Erich M. von Hornbostel *

American Anthropologist New Series, Vol. 38 No. 3, July-September, 1936.

Of Fuegian native music scarcely anything was known until recently, except a few poor notes scattered in ethnological literature, (1) and half a dozen musical examples recorded by ear and hence of doubtful reliability. (2) Colonel C. Wellington Furlong, on his visit to Tierra del Fuego in 1907-1908, was the first to take phonographic records of Ona and Yahgan songs, (3) copies of which were sent to the Berlin Archives and studied by the present writer. (4) The publication of this study, however, has been delayed first from outward reasons, then deliberately, the material available having been increased by a great number of records taken among Ona, Yahgan, and Alakaluf (5) by Professors M. Gusinde and W. Koppers in 1922-1923. Thus the conclusions already arrived at could be founded on a much broader basis. Considering the special interest of the Fuegians as preserving one of the most primitive culture types and, moreover, the frightful speed of their disappearance, the value of the phonographic documents collected in the eleventh hour cannot easily be overrated. In what follows I must content myself with outlining the main facts and conclusions concerning anthropology, and for a fuller discussion may refer to my contribution to Gusinde's comprehensive work on the Fuegians. (6) Colonel Furlong's singers were: among the Ona, two women, Ichjh (Phonograph Record 12) and the "doctor" Yoyo (5), both in Rio Fuego, and two men, Ishtone in Najmish (Via Monte) (3) and Tininisk, a medicine man in Haberton (9, 11); among the Yahgan, in Lauwi (Punta Remolino) the women Simoorwhilliss (Alice) and Weemanahakeepca (Gertie) (1), and the men Danushtana (Alfredo), Nyempenan (David) (13, 6), with Aselensinjiz (Charley), Wuroyinjez (William), and Calderon (13), or Chris (6) joining in; and, in Rio Douglas, Navarin Island, Edward (? 2, 7). (7). (7)

In the music printed here a few additional signs have been used: + sharpens,- flattens the note up to a quartertone, constant deviations throughout a song being indicated in the beginning. (8) A heavy line between two noteheads indicates glissando. Headless notes and notes in parentheses are indistinct, parts in square brackets are variants, shortens the note; bar-lines divide into groups, not into bars; (9) V marks a breathing space.

The women's songs were apparently distinguished neither as a partic ular category, nor musically, from the men's songs. Duties and rights, with the Fuegians, are almost equally distributed between the sexes. Thus, female shamans, as with some Siberian peoples, are not uncommon, and of course they will use the same kind of chants, and for the same magic purposes, as the medicine men (P. R. 5, 9a, 11a). (10) The Yahgan initiation ceremony is gone through by both boys and girls; consequently both know all the songs sung at this occasion mostly to accompany pantomimic (animal) dances (P.R. 1, 13, 6). (11) Like the Eskimo, the Ona medicine men sing invective songs in order to deride each other (P. R. 9b, 11b). Calls and answers voiced with a musical tone, i.e., rather sung than spoken, are used as signals by the Yahgan (P. R. 7c). The same custom has been reported from the primitive Kubu in the forests of Sumatra. (12) Somewhat similar in character to a call is the ff-phrase preceding and following the song proper in Musical Example 14. Again, such a framing, which very probably has (or originally had) a magical meaning, frequently occurs in native songs: cf., e.g., the stereotyped formulae of the Kubu (beginning and end), Uitoto Indians in Colombia (end flourish), (13) and Pangwe in the Cameroons (introduction). (l4) Realistic imitations of birds' cries -uttered by the bystanders, not the singers, during Yahgan pantomimic dance songs- bear no relation whatever, not even rhythmically, to the music, but testify to the mimic talent of these people. (15) (Cf. the coyote bark frequently added to the end of a song by North American Indians, e. g., the Pawnee.)

All Fuegian songs -as has likewise been ascertained for the Tehuelche (16)- lack (meaningful) words. The meaningless syllables, however, being unchangeably associated and handed down with the tunes, must be considered as survivals of original words which for one reason or another ceased to be understood, and then were liable to distortion. Such a process is evidenced by many examples From primitive as well as from civilized peoples. (17)

The songs -as in fact, the whole culture- of the Fuegians can be said to be characteristically American Indian and at the same time extremely primitive; the first statement applying more particularly to the Ona, the second to the Yahgan and Alakaluf.

