higher, are grandfathers and grandmothers; below, are sons and daughters and the grandsons and granddaughters. The same, of course, holds also for women. Thus, the system as a whole comprises five generations.
Now, it was maintained that this system could have arisen only out of a previous condition of general promiscuity. For, unless the actual father were universallyunknown, how could it be possible that a person would call by the name of father every man within the same kinship group who might, as regards age, be his father? If, however, we propose this argument, we immediately strike a weak point in the hypothesis, since all women of the older generation are called mother just as its men are called father. We should certainly expect that the real mother would be known, because the child derives its nourishment from her during a period which is especially long among primitive peoples, and because it grows up close to her. And, furthermore, the hypothesis is hardly reconcilable with the fact that, for the most part, Malayo-Polynesian languages differentiate relations by marriage even more sharply than do our own. An Hawaiian man, for example, calls the brother of his wife by a different name than does a woman the brother of her husband. Thus, in place of our word 'brother-in-law' they have two expressions. In any event, the term 'brother-in-law' is applied to an individual, and therefore implies marriage. To meet this point, we would be obliged to fall back on the supposition that these terms represent later additions to the original nomenclature of relationship. But even then the fact would remain that, in their direct reference, these terms are merely names for differences in age. It therefore remains an open question whether the terms also designate relationship; to the extent of our observation, this is certainly not the case. The native of Hawaii, so far as we know anything about him, knew his father and mother: what he lacked was merely a specific name for them. Whenever he did not call his father by his given name, he evidently called him by the same name that he applied to the older men of his immediate group. Among European peoples also, the terms 'father' and 'mother' are sometimes used in connection with men and women outside this relationship. For example, the Russians, particularly, have a custom of addressing as 'little father' and 'little mother' persons who are not in the least related to them. That which makes it highly probable that in the so-called Malayan system of relationship we are dealing notwith degrees of relationship but with age-periods, is, in the last event, a different phenomenon—one that has hitherto been overlooked in connection with these discussions. In the very regions whose languages employ this nomenclature, custom prescribes that the youths and men live in separation from the women and children from their earliest years on. This is the institution of the men's club with its age-groups. Its social rôle is an important one, crowding even the family association into the background. Under such circumstances, the individual is naturally interested first of all in his companions of the same age-group, for each of these usually occupies a separate apartment in the men's house. Thus, the so-called Malayan system of relationship is really not a system of relationship at all, but a nomenclature of age-groups based on social conditions. These conditions bring it about that companions of the same sex are more closely associated than are men and women. In the men's houses a companion of the same group is a brother, one of the next older group, a father. Together with these men the individual goes to war and to the hunt. Thus, these phenomena cannot be said to belong to the lowest stage of culture. Nor, obviously, does this terminology, which has reference to differences of age, exclude any particular form of marriage. In this case it is a mistake to associate the names 'father,' 'mother,' 'brother,' etc., with the concepts that we attach to these words.
The hypothesis that the family, whether of monogamous or of polygamous organization, was preceded by a state of unrestricted sexual intercourse, so-called agamy or promiscuity, is, however, as was remarked above, based not only on the fact of maternal descent and of the Malayo-Polynesian method of designating ages, but also on that of group-marriage. In this form of marriage, a number of men marry in common a number of women. This is interpreted as a transitional stage between an unrestricted sexual intercourse within the tribe and the limited marriage unions of later times. At first glance, indeed, this might appearprobable. In order, however, to decide whether such a transition could take place, and how it might occur, we must first of all consider the relation which group-marriage sustains, among the peoples who practise it, to the other forms of marriage. It then appears at once that it is a particular form of polygamy. True, it is not identical with the form of polygamy most familiar to us, in which one man possesses several wives. But there is also a second form, which, though less frequent, is of greatest importance for an interpretation of group-marriage. One woman may have several husbands. The two forms of polygamy may conveniently be calledpolygynyandpolyandry, and these terms should always be distinguished in any attempt at a precise account of polygamous marriage. Polygyny is very prevalent even in our day, occurring particularly in the Mohammedan world, but also among the heathen peoples of Africa, and in other regions as well. It was likewise practised by the ancient Israelites, and also by the Greeks, although the Indo-Germanic tribes for the most part adhered to monogamy from early times on. Polyandry is much less common, and is, indeed, to be found only among relatively primitive peoples. It occurs in Australia and, in the southern part of Hindustan, among the Dravidians, a tribe of people crowded back to the extreme end of the continent by peoples who migrated into India; it is found also far in the north among the Esquimos of Behring Strait and among the Tchuktchis and Ghilyaks of Siberia, and, finally, here and there in the South Sea Islands.
If, now, we wish to understand the relation of these two forms of polygamy to each other, we must first of all attempt to picture to ourselves the motives that underlie them, or, wherever the custom has become fixed through age, to bring to light the motives that were originally operative. In the case of polygamy, the immediate motive is evidently the sexual impulse of man which is more completely satisfied by the possession of several wives than by that of a single one. This motive, however, does not stand alone; as a rule other contributing circumstancesare present. Two such important factors, in particular, are property rights and the power of authority. Polygyny flourishes particularly wherever the general conceptions of property and of authority, and, connected with the latter, that of the supremacy of man within the family, have attained undue importance. Under the co-operation of these motives, the wife becomes the absolute property of the husband, and may, therefore, wherever polygyny prevails among barbaric peoples, be given away or exchanged. Bound up with this, moreover, is the fact that, wherever there are considerable social differences, dependent on differences in property and rank, it is principally the wealthy or the aristocratic man who possesses many wives. In the realm of Islam, the common man is, as a rule, content with asinglewife, so that monogamy here prevails in the lowest stratum of society.
With polyandry the case is essentially otherwise. In it, entirely different motives are operative; it might, indeed, be said that they are the exact opposite of those that bring about polygyny. It is particularly significant that polyandry is found in regions where there is a scarcity of women. This scarcity, however, is, in turn, generally due to an evil custom of barbaric culture, namely, infanticide. In Polynesia, where polyandry was very prevalent, this custom was at one time fairly rampant. Even to-day infanticide still appears to be practised by some of the Dravidian tribes of Hindustan. Similar conditions prevail among the Australians. In Polynesia, however, and probably in other localities as well, it was chiefly the female children who were the victims of infanticide. The natural result was a decrease in women and a striking numerical disproportion between the sexes. Thus, Ellis, one of the older English investigators of conditions in these territories, estimated the relation of men to women as about six to one. Under such circumstances the custom of polyandry is intelligible without further explanation. It was not possible for every one to possess a wife of his own, and so several men united to win one wife in common.
