McDermott - Self-Representation in Upper Paleolithic Female Figurines

Previous: Stylistic Variability and Choices in Visual Information 

Comparing Modern Bodies with Prehistoric Artifacts

There is an obvious relationship to be seen between the stylistic attributes of Upper Paleolithic representations of the female body in general and PKG-style images in particular and the structural regularities of form and content contained in those minimal viewpoints needed by a woman to see her own body. Personal experimentation will demonstrate that, without external technological assistance, a reasonably inclusive inventory requires at least five or six primary vistas: (1) head and face, (2) superior anterior or upper frontal surface of body, (3) inferior anterior or lower frontal surface of body, (4) inferior lateral or lower side surface of body, and (5) inferior posterior surface of body, including (a) under-the-arm views and (b) an over-the-shoulder view. 

1. Faceless heads. Although the seat of visual self-awareness, the objective appearance of the head and face is simply not visible from a self-viewing perspective. This logically explains why -- although there are regional variations in shape, size, and position in the heads of PKG-style pieces -- virtually all are rendered without facial features and most seem turned down toward the body as if to bring it into view12. The absence of direct visual knowledge may also explain why the most commonly encountered form of head is a generalized round shape vaguely reminiscent of an emergent mushroom "cap" or "button." Not only is this form found on the best-preserved French, Austrian, and Russian figurines but it predominates among fragments, strongly indicating that most missing heads should be similarly reconstructed (Abramova 1967b:pls. 9 and 10). Its stylistic dominance is further supported by its presence on several variant figurines made from mammoth phalanges or metacarpals, thought to represent squatting pregnant women, from Predmosti and Avdeevo (Jelinek 1975:figs.642, 643) 

With the head held upright, the body is absent from the visual field.l3 This discontinuity, in conjunction with the elemental fact that the human eye and self-consciousness alike reside in the head, reinforces the identification of numerous European Upper Paleolithic pieces, sometimes consisting of little more than a rounded button or caplike "head" at one end of a rod or tusk, as either abbreviated or incomplete human figures. Three lines of evidence support this possibility. First, similar undefined button-like heads at the ends of suggestively shaped rods of Aurignacian provenance, such as those from Abri Cellier and Vogelherd, could be earlier efforts at creating a full-length image of the human body (Delporte 1993a:fig. 121; White 1989:98). Second, on the basis of decorative motifs shared with finished figurines, Gvozdover has convincingly identified Kostenkian rods with stylistically similar rounded ends as abstracted or schematic female images (I989a). The third is the frequent identification of what are seen as preliminary sketches that could easily be figurines interrupted or abandoned at some stage prior to completion (McDermott 1985:270). In fact, Praslov (1985:182) claims that sufficient unfinished examples have been found on the Russian Plain to allow him to follow the different stages in making eastern PKG-style figurines from "initial cutting to final polish." The existence of a common fabrication process which begins with the major horizontal divisions of the body rather than with its outline or silhouette could be logically related to the sequential bending of the body necessary for direct visual self-knowledge. 

Sometimes these sketches are little more than tusks with a possible head differentiated at the narrow end, such as Pavlov no. 32460 (B. Klima, personal communication, August 9, 1988) and Avdeevo no. 4 (Abramova 1967b:pl. 27), or ivory rods with a button or caplike "head" at one end as seen in earlier Aurignacian examples (see fig. 4). 

