Custom and Myth/The Early History of the Family

by Andrew Lang

THE EARLY HISTORY OF THE FAMILY.

What are the original forms of the human family? Did man begin by being monogamous or polygamous, but, in either case, the master of his own home and the assured central point of his family relations? Or were the unions of the sexes originally shifting and precarious, so that the wisest child was not expected to know his own father, and family ties were reckoned through the mother alone? Again (setting aside the question of what was 'primitive' and 'original'), did the needs and barbarous habits of early men lead to a scarcity of women, and hence to polyandry (that is, the marriage of one woman to several men), with the consequent uncertainty about male parentage? Once more, admitting that these loose and strange relations of the sexes do prevail, or have prevailed, among savages, is there any reason to suppose that the stronger races, the Aryan and Semitic stocks, ever passed through this stage of savage customs? These are the main questions debated between what we may call the 'historical' and the 'anthropological' students of ancient customs.

When Sir Henry Maine observed, in 1861, that it was difficult to say what society of men had not been, originally, based on the patriarchal family, he went, of course, outside the domain of history. What occurred in the very origin of human society is a question perhaps quite inscrutable. Certainly, history cannot furnish the answer. Here the anthropologist and physiologist come in with their methods, and even those, we think, can throw but an uncertain light on the very 'origin' of institutions, and on strictly primitive man.

For the purposes of this discussion, we shall here re-state the chief points at issue between the adherents of Sir Henry Maine and of Mr. M'Lennan, between historical and anthropological inquirers.

1. Did man originally live in the patriarchal family, or did he live in more or less modified promiscuity, with uncertainty of blood-ties, and especially of male parentage?

2. Did circumstances and customs at some time compel or induce man (whatever his original condition) to resort to practices which made paternity uncertain, and so caused kinship to be reckoned through women?

3. Granting that some races have been thus reduced to matriarchal forms of the family—that is, to forms in which the woman is the permanent recognised centre—is there any reason to suppose that the stronger peoples, like the Aryans and the Semites, ever passed through a stage of culture in which female, not male, kinship was chiefly recognised, probably as a result of polyandry, of many husbands to one wife?

On this third question, it will be necessary to produce much evidence of very different sorts: evidence which, at best, can perhaps only warrant an inference, or presumption, in favour of one or the other opinion. For the moment, the impartial examination of testimony is more important and practicable than the establishment of any theory.

(1.) Did man originally live in the patriarchal family, the male being master of his female mate or mates, and of his children? On this first point Sir Henry Maine, in his new volume,^[1] may be said to come as near proving his case as the nature and matter of the question will permit. Bachofen, M'Lennan, and Morgan, all started from a hypothetical state of more or less modified sexual promiscuity. Bachofen's evidence (which may be referred to later) was based on a great mass of legends, myths, and travellers' tales, chiefly about early Aryan practices. He discovered Hetärismus, as he called it, or promiscuity, among Lydians, Etruscans, Persians, Thracians, Cyrenian nomads, Egyptians, Scythians, Troglodytes, Nasamones, and so forth. Mr. M'Lennan's view is, perhaps, less absolutely stated than Sir Henry Maine supposes. M'Lennan says^[2] 'that there has been a stage in the development of the human races, when there was no such appropriation of women to particular men; when, in short, marriage, as it exists among civilised nations, was not practised. Marriage, in this sense, was yet undreamt of.' Mr. M'Lennan adds (pp. 130, 131), 'as among other gregarious animals, the unions of the sexes were probably, in the earliest times, loose, transitory, and, in some degree, promiscuous.'

Sir Henry Maine opposes to Mr. M'Lennan's theory the statement of Mr. Darwin: 'From all we know of the passions of all male quadrupeds, promiscuous intercourse in a state of Nature is highly improbable.'^[3] On this first question, let us grant to Sir Henry Maine, to Mr. Darwin, and to common sense that if the very earliest men were extremely animal in character, their unions while they lasted were probably monogamous or polygamous. The sexual jealousy of the male would secure that result, as it does among many other animals. Let the first point, then, be scored to Sir Henry Maine: let it be granted that if man was created perfect, he lived in the monogamous family before the Fall: and that, if he was evolved as an animal, the unchecked animal instincts would make for monogamy or patriarchal polygamy in the strictly primitive family.

