure, and from breaking away from a given ethnic group or national language; it 1s sufficient to point to the Negroes in America who have adopted French or English; but when such a change takes place, there is probably a change also in the frame of mind of the people. On the other hand, the morphological structure of one language may combine with the material of another language (of course, only through borrowing); an example of this is the combination of Chinese structure with Russian material discussed in an article of S. I. čerepanov, “Kjaxtinskoe kitajskoe narečie russkogo jazyka,” in Izvestija imp. Akademii nauk po otdeleniju russkogo jazyka i slovesnosti, II, pp. 370-77.

40. However, this may be due to our inadequate and narrow morphological classification. Given a more precise and less narrow classification of the structure of various languages, the distinctions may turn out to be more numerous than has generally been assumed until now.

41. These groups are: (1) Indo-European, (2) Semitic, (3) Hamide, (4) FinnoTartar or Uralo-Altaic, (5) Dravidian, (6) Polynesian or Oceanic (Malayan), (7) South African.

42. The original basic difference between the Indo-European and Semitic languages is matched by an original basic difference in the religious beliefs of their speakers or nations. The seeds of their religious differences must have been planted at a time when man was beginning to be man, beginning to speak: the form in which the deity enters into contact with man is personified for the Indo-Europeans and is endowed with prophetic powers for the Semites

43. The close kinship between the Balto-Slavs and the Germans, posited by Grimm and Schleicher and now accepted by most scholars, was, surprisingly enough, expressed as early as the 13th century by a Polish chronicler, Boguchwał: “sic et Theutonici cum Slavic regna contigua haventes simul conversation e incedunt, пес aliqua gens in mundo est, sibi tam communis et familiaris, veluti Slavi et Theutonici” (cf. my O drevnepol’skom jazyke do XIV stoletija, dictionary entry slavjanin). The close kinship of the Slavs and the Lithuanians was recognized by the Poles of the 17th century. Here is what the well-known Polish writer Pasek writes on this subject: “Taka właśnie roż nica mowy Jutlandczyków od niemieckiej, jak Łotwy albo Żmudzi od Polaków” [There is the same difference in speech between the Jutlanders and the Germans, as between the Latvians or Lithuanians and the Poles.] (Pamiętniki Paska [Petersburg, i860], p. 19).

44. Cf. the attempts to explain the biblical tale of the Tower of Babel. “In dem letzten Worte sehen wir sogar, wie die Sage von dem Bischof Hatto durch die Volksetymologie veranlasst wird, ähnlich wie man im Pentateuch mehrere dergleichen etymologischen Sagen, die Mythe vom babylonischen Thurmbau (die bekanntlich nur auf der falschen Anknüpfung des Namens Babel an hebr. balal beruht) an der Spitze, längst erkannt hat” (Förstemann, “Über deutsche Volksetymologie” in Zeitschrift für Vergleichende Sprachforschung. . . . , ed. by Aufrecht and Kuhn, I, 6).

45. Here one must remember the dictum omnis definitio periculosa and forgo a real definition that would implicitly include all properties of language in one condensed formula, since these properties can be known only through a study of details; it is sufficient to give a nominal definition that indicates only the subject of study without defining a priori all its properties and peculiarities.

46. We can observe an organism with our eyes, but language only with our hearing; what appears to our eyes in books is not language, but its graphic representation (e.g., an alphabet). An organism appears to us globally; it has continuous existence from the time of its birth until death. Language as a whole exists only in potentia. Words are not bodies and not elements of a body: they appear as complexes of significant sounds and sound groups only when man speaks, and they exist as representations of significant sound groups only in the brain, only when man thinks in them.

47. The word presents primarily two sides to observation: the sound form and the function which, as body and spirit in nature, never appear separately; even in actuality it is impossible to separate them without destroying the other side (cf. Dr. Johannes Schmidt, Die Wurzel AK im Indogermanischen . . . , [Wiemar, 1865], p. 2). Form and content, sound and thought, are so inseparably linked that neither can be changed without producing a corresponding change in the other (ibid., p. 1). In this view of the nature of language there is obviously something missing that might link sound and meaning; namely, the conception of sound as an interpretation of the external side of the word. This shortcoming is the result of treating language apart from man. For where is the sound-form of thinking and writing, the processes which require the socalled function of words? These processes are, in fact, carried out by combining concepts of the object (meaning) with representations of sound (in writing, with representations of visual signs), and not with the audible production of sounds. A case in point is the deaf mute whose “speech״ produces sound-waves and corresponding sound impressions that are perceived only by the hearer; the deaf mute associates the so-called function, not with sounds and soundrepresentations, but with certain movements of the organs and their representations; the effect of these movements on the air and on the ear of the hearer is to him incomprehensible. Some people are also known to have learned English (whose sounds are represented by a very difficult orthography) by sight without the help of a teacher. For such people the so-called function of English words is associated not with the sound form of the words, but with their graphic signs (compare the substitution of visual musical notes by tactile ones in teaching the deaf to play music). And could it be claimed that meaning (function) is, on the other hand, inseparable from sound for people endowed with a good ear who listen to a foreign language which they do not know? In all these cases, there is supposed to be some mystical association of the sound form with the so-called function without the participation of the individuals (the speakers who do not hear, the people who only read, or who hear but do not understand). The views referred to above are based on a narrow, falsely interpreted monism which applied consistently would vitiate the notions of birth, life, and death of the organism, and of the organism itself. For a dead organism retains its external form more or less (appearance and bodily structure), but loses its essential functions which give way to functions characteristic of a different organism.

48. Perhaps I shall soon have the possibility of treating more closely and critically the misleading view that language is an organism and similar views, such as that linguistics is a natural science (like botany and zoology), that it is completely different from philology, that language and history are opposed to and incompatible with each other, etc. I shall also try to point out some other errors and false notions about language and linguistics, which are in part unconsciously shared by the public and in part consciously fostered by scholars.

49. The language of an individual can be examined in terms of quality (way of pronunciation, choice of words, forms and phrases peculiar to a given individual, etc.) and in terms of quantity (the store of words and expressions used by the given individual). As for the latter, cf. Max Müller, Vorlesungen über die Wissenschaft der Sprache . . . Für das deutsche Publikum bearbeitet von Dr. Carl Böttger (2d ed.; Leipzig, 1866), pp. 227-28. One must also pay attention to the distinction between ceremonial and daily language, family and social language, and generally to linguistic diversity under various circumstances of life, to the distinction between the common language and the language of specialists, to varieties of language according to mood: the language of feeling, imagination, cerebration, etc.

50. From this point of view, language (or dialect, even individual speech) does not form a whole, but is a generic concept, a category into which one may fit a complex of actual phenomena. Cf. also the concept of science as an ideal, as the complex of all scientific data and inferences, as opposed to science as a repeated scientific process.

51. Language is one of the functions of the human organism in the broadest sense of the word.

*The expressions “chance,” “chance similarity,” etc., as used by scientists signify either that the cause of the phenomenon is not known or that there is a similarity between various phenomena which have no genetic or natural connection.