The most conspicuous factor contributing to the character of Indian music is the manner of singing. It will be readily apprehended when hearing Indian singers (or records); but it is almost impossible to analyze the immediate impression and to convey to someone else a clear notion by enumerating its "elements." Bearing this in mind, we may describe Indian singing by such epithets as emphatic, pathetic, impressive, grave, solemn, dignified, weighty, stern, etc., (which would also apply equally to the Indian's dancing, general motor behavior, and temper). Among the features responsible for this impression will be found strong accents, often further increased by audible expiration, on almost every crochet; a tendency to connect the notes by a portato and to subdivide lengthened notes by pulsations; the time being moderate or rather slow and remaining constant throughout the song. Thus, the characteristic Indian movement is never sprawling or jumping (as with African Negroes), nor gliding or softly waving to and fro (as frequently in Oceania), but rather striding in a heavy regular pace. Perhaps the most important knowledge to be derived from Fuegian songs is the fact that the "emphatic" singing manner, though as strongly conspicuous among the Ona as among any other Indians, is entirely absent among the canoe-faring tribes, the Yahgan and Alakaluf.

Another basic feature hardly to be separated; from the one already mentioned, is the Indian's rhythmic feeling, which is strongly contrasting with our own. Our normal rhythmic groups are iambics, rising from a light (unaccented, short) beat to the heaviest accent of the group (which may be lengthened). In our usual notation the two beats are separated by the bar-line and the notes preceding the bar-lines are called "upbeat" (arsis). Indian rhythm, on the contrary, may be described as lacking an upbeat. The typical group is a "dynamic trochee." It begins with the strongest beat -which is generally shortened and followed by a breathing space, its strength thus being further increased- and then drops off through a series of, as it were, rebounding" accents. Though the length of this series is variable, even in one and the same song (cf., e.g., M. E. 6, 13), the second pair of a group is, as a rule, longer than the first: triplets are commonly of the 1+2 type (M. E. 1, 8, 10), and analogously four beats are grouped into 1+3 (M. E. 6), five into 1+4 (M. E. 11, beginning), etc. The same principle governs the subdivisions and, naturally, the structure of the whole song. In M. E. 9 the Western hearer will feel the phrases as consisting of three 6/8 bars, i. e., groups of 3+3; whereas they are almost certainly meant as 1+2 groups of 2+ (1+3) quavers each, the three pulsations on the long note again being a 1+2 grouplet.

The characteristic type of-rhythm here described prevails in Yahgan and Alakaluf as well as in Ona music and, in fact, in any Indian music from North or South America. Deep rooted in general motor behavior as it is, it reappears in the most common Indian dance movement: one foot, after a wide step aside, is stamped on the ground; then the other foot is drawn after and stamped down, though less heavily, near the first one; a short rest may be filled with treading on the spot, then the movement is repeated.

A characteristic of Fuegian music is its narrowness; this appears both in tonal range and length of melodic units. The backbone of the melodic structure is a pair of notes, limiting a distance which rarely exceeds a whole tone (M. E. 1, 3b) and in many cases is confined to a semitone or less (M. E. 8, 4, 14), or is merely the result of fluctuations of a single tone, following the fluctuations in dynamic stress. (18)

Sometimes the whole song is limited to the principal "second" (M. E. 1). A third tone, however, is frequently added when the higher principal tone is sharpened by an accent (M. E. 3aï; 4, 8, 11, 12 in the beginning), or when the lower principal tone is further dropped on a weak beat (M. E. 10b, 12, 7c). Moreover, a two-tone motive may be shifted downwards, its lower note becoming the higher one in the repetition (M. E. 4, 11, 14). The range thus being enlarged, consonant intervals are aimed at as a frame to the melody, viz., a fourth (M. E. 10, 11, 12, 13; 3, 5, 6, 2), or, with the Ona, a fifth (M. E. 7, 8) or even an octave (M. E. 9). In Ona songs the fourth and fifth also occur as melodic progressions (M. E. 2, 6, 7, 9), and in one instance (M. E. 8) the motive is transposed downward by a fifth. (l9)

By such means Ona melody escapes the primitive narrowness and approaches the Indian type. Usually an Indian melody starts on the highest note with utmost strength, descends by shifting the motive from step to step -hence this type has been styled "stair-pattern melody"- at the same time decreasing in loudness, and finally dies away on the lowest note, which in some cases is more than two octaves below the initial pitch; the secondly verse repeats the first one, starting from the next to the highest level, and so forth, the last verse being confined to pulsations on the lowest note. Approximations to this form may be found in the Ona songs (M. E. 7, 8, 9).

