The needle cases of the Alaskan Eskimo offer an excellent example of an
elaboration of geometrical into realistic forms. It must be remembered
that all the Eskimo tribes, east and west, are very fond of carving and
that they produce many small animal figures that serve no practical
purpose but which are made for the pleasure of artistic creation, and
that many of their small implements are given animal forms. The mind of
the worker in ivory is imbued with the idea of animal representation.
The Alaskan needle cases have a stereotyped form to which the bulk of
the specimens conform (fig. 119). The type consists of a tube slightly
bulging in the middle, with flanges at the upper end, with lateral
knobs under the flanges on opposite sides. On the body of the tubing,
between the flanges, is a long narrow concave face, set off from the
flanges and the body of the tube by parallel lines with small forks at
the lower ends. Lines border the sides and ends of the flanges and the
upper end of the concave face, and an alternate-spur band is found at
the lower end of the tube. This type is presumably related to the
ancient type of needle case from Hudson Bay, which is characterized by
a tubular form, flanges at the upper end and a pair of large wings in
the middle part of the tube. These, I believe are the prototypes of the
small knobs at the sides of the Alaskan type. These are sometimes so
small that they can hardly be seen, although they may be felt when the
fingers glide lightly over the surface of the tube. However this may
be, the majority of Alaskan specimens are of the type here described.
The variations in geometrical form are very slight.[71]
[71] See Franz Boas, Decorative Designs on Alaskan Needlecases,
Proceedings of the United States National Museum, Volume XXXIV, (1908),
pages 321-344, Washington, D. C.
[Illustration: Fig. 119. Alaskan needle cases.]
In a number of cases part of the needle case or the whole object is
elaborated in representative forms. In a few specimens the knobs
have been given the form of seal heads. A slight enlargement and the
addition of eyes and mouth are sufficient to bring about this effect
(fig. 119 _c_). In other cases the flanges have been developed into
walrus heads (fig. 119 _d_) or into other animal forms that fit the
form of the flange (fig. 119 _e_). By adding a point for the eye and by
cutting out the inner part of the flange the outer form is essentially
preserved and the form of the bulky walrus head with its large tusks is
successfully attained; or the walrus motive may be repeated so that the
original form of the needle case is considerably obscured. By adding a
seal’s head at the lower end the lower part of the object receives more
or less the form of a seal.
It is important to note that in all specimens of this type the double
spur ornament persists, although it interferes with the seal’s head
that is often added at the lower end. It seems very unlikely that
the many animal forms attached to the needlecases were the primary
forms from which the geometrical form developed. On the contrary, the
wide distribution and the great frequency of the geometrical forms,
their agreement with Eskimo geometrical style and the occurrence of
this geometric style on specimens that do not imitate animal forms
are entirely in favor of the assumption that the earlier form is
geometrical, whatever the origin of the flanges and knobs may have
been. Presumably the habit of carving animal forms has induced the
artist to produce the variants described here.
[Illustration: Fig. 120. Reels of Alaskan Eskimo.]
Equally interesting are the reels used by the Alaskan Eskimo. Their
prototype is a piece of caribou antler cut as shown in fig. 120.
This, however has been developed into animal forms. One specimen has
a geometrical decoration on the body, while three of the ends are
elaborated as animal heads. In another one the two prongs on one side
are given the form of seal heads which are connected by a long line
with the flippers which are shown on the opposite side. Still another
specimen has been treated like a much distorted form of a wolf. The
head is at the prong of one end, the forelegs at the other prong of
the same end; the two hind legs are carved on the two prongs at the
opposite end.
Similar observations may be made on arrow shaft straighteners and snow
knives of the Alaskan Eskimo.[72]
[72] E. W. Nelson, The Eskimo about Bering Strait, Annual Report Bureau
of American Ethnology, Volume XVIII, pl. 40, 48, 94.
[Illustration: Fig. 121. Eye shades and vessel, Ammassalik.]