(2.) Did circumstances and customs ever or anywhere compel or induce man (whatever his original condition) to resort to practices which made paternity uncertain, and so caused the absence of the patriarchal family, kinship being reckoned through women? If this question be answered in the affirmative, and if the sphere of action of the various causes be made wide enough, it will not matter much to Mr. M'Lennan's theory whether the strictly primitive family was patriarchal or not. If there occurred a fall from the primitive family, and if that fall was extremely general, affecting even the Aryan race, Mr. M'Lennan's adherents will be amply satisfied. Their object is to show that the family, even in the Aryan race, was developed through a stage of loose savage connections. If that can be shown, they do not care much about primitive man properly so called. Sir Henry Maine admits, as a matter of fact, that among certain races, in certain districts, circumstances have overridden the sexual jealousy which secures the recognition of male parentage. Where women have been few, and where poverty has been great, jealousy has been suppressed, even in the Venice of the eighteenth century. Sir H. Maine says, 'The usage' (that of polyandry—many husbands to a single wife) 'seems to me one which circumstances overpowering morality and decency might at any time call into existence. It is known to have arisen in the native Indian army.' The question now is, what are the circumstances that overpower morality and decency, and so produce polyandry, with its necessary consequences, when it is a recognised institution—the absence of the patriarchal family, and the recognition of kinship through women? Any circumstances which cause great scarcity of women will conduce to those results. Mr. M'Lennan's opinion was, that the chief cause of scarcity of women has been the custom of female infanticide—of killing little girls as bouches inutiles. Sir Henry Maine admits that 'the cause assigned by M'Lennan is a vera causa—it is capable of producing the effects.'^[4] Mr. M'Lennan collected a very large mass of testimony to prove the wide existence of this cause of paucity of women. Till that evidence is published, I can only say that it was sufficient, in Mr. M'Lennan's opinion, to demonstrate the wide prevalence of the factor which is the mainspring of his whole system.^[5] How frightfully female infanticide has prevailed in India, everyone may read in the official reports of Col. M'Pherson, and other English authorities. Mr. Fison's 'Kamilaroi and Kurnai' contains some notable, though not to my mind convincing, arguments on the other side. Sir Henry Maine adduces another cause of paucity of women: the wanderings of our race, and expeditions across sea.^[6] This cause would not, however, be important enough to alter forms of kinship, where the invaders (like the early English in Britain) found a population which they could conquer and whose women they could appropriate.

Apart from any probable inferences that may be drawn from the presumed practice of female infanticide, actual ascertained facts prove that many races do not now live, or that recently they did not live, in the patriarchal or modern family. They live, or did live, in polyandrous associations. The Thibetans, the Nairs, the early inhabitants of Britain (according to Cæsar), and many other races,^[7] as well as the inhabitants of the Marquesas Islands, and the Iroquois (according to Lafitau), practise, or have practised, polyandry.

We now approach the third and really important problem—(3.) Is there any reason to suppose that the stronger peoples, like the Aryans and the Semites, ever passed through a stage of culture in which female, not male, kinship was chiefly recognised, probably as a result of polyandry?

Now the nature of the evidence which affords a presumption that Aryans have all passed through Australian institutions such as polyandry, is of extremely varied character. Much of it may undoubtedly be explained away. But such strength as the evidence has (which we do not wish to exaggerate) is derived from its convergence to one point—namely, the anterior existence of polyandry and the matriarchal family among Aryans before and after the dawn of real history.

For the sake of distinctness we may here number the heads of the evidence bearing on this question. We have—

1. The evidence of inference from the form of capture in bridal ceremonies.

2. The evidence from exogamy: the law which forbids marriage between persons of the same family name.

3. The evidence from totemism—that is, the derivation of the family name and crest or badge, from some natural object, plant or animal.^[8] Persons bearing the name may not intermarry, nor, as a rule, may they eat the object from which they derive their family name and from which they claim to be descended.

4. 4. The evidence from the gens of Rome, or yενος of ancient Greece, in connection with Totemism.

5. The evidence from myth and legend.

6. The evidence from direct historical statements as to the prevalence of the matriarchal family, and inheritance through the maternal line.

To take these various testimonies in their order, let us begin with

(1.) The form of capture in bridal ceremonies. That this form survived in Sparta, Crete, in Hindoo law, in the traditions of Ireland, in the popular rustic customs of Wales, is not denied.

If we hold, with Mr. M'Lennan, that scarcity of women (produced by female infanticide or otherwise) is the cause of the habit of capturing wives, we may see, in survivals of this ceremony of capture among Aryans, a proof of early scarcity of women, and of probable polyandry. But an opponent may argue, like Mr. J. A. Farrer in 'Primitive Manners,' that the ceremony of capture is mainly a concession to maiden modesty among early races. Here one may observe that the girls of savage tribes are notoriously profligate and immodest about illicit connections. Only honourable marriage brings a blush to the cheek of these young persons. This is odd, but, in the present state of the question, we cannot lean on the evidence of the ceremony of capture. We cannot demonstrate that it is derived from a time when paucity of women made capture of brides necessary. Thus 'honours are easy' in this first deal.

(2.) The next indication is very curious, and requires much more prolonged discussion. The custom of Exogamy was first noted and named by Mr. M'Lennan. Exogamy is the prohibition of marriage within the supposed blood-kinship, as denoted by the family name. Such marriage, among many backward races, is reckoned incestuous, and is punishable by death. Certain peculiarities in connection with the family name have to be noted later. Now, Sir Henry Maine admits that exogamy, as thus defined, exists among the Hindoos. 'A Hindoo may not marry a woman belonging to the same gotra, all members of the gotra being theoretically supposed to have descended from the same ancestor.' The same rule prevails in China. 'There are in China large bodies of related clansmen, each generally bearing the same clan-name. They are exogamous; no man will marry a woman having the same clan-name with himself.' It is admitted by Sir Henry Maine that this wide prohibition of marriage was the early Aryan rule, while advancing civilisation has gradually permitted marriage within limits once forbidden. The Greek Church now (according to Mr. M'Lennan), and the Catholic Church in the past, forbade intermarriages 'as far as relationship could be known.' The Hindoo rule appears to go still farther, and to prohibit marriage as far as the common gotra name seems merely to indicate relationship.