We might ask why it was chiefly girls who fell victims to this murder. That children in general should be sacrificed, under the rough conditions of nature, is not inexplicable. It is due to the struggle for the necessities of life and to the indolence that shrinks from the labour of raising children. The desire is to preserve the lives of only a limited number; the remainder are killed immediately after birth. In Polynesia, the murder was forbidden if the child had lived but a single hour. Occasionally, magical motives are operative, as in the case of the horror which the man of nature feels towards deviations from the normal and towards the birth of twins. That male children are more often spared than female, however, can scarcely be explained otherwise than on the ground that a particular value is placed on men. The man is a companion in sport and in the chase, and is regarded as more valuable for the further reason that he aids in tribal warfare. This higher value reverts back even to the child. It is evidenced also in the fact that, in the case of women, the arrival of adolescence is not celebrated with the same solemn ceremonies as are held in the case of young men. Whereas great celebrations are held when the youth reaches the age of manhood, little notice is taken, as a rule, of the maiden's entrance into womanhood. By means of these celebrations, the youths are received into the society of men, and, together with companions of their own age, are initiated into the traditional ceremonies. In these ceremonies women are not allowed to participate.
Though the causes of polyandry are thus entirely different from those of polygyny, it does not at all follow that these forms of marriage are mutually exclusive. On the contrary, they may very well exist side by side, as, indeed, they actually do in many places. But how, then, is so-called group-marriage related to these two forms? It is obviously nothing but a combination of polyandry and polygyny. In fact, whenever a group of men marries a group of women, these two forms of polygamy are both involved. Every man has several wives, and every wife has several husbands. Only, indeed, on the basis of apurely external and superficial consideration could one look upon polygyny and polyandry as unconjoinable, because they are, in a certain sense, opposing ideas. As a matter of fact, they do not really exclude each other. If we bear in mind the causes mentioned above, it is obvious that under certain conditions of life, such as occur particularly in a more primitive environment, their combination is more probable than their mutual exclusion. If, especially among tribes who have not yet developed sharply defined distinctions based on property and power, as, for example, among the Australians, every man strives to obtain several wives (which is the state of polygyny), while, on the other hand, there actually exists a dearth of women (which means that motives to polyandry are present), the two forms naturally combine with each other. This is frequently verified, moreover, whenever we are able to gain any degree of insight into the particular conditions surrounding the origin of such group-marriages, and also whenever their forms undergo a modification of details. Among Australian tribes, for example, particularly in the southern part of the continent, there is a common form of group-marriage, in which a man possesses either one or several chief wives, together with secondary wives; the latter are the chief wives of other men, whereas his own chief wife is in turn the secondary wife of those men or of others. This custom is very similar to what is probably the most common form of polygyny, namely, the possession by a man of onlyonechief wife in addition to several secondary wives,—a form of marriage that is obviously derived from monogamy. One agency that is particularly apt to bring about such a form of marriage, transitional between monogamy and polygyny, is war. We know from the Iliad that in barbaric times woman was the booty of the conqueror, and became his slave or secondary wife. So also, according to the Biblical legend, Abraham possessed a chief wife, Sarah, who belonged to his own tribe, but also a secondary wife, Hagar, who was an Egyptian slave. Wherever the concept of property became prominent, the purchase of womenproved to be a further source of polygyny. In this case also, there was generallyonechief wife, wherever polyandry did not interfere. When the Mohammedan of modern times calls his chief wife 'favourite,' it is merely another indication that this form of polygyny developed from monogamy, since, according to the old custom, there was butonechief wife. Here, however, the chief wife is no longer necessarily the wife belonging to a man's own tribe, as was the case among the ancient Israelites; the favour of the master determines which wife shall be given the privileged place.
Thus, from whatever angle we view group-marriage, its polygyny and its polyandry seem to rest on monogamy. This is true also of forms of group-marriage other than those mentioned above. Where the theft of women still continues to be a practice more serious than are the somewhat playful survivals that occur in the marriage ceremonies of cultural peoples, the one who wishes to steal a wife not infrequently secures confederates for his undertaking. Custom then commonly gives these companions a certain right to the stolen woman. This right, of course, is for the most part temporary, but it may nevertheless come to approximate the conditions of group-marriage in case the first man assists his confederates in the same way in which they have aided him. There is still another and a related motive that may lead to the same result. When a woman enters into marriage with a man of a certain tribe, she at once enters into very close relations with the tribe itself. Where tribal association has gained a preponderant importance, custom sometimes grants to all the male members of the tribe certain transient rights with respect to the woman on the occasion of her marriage. This occurs particularly when the man and woman belong to different tribes—that is, in the case of exogamy, an institution characteristic of the totemic age and to be considered later. For, the lively consciousness of kinship differences naturally tends to strengthen the right of appropriation belonging to the entire tribe. A similar thought is reflected in the mediævaljus primæ noctisof certain provinces of France and Scotland,except that in place of the right of the kinship group to the possession of the individual we here find the authority of the lord over his vassals.
Thus, all these phenomena, belonging in part to the transitional stage between monogamy and polygamy and in part to a combination of the two forms of polygamy, namely, polygyny and polyandry, point to monogamy as the basal form of marriage, and that form from which, under the influence of particular conditions, all others have developed. Whether or not we regard it as probable that the system of maternal descent was at one time universal, no argument for the existence of an original promiscuity can be based upon it. If we call to mind the close association of the youths and men of the kinship group in the men's house, it will be apparent that such conditions of social intercourse make for a particularly intimate bond between mother and children. Before his entrance into the community of men, the boy lives in the company of the women. This close association between mother and children is sufficient to account for the origin of maternal descent. But, owing to the gradual change of cultural conditions, it is to be expected that maternal descent should pass over into paternal descent as soon as more positive conceptions of authority and property are formed. Moreover, the possibility also remains that among some tribes paternal descent prevailed from the very outset; positive proof is here not available. We cannot, of course, deny the possibility that under certain cultural conditions man exercised the decisive influence from the very beginning, as early, indeed, as one may speak of clan membership and hereditary succession. The most primitive stage of culture, as we shall see in the following discussion, lacks the conditions for either maternal or paternal descent, inasmuch as it possesses neither clearly defined clans nor any personal property worth mention.
Thus, the arguments based on the existing conditions of primitive peoples, and contending that the original condition of mankind was that of a horde in which both marriage and the family were lacking, are untenable. Onthe contrary, the phenomena, both of group-marriage, valued as the most important link in the chain of proof, and of the simpler forms of polygamy, everywhere point tomonogamyas their basis. Furthermore, these arguments all rest on the assumption that the peoples among whom these various phenomena occur, particularly the combination of polygyny and polyandry in group-marriage, occupy a primitive plane of social organization. This presupposition also has proven fallacious, since it has become evident that this organization, especially among the Australian tribes, is an extremely complicated one, and points back to a long history involving many changes of custom.