The examples which I have given demonstrate beyond cavil that the
process of reading in exists and accounts for the significance of
many geometrical forms; that it is not necessary to assume in each
and every case that geometrical ornament is derived from realistic
representations. Hjalmar Stolpe to whose researches on primitive art we
owe so much had an exaggerated opinion of the importance of realistic
designs, for he believed that every geometric form must have been
derived from a representation. Stephan expresses the same opinion. I
may perhaps illustrate Stolpe’s point of view by one example. Among the
Eskimo of East Greenland is found an eye-shade decoration consisting
of notched pieces of ivory placed side by side. Stolpe explained this
form as a derivation from carvings representing seals, abbreviated due
to the rhythmic repetition of the seal design (fig. 121). However, we
know that the Eskimo when making small blocks of ivory, such as are
used for the manufacture of toggles or other similar objects, cut the
walrus tusk just in this manner, so that through a technical process
they have become familiar with the ornamental form. Therefore it is not
indispensable to assume a realistic origin for the design.
It is interesting to compare the process with which we are dealing with
other ethnic phenomena of similar import. The essential conclusion
drawn from our observations is that the same form may be given
different meanings, that the form is constant, the interpretation
variable, not only tribally but also individually. It can be shown
that this tendency is not by any means confined to art, but that it is
present also in mythology and in ceremonialism, that in these also the
outer form remains, while the accompanying interpretations are widely
different. In the case of mythology we may observe that in the same way
as patterns have a wide distribution, so also motives of myths, or even
whole plots are found over extended areas. For example, the well known
tale of the magic flight, which is one of the most widely distributed
fairy tales, occurs in a number of North American localities. Among
the Eskimo it is given as accounting for the origin of fog; in British
Columbia it is made to account for the origin of certain cannibalistic
ceremonials. Another example is the so-called star husband story, a
tale of two girls who were taken up to the sky by the stars. The story
is widely spread and is told as the explanation of a great variety of
phenomena. The child of one of the women becomes the culture hero and
destroys the monsters infesting the world. Customs, ceremonials, and
prerogatives of clans are explained by this tale, in accordance with
the chief interests of the tribes concerned. The nature myths collected
from all parts of the world by Dähnhardt[73] are full of examples
of this kind: the most diverse features are explained by the same
tales. This shows that it is quite untenable to assume that the tales
developed through the immediate effect of viewing natural phenomena,
that much rather the tales preexisted and the explanatory part was
tacked on at a later time, precisely as the art forms preexisted
and their meaning was tacked on according to the peculiar mental
disposition of the individual or the tribe.
[73] Oskar Dähnhardt, Natursagen, Vol. 4. Leipzig 1912.
The same observation has been made in regard to ceremonials. The
so-called sun dance is performed by almost all the tribes of the
great plains. The general run of the ceremonial is essentially the
same everywhere, although different tribes have added on special
characteristic details. The meaning, on the other hand, presents far
greater variations. In one case the performance is made in fulfilment
of a vow made when supernatural aid is invoked in time of distress; in
another case it is undertaken to purchase the right of ownership in a
sacred bundle, and then again it is a tribal, seasonal ceremonial.[74]
[74] Leslie Spier, Anthropological Papers of the American Museum of
Natural History, Vol. XVI, (1921), pp. 457 et seq.
These three examples illustrate that the psychological explanation of a
custom and its historical development are not by any means the same; on
the contrary, owing to secondary interpretations that in course of time
set in and the general character of which depends upon the cultural
interest of the people, the psychological explanation is much more
likely to be quite independent of the actual historical happenings. The
mere fact that a tribe explains forms according to a certain pattern
does not prove, that they have developed from actual representations of
the objects they now represent.
