CTLF Corpus de textes linguistiques fondamentaux • IMPRIMER • RETOUR ÉCRAN
CTLF - Menu général - Textes

Boas, Franz. Race, Language and Culture – T56


Decorative Designs of Alaskan Needlecases:
a Study in the History of
Conventional Designs, based
on Materials in the U. S.
National Museum 11

In 1878 and the following years, Professor F. W. Putnam 22 described
in detail the decorative designs found in the pottery of the Chiriqui
Indians, and was the first, I believe, to propound clearly the theory
that conventional designs develop from attempts at realistic representations,
which gradually degenerate so that ultimately a purely conventional
design remains in which the realistic origin can hardly be
recognized. Since that time this theory has been independently stated
by a number of investigators, particularly by H. Stolpe 33 and H. Balfour. 44
It has been applied extensively to explanations of primitive
ornamental art. The most noteworthy contributions on this subject are
those by Karl von den Steinen, 55 on the art of the Brazilian Indians,
and by A. C. Haddon, 66 on the art of the natives of New Guinea.

Opposed to this view has been the theory propounded by Semper,
who emphasizes the influence of material upon the development of the
design, and that proposed by Cushing and Holmes, 77 who emphasize
the importance of technique upon the development of geometrical
design. More recently Karl von den Steinen 88 has also emphasized the
importance of technical conditions upon the development of design,
and his arguments have been followed and elaborated by Max Schmidt
564in discussions of South American designs. 19 Th. Koch-Grünberg 210 follows
in the same line of argument, showing that at least in Brazil a
considerable number of cases may be found in which designs that have
developed from technical motives receive a realistic significance.

From a wider point of view, the secondary development of motives
and their re-interpretation as realistic designs have been claimed by
Heinrich Schurtz 311 and by Professor Hamlin 412 in a discussion of the
development of architectural decorative forms. The secondary character
of symbolic interpretation has also been set forth by A. L. Kroeber, 513
Clark Wissler, 614 and by myself. 715

We have therefore at the present time three distinct theories regarding
the development of decorative design: First, the theory of the
realistic origin of conventional motives; second, that of the technical
origin of conventional motives; and, third, the theory that the explanations
of conventional motives are essentially secondary in character,
and due to a later association of the existing decorative forms with
realistic forms.

I shall discuss in the following pages the decorative designs of Alaskan
needlecases, largely from the region between the mouth of the Yukon
River and the western part of Norton Sound, which seem to throw
considerable light upon the history of decorative design, and illustrate
the applicability of these various theories.

Among the carvings of Alaskan Eskimo we find a very large number
of needlecases of peculiar form. They are of the characteristic tubular
type of the Eskimo needlecase, in which the needle is inserted in a
strip of skin pulled into a tube, which protects the needle against
breakage. The peculiar type to which I here refer has, on the whole,
a tube slightly bulging in the middle, and expanding into two wings
or flanges at the upper end. It is characteristic of almost all these
specimens that at a short distance below the flanges there are two
small knobs on opposite sides of the tube. In some cases these are
well marked, while in others they are so diminutive that they cannot
565be seen at all, although they can be felt when moving the finger gently
over the surface of the tube. They are a characteristic feature of this
type, which is so well defined, and whose distribution is so restricted
that there cannot be the slightest doubt as to the unity of its origin.

These needlecases have also a characteristic decoration. On the
whole there is a tendency to set off a slightly concave surface, which
extends along the faces of the tube, between the flanges and farther
down. This concave face may be observed on the needlecases shown on
figures 1 and 2. The flanges and the upper border of the tube are


Fig. 1.

generally decorated by a design consisting of a number of parallel
lines, which is repeated near the lower end of the flanges, where the
parallel lines almost always slope slightly downward towards the tube.
Similar line designs are also found on the concave face of the tube. In
many cases these lines meet the lower lines on the flanges at an angle,
being incised so that they slope downward from the middle line of the
tube outward. In other cases they continue in the same direction as
the lines on the flanges (figs. 1 b; 2 a, b). Many of the needlecases are
so much polished and rubbed off by use that the design lines cannot
be recognized distinctly. In other cases broken ends have been cut off
(fig. 1 b), with the result that some of the characteristic decorative
566traits have become obscure. It would seem, however, that in all the
better specimens of this simple type the central concave face of the
needlecase is set off by two parallel incised lines, which extend downward
to about the middle of the tubing, and which end at this place
in two or three small spurs. The border designs on the flanges do not
extend beyond these lines inward, so that they form the limits of the
flanges. The parallel lines near the lower border of the flanges also
generally end at the vertical lines.

Another characteristic decorative design of these needlecases is a
narrow band extending around the lower end. This consists always,


Fig. 2.


Fig. 3.

wherever it can be distinctly recognized, of two parallel lines with
short alternating spurs directed toward the space between the two lines.
Whenever these spurs are given a greater width this design assumes more
or less the form of a zigzag band. A study of a considerable number
shows clearly that the primary idea is not the zigzag band, but rather
the two lines with alternating spurs. This is best shown by the fact that
in those cases in which the lines are thin the alternation is often quite
irregular. This may be observed, for instance, in the specimen shown
in figure 3. On the whole, however, an alternation is observed. Bands
of this kind may be recognized clearly in figures 1 a, c, d. Sometimes
567the band at the lower end appears doubled, or elaborated by the addition
of short vertical lines with short spurs at their ends (figs, 1c, 2 b, 4).
These lines are usually four in number. In the specimen shown in figure
5 a, two of these lines are absent, because their space is occupied by a
long alternate-spur band which runs down the whole side of the needlecase.
In another specimen one is absent, probably because the ivory
at the place where it would be shows the soft inner part of the tusk, and
has besides other defects. In figure 5 b, there is one of these lines on each
side of the needlecase. In one specimen in the Ethnographical Museum
of Berlin (IV A 5892) the number of these lines is more than four.