Meanwhile, primitive man, in so far as we may speak of him in the relative sense already indicated, has really been discovered. But the Australian does not belong to this class, nor, even less, can many of the peoples of Oceania be counted within it. It includes only those tribes which, having probably been isolated for many centuries and cut off from the culture of the rest of the world, have remained on the same primitive level. We have become familiar with them in the preceding account of the external culture of primitive man. We find them to be forest peoples who have, for the most part, been crowded back into inaccessible territory and who have entered but slightly into intercourse with the outside world, inasmuch as their needs are limited. They generally call themselves, whether rightly or wrongly we need not inquire, the original inhabitants of these regions, and they are regarded as such by their neighbours. They include, in addition to several tribes of Hindustan (as yet insufficiently studied), particularly the Semangs and Senoi of the interior of the Malay Peninsula, the Veddahs of Ceylon, the Negritos of the Philippines and Central Africa, and, finally, to some extent, also the Bushmen. This is certainly a considerable number of peoples, some of whom live at great distances from the others. In spite of this, however, even their external culture is largely the same. Considering the primitive character of their social institutions and customs, it would seem safe tosay that without doubt they approach the lowest possible level of human culture. Besides bow and arrow they have scarcely a weapon, no vessels of clay, and practically only such implements as are presented directly by nature herself. At this stage there is scarcely anything to distinguish man from the animal except the early discovered art of kindling fire, with its influence on the utilization of the food that is gathered. Briefly summarized, these are the main traits of primitive culture that are known to us.
What, now, is the status of marriage and the family at this period? The answer to this question will come as a surprise to those who are imbued with the widespread hypotheses that presuppose the primitive state to be that of the horde. And yet, if these hypotheses be regarded in the proper light, our answer might almost be expected. Among the primitive tribes that we have mentioned, monogamy is everywhere found to be not only the exclusive mode of marriage, but that which is always, so to speak, taken for granted; and this monogamy, indeed, takes the form of single marriage. It is but rarely that related families live together more or less permanently, forming the beginning of the joint family. The Bushmen alone offer something of an exception to this rule. Among them, polygyny, together with other practices, has been introduced. This is probably due to the influence of neighbouring African peoples, such as the Hottentots and the Bantus. Elsewhere conditions are different. This is true especially of the Semangs and Senoi, whose isolation has remained more complete, and of the Veddahs of nature, as the Sarasin cousins call them in distinction from the surrounding Veddahs of culture. Among these peoples, monogamy—indeed, lifelong monogamy—has remained the prevailing form of marriage. Connected with it is found the original division of labour, which is based on sex. Man provides the animal food by hunting; woman gathers the vegetable food—fruits, tubers, and seeds—and, by the employment of fire, if necessary, renders both it and the game edible. This basis of division of labour, which appears natural andin harmony with the endowment of the sexes, contrasts with the conditions of later culture in that it indicates an approximate equality of the sexes. Furthermore, Rudolf Martin and the two Sarasins, investigators of the primitive Asiatic tribes of Malacca and Ceylon, commend the marriage of these peoples as being a union of husband and wife strictly guarded by custom. In forming a moral estimate of these conditions, it should not be overlooked that the exclusive possession of the wife is probably due to jealousy as much as it is to mutual faithfulness. Among the Veddahs, the intruder who threatens this possession is struck to earth by a well-aimed arrow shot from behind ambush, and custom approves this act of vengeance as a justifiable measure on the part of the injured man. Therefore, even though a French traveller and investigator may, to a certain extent, have confused cause and effect when he stated that the monogamy of these tribes had its origin in jealousy, the exercise of the right of revenge may, nevertheless, have helped to strengthen the custom. But, of course, in view of the primitive state of culture that here prevails, this custom of revenge is itself merely an indication of the undisputed supremacy of monogamy. Even as the individual, and not the clan, exercises this vengeance, so also does marriage continue to be restricted to single marriage. Of the formation of joint families, which arise out of the union of immediate blood relations, we find at most, as has been remarked, only the beginnings.
The more extensive social groups generally result from the fact that during the rainy season families withdraw into caves among the hills. The larger caves are frequently occupied in common by a number of families, particularly by such as are most closely related. Yet the groups of co-dwellers are not so much determined by considerations of kinship as by the size of the places of refuge; a single family occasionally occupies a small cave by itself.Nevertheless, this community life plainly furnishes the incentive to a gradual formation of wider social groups. This, no doubt, accounts for the fact that during the favourable season of the year several families of the Veddahs claim for themselves a specific plot of ground, whose supply of game, as well as of the products of the soil, which the women gather, belongs exclusively to them. Thus, there is a division of the people into districts, and these are determined geographically rather than ethnologically. Every one is entitled to obtain his food, whether game or products of the soil, from a specified territory. Custom strictly guards this communal property, just as it protects the single marriage. The Veddah, for example, who encroaches upon the territory belonging to a group other than his own, is in no less danger of falling a victim to an arrow shot from an ambush than is the one who trespasses on marriage ties.
These various institutions form the beginnings of social organization, but as yet they do not represent developed clan groups or established joint families of the patriarchal type. On the contrary, as they arise through the free association of individuals, so also may they be freely dissolved. Each man has exclusive possession of his wife. Without interference on the part of his clan, moreover, he exercises absolute control over his children, who remain with the individual family just as in the case of a developed monogamy. There is no trace of sex-groups, such as are later to be found in the case of the men's houses and the age-groups. Only temporarily, on the occasion of common undertakings, such as the hunting of large animals, which requires a considerable measure of strength, or when new hunting-grounds are being sought, is a leader appointed from among the older men. His leadership, however, ceases with the completion of the undertaking. There are no permanent chiefs, any more than there are clans or tribal organizations.
Thus, in summary, we might say: Whenever the social organization of primitive man has remained uninfluencedby peoples of a higher culture, it consists in a firmly established monogamy of the form ofsingle marriage—a mode of existence that was probably carried over from a prehuman stage resembling that of the present-day anthropoids. There are also scanty beginnings of social groups. If we consider these tribes as a whole, they still continue to lead the life of ahorde, meaning by this an unorganized, in contrast to an organized, tribe of people. Indeed, it was through a curious change of meaning that this word acquired its present significance. It is supposed to have originated in a Mongolian idiom, whence it found its way first into the Russian and later into other European languages. The Tartars called a division of warriors ahorda. First used in this sense, the word apparently did not receive its present meaning in Germany until the beginning of the eighteenth century. Having in mind the "Golden Horde" of the Tartars, a horde was understood to mean a particularly dreaded division of warriors. The furious force of these Asiatic hordes, and the terror which they spread, later caused the concept to be extended to all unorganized, wild, and unrestrained masses of men. Taking the word in this wider significance, we may now say that the horde, as a fairly large social group in which only very meagre suggestions of an organized tribal system occur, is characteristic of primitive times, no less than are the isolated single family and the beginnings of the joint family. Thus defined, however, the horde does not differ essentially from the animalherd, in the meaning which the latter concept would possess when applied to human-kind. And it is not impossible that in the extension of the meaning of the term 'horde,' this association of the foreign word with the original Germanic word 'herd' played a part. A horde, we might say, is a human herd, but it is precisely ahumanherd. Between the members of a horde, therefore, there exists a relation that is lacking in the animal herd, in flocks of migratory birds, for example, or in herds of sheep and cattle. This relation is established and preserved through a community of language. Herder,therefore, truthfully remarks that man was from the beginnings a 'herding animal,' in so far as he possessed social instincts. Even in the formation of language these social instincts were operative. Without a community life, and, we may add, without the mental interaction of individuals, language would be impossible. Language, however, in turn, strengthened this community life, and elevated it above the status of the animal herd or of an association concerned merely with momentary needs.