It might perhaps be admitted that in the special case of the North
American triangle which we have discussed the argument is convincing
because no transitional forms exist, that, however, when an almost
continuous series of forms is found, beginning with the most realistic
and ending with a purely geometrical one, the conclusion cannot be
avoided that the development has been in the direction from realism
to conventionalism. I have already indicated that as long as there is
no historic proof, the sequence might as well be reversed. In all the
cases that have been described and that are based on direct inquiry
among primitive people regarding the meaning of designs, realistic and
conventional forms have been found to occur at the same time. It would
therefore be necessary to give a reason why some excellent artists
should use the one style, others, equally good ones the other; or why
even the same artist should combine the two styles.
It is generally claimed that slovenly execution brings about
deterioration of pattern and through this, causes misunderstandings.
I do not consider this explanation as tenable under the conditions of
life prevailing among primitive people, because there is no slovenly
execution among natives who make utensils for their own use. The
pottery and paintings of the South American Indians, observed by Von
den Steinen, the crocodile arrows of New Guinea, the frigate bird
carvings, are all carefully made. We find often that people will use
inferior wares for exchange with neighboring tribes, while they keep
the good material for themselves. The natives of Vancouver Island
for instance, adulterate with elderberries berry cakes that are made
for trade, while those made for home consumption are made of the
more valuable berries, such as blueberries or salmonberries without
the addition of other material of inferior value. This tendency,
particularly when combined with the desire to manufacture quickly
large masses of material, leads to poor work. The question then
arises, what happens in cases of this kind. Does slovenly work lead to
misunderstandings and to conventionalism? Balfour has tried to make the
process clear. He let an individual copy a design and used the first
reproduction as original for a second copy. By continuing in this way
he received the most astonishing transformations. Such results may
occur when work executed in a highly developed technique is imitated by
people of lesser accomplishments. The standard example is that of the
degeneration of Greek coins when copied by Keltic imitators[75] which
led to a complete disruption of the original design. However, this
example is not to the point, because we are ordinarily not dealing with
copies of designs borrowed from people of higher technical development,
but with those belonging to a single tribe.
[75] See Max Ebert, Reallexikon der Vorgeschichte, Vol. 6, pp. 301 et
seq.
I have had occasion to observe the effect of factory production and of
slovenly execution in some Mexican material. In western Mexico dishes
are made of tree calabashes which are covered with an orange laquer. By
the lost-color process these are overlaid with designs in green laquer.
The industry is probably of Spanish origin. Old specimens made of wood
are of excellent workmanship. These are decorated mostly with animal
forms, deer, fish, and so on. At present the work is of much poorer
execution and the ware is sold in market places, as in Oaxaca (fig.
122). On some specimens the fish designs of ancient type are still
used, but we find also, even more frequently, leaf patterns and we may
here apparently observe just that kind of misunderstanding described
by Balfour. The gill region becomes the base of the leaf. The head of
the fish corresponds to the base of the leaf; the fins to the marginal
indentations, and the ribs of the fish to the veins of the leaf. Since
both the fish and the leaf occur in modern types it may be doubted
whether we are dealing with an actual transformation, with a real
misunderstanding. It might be as well that the psychological process
involved has been rather the substitution of new subject matter for the
old, in which process the new subject was rigidly controlled by the
old, stereotyped form. We shall see that conditions of this kind are
often very potent.
[Illustration: Fig. 122. Designs from vessels made of tree calabashes,
Oaxaca.]
[Illustration: Fig. 123. Fragments of pottery vessels, Texcoco.]
[Illustration: Fig. 124. Designs from pottery vessels, Culhuacan.]
Another good example is the pottery made for sale by the ancient
inhabitants of the Valley of Mexico. Pottery was made here in great
masses and, as described by Sahagun,[76] sold in the market places.
The vessels show plainly the effect of factory production and of the
resultant slurring. The Aztec pottery is fundamentally of uniform type.
In Culhuacan, a small village at the foot of the Sierra de la Estrella,
a coarse kind of this ware was made. Great masses of potsherds are
found in the swampy soil which was used in early times as garden beds.