As to the ancient Romans, Plutarch says: Formerly they did not marry women connected with them by blood, any more than they now marry aunts or sisters. It was long before they would even intermarry with cousins.' Plutarch also remarks that, in times past, Romans did not marry συyyενιδας, and if we may render this 'women of the same gens,' the exogamous prohibition in Rome was as complete as among the Hindoos. I do not quite gather from Sir Henry Maine's account of the Slavonic house communities (pp. 254, 255) whether they dislike all kindred marriages, or only marriage within the 'greater blood'—that is, within the kinship on the male side. He says: 'The South Slavonians bring their wives into the group, in which they are socially organised, from a considerable distance outside. . . . Every marriage which requires an ecclesiastical dispensation is regarded as disreputable.'

On the whole, wide prohibitions of marriage are archaic: the widest are savage; the narrowest are modern and civilised. Thus the Hindoo prohibition is old, barbarous, and wide. 'The barbarous Aryan,' says Sir Henry Maine, 'is generally exogamous. He has a most extensive table of prohibited degrees.' Thus exogamy seems to be a survival of barbarism. The question for us is, Can we call exogamy a survival from a period when (owing to scarcity of women and polyandry) clear ideas of kinship were impossible? If this can be proved, exogamous Aryans either passed through polyandrous institutions, or borrowed a savage custom derived from a period when ideas of kinship were obscure.

If we only knew the origin of the prohibition to marry within the family name all would be plain sailing. At present several theories of the origin of exogamy are before the world. Mr. Morgan, the author of 'Ancient Society,' inclines to trace the prohibition to a great early physiological discovery, acted on by primitive men by virtue of a contrat social. Early man discovered that children of unsound constitutions were born of nearly related parents. Mr. Morgan says: 'Primitive men very early discovered the evils of close interbreeding.' Elsewhere Mr. Morgan writes: 'Intermarriage in the gens was prohibited, to secure the benefits of marrying out with unrelated persons.' This arrangement was 'a product of high intelligence,' and Mr. Morgan calls it a 'reform.'

Let us examine this very curious theory. First: Mr. Morgan supposes early man to have made a discovery (the evils of the marriage of near kin) which evades modern physiological science. Modern science has not determined that the marriages of kinsfolk are pernicious. Is it credible that savages should discover a fact which puzzles science? It may be replied that modern care, nursing, and medical art save children of near marriages from results which were pernicious to the children of early man. Secondly: Mr. Morgan supposes that barbarous man (so notoriously reckless of the morrow as he is), not only made the discovery of the evils of interbreeding, but acted on it with promptitude and self-denial. Thirdly: Mr. Morgan seems to require, for the enforcement of the exogamous law, a contrat social. The larger communities meet, and divide themselves into smaller groups, within which wedlock is forbidden. This 'social pact' is like a return to the ideas of Rousseau. Fourthly: The hypothesis credits early men with knowledge and discrimination of near degrees of kin, which they might well possess if they lived in patriarchal families. But it represents that they did not act on their knowledge. Instead of prohibiting marriage between parents and children, cousins, nephews and aunts, uncles and nieces, they prohibited marriage within the limit of the name of the kin. This is still the Hindoo rule, and, if the Romans really might not at one time marry within the gens, it was the Roman rule. Now observe, this rule fails to effect the very purpose for which ex hypothesi it was instituted. Where the family name goes by the male side, marriages between cousins are permitted, as in India and China. These are the very marriages which some theorists now denounce as pernicious. But, if the family name goes by the female side, marriages between half-brothers and half-sisters are permitted, as in ancient Athens and among the Hebrews of Abraham's time. Once more, the exogamous prohibition excludes, in China, America, Africa, Australia, persons who are in no way akin (according to our ideas) from intermarriage. Thus Mr. Doolittle writes:^[9] 'Males and females of the same surname will never intermarry in China. Cousins who have not the same ancestral surname may intermarry. Though the ancestors of persons of the same surname have not known each other for thousands of years, they may not intermarry.' The Hindoo gotra rule produces the same effects.

For all these reasons, and because of the improbability of the physiological discovery, and of the moral 'reform' which enforced it; and again, because the law is not of the sort which people acquainted with near degrees of kinship would make; and once more, because the law fails to effect its presumed purpose, while it does attain ends at which it does not aim—we cannot accept Mr. Morgan's suggestion as to the origin of exogamy. Mr. M'Lennan did not live to publish a subtle theory of the origin of exogamy, which he had elaborated. In 'Studies in Ancient History,' he hazarded a conjecture based on female infanticide:—

   'We believe the restrictions on marriage to be connected with the
   practice in early times of female infanticide, which, rendering women
   scarce, led at once to polyandry within the tribe, and the capturing
   of women from without. . . . Hence the cruel custom which, leaving
   the primitive human hordes with very few young women of their own,
   occasionally with none, and in any case seriously disturbing the
   balance of the sexes within the hordes, forces them to prey upon one
   another for wives. Usage, induced by necessity, would in time
   establish a prejudice among the tribes observing it, a prejudice
   strong as a principle of religion—as every prejudice relating to
   marriage is apt to be—against marrying women of their own stock.'