Thus, these reflections concerning the social relations of primitive man lead us to a further field of phenomena which likewise affords a glimpse into the mental characteristics of primitive peoples. For that which differentiates the horde from the herd is thelanguageof primitive man, together with the activity most closely bound up with language, namely,thinking.
Our knowledge of those peoples whom we, avoiding the errors of the past, may now regard as primitive, led to the conviction that the Asiatic and African tribes described above were actually primitive, in the above-mentioned relative sense of the word. Naturally the question concerning the language of these peoples then began to arouse considerable attention, on the part, not only of ethnologists, but also of those interested in philology. The question is of equal importance, to say the least, for the psychologist. For language is bound up with thought. From the phenomena of language, therefore, we may draw inferences concerning the most general characteristics of thought. Such fundamental differences of language as exist, for example, between the Chinese and the Indo-Germanic tongues do not, of course, allow the direct conclusion that there are quantitative differences in mental culture. They do imply, however, that there are divergent directions and forms of thought. In their ceaseless change, the latter react upon language, and this, in turn, again influences mental characteristics. We cannotsuppose that, in the period of Old High German, much less in that of the original German, our ancestors employed the same forms of thought with which we of to-day are familiar. To a lesser degree, similar changes have undoubtedly transpired within much shorter spaces of time.
These considerations make the question concerning the language of primitive man of the utmost psychological importance. Linguistic investigations, however, so far as they, in their early attempts, had been able to survey the field, had brought to light a fact which discouraged all efforts to discover an original language. Indeed, it was inevitable that at first glance this discovered fact should have appeared exceedingly strange, particularly when viewed in connection with the life of primitive man. It appeared that, for the most part, the original languages of primitive tribes no longer exist. It is true that in the vocabularies of the Semangs and Senoi of Malacca, of the Veddahs of Ceylon, of the Negritos of the Philippines, and in other vocabularies that have been collected, single words may be found which do not occur in the languages of the neighbouring tribes; and it is noteworthy that the bow and arrow are the objects most frequently designated by such words, a proof of the fact that these are really relatively primitive inventions. On the whole, however, the Veddahs speak the language of the Singhalese and Tamils; the Semangs and Senoi, as well as the Negritos of the Philippines, that of their neighbours, the Malays; similarly, among the African tribes, the Pygmies of Central Africa have apparently appropriated the language of the Monbuttus and other negro races, and the Bushmen that of the Hottentots.
How may this remarkable fact be explained? That these tribes formerly possessed languages of their own can scarcely be doubted. For, as respects physical characteristics, they are absolutely distinct races. Considering their characteristics as a whole, moreover, it is utterly impossible that they could have lacked language before coming into contact with the peoples who entered the country at a later period. How, then, did these people comeapparently to lose their original language? To this we may briefly reply that there here transpired what always occurs when the well-known principle of the struggle for existence is applied to the field of mental phenomena. The stronger race crowded out the most important mental creation of the weaker, its language. The language of the weaker race, which was probably very meagre, succumbed to a language that was more highly developed. At first glance, this explanation would appear to contradict what we know concerning the life of these primitive tribes. With what anxiety they isolate themselves from their neighbours! A striking proof of this is offered by the practice of secret barter, in which primitive man sets out from the forest, if possible by night, and deposits his captured game at a place which custom has set apart for this purpose, returning the next night to take whatever the more civilized neighbouring tribes have left in exchange—iron implements and weapons, material for clothing, and especially articles of adornment. The participants in this barter do not see each other, much less speak with each other. But where such seclusion exists, how is it possible for a strange language to penetrate? This problem appears almost insoluble. Nevertheless, a solution that appears at least probable was suggested by the investigations of Kern, an able Dutch scholar. His studies were based mainly on the development of the various Malayan idioms. A remarkable exception to the rule that primitive tribes have adopted the language of their more civilized neighbours came to light in the case of the Negritos of the Philippines. Their neighbours, as well as those of the tribes of the interior of Malacca, belong to that much-migrating race, the Malayans. If we compare the Negrito word-formations that have been collected during the past forty years with the vocabulary of the neighbouring Malayans, it is evident that all the words are entirely different, or at least seem to be so with few exceptions. When, however, Kern traced the probable development of these words, and compared them, not with the present-day usage of the Malays but with older stagesof their language, he found that the latter invariably contained the counterparts of the Negrito words. Thus, while these Negritos have remained untouched by the present-day Malays, who probably entered the country at least several centuries ago, they have evidently derived their language from a Malayan influx that occurred much earlier still. To this may be added the demonstrable fact, gleaned from another source, that from very early times the Malayan tribes undertook migrations at widely separated intervals. Traversing the seas in their unsteady boats, they at various times peopled such islands, in particular, as were not too remote from the mainland. Now the testimony of language, to which we have referred, demonstrates that there were at least two such migrations to the Philippines, and that they occurred at widely different times. The original Malayan dialect, which has now become extinct or unknown to the modern Malays, was assimilated by the Negrito peoples, who probably occupied this territory before the arrival of any of the Malays. But this leads to a further inference. If the language was appropriated in prehistoric times and if the conditions of the present are such as would make this scarcely possible, we must conclude that the interrelations of the immigrants and the original inhabitants were formerly not the same as those that now prevail. And, as a matter of fact, this seems altogether probable, if we call to mind the descriptions which modern travellers give of their experiences among these primitive peoples. The traits of character that particularly distinguish them are fear and hatred of their more civilized neighbours; corresponding to this, is the contempt felt by the latter, because of their higher culture, for the more primitive peoples. The only thing that restrains the immigrant people from waging a war of extermination against the original inhabitants is the fear of the poisoned arrow which the Negrito directs against his enemy from behind an ambush. In view of these facts it is not difficult to understand the almost universal isolation of primitive man at the present time. On the other hand,travellers who have been admitted into the lives of the primitive tribes of Malacca and Ceylon and have sought to gain their friendship, unanimously assure us that, whenever a person has once succeeded in coming close to these people and in overcoming their distrust, he finds their outstanding characteristics to be good nature and readiness to render assistance. We may, therefore, be justified in assuming that the seclusion of primitive man was not an original condition, but that it grew up, here and elsewhere, as a result of the war of extermination to which he was exposed on the part of the races attempting to crowd him out of a large part of his territory. Before this state of affairs arose, barter also could scarcely have possessed that character of secrecy which only fear and hatred could give it. In all probability the intercourse which necessarily took place in early times between the older inhabitants and the newer peoples, led to a competition of languages in which the poorer and less developed language of primitive man inevitably succumbed. Nevertheless, the primitive language may also have quietly exercised a reciprocal influence upon the more advanced language. An observation that we cannot escape, even on far higher stages of linguistic development, is the fact that, in such a struggle between a superior minority and a less civilized majority, the former determines the main stock of words, and even, under favourable conditions, the grammatical form, whereas the latter exercises a decisive influence on pronunciation. That a similar process occurred in connection with the displacement of primitive languages, the language of the Bushmen offers proof. This is essentially a Hottentot dialect, even though it is characterized by certain traits of primitive thought. The Hottentots, however, have derived their well-known clacking sounds from the Bushmen, who also gave these sounds to the languages of the Bantu peoples.