The pottery is thick, dark orange, painted black. It is a pronounced
local form, darker than the light ware of Texcoco (fig. 123); the
painted lines are broad and coarse, while those of Texcoco are very
delicate. The patterns are fixed, but the rapidity of manufacture has
developed a definite style, analogous to the styles of handwriting.
Each painter had his own method of handling the brush, with the result
that his individuality may easily be recognized.[77] I select a few
examples here in order to show the effect of slurring upon design. It
is essential to note that on the same specimen the lines are always
drawn in the same way; that is, that the hand of the painter followed
very definite motor habits.
[76] Bernardino de Sahagun, Historia general de las cosas de Nueva
España, ed. C. M. de Bustamente, México, 1830, Vol. 3, p. 56.
[77] See, Franz Boas y Manuel Gamio, Album de colecciones
arqueológicas, México 1912.
One of the simplest designs is instructive in regard to the effect of
slurring and the development of individual style. One of the decorative
elements consists of a series of S shaped, interlocking figures (fig.
124). It will be noticed that, in fig. 124 _b_ the lines which seem to
have developed from the S shaped forms, are degenerating into simple
curves, while in _c_ the lines are drawn more carefully. Fig. 124 _d_
represents the decoration from the outer side of a vessel and here the
S shaped curves have been carefully connected and are developing into a
new pattern.
[Illustration: Fig. 125. Designs from pottery vessels, Culhuacan.]
Another simple pattern which illustrates the effect of rapid production
consists of a regular repetition of a curve and two vertical strokes.
It seems likely that the form is derived from a circle followed by
vertical strokes such as are also found in the finely painted pottery
of Texcoco (Fig. 125 _a_). The corresponding Culhuacan specimens are
shown in fig. 125 _b_, _c_. While in a few cases the curve is a more or
less carefully drawn circle (_b_, _c_), it has generally the shape of
a spiral. One of the most striking characteristics of the development
of this pattern is the tendency to change the direction of the vertical
strokes to an inclined position (fig. 126). According to the tendency
of the painter the strokes incline either strongly from right above to
left below (_a_, _b_ inside, _c_, _d_ outside), while in _e_ and _f_
the strokes run from the left to the right. In the rapid creation of
these forms the vertical lines are transformed into hooks. They are so
characteristic that it seems perfectly feasible to recognize the same
hand in these specimens.
There is still another way the design is treated. In fig. 127 _a_ the
spiral lies in the direction opposite to the one usually given to it
and a single dividing vertical line is placed between the two spirals;
the end of the first spiral taking the place of the second vertical
line. Here again the method of treatment is consistent. In a few cases
the spiral line instead of terminating abruptly is continued under the
vertical strokes (_b_). In still another case, a separate single or
double horizontal line is placed under the spiral (_c_).
[Illustration: Fig. 126.]
[Illustration: Fig. 127.]
[Illustration: Fig. 128. Designs from pottery vessels, Culhuacan.]
A frequent design of which many specimens have been collected consists
of a spiral with attached curves. Presumably it is derived from a
circle and straight lines (fig. 128 _a_). In rapid execution, the
center of the circle is connected with the circumference so that the
whole line is transformed into a spiral. Sometimes the spirals develop
into a simple hook (_b_), and the lowest point of the circumference
of the circle is continued into the adjoining horizontal line. In by
far the majority of specimens the distinction between the circle and
the adjoining line may be recognized by a sudden turn, or at least an
indention in the lowest horizontal line (_c_). The form developing
from the original pattern depends entirely upon the peculiar turn of
the brush used by the artist. In many cases (_b_, _c_) the spiral,
continuing into the lowest horizontal line and turning back into the
central horizontal line, is made in one stroke and the upper horizontal
line is added on by a separate movement, adjoining the middle and
forming a hook. In other cases (_d_, _e_), the spiral and lowest
horizontal are drawn in the same way. The horizontal lines, however,
are made separately by drawing a right angle, probably from the top
downward and adding a central line. In these cases, the horizontal ends
at the far side in a sharp angle. In fig. 128 _f_, the two horizontal
lines are separated from the curve; in _g_, they are made very long
and apparently the middle one is made in one stroke with the spiral.