Fig. 4.


Fig. 5.

A partial doubling of the spur band may be observed on figs. 3 and
5 a.

The features here enumerated comprise those of the most generalized
type of these needlecases. They may be briefly summed up as (1) a tube
slightly bulging in the middle, (2) flanges at the upper end, (3) small
knobs under the flanges, (4) long concave faces on opposite sides at the
upper end of the tube, (5) long parallel lines with small forks at their
lower ends setting off the concave faces, (6) border designs consisting
of lines at the upper and lower ends of the flanges and on the concave
face, and (7) an alternate-spur band at the lower end of the tube.568

In order to understand the significance of this peculiar type of needlecase,
we must bear in mind that the two design elements which are most
characteristic of these specimens — namely, the line design with short
branches and the alternate-spur design — are characteristic Eskimo motives
over the greater part of the Arctic coast. The alternate spur band
design is found on a number of very old specimens from Southampton
Island and Lyons Inlet, collected by Capt. G. Comer, which are reproduced
here in fig. 6 a and c. In the same region the forked-line design
is found on bone engravings (fig. 6 b). It may be observed in a few of
the specimens found by Parry in Fury and Hecla Straits in 1820 and on


Fig. 6. a, Ivory attachment to line, west coast of Hudson Bay. b, Creaser, Iglulik.
c, Design of needlecase, King William Land.

an ancient pair of ivory snow-goggles from Cumberland Sound (Ethn.
Museum Berlin IV A 6833). It is also commonly found in tattooings.
It occurs in the tattooings from the west coast of Hudson Bay, as well as
in those from Baffin Land (fig. 7). Unfortunately I have not had opportunity
to examine extensive collections from Greenland, in order to
ascertain the occurrence of these designs. In view of their wide distribution
over the whole Eskimo area, it seems justifiable to consider them as
a very old possession of the Eskimo, and to assume that originally they
bore no relation to the needlecases on which they are found with such
great regularity. Incidentally it may be remarked that the explanations
of these forms as bushes and whales' tails, which are given by the
569Alaskan Eskimo, appear so one-sided that they cannot be accepted as a
general interpretation.

The designs here mentioned do not seem to occur in parts of America
or Asia which are outside of Eskimo influence. I have not been able to
discover them on any objects of Indian manufacture except on a few
specimens from the Yukon River made by Athapascan tribes directly
under Eskimo influence. In Asia similar designs occur among the
Koryak and Chukchee (fig. 8), while farther to the west and south I
have not been able to find them. I am not certain whether the alternate-spur-line
design does occur in the art of the Samoyed, but I have not
discovered a single example in a large collection of Yakut specimens


Fig. 7. Tattooings from the west coast of Hudson Bay
and Hudson Strait.


Fig. 8. Ear-spoon,

brought together by Mr. Jochelson; and it does not seem to occur
among the Gilyak, Ainu, and southeastern Tungus tribes. It seems that
the design occurs occasionally in Polynesian and Micronesian art, but
I should not venture to conclude from this an historical relation, notwithstanding
the rather large number of peculiar analogies between the
northeast coast of the Pacific Ocean and the islands northeast of Australia.

Considering the continuous area in which the two designs occur, we
may say that their essential home seems to be the Eskimo region, beginning
with Alaska, and extending eastward and northeastward to Hudson
Bay and Smith Sound, and that a few of the neighboring Indian
570tribes may have adopted them, and that they also occur among the
neighboring Chukchee and Koryak.

One needlecase that has been found in the region of Southampton
Island seems to me of particular importance in this connection (fig.
11 a). It will be seen that it also consists of a tube, like most Eskimo
needlecases; that it expands widely near its upper end, the whole tube
being flattened; and that near the middle there are two large wings,
which correspond in their position to the small knobs of the Alaskan
needlecases. This specimen has also the characteristic alternate-spur
band of the Alaskan needlecases at its upper end, and the decoration is


Fig. 9.


Fig. 10.

repeated here in two parallel lines. Attention may be called to the occurrence
of the same pattern at the same place in a number of the more
complex specimens from Alaska, shown in figures 4 b, 5 b, 9. These and
other similar occurrences show that the Eskimo often substituted this
design for the single parallel lines.

The alternate-spur-band design is releated to the single spurred line, a
pattern which is very common in many parts of the world. In the decorative
art of the Eskimo it appears often in place of the alternate-spur-band;
for instance, on some needlecases of the type here discussed (see
figs. 2 c, 10). In other cases the alternate-spur band is replaced by a
571ladder design (fig. 15 a) which on account of its rarity, may be considered
as a degenerate form of the alternate-spur band.

A group of needlecases similar to the one just described from Southampton
Island has been found in the district between Southampton
Island and Smith Sound (figs. 11, 12). The only type of needlecase
known from Smith Sound has this peculiar character. Unfortunately
the specimens which I have seen are all exceedingly rough; but they
all consist of a flattened tube, very wide at the upper end, and small


Fig. 11. Needlecases. a, b, d, e, f, Frozen
Strait, c, Pond's Bay.