Mr. M'Lennan describes his own hypothesis as 'a suggestion thrown out at what it was worth.'^[10] In his later years, as we have said, he developed a very subtle and ingenious theory of the origin of exogamy, still connecting it with scarcity of women, but making use of various supposed stages and processes in the development of the law. That speculation remains unpublished. To myself, the suggestion given in 'Studies in Ancient History' seems inadequate. I find it difficult to conceive that the frequent habit of stealing women should indispose men to marry the native women they had at hand. That this indisposition should grow into a positive law, and the infringement of the law be regarded as a capital offence, seems still more inconceivable. My own impression is, that exogamy may be connected with some early superstition or idea of which we have lost the touch, and which we can no longer explain.

Thus far, the consideration of exogamy has thrown no clear light on the main question—the question whether the customs of civilised races contain relics of female kinship. On Sir Henry Maine's theory of exogamy, that Aryan custom is unconnected with female kinship, polyandry, and scarcity of women. On Mr. M'Lennan's theory, exogamy is the result of scarcity of women, and implies polyandry and female kinship. But neither theory has seemed satisfactory. Yet we need not despair of extracting some evidence from exogamy, and that evidence, on the whole, is in favour of Mr. M'Lennan's general hypothesis. (1.) The exogamous prohibition must have first come into force when kinship was only reckoned on one side of the family. This is obvious, whether we suppose it to have arisen in a society which reckoned by male or by female kinship. In the former case, the law only prohibits marriage with persons of the father's, in the second case with persons of the mother's, family name, and these only it recognises as kindred. (2.) Our second point is much more important. The exogamous prohibition must first have come into force when kinship was so little understood that it could best be denoted by the family name. This would be self-evident, if we could suppose the prohibition to be intended to prevent marriages of relations. Had the authors of the prohibition been acquainted with the nature of near kinships, they would simply (as we do) have forbidden marriage between persons in those degrees. The very nature of the prohibition, on the other hand, shows that kinship was understood in a manner all unlike our modern system. The limit of kindred was everywhere the family name: a limit which excludes many real kinsfolk and includes many who are not kinsfolk at all. In Australia especially, and in America, India, and Africa, to a slighter extent, that definition of kindred by the family name actually includes alligators, smoke, paddy melons, rain, crayfish, sardines, and what you please.^[11] Will anyone assert, then, that people among whom the exogamous prohibition arose were organised on the system of the patriarchal family, which permits the nature of kinship to be readily understood at a glance? Is it not plain that the exogamous prohibition (confessedly Aryan) must have arisen in a stage of culture when ideas of kindred were confused, included kinship with animals and plants, and were to us almost, if not quite, unintelligible? It is even possible, as Mr. M'Lennan says,^[12] 'that the prejudice against marrying women of the same group may have been established before the facts of blood relationship had made any deep impression on the human mind.' How the exogamous prohibition tends to confirm this view will next be set forth in our consideration of Totemism.

The Evidence from Totemism.—Totemism is the name for the custom by which a stock (scattered through many local tribes) claims descent from and kindred with some plant, animal, or other natural object. This object, of which the effigy is sometimes worn as a badge or crest, members of the stock refuse to eat. As a general rule, marriage is prohibited between members of the stock—between all, that is, who claim descent from the same object and wear the same badge. The exogamous limit, therefore, is denoted by the stock-name and crest, and kinship is kinship in the wolf, bear, potato, or whatever other object is recognised as the original ancestor. Finally, as a general rule, the stock-name is derived through the mother, and where it is derived through the father there are proofs that the custom is comparatively modern. It will be acknowledged that this sort of kindred, which is traced to a beast, bird, or tree, which is recognised in every person bearing the same stock-name, which is counted through females, and which governs marriage customs, is not the sort of kindred which would naturally arise among people regulated on the patriarchal or monandrous family system. Totemism, however, is a widespread institution prevailing all over the north of the American continent, also in Peru (according to Garcilasso de la Vega); in Guiana (the negroes have brought it from the African Gold Coast, where it is in full force, as it also is among the Bechuanas); in India among Hos, Garos, Kassos, and Oraons; in the South Sea Islands, where it has left strong traces in Mangaia; in Siberia, and especially in the great island continent of Australia. The Semitic evidences for totemism (animal-worship, exogamy, descent claimed through females) are given by Professor Robertson Smith, in the 'Journal of Philology,' ix. 17, 'Animal Worship and Animal Tribes among the Arabs, and in the Old Testament.' Many other examples of totemism might be adduced (especially from Egypt), but we must restrict ourselves to the following questions:—

(1.) What light is thrown on the original form of the family by totemism? (2.) Where we find survivals of totemism among civilised races, may we conclude that these races (through scarcity of women) had once been organised on other than the patriarchal model?