The method of producing _h_ is the same as that applied in _b_ and _c_
but the upper horizontal line is very much reduced in size. The same
method is applied in _i_ with the difference that the lower horizontal
is turned up and ends rather abruptly, and that the upper and middle
horizontals are made in the form of a single horseshoe curve. In _j_
and _k_ quite a different method is applied, the spiral remains as
before but the horizontal lines are made separately in the form of a
3. In _l_ and _m_ the whole curve is made in one continuous line which
has led to the doubling of the middle horizontal in the form of a
loop. In _n_ the three horizontals are treated quite differently; the
long S shaped figure starting above on the left being substituted for
the lines attached to the lowest horizontal. A similar principle, but
beginning on the opposite side, is applied in _o_, in which specimen
the three horizontals have taken the form of a spiral ending below in
an almost vertical spur. In _p_ we recognize a form in which the S
shaped curve is made separately; in the middle an additional horizontal
line is added and furthermore, the lowest part of the S shaped curve is
connected with the lower horizontal. The form _q_ is reversed and by
attaching the S shaped curve to the center of the spiral, a divergent
form is developed.
[Illustration: Fig. 129. Alligator designs from Chiriqui pottery.]
I believe that many of the highly irregular forms that occur in painted
pottery must be explained in the same way. Another instance of this
kind is represented by the so-called alligator ware of the Chiriqui
(fig. 129) on which W. H. Holmes based his arguments of gradual
degeneration of realistic forms into conventional forms. Although
the pot itself is well made, the painting is almost always slovenly;
evidently the result of mass production. The most characteristic trait
of decoration of this ware is line and dot work. All the designs are
characterized by the use of black and red lines interspersed with
dots. The geometrical designs, as well as the animal forms are crudely
executed. Professor Holmes has called attention to the fact that the
dots are used to indicate the scales of the alligators, but this fact
does not prove that all the dots are derived from alligator scales.
Forms like those shown in fig. 129 to the right above may well be
understood as attempts in general decoration in black and red lines
and dots, badly executed. This seems more probable since the dotted
triangle has a much wider distribution than the alligator motives. A
representation of the alligator might then be explained as executed
in accordance with the technique applied to geometrical motives.
Owing to slovenly execution the animal form may degenerate according
to the motor habits of the individual artist. However, this would
not prove that the alligator, as such, is older than the line and dot
decoration. It must be borne in mind also, that the upturned snout of
the alligator, of which much is made as a means of identification,
is a character of much wider distribution than the alligator motive
itself. Representations of monkeys have it and we find it also in
representations from the interior of Costa Rica and from some parts of
South America. All this is also true of the curious nuchal appendage
which occurs in Costa Rica as well as in South America.
Equally instructive is the application of small nodes and fillets
to pottery which has been explained by Dr. G. G. MacCurdy as
resulting from degeneration of armadillo figures.[78] The essential
characteristic of all this ware is the use of small nodes and fillets
applied to the surface of the vessel or to some of its parts, like
feet, neck, shoulder or handle. These attachments are decorated by a
series of short parallel impressions. An oval node with single medial
line or lines is often used to indicate an eye; a similar nodule with
several parallel lines indicates the foot; a series of parallel, short
fillets with short parallel crosslines, are applied to forms that
represent animals, but they are also found on the bodies of vases. Dr.