Fig. 12. Needlecases. a, Smith
Sound, b, Rawlings Bay, west coast
of Smith Sound.

and round at the lower end, provided at the sides with two characteristic
wings (fig. 12 a). The same type with some dot decorations has been
collected at Ponds Bay in the northern part of Baffin Land (fig. 11 c),
while the older specimens from the northern part of Hudson Bay are
much more elongated, and have the wings and flanges set off more
clearly from the body of the needlecase (fig. 11 d, e).

A specimen in the British Museum from Rawlings Bay on the West
572coast of Smith Sound is of similar type (fig. 12 b). 116 It proves that this
type extended from Hudson Bay northward over Grinnell Land to
north Greenland.

It seems plausible that the Alaskan type and the Eastern type represent
specialized developments of the same older type of needlecase, and
that the flanges and diminutive knobs of the Alaskan specimens are
homologous to the flanges and large wings of the Eastern specimens.
When the first specimens of this kind were collected, Prof. O. T. Mason,
according to information which he has kindly given to me, was inclined
to believe that they were of foreign origin. In a note on the specimen
shown in fig. 9 b, he wrote at that time:

“This specimen is a needlecase from St. Michael, Alaska. It is made of
walrus ivory and carved in a form which suggests the butt end of an arrow,
with two feathers projecting from opposite sides on the shaft. The likeness
is made more striking by the fluting on the butt end, which resembles the
nock of the arrow. A little in front of the two feathers are projecting bosses.
The tube of the needlecase is slightly expanded in the middle and contracted
at the smaller end. The ornamentation consists of narrow bands
across the shaft, and the feathers at their extremities cut out in zigzag line
very much in the style of Polynesian ornamentation. At the smaller end
there is also a similarly ornamented band from which rise four symbols of
shrubs. An exactly similar piece is figured in N. A. E. Nordenskiöld 217 and
labeled “knife handle from Port Clarence.” There are four of these objects
in the United States National Museum, and, compared with hundreds of
others, they place themselves unmistakably in the class of needlecases.
There is no doubt that these six specimens — five in the United States National
Museum and one shown by Nordenskiöld — are not aboriginal in form
or ornament; that they belong to a style of art introduced into Alaska after
the advent of the Russians.

In Seebohm 318 will be seen the figure of a Samoyed needlecase with a
tube of metal, inclosed at its top in a belt, and riveted along the side. The
suggestion is here thrown out that the Eskimo artist has endeavored to
reproduce in ivory a facsimile of this metal tube and a portion of the
leather belt, even to the projecting rivets. The Nordenskiöld specimen has,
in addition, walrus heads and seals carved on the side of the tube.

This Polynesian style of ornamentation is common on hundreds of
Eskimo objects in and about St. Michael; for, after the advent of the Russians
and intercourse with sailors of the Pacific Ocean, the arts of the two
areas became very much entangled.”573

Considering the antiquity of the eastern specimens, it does not seem
plausible that the Alaskan specimens are a newly developed type. Their
great frequency and the fixity of the type are also not in favor of this

It might perhaps also be argued that the knobs serve for firmly attaching
the needlecase to a skin strap, but there is no evidence whatever that
the needlecases were thus suspended. They seem to have been carried
like all other Eskimo needlecases, by an attachment to the strip of skin
into, which the needles are inserted.

It seems certain, therefore, that the diminutive knob of the Alaskan
needlecase serves no practical end whatever, and that it is a purely conventional
feature in the form of the utensil. It is true that the large
wings and flanges of the Eastern needlecases also serve no practical end;
but it seems well to bear in mind the close resemblance of the two types.

It is interesting to compare the simple types heretofore described
with a number of more complex needlecases which clearly belong to the
same type.

It would seem that, first of all, the strong inclination of the Alaskan
Eskimo to decorate carved objects by means of incised designs has led to
further developments of the patterns heretofore described. Examples of
this kind may be observed in many specimens. In fig. 4 b, the same
typical arrangement of flanges, knobs, and faces may be observed; but
the concave face and vertical line are further decorated by oblique spurs
placed in pairs, and the lower border design of the flange is elaborated
as a single line with double oblique spurs also. On both sides of the
needlecase, and surrounded by the line running downward along the
lower border of the flange and on the body of the tube, is a design of
what seems to be a human being with a caribou head, which stands on
a line extending across the side of the needlecase, just over two knobs,
the single knob on each side being doubled in this case. On the lowest
point of the line surrounding the concave face stands a quadruped with
long body and bent legs. Another type of elaboration and modification
of the design is shown in figure 4 a, where the lines with pairs of oblique
spurs have also been made use of. Many needlecases are so much worn
down that the designs have become quite indistinct; but often the
middle concave face was never well marked. These specimens resemble
in general shape the characteristic designs, the forms are rounded off,
and have lost many of their decorative traits. On the reverse side of one
of these specimens (fig. 13), a double line with oblique spurs running
574outward from the lines is shown, but not in the middle of the needlecase.
Its position is so irregular that it cannot be compared with the decorations
of the specimens heretofore described. The specimen has quite
an irregular line decoration on the flanges. A modern, roughly finished
specimen (U.S.N.M. 33697) of the usual type has the concave face
hollowed out deeply, and the flanges are set off more markedly than in
the majority of specimens. The knobs have been moved very far downward.
Fig. 9 a also seems to be a modern specimen, in which the vertical
border lines of the concave face have been moved toward the border
of the flanges, and where the upper border is replaced by an alternate-spur
band. In this specimen the knobs are also moved very far downward.
The specimen shown in figure 14, resembles in many respects


Fig. 13.


