As a science, Australian philology does not exist. In 1880 Mr. Fison wrote, "It is simply impossible to ascertain the exact meaning of these words" (changes of name and grade conferred at secret ceremonies), "without a very full knowledge of the native dialects," and without strong personal influence with the blacks.... "In all probability there are not half-a-dozen men so qualified in the whole Australian continent."[1]
The habit of using, in the case of the initiate, mystic terms even for the everyday names of animals, greatly complicates the problem. It does not appear that most of the recorders of the facts know even one native dialect as Dr. Walter Roth knows some dialects of North-West Central Queensland. In the south-east, Kamilaroi was seriously studied, long ago, by Mr. Threlkeld and Mr. Ridley, who wrote tracts in that language. Sir George Grey and Mr. Matthews, with many others, have compiled vocabularies, the result of studies of their own, and Mr. Curr collected brief glossaries of very many tribes, by aid of correspondents without linguistic training.
Into this ignorance as to the meanings of the names of matrimonial classes, Mr. Howitt brings a faint little gleam of light In a few cases, he thinks, the meaning of class and "sub-class" names is ascertained. Among the Kuinmurbura tribe, between Broad Sound and Shoal water Bay, the "sub-classes" (our "matrimonial classes") "were totems." By this Mr. Howitt obviously means that the classes bore animal names. They meant (i.) the Barrimundi, (ii.) a Hawk, (iii.) Good Water, and (iv.) Iguana.[2] For the Annan River tribe, he gives "sub-classes" (our "matrimonial classes"), (i.) Eagle Hawk, (ii.) Bee, (iii.) Salt-Water-Eagle Hawk, (iv.) Bee.[3] This is not very satisfactory. In previous works he gave so many animal names for his "sub-classes," Mr. Frazer's "sub-phratries" (our "matrimonial classes"), that Mr. Frazer wrote, "It seems to follow that the sub-phratries of the Kamilaroi (Muri, Kubi, Ipai, and Kumbo) have, or once had, totems also," that is, had names derived from animals or other objects.[4]
Mr. Howitt himself at one time appeared to hold that the names of the matrimonial classes are often animal names. His phraseology here is not very lucid. "The main sections themselves are frequently, probably always, distinguished by totems." Here he certainly means that the phratries have usually animal names, though we are not told that the phratries, as such, treat their name-giving animal, even when they know the meaning of its name, "with the decencies of a totem." Mr. Howitt goes on, "The probability is that they are all" (that all the classes are) "totems."[5] By this Mr. Howitt perhaps intends to say that all the "classes" (both the phratries and the matrimonial classes) probably have animal or other such names.
Again, the class names of the Kiabara tribe were said to denote four animals—Turtle, Bat, Carpet Snake, Cat.[6] But now (1904) the Kiabara class names are given without translation, and the four animals are thrown into the list of totems, with Flood Water and Lightning totems (which names were previously given as translations of Kubatine and Dilebi, the phratry names).[7] Doubtless Mr. Howitt has received more recent information, but, if we accept what he now gives us, the meanings of his "sub-class" names are only ascertained in the cases of two tribes, and then are names of animals.
I spent some labour in examining the class names of the tribes studied by Messrs. Spencer and Gillen, from the Arunta in the centre to the Tingilli at Powell's Creek, after which point our authors no longer marched due north, but turned east, at a right angle, reaching the sea, and the Binbinga, the Mara, and Anula coast tribes, on or near the MacArthur River. The class names of these coastal tribes did not resemble those of the central tribes. But if Messrs. Spencer and Gillen had held north by west, in place of turning due east from Newcastle Waters, they would have found, as far as the sea at Nichol Bay, four classes whose names closely resemble the class names of the central tribes, and are reported as Paljarie, or Paliali, or Palyeery (clearly the Umbaia and Binbinga Paliarinji), Kimera or Kymurra, (obviously Kumara), Banigher, or Bunaka, or Panaka (Panunga, cf. Dieri Kanunka = Bush Wallaby),[8] and Boorungo, or Paronga.[9]
It thus appears scarcely doubtful that, from the Arunta in the centre, to the furthest north, several of the class names are of the same linguistic origin, and—whether by original community of speech, or by dint of borrowing—had once the same significance. Now we can show that some of these names, in the dialects of one tribe or another, denote objects in nature. Thus Warramunga Tj-upila' (Tj being an affix) at least suggests the Dieri totem, Upala, "Cloud." Biliarinthu, in the same way, suggests the Barinji Biliari, "Eagle Hawk," or the Umbaia Paliarinji. Ungalla, or Thungalla, is Arunta Ungilla, "Crow," the Ungōla, or Ungăla, "Crow" of the Yaroinga and Undekerabina of North-West Queensland,[10] while Panunga, Banaka, Panaka, resembles Dieri Kanunka—"Bush Wallaby," or Kanunga, "Kangaroo Rat."
The process of picking out animal names in one tribe corresponding to class names in other tribes, is not so utterly unscientific as it may seem, for the tribes have either borrowed the names from each other, or have a common basis of language, and some forms of dialectical change are obvious. We lay no stress on the "equations" given above, but merely offer the suggestion that class names have often been animal names, and hint that inquiry should keep this idea in mind.
I do not, then, offer my "equations" as more than guesses in a field peculiarly perilous. The word which means "fire" in one tribe, means "snake" in another. "What fools these fellows are, they call 'fire' 'snakes,'" say the tribesmen. However, if we guess right, we find Eagle Hawk, Crow, Cloud, and Kangaroo Rat, as class names, over an enormous extent of Central and Northern Australia.[11]
About the deliberate purpose of the classes there can be no doubt. They were introduced to bar marriages, not between parents and children, for these are forbidden in primitive tribes, but between persons of the parental and filial generations. Or the names were given to stereotype classes, already existing, but hitherto anonymous, within which marriage was already prohibited. To make the distinction permanent, it was only necessary to have a linked pair of classes of different names in each phratry, the child never taking the maternal class name, but always that of the linked class in her phratry (under a system of female descent). The names Red, Blue, Green, Yellow, would have served the turn as well as any others. If a tribe had two words for young, and two for old, these would have served the turn; as