As to the first question, we must remember that the origin and determining causes of totemism are still unknown. Mr. M'Lennan's theory of the origin of totemism has never been published. It may be said without indiscretion that Mr. M'Lennan thought totemism arose at a period when ideas of kinship scarcely existed at all. 'Men only thought of marking one off from another,' as Garcilasso de la Vega says: the totem was but a badge worn by all the persons who found themselves existing in close relations; perhaps in the same cave or set of caves. People united by contiguity, and by the blind sentiment of kinship not yet brought into explicit consciousness, might mark themselves by a badge, and might thence derive a name, and, later, might invent a myth of their descent from the object which the badge represented. I do not know whether it has been observed that the totems are, as a rule, objects which may be easily drawn or tattooed, and still more easily indicated in gesture-language. Some interesting facts will be found in the 'First Annual Report of the Bureau of Ethnology,' p. 458 (Washington, 1881). Here we read how the 'Crow' tribe is indicated in sign-language by 'the hands held out on each side, striking the air in the manner of flying.' The Bunaks (another bird tribe) are indicated by an imitation of the cry of the bird. In mentioning the Snakes, the hand imitates the crawling motion of the serpent, and the fingers pointed up behind the ear denote the Wolves. Plainly names of the totem sort are well suited to the convenience of savages, who converse much in gesture-language. Above all, the very nature of totemism shows that it took its present shape at a time when men, animals, and plants were conceived of as physically akin; when names were handed on through the female line; when exogamy was the rule of marriage, and when the family theoretically included all persons bearing the same family name, that is, all who claimed kindred with the same plant, animal, or object, whether the persons are really akin or not. These ideas and customs are not the ideas natural to men organised in the patriarchal family.

The second question now arises: Can we infer from survivals of totemism among Aryans that these Aryans had once been organised on the full totemistic principle, probably with polyandry, and certainly with female descent? Where totemism now exists in full force, there we find exogamy and derivation of the family name through women, the latter custom indicating uncertainty of male parentage in the past. Are we to believe that the same institutions have existed wherever we find survivals of totemism? If this be granted, and if the supposed survivals of totemism among Aryans be accepted as genuine, then the Aryans have distinctly come through a period of kinship reckoned through women, with all that such an institution implies. For indications that the Aryans of Greece and India have passed through the stage of totemism, the reader may be referred to Mr. M'Lennan's 'Worship of Plants and Animals' ('Fortnightly Review,' 1869, 1870). The evidence there adduced is not all of the same value, and the papers are only a hasty rough sketch based on the first testimonies that came to hand. Probably the most important 'survival' of totemism in Greek legend is the body of stories about the amours of Zeus in animal form. Various noble houses traced their origin to Zeus or Apollo, who, as a bull, tortoise, serpent, swan, or ant, had seduced the mother of the race. The mother of the Arcadians became a she-bear, like the mother of the bear stock of the Iroquois. As we know plenty of races all over the world who trace their descent from serpents, tortoises, swans, and so forth, it is a fair hypothesis that the ancestors of the Greeks once believed in the same fables. In later times the swan, serpent, ant, or tortoise was explained as an avatar of Zeus. The process by which an anthropomorphic god or hero succeeds to the exploits of animals, of theriomorphic gods and heroes, is the most common in mythology, and is illustrated by actual practice in modern India. When the Brahmins convert a pig-worshipping tribe of aboriginals, they tell their proselytes that the pig was an avatar of Vishnu. The same process is found active where the Japanese have influenced the savage Ainos, and persuaded them that their bear- or dog-father was a manifestation of a deity. We know from Plutarch ('Theseus') that, in addition to families claiming descent from divine animals, one Athenian yενος, the Ioxidæ, revered an ancestral plant, the asparagus. A vaguer indication of totemism may perhaps be detected in the ancient theriomorphic statues of Greek gods, as the Ram-Zeus and the Horse-headed Demeter, and in the various animals and plants which were sacred to each god and represented as his companions.

The hints of totemism among the ancient Irish are interesting. One hero, Conaire, was the son of a bird, and before his birth his father (the bird) told the woman (his mother) that the child must never eat the flesh of fowls. 'Thy son shall be named Conaire, and that son shall not kill birds.'^[13] The hero Cuchullain, being named after the dog, might not eat the flesh of the dog, and came by his ruin after transgressing this totemistic taboo. Races named after animals were common in ancient Ireland. The red-deer and the wolves were tribes dwelling near Ossory, and Professor Rhys, from the frequency of dog names, inclines to believe in a dog totem in Erin. According to the ancient Irish 'Wonders of Eri,' in the 'Book of Glendaloch,' 'the descendants of the wolf are in Ossory,' and they could still transform themselves into wolves.^[14] As to our Anglo-Saxon ancestors, there is little evidence beyond the fact that the patronymic names of many of the early settlements of Billings, Arlings, and the rest, are undeniably derived from animals and plants. The manner in which those names are scattered locally is precisely like what results in America, Africa, and Australia from the totemistic organisation.^[15] In Italy the ancient custom by which animals were the leaders of the Ver sacrum or armed migration is well known. The Piceni had for their familiar animal or totem (if we may call it so) a woodpecker; the Hirpini were like the 'descendants of the wolf' in Ossory, and practised a wolf-dance in which they imitated the actions of the animal.