Fig. 14.

the one show in fig. 9 a, particularly in the depth of the concave faces
and in the sharp angle formed by the flanges where they are set off from
the body of the tube.

In figure 9 b, the border lines of the concave face consist of two forked
lines on each side, and the border lines of the flanges have been transformed
into alternate-spur bands. In fig. 10 c the same lines are etched
as spur bands, and the forked-line design is placed on the flanges. There
is no indication of knobs. In the specimen shown in fig. 5 a we find on
the concave face of the tube an alternate-spur band added, which ends
below, on the reverse and the obverse, with two parallel cross-lines. On
the lower part of the flanges is shown, on one side a double alternate-spur
band. The opposite side is laid out on the same plan, with the only
575exception that the cross-lines between the lowest pair of border lines are
drawn right across (as in the bands in fig. 15 a). The whole side of this
needlecase is flattened, beginning under the flange, down to the lower
border. This flat field is occupied in its whole length by an alternate-spur

The bands in fig. 15 a are occupied by ladder designs instead of alternate-spur
designs. Presumably this is the result of careless execution of
the older spur design. Fig. 156 shows a very careful technique, and it is
characterized by a strict adherence to the general type, extreme smallness
of the knobs, and elaboration of the decorative motives. Thus the


Fig. 15.

upper border consists here of two alternate-spur bands; the lower border
of the flanges of a number of parallel lines which are very close together.
The same kind of lines occur on the middle field. The decorative band
at the lower end is also doubled, and repeated at a short distance above
the lower end.

Other modifications are found in the following specimens. In fig. 15 c
there is no middle concave face, but in its place we find two parallel lines
which are carried down to the lower border. There are also two parallel
lines on each side running down from the flanges to the lower border,
and to the upper and lower border lines of the flanges are added vertical
576border lines, so that the whole flanges appear framed. The cross-section
of this specimen is angular. In fig. 15 d two parallel lines are substituted
for the concave face, as in the specimen just described. The sides of this
needlecase are also fiat while the back shows no vertical design and a
rounded surface. Its only decoration consists of a continuous alternate-spur
band design on top and at the lower end of the small flanges, continuing
the corresponding bands on the front of the specimen and on
the narrow sides of the flanges. This specimen has no indication of

Fig. 15 e is in many respects peculiar, particularly in so far as the
two small knobs are not on the same level. The middle concave field is
carried down to the lower end of the needlecase, as in the two preceding
cases, and the whole needlecase is angular in cross-section. It has eight
faces, which taper down toward the lower end. On the three faces on
the right-hand side is the small double-angle decoration which has been
indicated in our illustration. A double angle turned with its apex downward
is also found on the lateral face on the right-hand side. As shown
in the illustration, the flanges do not extend up to the top of the needlecase,
as is the case in most specimens.

The needlecase represented in fig. 15 f illustrates a very peculiar reduction
in the general form. The flanges have almost disappeared, and
with them the upper and lower decorative border, as well as the border
at the lower end of the tube; and all that remains to remind us of the
form here discussed are four parallel forked lines, which, however, are
continued beyond the forks. Nevertheless the impression given by the
specimen in connection with the whole series is such that I do not doubt
for a moment that it belonged originally to the series under discussion.

The series represented in the following figures seem to me of special
interest from a theoretical point of view. The identity of the types of
needlecases here shown with the preceding ones is perfectly obvious. The
specimens collected in figures 16 and 17 show with perfect distinctness
the bulging tube, the flange with its decoration, the knobs, and the concave
face of the tube. Here part of the specimen seems to be conceived
of as an animal. The bulging tube is the body of the animal, whose
head has been added at the lower end. Although the transformation of
the lower end of the needlecase into an animal has been perfected; it does
not seem likely that the whole object was conceived as an animal form.
If this were the case, the flanges, when transformed into the tail of a sea577


Fig. 16.

mammal, would probably have been modified, and the position of the
head would be so changed as to be in proper relation to the tail.

It seems entirely artificial to assume that in this case the animal form
as such could possibly have preceded the typical needlecase as before
described, but that we are dealing here evidently with a secondary interpretation


Fig. 17.578

of the design, which finds expression in the addition of the
animal head and in other later additions to the whole form. In figs.
16 b, c and 17 a-c, the entire old design may be recognized in all its
details; even the alternate-spur band remains, although it interferes
with the form of the seal's head which has been added. In fig. 16 c the
head of the animal has been turned, so that the lower part of the needlecase
looks like a sea animal swimming on its back. A similar specimen
from the Ethnographical Museum in Berlin is shown in fig. 18. It has a
unilateral small knob. In fig. 16 a we find what may be a still further
development of the original design here described. The seal's head has


Fig. 18.


Fig. 19.

disappeared again, and in its place we find a simple knob. There are
three parallel lines near the lower ends of this knob, which makes the
whole area, seen from the top, look a little like a small crustacean. The
knobs in this specimen are very small. I consider it quite possible that
here we may have a case where, under the stress of older forms of the
needlecases, a partial reversion to the original type has taken place.