But are we deprived of all knowledge concerning the most primitive grammatical forms and concerning the related question of the origin of language, by virtue of the fact that the languages of primitive peoples have, withthe exception of meagre remnants, apparently been lost? There is a consideration touching the question of primitive forms of thought and language that enables us, in spite of the difficulty suggested, to answer this question in thenegative. The development of language does not at all keep pace with that of the other forms of culture. Primitive forms of thought especially, and the corresponding expression which they receive in language, may long persist after external culture is relatively advanced. And thus, among tribes that are in general far beyond the primitive stage, linguistic forms may still be found which are exact counterparts of phenomena that, from a psychological point of view, must be regarded as primitive. As regards this point, it is especially the African languages of the Soudan that offer a typical field for linguistic study. If we analyse the syntax of such a language and the forms of thought which the sentence structure allows us to infer, we gain the impression that it is hardly possible to imagine a form of human thought whose essential characteristics could be more primitive. This is clearly apparent from a consideration of the Ewe language of the peoples of Togo, a German colonial possession. This is a Soudan language, on whose grammar D. Westermann, a German missionary, has given us a valuable treatise. While the Ewe language does not contain all the essential features apparently characteristic of relatively primitive thought, it does exemplify some of them. We are led to this conclusion particularly when we compare it, together with other Soudan languages, with a form of language which, though it arises under highly advanced cultural conditions, may nevertheless be regarded as primitive, since it is actually formed anew before our very eyes. I refer togesture-language. In this case, it is not sounds, but expressive movements, imitative and pantomimic, that form the means by which man communicates his thoughts to man. Though we may regard gesture-language as an original form of language, in so far as we can observe it at the moment of its creation, we must not, of course, forget that the genesis of the forms ofgesture communication familiar to us belongs to a higher culture whose conditions differ widely from those of primitive thought.
Now, of the various forms of gesture-language, the one that is least subject to change is doubtless the means of communication employed by those who are bereft of hearing, and therefore of speech as well, namely, thedeaf and dumb. A similar means of communication through signs and gestures may also be observed among peoples of low culture. Especially when they consist of tribes with markedly different dialects, do such peoples make use of gestures in communicating with one another. Investigations of the spontaneously arising gesture-language of deaf-mutes date largely from the first half of the nineteenth century. More recent studies have been made of the gestures of the North American Indian tribes, and similar, though less complete, observations have been reported concerning the Australians. In these cases, however, gestures sometimes serve also as a sort of secret language. This is even more true of certain signs that occur among some of the peoples of southern Europe, as, for example, among the Neapolitans. In considering the question before us, such cases must, of course, be excluded, since the motive of communicating ideas may here be entirely displaced by that of keeping them secret; instead of a language that arises spontaneously, we have a means which is, on the whole, consciously elaborated for purposes of mutual understanding. If we disregard these cases, which belong to an entirely different order of facts, and examine the data gathered from widely different parts of the earth and from very diverse conditions of culture, we find a remarkable agreement. In certain details, of course, there are differences. The ideas of the Indian are not in all respects like those of the civilized European or those of the Australian. Nevertheless, the gestures that refer to specific concrete objects are frequently so similar that many of the signs employed by the gesture-language of the deaf-mutes of Europe may be found among the Dakota Indians.Could we transfer one of these deaf and dumb persons to this group of Indians, he would probably have no difficulty at all in communicating with them. In more recent times the opportunity of investigating spontaneous gesture-language has not been so great, because deaf-mutes have become more and more educated to the use of verbal language. The principal material for the study of the natural gesture-language of deaf-mutes is, therefore, still to be found in the older observations of Schmalz (1838, 2nd ed. 1848), a German teacher of people thus afflicted, and in the somewhat later reports of an Englishman by the name of Scott (1870).
What, now, do these observations teach us concerning the origin of gesture-language, and therefore probably also concerning the factors underlying the origin of language in general? According to the popular notion, a so-called impulse for communication or, perhaps, certain intellectual processes, voluntary reflections, and actions, account for the fact that the contents of one's own consciousness come to be communicated to other individuals. If, however, we observe gesture-language in its origin, we obtain an entirely different view. This mode of communication is not the result of intellectual reflections or conscious purposes, but of emotion and the involuntary expressive movements that accompany emotion. Indeed, it is simply a natural development of those expressive movements of human beings that also occur where the intention of communicating is obviously absent. As is well known, it is not only emotions that are reflected in one's movements, particularly in mimetic movements of the face, but also ideas. Whenever ideas strongly tinged with feeling enter into the course of emotions, the direct mimetic expressions of the face are supplemented by movements of the arms and hands. The angry man gesticulates with movements which clearly indicate the impulse to attack that is inherent in anger. Or, when we have an ideational process of an emotional nature, and ideas arise referring to objects that are present to us, we point to the objects, even though there be no intention of communicating the ideas. Directions in space, likewise, as well as past time and futurity,are involuntarily expressed by means of backward and forward pointing movements; 'large' and 'small' are expressed by the raising and lowering of the hands. When further movements are added, indicating the form of an object by describing its image in the air with the hands, all the elements of a gesture-language are complete. What is lacking is only that the emotionally coloured idea be not a mere expression of one's own emotion, but that it evoke the same emotion and, through this, the same idea, in the minds of others. Under the influence of the emotion aroused within them, those addressed must then reply with the same, or slightly different, expressive movements. When this occurs, there is developed a common thinking in which impulsive movements are more and more displaced by voluntary actions, and ideational contents, together with the corresponding gestures, enter into the foreground of attention. By virtue of this ideational content, movements expressive of emotions come to be expressions of ideas; the communication of an individual's experiences to others results in an exchange of thought—that is, in language. This development, however, is influenced by that of all the other psychical functions, and especially by the transition of emotional and impulsive movements into voluntary actions.
Of what nature, now, is the content of such a gesture-language as arises independently within a community, and which may, in so far, be regarded as primitive? To this we may briefly reply that all elements of this language are perceptible to the senses, and therefore immediately intelligible. Hence it is that deaf-mutes, though of different nationalities, can make themselves understood without difficulty, even upon meeting for the first time. This intelligibility of gesture-language, however, rests upon the fact that the signs it employs—or, translated into the terminology of spoken language, its words—are direct representations of the objects, the qualities, or the events referred to. Whenever the object discussed is present, the gesture ofpointingwith the hand and finger is itself the clearest way of designating the object. Thus, for instance, 'I' and 'you' are expressedby the speaker's pointing to himself or to the other person. This suggests a similar movement to designate a 'third person' who is not present. The sign in this case is a backward movement of the finger. Whenever the objects of conversation are present in the field of vision, the dumb person, as a rule, dispenses with every other form of representation but that of merely pointing to them.