C. V. Hartman[79] and S. K. Lothrop[80] describe the same technical
motives from other parts of Costa Rica. In technical character these
are so much like the Chiriqui specimens that we can hardly doubt that
they are derived from the same device. This method of decoration is
widely spread. Its use extends over Central America and the West
Indies. It is most characteristic of archaic ware, particularly
figurines of this early period are always modelled by means of
attaching nodules and fillets. During this period the eye is regularly
represented by a nodule with one or several incisions. This technique
is also found in Ecuador[81] up to more recent times. In the Toltec
period the adornment of vessels by attached nodule decoration reached
its highest development. In North America it is not common. Incised
fillets occur in remains from the middle Mississippi region but even
here they are not one of the pronounced features. In contrast with its
frequency in the highly developed pottery of early Central America its
rarity may be noted in Africa, where highly developed forms are by no
means absent, and where lids with animal figures might seem to suggest
readily the application of the device.[82] This is true also of the
prehistoric pottery of Europe. The nodule appears in the pottery of
Michelsberg, in Jaispitz (Moravia) and in a few other late localities.
Only in the slip (barbotine) decorations of the terra sigillata do we
find anything resembling the American appliqué ornamentation, but since
the material is applied in a semifluid state, it does not attain the
same freedom of treatment. Nodes that do occur in European prehistoric
pottery were apparently made rather in imitation of punched bronze
decorations and belong to a late period. Attached animal figures, made
in clay, like those found at Gemeinlebarn, also seem to be imitations
of metal work and have never reached that development which is so
characteristic of Central American ceramic art.
[78] George Grant MacCurdy, A Study of Chiriquian Antiquities, Memoirs
Connecticut Academy of Arts and Sciences, Vol. 3 (1911), pp. 48 et seq.
[79] C. V. Hartman, Archaeological Researchers in Costa Rica,
Stockholm, 1901.
[80] Samuel Kirkland Lothrop. Pottery of Costa Rica and Nicaragua, New
York, Museum of the American Indian, Heye Foundation, 1926.
[81] Marshall H. Saville, The Antiquities of Manabi, Ecuador, New York,
1910.
[82] See a relief ornament on a red ware vessel from Banana, Belgian
Congo, Annales du Musée du Congo; Notes analytiques sur les collections
ethnographiques, Vol. 2, Brussels, 1907, Les industries indigènes, Pl.
III, fig. 34.
The characteristic slit rattle feet of Chiriqui pottery prove even more
conclusively than the application of fillets and nodes, that the art
forms of this province must be considered as a special development of
forms characteristic of a much wider area. This type of foot is widely
spread beyond the territory in which the fish forms prevailed.[83]
[83] See Franz Boas y Manuel Gamio, Album de colecciones arqueológicas,
México, 1912, Plates 36, 42, 51.
We are thus led to the conclusion that the use of the nodes and fillets
for building up armadillo motives, are historically related to the
method of decorating vessels by means of the attachment of separate
pieces. The armadillo motive can then be only a specialized application
of building up animal motives from the elements in question. The
elements themselves must not be considered primarily as symbols of
the armadillo, nor can all the animals built up of these elements be
interpreted as armadillos.
The essential point of this consideration lies in the wide distribution
of technical and formal motives over large areas, although differing
in details in various localities. These technical and formal motives
are the materials with which the artist operates and they determine
the particular form which a geometrical motive or a life motive
takes. If the notched fillet and node are the material with which the
hand and the mind of the artist operate, they will occur in all his
representations.
The investigators who have tried to prove that conventionalized forms
originate through a process of degeneration from representations, have
generally overlooked the strong influence of motor habits and of formal
arrangements upon the resultant style. In those cases in which there
is a tendency to organize decorative motives in rectangular panels, in
circular areas, or in fields defined in other ways, the result must be
quite different from others in which the artist habitually arranges
his material in large fields or in continuous bands. The habit of
decorating pottery by moulding and by adding on relief forms, must
lead to results different from those which are obtained by painting
or engraving. The use of lines and the habit of using dots or circles
will also effect the resultant style. I think there can be very little
doubt that if an artist is in the habit of using dot designs combined
with lines and if, later on, he tries to represent an animal, this
particular method will be applied in the representation. The origin of
the motor habits must probably be looked for in technical processes,
that of arrangement in the same processes and in the forms of familiar
utensils.