The strong tendency of the Eskimo to utilize animal motives has
found expression in another manner in the specimen represented in
fig. 16. d. Here the small lateral knobs have been considerably enlarged
and have been given the form of seals' heads. I believe that here also
there can be no doubt in regard to the question whether the seal's head
579or the knobs are older. If the knob had to be considered as a degenerate
form of the seal's head, it would hardly be intelligible why only one or
two specimens out of a great number should retain the heads in this
place, while in practically all other cases the reduction to a simple knob,
sometimes so small that it can hardly be felt, should have occurred. It


Fig. 20.

seems quite evident that in this case the imagination of the artist was
stimulated by the traditional knob, and that it has been developed,
owing to a desire to further decorate the utensil, into a seal's head. The
modification of the central concave face of this specimen is quite in
accord with other modifications of the same surface, which have been
described before. On the reverse of this needlecase the pairs of oblique
580spurs attached to the converging lines are directed toward the upper
part of the needlecase.

In the following figures a number of specimens have been collected,
in which another part of the needlecase has been modified through the
general tendency of the Eskimo artists to introduce animal designs. Instead
of the lower end, the flange has been thus developed. The procedure
appears perhaps clearest in the specimen shown in fig. 19 a,
where on one side the flange shows a number of perforations and
modifications, by means of which it has been developed into a quadruped,
while on the other side a walrus head has been developed by
making a long slit along the body of the tube and by inserting an eye,
and lines indicating nostrils and mouth, near the upper border. Thus
the outer sides of the flanges form the tusks of the walrus head, while the


Fig. 21.

top forms the head itself. The specimen here referred to
shows clearly its close relation to the original type of
needlecase. The decoration of the lower part and the concave
face may still be observed. The characteristic decorations
of the concave face are also indicated. In fig. 20 we
find the same type of needlecase with a double walrus head
at the top. In most of the specimens the tusks have been
broken off. In fig. 20 a traces of the vertical forked lines
bordering the middle field also remain. In the specimens
shown in fig. 20 a and b the middle concave face is quite
distinct. In figs. 20 b and c two specimens are represented
which combine a modification of the lower end of the
needlecase with that of the upper end. At the lower end
a seal head is shown on one side, as in the specimens previously
discussed, while at the upper end the double walrus head is
found. In one of these specimens the middle concave face is well
marked, although it is not bordered by an incised line.

The next group of modifications of the old type of needlecase follows
the same direction as those just desribed, the flanges being modified
so as to represent an animal on each side. A specimen of this type is
shown in fig. 21, where a walrus with head stretched forward is shown.
The tusks touch the upper end of the tube, while the two flippers are
shown at the lower end. Two seals are shown in the same position on
fig. 22 g, while two quadrupeds occupy the position of the flanges in figs.
22 a, f. In fig. 22 d the quadrupeds appear doubled; and in fig. 22 h
the seals have so much increased in size that they occupy the whole side
581of the needlecase. However, in this case also, the close relation between
all these types can easily be demonstrated by an examination of figs. 22
d and g which retain all the characteristic traits of the simple type. The
two animals in fig. 22 f seem to represent lemmings. They are placed
somewhat differently from the ordinary form of the flanges, but are


Fig. 22.582

evidently developed from forms like those shown in figs. 22 a and g. A
specimen in which the one side of the needlecase retains the ordinary
shape, while the opposite side of the flange has been transformed into
an animal, is represented in fig. 17 d.

In this case also it would seem exceedingly difficult to interpret the
simple geometrical form of the needlecase as a later development from
the animal representations here discussed. In this case, similarities of
the decorative designs on the tubings would be entirely unintelligible,
while the assumption that the animal forms have developed from the
geometrical forms seems to give a very plausible explanation of the

The specimens in which the upper end has been so modified as to
become a double walrus head lead us to another group in which the
walrus head is repeated a number of times along the sides of the tube.
Specimens of this kind are represented in figs. 22 b and c. In both of
these traces of the old upper and lower border decoration remain, and
fig. 22 c also shows the typical oblique spurs in pairs in the same position
which has been described several times. It therefore seems perfectly
natural to interpret these forms as the result of repetitions of the animal
design, which was first developed from the flanges of the needlecase.
Figs. 22 c and e differ from other specimens of their kind in that they
have the walrus head developed only on one side, while on the opposite
side the flange is suppressed.

As has been indicated, the geometrical decorations of the typical
flanged needlecases reappear in many of these highly modified specimens.
Attention may also be called to the forked-line designs which
rise from the lower border in the usual number in the specimens shown
in figs. 22 d, f and g. In the last named specimen the number of these
lines is five. The deviation in number may be due to inaccuracy in laying
out the ornament. In fig. 22 d there are two forked designs on opposite
sides, while on each side from the tails of the animals down to the
lower border runs an alternate-spur band. Between the alternate-spur
bands and the long forked lines there are short forked lines, as indicated
in the illustration. Only in fig. 22 h do we find an important modification
of the lower end of the needlecase, which forms a ring. That in our
specimen has been broken. The backs of the two needlecases shown in
figs. 22 c and e are somewhat flat. It is of interest to compare the line
decoration of the latter needlecase with the one shown in fig. 4 b, which
is a simple modification of the fundamental type.583