Since the objects under discussion are, on the whole, only rarely present, there is a second and important class of gestures, which, for the sake of brevity, we may callgraphic. The deaf-mute, as also the Indian and the Australian, represents an absent object by pictures outlined in the air. What he thus sketches in only very general outlines is intelligible to one practised in gesture-language. Moreover, there is a marked tendency for such gestures to become permanent within a particular social group. For the word 'house,' the outlines of roof and walls are drawn; the idea of walking is communicated by imitating the movements of walking with the index and middle fingers of the right hand upon the left arm, which is held out horizontally; the idea of striking is represented by causing the hand to go through the movements of striking. Not infrequently, however, several signs must be combined to make a gesture intelligible. In the German and English deaf and dumb language, the word 'garden,' for example, is expressed by first describing a circle with the index finger to indicate a place, and by then lifting the thumb and the index finger to the nose as the gesture for smelling. 'Garden,' thus, is, as it were, a place where there are flowers to smell. The idea 'teacher' cannot, of course, be directly represented or pictured; it is too complicated for a language of representation. The deaf and dumb person, therefore, is likely to proceed by first making the gesture for man. For this purpose, he singles out an incidental characteristic, his gesture being that of lifting his hat. Since women do not remove their hats in greeting, this gesture is highly typical. The distinctive sign for woman consists in laying the hands upon the breast. Now, in order to communicatethe idea of 'male teacher,' the hat is first lifted as the above gesture for man, and then the index finger is raised. This is done either because pupils in school raise the index finger to indicate their knowledge of a certain thing or, perhaps, because the teacher occasionally raises his finger when he wishes to command attention or to threaten punishment.
Pointing and graphic gestures thus represent the two means which gesture-language employs. Within the second of these classes of gestures, however, we may distinguish a small sub-group that may be calledsignificant; in this case, the object is not represented by means of a direct picture, but by incidental characteristics—man, for example, is expressed by lifting the hat. The signs are all directly perceptible. The most important characteristic of gesture-language, as well as the most distinctive feature of an original language, is the fact that there is no trace of abstract concepts, there being merely perceptual representations. And yet some of these representations—and this is a proof of how insistently human thought, even in its beginnings, presses on to the formation of concepts—have acquired a symbolical meaning by virtue of which they become sensuous means, in a certain sense, of expressing concepts which in themselves are not of a perceptual nature. We may here mention onlyonesuch gesture, noteworthy because it occurs independently in the language of the European deaf and dumb and in that of the Dakota Indians. 'Truth' is represented by moving the index finger directly forward from the lips, while 'lie' is indicated by a movement towards the right or left. The former is thus held to be a straight speech and the latter a crooked speech, transcriptions which also occur, as poetical expressions, in spoken language. On the whole, however, such symbolical signs are rare if the natural gesture-language has not been artificially reconstructed; moreover, they always remain perceptual in character.
Corresponding to this feature is also another characteristic which all natural gesture-languages will be found to possess. In vain we search them for the grammaticalcategories either of our own or of other spoken languages—none may be found. No distinction is made between noun, adjective, or verb; none between nominative, dative, accusative, etc. Every representation retains its representative character, and that to which it refers may exemplify any of the grammatical categories known to us. For example, the gesture for walking may denote either the act of walking or the course or path; that for striking, either the verb 'to strike' or the noun 'blow.' Thus, in this respect also, gesture-language is restricted to perceptual signs expressing ideas capable of perceptual representation. The same is true, finally, of the sequence in which the ideas of the speaker are arranged, or, briefly, of the syntax of gesture-language. In various ways, depending on fixed usages of language, our syntax, as is well known, permits us to separate words that, as regards meaning, belong together, or, conversely, to bring together words that have no immediate relation. Gesture-language obeys butonelaw. Every single sign must be intelligible either in itself or through the one preceding it. It follows from this that if, for example, an object and one of its qualities are both to be designated, the quality must not be expressed first, since, apart from the object, it would be unmeaning; its designation, therefore, regularly occurs after that of the object to which it belongs. Whereas, for example, we say 'a good man,' gesture-language says 'man good.' Similarly, in the case of verb and object, the object generally precedes. When, however, the action expressed by the verb is thought of as more closely related to the subject, the converse order may occur and the verb may directly follow the subject. How, then, does gesture-language reproduce the sentence 'The angry teacher struck the child'? The signs for teacher and for striking have already been described; 'angry' is expressed mimetically by wrinkling the forehead; 'child' by rocking the left forearm supported by the right. Thus, the above sentence is translated into gesture-language in the following manner: First, there are the two signs for teacher, lifting the hat and raisingthe finger; then follows the mimetic gesture for anger, succeeded by a rocking of the arm to signify child, and, finally, by the motion of striking. If we indicate the subject of the sentence by S, the attribute by A, the object by O, and the verb by V, the sequence in our language is ASVO; in gesture-language it is SAOV, 'teacher angry child strikes,' or, in exceptional cases, SAVO. Gesture-language thus reverses the order of sequence in the two pairs of words. A construction such as 'es schlug das Kind der Lehrer(VOS), always possible in spoken language and occurring not infrequently (for example, in Latin), would be absolutely impossible in gesture-language.