Such is a summary of the evidence which shows that Aryans had once been totemists, therefore savages, and therefore, again, had probably been in a stage when women were scarce and each woman had many husbands.

Evidence from the Gens or yενος.—There is no more puzzling topic in the history of the ancient world than the origin and nature of the community called by the Romans the gens, and by the Greeks the yενος. To the present writer it seems that no existing community of men, neither totem kin, nor clan, nor house community, nor gotra, precisely answers to the gens or the yενος. Our information about these forms of society is slight and confused. The most essential thing to notice for the moment is the fact that both in Greece and Rome the yενος and gens were extremely ancient, so ancient that the yενος was decaying in Greece when history begins, while in Rome we can distinctly see the rapid decadence and dissolution of the gens. In the Laws of the Twelve Tables, the gens is a powerful and respected corporation. In the time of Cicero the nature of the gens is a matter but dimly understood. Tacitus begins to be confused about the gentile nomenclature. In the Empire gentile law fades away. In Greece, especially at Athens, the early political reforms transferred power from the yενος to a purely local organisation, the Deme. The Greek of historical times did not announce his yενος in his name (as the Romans always did), but gave his own name, that of his father, and that of his deme. Thus we may infer that in Greek and Roman society the yενος and gens were dying, not growing, organisations. In very early times it is probable that foreign gentes were adopted en bloc into the Roman Commonwealth. Very probably, too, a great family, on entering the Roman bond, may have assumed, by a fiction, the character and name of a gens. But that Roman society in historical times, or that Greek society, could evolve a new gens or yενος in a normal natural way, seems excessively improbable.

Keeping in mind the antique and 'obsolescent' character of the gens and yενος, let us examine the theories of the origin of these associations. The Romans themselves knew very little about the matter. Cicero quotes the dictum of Scævola the Pontifex, according to which the gens consisted of all persons of the same gentile name who were not in any way disqualified.^[16] Thus, in America, or Australia, or Africa, all persons bearing the same totem name belong to that totem kin. Festus defines members of a gens as persons of the same stock and same family name. Varro says (in illustration of the relationships of words and cases) 'Ab Æmilio homines orti Æmilii sunt gentiles.' The two former definitions answer to the conception of a totem kin, which is united by its family name and belief in identity of origin. Varro adds the element, in the Roman gens, of common descent from one male ancestor. Such was the conception of the gens in historical times. It was in its way an association of kinsfolk, real or supposed. According to the Laws of the Twelve Tables the gentiles inherited the property of an intestate man without agnates, and had the custody of lunatics in the same circumstances. The gens had its own sacellum or chapel, and its own sacra or religious rites. The whole gens occasionally went into mourning when one of its members was unfortunate. It would be interesting if it could be shown that the sacra were usually examples of ancestor-worship, but the faint indications on the subject scarcely permit us to assert this.

On the whole, Sir Henry Maine strongly clings to the belief that the gens commonly had 'a real core of agnatic consanguinity from the very first.' But he justly recognises the principle of imitation, which induces men to copy any fashionable institution. Whatever the real origin of the gens, many gentes were probably copies based on the fiction of common ancestry.

On Sir Henry Maine's system, then, the gens rather proves the constant existence of recognised male descents among the peoples where it exists.

The opposite theory of the gens is that to which Mr. M'Lennan inclined. 'The composition and organisation of Greek and Roman tribes and commonwealths cannot well be explained except on the hypothesis that they resulted from the joint operation, in early times, of exogamy, and the system of kinship through females only.'^[17] 'The gens', he adds, 'was composed of all the persons in the tribe bearing the same name and accounted of the same stock. Were the gentes really of different stocks, as their names would imply and as the people believed? If so, how came clans of different stocks to be united in the same tribe? . . . How came a variety of such groups, of different stocks, to coalesce in a local tribe?' These questions, Mr. M'Lennan thought, could not be answered on the patriarchal hypothesis. His own theory, or rather his theory as understood by the present writer, may be stated thus. In the earliest times there were homogeneous groups, which became, totem kin. Let us say that, in a certain district, there were groups called woodpeckers, wolves, bears, suns, swine, each with its own little territory. These groups were exogamous, and derived the name through the mother. Thus, in course of time, when sun men married a wolf girl, and her children were wolves, there would be wolves in the territory of the suns, and thus each stock would be scattered through all the localities, just as we see in Australia and America. Let us suppose that (as certainly is occurring in Australia and America) paternal descent comes to be recognised in custom. This change will not surprise Sir Henry Maine, who admits that a system of male may alter, under stress of circumstances, to a system of female descents. In course of time, and as knowledge and common sense advance, the old superstition of descent from a woodpecker, a bear, a wolf, the sun, or what not, becomes untenable. A human name is assumed by the group which had called itself the woodpeckers or the wolves, or perhaps by a local tribe in which several of these stocks are included. Then a fictitious human ancestor is adopted, and perhaps even adored. Thus the wolves might call themselves Claudii, from their chief's name, and, giving up belief in descent from a wolf, might look back to a fancied ancestor named Claudius. The result of these changes will be that an exogamous totem kin, with female descent, has become a gens, with male kinship, and only the faintest trace of exogamy. An example of somewhat similar processes must have occurred in the Highland clans after the introduction of Christianity, when the chief's Christian name became the patronymic of the people who claimed kinship with him and owned his sway.