The illustrations, figs. 23 a and b, of two needlecases in human form,
are not quite as convincing as the specimens themselves; but a comparison
of these forms with the other needlecases of this series seems to me to
suggest with great force that the human figures here shown are related
to the same type of needlecase that we are discussing. The whole human


Fig. 23.584

figure is treated as a tube, and it is my opinion that the bulging hips
correspond to the bulging middle part of the needlecase, while the arms
correspond to the flanges, and perhaps more particularly to developments
of the flanges similar to the walrus-head developments, while the
head is a later development of the upper border, suggested by the perception
of the similarity of the whole form to a human figure. I do not
wish to imply that the human figure in this case has necessarily developed
from the type of needlecase first discussed; but it seems plausible to
me that an assimilation between the human figure and this type has
taken place in the two specimens here illustrated. It seems likely that the
animal figure shown in fig. 24 must be considered in a similar manner.
There is no doubt that the vivid representation of the animal lying down
has very little to do with our type of needlecase; nevertheless I cannot


Fig. 24.

free myself from the impression that the artist, in his treatment
of the subject, has been influenced by the treatment of
the flanges of needlecases and by the general form of this
utensil. There is a certain similarity between the position of
the feet and the positions of the walrus tusks shown in figs.
19 and 20, which is not explained by a realistic treatment
of the animal alone; and the same is true of the position of
the neck and head and of the curves in the hind part of the

The similarities which I am discussing here are even less
clear in some of the other specimens represented here. Fig.
23 c evidently represents a human leg, the design on one side
being a representation of tattooing. In this case faint traces
of the upper border design and of the lower border design remain,
and the outline of the whole specimen still recalls to a certain extent
the bulging tube below and the wider part with its flanges above.
If we agree to accept this specimen as belonging to the present series,
the specimen shown in fig. 23 e, must be considered as belonging
here also. There is no doubt that fig. 23 d belongs to our series.
The tube and the knobs are the same as those occurring in the
most typical specimens. Instead of the concave faces, we have merely
flat surfaces, and the flanges have been much reduced in size, but are
perfectly distinct and sharply set off. The ornamentation, however,
differs on the flanges and concave faces from the ordinary decoration.
Besides the designs shown in the illustration, we have on the back of the
flange to the right, a line with two pairs of one-sided oblique spurs running
585downward and a forked line running down from the black ring,
like the one shown on the right-hand side of the illustration. On the
right-hand side of the lower part of the needlecase an etched design,
representing a quadruped with long tail, will be observed. Fig. 23 f
shows a simple tube with four groups of knobs, which may have been
suggested by the knobs of the specimens here described. Figs. 23 g and h
represent a needlecase, which on one side shows the typical form of the
flanged specimens, while on the opposite side the head, neck, and forepaws
of an animal are set off.

Another geometrical development of the ordinary type is represented
in fig. 25 a. In this specimen the general outline of the flanged tube is


Fig. 25. a, c, d, Needlecases. b, Awl, Alaska.

readily recognized, but all the other characteristic features have disappeared.

In fig. 26 specimens are illustrated the relations of which to the
flanged type are very doubtful. The knobs in fig. 26 a, which are
doubled in the axial direction and appear on four sides of the tube, are
analogous to those shown in fig. 23 f; and these two types are undoubtedly
closely related. Attention may be called to the awl-like implement
illustrated in fig. 25 b, which shows the same four knobs here described,
and which therefore in its origin may well be related to the decorative
designs on the needlecases. The animal types figs. 26 b-e and fig. 25 c,
diverge so much from the flanged type that their relationship seems very
doubtful. Still I cannot free myself of the impression of a certain586


Fig. 26.587

influence of the flanged types upon these forms also. This becomes
apparent by a comparison of the needlecase shown in fig. 25 d with the
animal types here discussed. It is quite evident that this specimen has
its affiliations both with the animal types and the walrus-head types
shown in figs. 19 and 20. It is, however, also possible that its form has
originated by assimilation of two distinct types.

The conclusion which I draw from a comparison of the types of
needlecases here represented is that the flanged needlecase represents an
old conventional style, which is ever present in the mind of the Eskimo
artist who sets about to carve a needlecase. The various parts of the
flanged needlecase excite his imagination; and a geometrical element
here or there is developed by him, in accordance with the general
tendencies of Eskimo art, into the representation of whole animals or of
parts of animals. In this manner, the small knobs or the flanges are developed
into heads of animals. After this modification has once set in, the
animal figures may be repeated on other parts of the implement. Besides
this, associations between animal forms and the form of the whole
needlecase seem to have taken place, which have to a certain extent
modified the manner of representing animal forms which were adapted
to use as needlecases; so that the old form and style of the needlecase
determined the treatment of the animal form.

If we were to apply to the present series the theory of the origin of
conventional form from realistic motives, it would be exceedingly difficult
to account for the general uniformity of the fundamental type. It
seems to me that, on the basis of this theory, we could not account for
the diversity of realistic forms and the uniformity of general type.
Neither does it seem possible to account for the series of types by assumption
of any influence of technique; and my impression is that the only
satisfactory explanation lies in the theory that the multifarious forms
are due to the play of the imagination with a fixed conventional
form, the origin of which remains entirely obscure. This I freely
acknowledge. If, however, we are to form an acceptable theory of the
origin of decorative designs, it seems a safer method to form our judgment
based on examples the history of which can be traced with a fair
degree of certainty, rather than on speculations in regard to the origin
of remote forms for the development of which no data are available.