If, then, gesture-language affords us certain psychological conclusions regarding the nature of a primitive language, it is of particular interest, from this point of view, to compare its characteristics with the corresponding traits of the most primitive spoken languages. As already stated, the so-called Soudan languages typify those that bear all the marks of relatively primitive thought. These languages of Central Africa obviously represent a much more primitive stage of development than do those of the Bantu peoples of the south or even those of the Hamitic peoples of the north. The language of the Hottentots is related to that of these Hamitic peoples. It is, in fact, because of this relationship, and also because of characteristics divergent from the negro type, that the Hottentots are regarded as a race that immigrated from the north and underwent changes by mixture with native peoples. If, now, we compare one of the Soudan languages, the Ewe, for example, with gesture-language, one difference will at once be apparent. The words of this relatively primitive spoken language do not possess the qualities of perceptibility and immediate intelligibility that characterize each particular gesture-sign. This is readily explicable as a result of processes of phonetic change, which are never absent, as well as of the assimilation of foreign elements and of the replacement of words by conceptual symbols that are accidental and independent of the sound. These changes occur in thehistory of every language. Every spoken language is the outcome of recondite processes whose beginnings are no longer traceable. And yet the Soudan languages, particularly, have preserved characteristics that show much more intimate connections between sound and meaning than our cultural languages possess. The very fact is noteworthy that certain gradations or even antitheses of thought are regularly expressed by gradations or antitheses of sound whose feeling tone plainly corresponds to the relation of the ideas. While our words 'large' and 'small,' 'here' and 'there,' show no correspondence between the character of the sound and the meaning, the case is entirely different with the equivalent expressions in the Ewe language. In this language large and small objects are designated by the same word. In the one case, however, the word is uttered in a deep tone, while in the other a high tone is used. Or, in the case of indicative signs, the deep tone signifies greater remoteness, the high tone, proximity. Indeed, in some Sudan languages three degrees of remoteness or of size are thus distinguished. 'Yonder in the distance' is expressed by a very deep tone; 'yonder in the middle distance,' by a medium tone; and 'here,' by the highest tone. Occasionally, differences of quality are similarly distinguished by differences of tone, as, for example, 'sweet' by a high tone, 'bitter' by a deep tone, 'to be acted upon' (that is, our passive) by a deep tone, and activity (or our active) by a high tone. This accounts for a phenomenon prevalent in other languages remote from those of the Soudan. In Semitic and Hamitic languages, the letter 'U,' particularly, has the force of a passive when occurring either as a suffix to the root of a word or in the middle of the word itself. For example, in the Hebrew forms of the so-called 'Pual' and 'Piel,' as well as 'Hophal' and 'Hiphil,' the first of each pair is passive, and the second, active in meaning. It was frequently supposed that this was accidental, or was due to linguistic causes of phonetic change other than the above. But when we meet the same variations of sound and meaning in other radically differentlanguages, we must stop to ask ourselves whether this is not the result of a psychological relationship which, though generally lost in the later development of language, here still survives in occasional traces. In fact, when we recall the way in which we relate stories to children, we at once notice that precisely the same phenomenon recurs in child-language—a language, of course, first created, as a rule, by adults. This connection of sound and meaning is clearly due to the unconscious desire that the sound shall impart to the child not merely the meaning of the idea, but also its feeling-tone. In describing giants and monsters, she who relates fairy-tales to the child deepens her tone; when fairies, elves, and dwarfs appear in the narrative, she raises her voice. If sorrow and pain enter, the tone is deepened; with joyous emotions, high tones are employed. In view of these facts, we might say that this direct correlation of expression and meaning, observed in that most primitive of all languages, gesture-language, has disappeared even from the relatively primitive spoken languages; nevertheless, the latter have retained traces of it in greater abundance than have the cultural languages. In the cultural languages they recur, if at all, only in the onomatopoetic word-formations of later origin. We may recall such words assausen(soughing),brummen(growling),knistern(crackling), etc.
The question still remains how the other characteristics of gesture-language, particularly the absence of grammatical categories and a syntax which follows the principle of immediate and perceptual intelligibility, compare with the corresponding characteristics of the relatively primitive spoken languages. These characteristics, indeed, are of incomparably greater importance than the relations of sound and meaning. The latter are more strongly exposed to external, transforming influences. Word-formations, however, and the position of the words within the sentence, mirror the forms of thought itself; whenever the thought undergoes vital changes, the latter inevitably find expression in the grammatical categories of the language, and in the laws of syntax which it follows.
From the point of view just developed, the investigation of the grammatical forms of primitive language is of particular importance for the psychology of primitive man. True, as has already been remarked, the languages of the most primitive tribes have not been preserved to us in their original form. And yet it is in this very realm of grammatical forms, far more even than in that of sound pictures and onomatopoetic words, that the Soudan languages possess characteristics which mark them as the expression of processes of thought that have remained on a relatively primitive level. This is indicated primarily by the fact that these languages lack what we would call grammatical categories. As regards this point, Westermann's grammar of the Ewe language is in entire agreement with the much earlier results which Steinthal reached in his investigation of the Manda language, which is also of the Soudan region. These languages consist of monosyllabic words which follow one another in direct succession without any intermediate inflectional elements to modify their meaning. Philologists usually call such languages 'root-languages,' because a sound complex that carries the essential meaning of a word, apart from all modifying elements, is called by their science a verbal root. In the Latin wordfero, fer, meaning 'to bear,' is the root from which all modifications of the verbferre(to carry) are formed by means of suffixal elements. If, therefore, a language consists of sound complexes having the nature of roots, it is called a root-language. As a matter of fact, however, the languages under discussion consist purely of detached, monosyllabicwords; the conception 'root,' which itself represents the product of a grammatical analysis of our flectional languages, may only improperly be applied to them. Such a language is composed of detached monosyllabic words, each of which has a meaning, yet none of which falls under any particular grammatical category. One and the same monosyllabic word may denote an object, an act, or a quality, just as in gesture-languagethe gesture of striking may denote the verb 'to strike' and also the noun 'blow.' From this it is evident to what extent the expressions 'root' and 'root-language' carry over into this primitive language a grammatical abstraction which is entirely inappropriate in case they suggest the image of a root. This image originated among grammarians at a time when the view was current that, just as the stem and branches of a plant grow out of its root, so also in the development of a language does a word always arise out of a group of either simple or composite sounds that embody the main idea. But the component parts of a language are certainly not roots in this sense; every simple monosyllabic word combines with others, and from this combination there result, in part, modifications in meaning, and, in part, sentences. Language, thus, does not develop by sprouting and growing, but by agglomeration and agglutination. Now, the Soudan languages are characterized by the fact that they possess very few such fixed combinations in which the individual component parts have lost their independence. In this respect, accordingly, they resemble gesture-language. The latter also is unfamiliar with grammatical categories in so far as these apply to the words themselves; the very same signs denote objects, actions, and qualities—indeed, generally even that for which in our language we employ particles. This agreement with gesture-language is brought home to us most strikingly if we consider the words which the primitive spoken languages employ for newly formed ideas—such, for instance, as refer to previously unknown objects of culture. Here it appears that the speaker always forms the new conception by combining into a series those ideas with which he is more familiar. When schools were introduced into Togo, for example, and a word for 'slate-pencil' became necessary, the Togo negroes called it 'stone scratch something'—that is, a stone with which we scratch something. Similarly 'kitchen,' an arrangement unknown to these tribes, was referred to as 'place cook something'; 'nail,' as 'iron head broad.' The single word alwaysstands for a sensibly perceptual object, and the new conception is formed, not, as epistemologists commonly suppose, by means of a comparison of various objects, but by arranging in sequence those perceptual ideas whose combined characteristics constitute the conception. The same is true with regard to the expressions for such thought relations as are variously indicated in our language by the inflections of substantive, adjective, and verb. The Soudan languages make no unambiguous distinction between noun and verb. Much less are the cases of the substantive, or the moods and tenses of the verb, distinguished; to express these distinctions, separate words are always used. Thus, 'the house of the king' is rendered as 'house belong king.' The conception of case is here represented by an independent perception that crowds in between the two ideas which it couples together. The other cases are, as a rule, not expressed at all, but are implied in the connection. Similarly, verbs possess no future tense to denote future time. Here also a separate word is introduced, one that may be rendered by 'come.' 'I go come' means 'I shall go'; or, to mention the preterit, 'I go earlier' means 'I went.' Past time, however, may also be expressed by the immediate repetition of the word, a sensibly perceptual sign, as it were, that the action is completed. When the Togo negro says 'I eat,' this means 'I am on the point of eating'; when he says 'I eat eat,' it means 'I have eaten.'