Are there any traces at all of totemism in what we know of the Roman gentes? Certainly the traces are very slight; perhaps they are only visible to the eye of the intrepid anthropologist. I give them for what they are worth, merely observing that they do tally, as far as they go, with the totemistic theory. The reader interested in the subject may consult the learned Streinnius's 'De Gentibus Romanis,' p. 104 (Aldus, Venice, 1591).

Among well-known savage totems none is more familiar than the sun. Men claim descent from the sun, call themselves by his name, and wear his effigy as a badge.^[18] Were there suns in Rome? The Aurelian gens is thus described on the authority of Festus Pompeius:—'The Aurelii were of Sabine descent. The Aurelii were so named from the sun (aurum, urere, the burning thing), because a place was set apart for them in which to pay adoration to the sun.' Here, at least, is an odd coincidence. Among other gentile names, the Fabii, Cornelii, Papirii, Pinarii, Cassii, are possibly connected with plants; while wild etymology may associate Porcii, Aquilii, and Valerii with swine and eagles. Pliny ('H. N.' xviii. 3) gives a fantastic explanation of the vegetable names of Roman gentes. We must remember that vegetable names are very common in American, Indian, African, and Australian totem kin. Of sun names the Natchez and the Incas of Peru are familiar examples. Turning from Rome to Greece, we find the yενος less regarded and more decadent than the gens. Yet, according to Grote (iii. 54) the yενος had—(l) sacra, 'in honour of the same god, supposed to be the primitive ancestor.' (2) A common burial-place. (3) Certain rights of succession to property. (4) Obligations of mutual help and defence. (5) Mutual rights and obligations to intermarry in certain cases. (6) Occasionally possession of common property.

Traces of the totem among the Greek yενη are, naturally, few. Almost all the known yενη bore patronymics derived from personal names. But it is not without significance that the Attic demes often adopted the names of obsolescent yενη, and that those names were, as Mr. Grote says, often 'derived from the plants and shrubs which grew in their neighbourhood.' We have already seen that at least one Attic yενος, the Ioxidæ, revered the plant from which they derived their lineage. One thing is certain, the totem names, and a common explanation of the totem names in Australia, correspond with the names and Mr. Grote's explanation of the names of the Attic demes. 'One origin of family names,' says Sir George Grey (ii. 228), 'frequently ascribed by the natives, is that they were derived from some vegetable or animal being common in the district which the family inhabited.' Some writers attempt to show that the Attic yενος was once exogamous and counted kin on the mother's side, by quoting the custom which permitted a man to marry his half-sister, the child of his father but not of his mother. They infer that this permission is a survival from the time when a man's father's children were not reckoned as his kindred, and when kinship was counted through mothers. Sir Henry Maine (p. 105) prefers M. Fustel De Coulanges' theory, that the marriage of half-brothers and sisters on the father's side was intended to save the portion of the girl to the family estate. Proof of this may be adduced from examination of all the recorded cases of such marriages in Athens. But the reason thus suggested would have equally justified marriage between brothers and sisters on both sides, and this was reckoned incest. A well-known line in Aristophanes shows how intense was Athenian feeling about the impiety of relations with a sister uterine.

On the whole, the evidence which we have adduced tends to establish some links between the ancient yενος and gens, and the totem kindreds of savages. The indications are not strong, but they all point in one direction. Considering the high civilisation of Rome and Greece at the very dawn of history—considering the strong natural bent of these peoples toward refinement—it is almost remarkable that even the slight testimonies we have been considering should have survived.

(5.) On the evidence from myth and legend we propose to lay little stress. But, as legends were not invented by anthropologists to prove a point, it is odd that the traditions of Athens, as preserved by Varro, speak of a time when names were derived from the mother, and when promiscuity prevailed. Marriage itself was instituted by Cecrops, the serpent, just as the lizard, in Australia, is credited with this useful invention.^[19] Similar legends among non-Aryan races, Chinese and Egyptian, are very common.