I believe a considerable amount of other evidence can be brought
forward sustaining the point of view that I have tried to develop,
namely, that decorative forms may be largely explained as results of the
588play of the imagination under the restricting influence of a fixed conventional
style. Looking at this matter from a purely theoretical point of
view, it is quite obvious that in any series in which we have at one end
a realistic figure and at the other end a conventional figure, the arrangement
is due entirely to our judgment regarding similarities. If, without
further proof, we interpret such a series as a genetic series, we simply
substitute for the classificatory principle which has guided us in the
arrangement of the series a new principle which has nothing to do with
the principle of our classification. No proof whatever can be given that
the series selected according to similarities really represents an historical
sequence. It is just as conceivable that the same series may begin at the
conventional end and that realistic forms have been read into it, and we
might interpret the series, therefore, as an historical series beginning at
the opposite end. Since both of these tendencies are active in the human
mind at the present time, it seems much more likely that both processes
have been at work constantly, and that neither the one nor the other
theory really represents the historical development of decorative design.

The assumption of a development from realistic design to conventional
design also omits the consideration of one exceedingly important
element, namely, the style of convention that prevails in the types of art
of different areas. If geometrical designs developed from realistic motives
the world over, it still would remain to be proved why a certain
style of conventionalism belongs to one art and another style to another
art; and in order to explain in a satisfactory way the different styles of
art, we should have to accept these as given at a very early stage during
the process of conventionalization of realistic designs.

The attempt to explain the processes of conventionalization by the
theory of the influence of technical motives does not seem to offer an
entirely adequate solution of this problem. It is true that certain very
simple designs seem to be due almost entirely to the influence of technique
upon simple decorative tendencies. This influence, however, does
not reach so far as to determine in detail the character of design in the
same kind of material or in the same technique. As examples of such
differences may be mentioned, for instance, the designs in woven checkered
mattings from West Africa, where peculiar realistic figures alternate
with geometrical band designs; the designs of cedar-bark mattings
of the Ojibwa and of those of the North Pacific coast, and of designs
made in the same technique by the South American Indians. In all
these cases the technical conditions are practically the same, but the
589styles differ vastly. It seems necessary, therefore, to assume in the development
of design the existence of tendencies which are due to causes
different from the technique, and unrelated to the realistic motives
which may be current or may have been current.

I have no theory to offer in regard to the origin of these types of convention,
which presumably was connected with a whole series of activities
determining the perception and reproduction of forms; but it seems
desirable to illustrate by a number of instances the fixity of these conventional
forms and the deep influence that they have had even in apparently
realistic forms. I have pointed out in the discussion of the
designs of the blankets of the Chilkat Indians that a great many of the
older forms can be reduced to two fundamental types, and that, no
matter what animal may be represented in the art of the weaver, it is
almost always reduced to one of these two forms. 119 In the same place I
have shown that the treatment of the animal figure on carved boxes of
the Tlingit has other fixed conventional forms, which, although closely
related to the blanket design, are quite permanent and applied only in
the manufacture of boxes. 220

In a quite different region, among the Tungus tribes of the Amur
River, Dr. Berthold Laufer has shown that one of the essential types determining
the whole arrangement of decorative designs, which consist
of realistic figures as well as of curved lines, is based on the type of
“cocks combatant.” 321

It is also important to note that figures conforming to such fundamental
types may be interpreted in a great variety of ways by the people
who use them. I have pointed out such a similarity of type and fundamental
difference of interpretation in explanations given by the Huichol
Indians. 422 Here we find practically the same figures once interpreted as
the freshwater crab, and then as oak leaves and stems. Other more
extended series of such ambiguous interpretations may be found in the
art of the Plains Indians as well as in those of other parts of the world. 523

I have suggested before that in many cases these forms seem to compel
590us to assume that the interpretations of many simple forms are entirely
secondary; that often the forms have been borrowed; and that later on,
according to their use in the life of the people, they have been given a
fitting interpretation. 124

I think evidence can be brought forward also to show that the tendency
to play, and the play of the imagination with existing forms, have
deeply influenced the decorative art of primitive tribes as we find it at
the present time.

The first of these traits appears with particular clearness in the
tendency to use rhythmic repetitions of varying forms. Bead necklaces
are one of the most striking examples of the pleasure that man receives
through the use of rhythmic repetition of colors and forms. Among
primitive tribes the rhythmic and symmetrical order of such arrangements
are often exceedingly complex, — so complex, in fact, that they
can be recognized by us only by a close study of the arrangement. A
case of this kind occurs in the fringe on a pair of leggings collected
among the Thompson Indians, which I have described. 225 In this specimen
we have a fringe which hangs down in a very disorderly fashion, so
that the constituent elements cannot be seen distinctly. Nevertheless a
most painstaking arrangement of the component elements is adhered
to, the rhythmic unit consisting of five elements, — one string having
one glass bead and two bone beads in alternating order, one undecorated
string, one having alternating glass and bone beads, one undecorated,
and one having one glass bead and two bone beads in alternating order.
I have found still more complex rhythmic repetitions and symmetrical
arrangements on the embroidered borders of coats of the Koryak. These
contained sometimes ten and more elements in one group. 326 Still another
case of similar kind, from Peru, has been described by Mr. Mead. 427 Here
a rhythmic repetition of six units seems to be very common.