There are several features in Fuegian songs, on the other hand, which offer exceptions to the exclusively downward trend of Indian melody. Frequently, and with the Yahgan almost as a rule, motives or whole melodies end with a rise to the next to the lowest note (M. E. 11; 12; 13; 14, V, VII; 15a, b; lb; 4b2; 9a, c). A similar rise sometimes occurs in the beginning (M. E. 2; 6; 15d). Even the level, in strict opposition to Indian style, may be gradually shifted upwards from verse to verse. (20)

To summarize: Indian musical style is easily to be recognized by (1) its "dynamic trochee" rhythm which, contrary to our Western tendencies, avoids the upbeat; (2) a peculiar "emphatic" manner of singing which results from such factors as a certain voice-quality, strong accents on every time-unit, pulsation, slow and constant time; (3) a downward step-by-step shift of the main theme (stair-pattern) and a collapsing tendency of the melody, which from start to end continually decreases in pitch, intensity, and tonal range. This style prevails among the Indians of both Americas, including the Eskimo (also in Greenland), and among Siberian tribes who are related to the Indians, both somatically and culturally, as, e.g., the "Palaeo-asiatic" Chukchee and the Keto (Ostyak) on the Jenissei River, and among the semi-Tungus Orotchee on the lower Amur River, and in Korean folk-songs.

On the other hand, there are exceptions to the Indian style even in America. The Yahgan and Alakaluf, though conforming to the Indians in their rhythmic and general motor behavior, differ from them in their manner of singing which entirely lacks the Indian characteristics, and further in their melodies by several "primitive" features, viz., extreme narrowness, simplicity, and shortness of themes, a tendency to rise (or return) to an upper level contrary to the general downward trend. The same archaic traits, including non-emphatic singing manner, have been found in some other places in America, though they are buried under more recent layers of Indian culture and hence less frequent and less clearly pronounced. As the Ona in some of their songs approach the style of the canoe-faring tribes in narrowness and melodic rise -but not in singing manner- so do, if to a still lesser degree, their continental relatives, the Tehuelche. (21) As with the Fuegians, so equally with the Tehuelche the words sung have lost their meaning. (22) Again, songs of the primitive "Fuegian" type have been reported from the Uitoto in the virgin forest on the Colombian-Peruvian frontier. (23) (The old stratum to which this musical style belongs may be detected beneath still other South American cultures, e.g., that of the Botocudo and other tribes of the Geez stock, from whom unfortunately phonographic records are not yet available.)

In North America the primitive musical style survives in songs from Southern California Indians and tribes of the Yuma group closely related to them. (24) Their voices and singing manner lack the characteristic "Indian" quality; the "rise" in melody is even more apparent than with the Fuegians and has developed into various forms; though the tonal range is usually not very small, traces of original narrowness may be found in the fact that the two prominent notes are only a second apart from each other, and of about equal importance but for the higher note being occasionally stressed as a final goal. In many of the songs the meaning of the words is obsolete. The relation between Californian and Fuegian cultures is further confirmed by parallels in the realm of religious, social, and economic life. (25)

Considering, on the other hand, the uniformity of the Indian musical style extending over the whole area from Magellan Straits to the Arctic sea and from the east shore of Greenland to the Jenissei; and on the other hand, the strong contrast to this style shown by songs of the Yahgan and Alakaluf, we may feel inclined to distinguish the latter tribes culturally, if not somatically as belonging to a separate pre-Indian group. (26) These people then would have been literally the forerunners of the real Indians' immigration into the American continent. Here they were not only driven to the remotest borders, or places difficult of access, but in the course of time naturally have been subject to influence from their less primitive neighbors. Thus the Yahgan very probably have borrowed the mask plays, that are kept secret from women, and the medicine men's organization and ceremonial meetings from the Ona, who in turn apparently are not the originators but merely mediators of these customs, at least of the first.