(6.) There remains the evidence of actual fact and custom among Aryan peoples. The Lycians, according to Herodotus, 'have this peculiar custom, wherein they resemble no other men, they derive their names from their mothers, and not from their fathers, and through mothers reckon their kin.' Status also was derived through the mothers.^[20] The old writer's opinion that the custom (so common in Australia, America, and Africa) was unique, is itself a proof of his good faith. Bachofen (p. 390) remarks that several Lycian inscriptions give the names of mothers only. Polybius attributes (assigning a fantastic reason) the same custom of counting kin through mothers to the Locrians.^[21] The British and Irish custom of deriving descents through women is well known,^[22] and a story is told to account for the practice. The pedigrees of the British kings show that most did not succeed to their fathers, and the various records of early Celtic morals go to prove that no other system of kinship than the maternal would have possessed any value, so uncertain was fatherhood. These are but hints of the prevalence of institutions which survived among Teutonic races in the importance attached to the relationship of a man's sister's son. Though no longer his legal heir, the sister's son was almost closer than any other kinsman.

We have now summarised and indicated the nature of the evidence which, on the whole, inclines us to the belief of Mr. M'Lennan rather than of Sir Henry Maine. The point to which all the testimony adduced converges, the explanation which most readily solves all the difficulties, is the explanation of Mr. M'Lennan. The Aryan races have very generally passed through the stage of scarcity of women, polyandry, absence of recognised male kinship, and recognition of kinship through women. What Sir Henry Maine admits as the exception, we are inclined to regard as having, in a very remote past, been the rule. No one kind of evidence—neither traces of marriage by capture, of exogamy, of totemism, of tradition, of noted fact among Lycians and Picts and Irish—would alone suffice to guide our opinion in this direction. But the cumulative force of the testimony strikes us as not inconsiderable, and it must be remembered that the testimony has not yet been assiduously collected.

Let us end by showing how this discussion illustrates the method of Folklore. We have found anomalies among Aryans. We have seen the gens an odd, decaying institution. We have seen Greek families claim descent from various animals, said to be Zeus in disguise. We have found them tracing kinship and deriving names from the mother. We have found stocks with animal and vegetable names. We have found half-brothers and sisters marrying. We have noted prohibition to marry anyone of the same family name. All these institutions are odd, anomalous, decaying things among Aryans, and the more civilised the Aryans the more they decay. All of them are living, active things among savages, and, far from being anomalous, are in precise harmony with savage notions of the world. Surely, then, where they seem decaying and anomalous, as among Aryans, these customs and laws are mouldering relics of ideas and practices natural and inevitable among savages.

Footnotes

1. Early Law and Custom.
2. Studies in Ancient History, p. 127.
3. Descent of Man, ii. 362.
4. Early Law and Custom, p. 210.
5. Here I would like to point out that Mr. M'Lennan's theory was not so hard and fast as his manner (that of a very assured believer in his own ideas) may lead some inquirers to suppose. Sir Henry Maine writes, that both Mr. Morgan and Mr. M'Lennan 'seem to me to think that human society went everywhere through the same series of changes, and Mr. M'Lennan, at any rate, expresses himself as if all those stages could be clearly discriminated from one another, and the close of one and the commencement of another announced with the distinctness of the clock-bell telling the end of the hour.' On the other hand, I remember Mr. M'Lennan's saying that, in his opinion, 'all manner of arrangements probably went on simultaneously in different places.' In Studies in Ancient History, p. 127, he expressly guards against the tendency 'to assume that the progress of the various races of men from savagery has been a uniform progress: that all the stages which any of them has gone through have been passed in their order by all.' Still more to the point is his remark on polyandry among the very early Greeks and other Aryans; 'it is quite consistent with my view that in all these quarters (Persia, Sparta, Troy, Lycia, Attica, Crete, &c.) monandry, and even the patria potestas, may have prevailed at points.'
6. Early Law and Custom, p. 212.
7. Studies in Ancient History, pp. 140-147.
8. Totem is the word generally given by travellers and interpreters for the family crests of the Red Indians.
9. Domestic Manners of the Chinese, i. 99.
10. Fortnightly Review, June 1, 1877.
11. Kamilaroi and Kurnai. Natives call these objects their kin, 'of one flesh' with them.
12. Studies, p. 11.
13. O'Curry, Manners of Ancient Irish, l. ccclxx., quoting Trin. Coll. Dublin MS.
14. See also Elton's Origins of English History, pp. 299-301.
15. Kemble's Saxons in England, p. 258. Politics of Aristotle, Bolland and Lang, p. 99. Mr. Grant Allen kindly supplied me some time ago with a list of animal and vegetable names preserved in the titles of ancient English village settlements. Among them are: ash, birch, bear (as among the Iroquois), oak, buck, fir, fern, sun, wolf, thorn, goat, horse, salmon (the trout is a totem in America), swan (familiar in Australia), and others.
16. 'Gentiles sunt qui inter se eodem nomine sunt. Qui ab ingeniis oriundi sunt. Quorum majorum nemo servitutem servivit. Qui capite non sunt deminuti.'
17. Studies in Ancient History, p. 212.
18. Fortnightly Review, October 1869: 'Archæologia Americana,' ii. 113.
19. Suidas, 3102.
20. Herod., i. 173.
21. Cf. Bachofen, p. 309.
22. Compare the Irish Nennius, p. 127.