I consider it particularly important to observe that in the first of these
specimens the rhythmic repetition cannot be seen when the leggings are
in use, because this suggests strongly that the reason for the application
of the rhythmic repetition is not the aesthetic pleasure in the effect which
it produces, but the pleasure felt by the maker. If this is true, then we
591do not need to assume that in the other cases a much more highly developed
appreciation of complex rhythm is found among primitive people
than the one we possess. Corroborative evidence in regard to this point
is offered by the basketry of the Thompson and Lillooet Indians. I have
noticed that here, where in a fine imbricated technique color bands are
produced, the basket weavers tend to use with great regularity certain
groupings of the number of stitches belonging to each color, although,
owing to the irregularity of the size of the stitches, these modifications
can hardly be observed. 128 If these facts have a wider application, it
would seem that on the whole the pleasure given by much of the decorative
work of primitive people must not be looked for in the beauty of the
finished product, but rather in the enjoyment which the maker feels at
his own cleverness in playing with the technical elements that he is using.
In other words, one of the most important sources in the development of
primitive decorative art is analogous to the pleasure that is given by the
achievements of the virtuoso.

Examples may also be given illustrating the effect of the play of imagination
upon the development of design. One of the best examples of
this kind is offered by the decorated bag of the Thompson Indians illustrated
by Professor Farrand. 229 The analogy of this soft rectangular bag,
which is decorated with rows of large diamonds, to other similar bags
shows quite clearly that the rows of diamonds have the same origin as
the rows of diamonds which are painted on parfleches of the Plains
Indians. In this case the diamonds suggested the idea of ponds; and, in
order to emphasize this idea, which came to the mind of the woman who
used the bag, she added a number of birds flying toward these ponds.
Other examples of this kind have been mentioned by Doctor Koch-Grünberg
in his observations on the drawings of South American Indians.
The development of the triangles in the designs of the Plains
Indians to tent designs or cloud designs brings out similar points.

Thus it would seem that the development of decorative designs cannot
be simply interpreted by the assumption of a general tendency
toward conventionalism or by the theory of an evolution of technical
motives into realistic motives by a process of reading in, but that a considerable
number of other psychic processes must be taken into consideration
if we desire to obtain a clear insight into the history of art.592

11 Proceedings of the U. S. National Museum, vol. 34 (1908), pp. 321-344.

22 “Conventionalism in American Art,” Bulletin of the Essex Institute, vol. 18

33 “Entwicklungserscheinungen in der Ornamentik der Naturvölker,” Mitteilungen
der Anthropologischen Gesellschaft in Wien
, vol. 22 (1892), pp. 19 et seq.

44 The Evolution of Decorative Art (London, 1893).

55 Unter den Naturvölkern Zentral-Brasiliens (Berlin, 1894), pp. 258 et seq.

66 Evolution in Art (1895).

77 W. H. Holmes, “Textile Art in Relation to Form and Ornament,” 6th Annual
Report of the Bureau of Ethnology
(1884-85), p. 223.

88 Korrespondenzblatt der deutschen Gesellschaft für Anthropologie, Ethnologie
und Urgeschichte
, vol. 35 (1905), p. 126.

91 Indianerstudien in Zentralbrasilien (Berlin, 1905), pp. 330 et seq.

102 Anfänge der Kunst im Urwald (Berlin, 1905).

113 Urgeschichte der Kultur (Leipzig, 1900), p. 540.

124 The American Architect and Building News (1898).

135 “Decorative Symbolism of the Arapaho,” American Anthropologist, N.S., vol. 3
(1901), p. 329.

146 “Decorative Art of the Sioux Indians,” Bulletin American Museum of Natural
, vol. 18 (1904), pp. 231 et seq.

157 “The Decorative Art of the North American Indians,” Popular Science Monthly
(1903), pp. 481 et seq., see pp. 546 et seq. of this volume.

161 American Anthropologist, N.S., vol. II (1909), p. 135.

172 Voyage of the Vega (London, 1881), vol. 2, p. 241.

183 Siberia in Asia (London, 1882), p. 56.

191 G. T. Emmons, “The Chilkat Blanket,” Memoirs of the American Museum of
Natural History
, vol. 3 (1907), p. 355.

202 Ibid., pp. 357 et seq.

213 Publications of the Jesup North Pacific Expedition, vol. 4 (1902), pp. 22 et seq.

224 Carl Lumholtz, “Decorative Art of the Huichol Indians,” Memoirs of the
American Museum of Natural History
, vol. 3 (1900), p. 287 and Figs. 451 and 465.

235 A. L. Kroeber, “The Arapaho,” Bulletin of the American Museum of Natural
, vol. 18 (1902); Clark Wissler, “Decorative Art of the Sioux Indians,” ibid.,
vol. 18, part 3 (1904).

241 See p. 562 of this volume.

252 Publications of the Jesup North Pacific Expedition, vol. 1 (1906), p. 384,
fig. 313.

263 Publications of the Jesup North Pacific Expedition, vol. 6, part 2 (1908), pp.
689 et seq.

274 Boas Anniversary Volume (1906), pp. 193 et seq.

281 Publications of the Jesup North Pacific Expedition, vol. 2 (1906), p. 206.

292 Ibid., vol. 1, Plate xxxiii, fig. 1.