Our examination of the factory-made material shows that the process
of slurring, or slovenly execution leads first of all to the
development of individual characteristics that can best be compared
with handwriting. Pronounced mannerisms permit us to recognize the hand
of the artisan. It is only when an unusually careful and ingenious
person operates with this traditional material that new forms develop,
analogous to those described by Balfour. It will readily be seen that
these conditions are not often realized in primitive society.
[Illustration: Fig. 130. Chinese embroidery representing bats.]
I believe another cause is more potent in bringing about a modification
of design. Ornamental patterns must be adjusted to the decorative field
to which they are applied. It is not often that the artist is satisfied
with representing part of his subject and cutting it off where the
decorative field ends. He will much rather distort and adjust the
parts in such a way that they all fit in the field that he has at his
disposal. When a bird is represented with outspread wings, which would
occupy approximately a square field, and the space to be decorated is
long and narrow, the artist may twist body and tail about, and draw out
the wings and thus squeeze the design into the available space. Henry
Balfour[84] gives as an instance the adjustment of Chinese figures of
bats to an ornamental band (fig. 130).
[84] The Evolution of Art, 1893, p. 50.
The northwest coast Indians, who always take the greatest liberties
with the outer forms of animals, do not hesitate to distort them in
a way that allows the artist to adjust the animal to the decorative
field. Their method will be discussed in greater detail at another
place (see pp. 183 et seq.).
On shell discs from the mounds of Tennessee the rattlesnake is so
represented (fig. 131). The head of the rattlesnake with upturned upper
jaw is readily recognized in fig. 131 _a_. Behind the mouth, the eye
represented by a number of circles, will be seen. The body continues
along the lower rim towards the right and terminates on the left, in
a rattle. The analogy between fig. _b_ and _a_ is easily recognized;
the essential difference consists in the fact that the body in _b_
is undecorated; the rattle lies just over and behind the eye. Figure
_c_ still preserves the same form, but added to the decorations found
in the preceding specimens, is the long loop with small circles
surrounding the eye. The position in _d_ is slightly changed; the eye
will be easily recognized and just to the right of it, more upright
than in the preceding specimens, is the mouth with a fang. The body
is in the same position as before, following the rim of the disk and
ending in a rattle. In _e_ the mouth is very much shortened and the eye
is reduced to a single small circle, while the body and tail retain
their characteristic features.
[Illustration: Fig. 131. Shell with representation of rattle snake.]
The formal elements which were discussed in the beginning of this book
exert a far-reaching influence upon decorative forms. The exigencies of
symmetry within a decorative field require adjustments which may modify
the representative form considerably. A peculiar effect of inverted
symmetry may be seen in those Borneo shields in which the whole shield
represents the face of a demon; one half to the left of the vertical
middle line right side up, the other half upside down.[85]
[85] A. R. Hein, Die bildende Künste bei den Dayaks of Borneo, figs.
48, 49, 51.
Much more potent than the necessities of formal adjustment is the
symbolic tendency which is liable to lead to abbreviations in which
the representation is reduced to the slightest indications. In our
former discussions we have seen that symbolic representations are
very common and that it happens in many cases that the symbol itself
is represented in a more or less perspective way. Wherever the art of
the people wavers between the symbolic and representative modes of
delineation, opportunity arises for the occurrence of realistic and
abbreviated forms, side by side. To this class belong the face urns,
the prehistoric representations of human figures in stone, and even
our busts and portraits may be considered as continuing this practice,
for they are fragmentary in so far as they show only that part of the
body in which we find the character of the individual most clearly
expressed; in part, because the rest of the body is always covered by
inexpressive clothing which hides whatever individuality may exist.
The principal character of forms of this type will be the tendency to
suggest an object by the indication of a few of the most characteristic
traits. It would probably be erroneous to speak in these cases of a
gradual breaking up of the realistic form and the development from it
of a conventional form, for this is not what actually happens. The two
types occur side by side.