But ideas of such acts and conditions as are in themselves of a perceptual nature are also occasionally expressed by combining several elements which are obtained by discriminating the separate parts of a perceptual image. The idea to bring, for example, is expressed by the Togo negro as 'take, go, give.' In bringing something to some one, one must first take it, then go to him and give it to him. It therefore happens that the word 'go,' in particular, is frequently added even where we find no necessity for especially emphasizing the act of going. Thus, the Togo negro would very probably express the sentence,'The angry teacher strikes the child,' in the following way: 'Man-school-angry-go-strike-child.' This is the succession that directly presents itself to one who thinks in pictures, and it therefore finds expression in language. Whenever conceptions require a considerable number of images in order to be made picturable, combinations that are equivalent to entire sentences may result in a similar manner. Thus, the Togo negro expresses the concept 'west' by the words 'sun-sit-place'—that is, the place where the sun sits down. He thinks of the sun as a personal being who, after completing his journey, here takes a seat.
These illustrations may suffice to indicate the simplicity and at the same time the complexity of such a language. It is simple, in that it lacks almost all grammatical distinctions; it is complicated, because, in its constant reliance on sensibly perceptual images, it analyses our concepts into numerous elements. This is true not merely of abstract concepts, which these languages, as a rule, do not possess, but even of concrete empirical concepts. We need only refer to the verb 'to bring,' reduced to the form of three verbs, or the concept 'west,' for whose expression there is required not only the sun and the location which we must give it but also its act of seating itself. In all of these traits, then, primitive language is absolutely at one with gesture-language.
The same is true of the syntax of the two kinds of language. This also is no more irregular and accidental in the Soudan language than it is in gesture-language. As a rule, indeed, it is stricter than the syntax of our languages, for in the latter inflection makes possible a certain variation in the arrangement of words within a sentence according to the particular shade of meaning desired. In primitive language, the arrangement is much more uniform, being governed absolutely and alone by the same law as prevails in gesture-language—namely, the arrangement of words in their perceptual order. Without exception, therefore, object precedes attribute, and substantive, adjective. Less constant, however, is the relation of verb and object, in theEwe language; the verb generally precedes, but the object may come first; the verb, however, always follows the subject whose action it expresses. This perceptual character of primitive language appears most strikingly when we translate any thought that is at all complicated from a primitive language into our own, first in its general meaning, and then word for word. Take an illustration from the language of the Bushmen. The meaning would be substantially this: 'The Bushman was at first received kindly by the white man in order that he might be brought to herd his sheep; then the white man maltreated the Bushman; the latter ran away, whereupon the white man took another Bushman, who suffered the same experience,' The language of the Bushmen expresses this in the following way: 'Bushman-there-go, here-run-to-white man, white man-give-tobacco, Bushman-go-smoke, go-fill-tobacco-pouch, white man-give-meat-Bushman, Bushman-go-eat-meat, stand-up-go-home, go happily, go-sit-down, herd-sheep-white man, white man-go-strike Bushman, Bushman-cry-loud-pain, Bushman-go-run-away-white man, white man-run-after-Bushman, Bushman-then-another, this one-herd-sheep, Bushman-all-gone.' In this complaint of the man of nature against his oppressor, everything is concrete, perceptual. He does not say, The Bushman was at first kindly taken up by the white man, but, The white man gives him tobacco, he fills his pouch and smokes; the white man gives him meat, he eats this and is happy, etc. He does not say, The white man maltreats the Bushman, but, He strikes him, the Bushman cries with pain, etc. What we express in relatively abstract concepts is entirely reduced by him to separate perceptual images. His thought always attaches to individual objects. Moreover, just as primitive language has no specific means for expressing a verb, so also are change and action overshadowed in primitive thought by the concrete image. The thinking itself, therefore, may be calledconcrete. Primitive man sees the image with its separate parts; and, as he sees it, so he reproduces it in his language. It is for this very reason that he is unfamiliarwith differences of grammatical categories and with abstract concepts. Sequence is still governed entirely by the pure association of ideas, whose order is determined by perception and by the recollection of that which has been experienced. The above narrative of the Bushman expresses no unitary thought, but image follows upon image in the order in which these appear to consciousness. Thus, the thinking of primitive man is almost exclusively associative. Of the more perfect form of combining concepts, the apperceptive, which unites the thoughts into a systematic whole, there are as yet only traces, such as occur in the combination of the separate memory images.
Many analogues to the formal characteristics of primitive thought revealed in these linguistic phenomena may be met in child-language. There is a wide divergence, however, with respect to the very element which has already disappeared, with the exception of slight traces, from the language of primitive peoples. I refer to the close correlation of sound and meaning. As regards this feature, child-language is much more similar to gesture-language than is possible in the case of forms of speech that have undergone a long historical development. For, child-language, like gesture-language, is, in a certain sense, continually being created anew. Of course, it is not created, as is sometimes supposed, by the children themselves. It is a conventionalized language of the mothers and nurses who converse with the child, supplemented, in part, by the child's associates along the lines of these traditional models. The sound-complexes signifying animals, 'bow-wow' for the dog, 'hott-hott' for the horse, 'tuk-tuk' for the chicken, etc., as also 'papa' and 'mamma' for father and mother, are sounds that are in some way fitted to the meaning and at the same time resemble so far as possible the babbling sounds of the child. But this entire process is instituted by the child's associates, and is at most supplemented by the child himself to the extent of a few incidental elements. For this reason, child-language has relatively little to teach us concerning the development of speaking and thinking;those psychologists and teachers who believe that it affords an important source of information concerning the origin of thought are in error. Such information can be gained only from those modes of expressing thought which, like gesture-language, are originated anew by the speaker and are not externally derived, or from those which, like the spoken languages of primitive peoples, have retained, in their essential characteristics, primitive modes of thinking. Even in these cases it is only theformsof thought that are thus discoverable. The content, as is implied by the formal characteristics themselves, is, of course, also of a sense-perceptual, not of a conceptual, nature. And yet the particular character or quality of this content is not inherent in the forms of the language as such. To gain a knowledge of its nature we must examine the specific ideas themselves and the associated feelings and emotions.
Thus, then, the further question arises: Wherein consists the content of primitive thought?Twosorts of ideas may be distinguished. The one comprises that stock of ideas which is supplied to consciousness by the direct perceptions of daily life—ideas such as go, stand, lie, rest, etc., together with animal, tree (particularly in the form of individual animals and trees), man, woman, child, I, thou, you, and many others. These are objects of everyday perception that are familiar to all, even to the primitive mind. But there is also asecondclass of ideas. These do not represent things of immediate perception; briefly expressed, they originate in feeling, in emotional processes which are projected outward into the environment. This is an important and particularly characteristic group of primitive ideas. Included within it are all references to that which is not directly amenable to perception but, transcending this, is reallysupersensuous, even though appearing in the form of sensible ideas. This world of imagination, projected from man's own emotional life into external phenomena, is what we mean bymythologicalthinking. The things and processes given to perception are supplemented by other realities that are of a non-perceptiblenature and therefore belong to an invisible realm back of the visible world. These are the elements, furthermore, which very early find expression in theartof primitive man.