Outside America, the primitive pre-Indian musical style is closely paralleled by that of the Vedda (27) and the Andaman Islanders. (28) Their melodies are extremely narrow, frequently rising at the end of the lines; the number of time-units varies arbitrarily (or according to the number of syllables) in almost every repetition of one single phrase, though the time remains constant. (The Andaman Islanders also sing in octaves and fiiths instead of in unison, as sometimes do the Yahgan.)

The well-known fact that in really primitive cultures sound instruments are almost completely absent, though being of no great importance in itself, enhances the purport of those rare parallels which may be found in this realm. In their death dance, as witnessed by Colonel Furlong, (29) the Yamana women were tamping the ground with pairs of long thick posts (which may have been substituted for oars). Sachs, (30) quoting Playfair, has drawn attention to the similar custom of the Garo in Assam, where the women, when watching the dead, accompany their songs by beating or tamping on the floor with a piece of wood. Likewise among the Kurnai, one of the most primitive southeastern Australian tribes, in a ceremony preliminary to the initiation, the mothers of the novices keep time by tamping their yam-sticks on the ground. (31) (Note that in every instance it is the women who use the tamping sticks.) In their mask plays the Ona and (hence) the Yahgan make frightful noises by beating the ground with rolled-up hides in order to stage the raging of a wrathful spirit of the earth. (32) Again the same instrument is used by southeast Australian women, though not as a beater, but rather as a drum. (33) The Ona custom is further paralleled by southeast Australian men, when (in certain dances) they beat small mounds of earth or simply the ground with strips of bark. (34) Still more striking a coincidence is the belief, common to both southeast Australians and Fuegians, that the medicine men learn songs and spells in dreams from one of their defunct relatives or from other spirits. (35) (With both peoples in olden times the medicine men, or rather the spirits "talking in very curious voices" through the mouth of the shaman, used to inform their tribesmen of distant events, e g., of a whale stranded on the shore. (36)) As with the Yahgan, so likewise with the southeast Australians, the initiation ceremonies include pantomimes and dances in which the behavior and cries of (totem) birds and animals are realistically imitated. (37) Unfortunately no phonographic records were made of southeast Australian (and Tasmanian) songs before the curtain dropped forever on the history of these natives. Yet, from the competent remarks and notations of Wurunjerri songs contributed to A W. Howitt's work by Dr Torrance, (38) we may at least feel sure that southeast Australian melody was of a primitive narrow type. Every verse begins with a downward glide through a whole tone (or less) and continues, with strong accents on every beat, throughout on the lower note thus reached, until, at the beginning of the second verse, the level is shifted one step downward. Usually the song being then repeated, its entire compass does not exceed a (minor) third. In other songs there occur one or more downward shifts with every repetition yet this variant of the "stair-pattern"-which is used incidentally by the Papuans of Torres Straits in their ancient ceremonial songs (39)- may be a more recent importation from the north This seems the more probable as the songs of the central and west Australians (40) (whose culture is not primitive) actually belong to the "stair-pattern" type, and -but for the singing manner- so closely resemble American Indian songs, that, e.g., an Arunta song may easily be mistaken for, say, a Pawnee song, even by an expert. Thus the situation in Australia and South America as regards musical culture appears to be exactly homologous, and this fits well in the cluster of remarkable correspondences in cultural details, on which Professor Koppers (41) has based his hypothesis of a common origin of the southeast Australian and Andamanese cultures on the one hand, and Fuegian and Californian cultures on the other. According to such an hypothesis, the forefathers of the primitive tribes (Fuegians, Californians, southeast Australians and Tasmanians, Andamanese, Vedda) would have lived as neighbors somewhere in Asia in very remote times, and from there would have migrated under pressure of more advanced tribes (American Indians, Australians, Papuans, etc.) on divergent lines, until they reached their present habitats.

CAMBRIDGE, ENGLAND