FIG.4. Aurignacian (a, b) and PKG-style (c-f) button or caplike "headed" ivory rods' sketches, and unfinished figurines suggestive of a time-factored fabrication process beginning with the head. a, Abri Cellier; b, Vogelherd; c, Pavlov; d, Gagarino sketch; e, Brassempony "girl"; f, short figure from Gagarino ivory rod containing two unfinished figurines.
The latter category includes the "doll" sketche from Brassempouy, one of the sketches reported from Gagarino (Delporte 1993a:figs. 13, 187), and a similar piece from Pavlov (Marshack 1991a:fig. 163). Although long associated with finished female statuettes, such pieces actually possess no primary or secondary sexual characteristics. Marshack has argued that these and other sketches were made rapidly for a specific one-time use (1991a:287) and never intended to be finished. Although logical, such a conclusion implies a knowledge of motivation which we in fact do not have. It would be best to restrict questions of procedure to those pieces that clearly reflect a common process. What we know is that some pieces definitely represent unfinished female figurines at different stages of completion and that ivory rods or tusks with rounded buttons or caplike "heads" could represent an even earlier stage in this fabrication process. The unusual ivory rod containing two figurines joined at the head from Gagarino, for example, clearly shows different stages of carving in each figure (Tarassov 1971), with the shorter figure having legs and abdomen more differentiated than the taller. A comparable "in-process quality" is clearly seen in Kostenki statuette no. 5 and Khotylevo no. 3 in the east and the Brassempouy "girl" in the west (Delporte 1993a: figs. 5, 11, 170, 203). Similar roughed-out development is seen in fragments of the lower body preserved at Brassempouy and Gagarino (Delporte 1993a:figs. 6, 196). 

It is possible that fabrication of a human figurine involved first differentiating a "head" from a "body" of material and then following an essentially logical time-factored sequence which might remain unfinished. Both the autogenous hypothesis and the evidence of these pieces, if they are unfinished, predict that this emergent process began with the head, the seat of visual self-awareness, and then employed the sequential movements necessary for complete visual self-inspection with attention focused last on the central parts of a woman's body involved in reproduction. Pregnancy and self-inspection both involve sequential stages whose typical time-factored progress might well be revealed in the processes preserved in unfinished pieces. During pregnancy, some parts of the body change while others remain the same, and the parts which undergo the most change appear to be defined last in the fabrication process. 

2. Superior anterior or upper frontal surface of body. Standing erect with the head bowed presents to a woman's eye a strongly foreshortened view of the upper frontal surface of the thorax and abdomen, while the breasts, being close to the eyes, will loom large in the visual field. Creation from this perspective provides a parsimonious explanation for the voluminousness and distinctive pendulous elongation routinely observed in the breasts of PKG-style figurines.14 When looked at from above, as a woman observes herself, the breasts of PKG-style figurines assume the natural proportions of the average modern woman of childbearing age. For example, the dimensions of the breasts of the off-illustrated Venus of Willendorf are comparable to those of a 26-year-old mother-to-be with a 34C bust (see fig. 5). When foreshortened from above, even the apparent hypertrophic dimensions of the Venus of Lespugue and the best-preserved figurine from Dolní Vestonice enter into a reasonably normal, albeit buxom, range (see fig. 6). In addition, the fact that the true thickness of the upper body cannot be experienced by self-viewing is logically consistent with the abnormal thinness seen in the torsos of many PKG-style figurines (see fig. 3). 

FIG. 5. Autogenous visual information of the upper body. Top, photographic simulation of what a six-months-pregnant 26-year-old Caucasian female of average weight sees when looking down while standing erect; bottom, same view of Willendorf no. 1 (cast).
FIG. 6. Oblique aerial views of front body surfaces. Top, 30-year-old Caucasian female, four months pregnant; bottom, same view of figurine from Lespugue (cast).

When viewed from above, most other apparent anatomical distortions or omissions of the upper body undergo similar realistic transformations. For example, PKG-style figurines commonly have what seems to be only an ill-proportioned, sharply tapering fragment of the upper arm represented, with the forearm merging into the side of the body. However, in looking down with arms at the side, a woman does see only the foreshortened front surface of her upper arm, with the forearms normally occluded below the breasts. Another convention explained by the foreshortening and occluding effect of a self-viewing perspective is the unnaturally large, elliptical navel located too close to the pubic triangle in several figurines.15 The annular depression surrounding the navel proper, seen obliquely from above, projects just this size ellipse, and when pregnant a woman cannot see the abdomen below the navel. 

Also, the dual role of hands and arms as both agent of fabrication and model could relate to their variability and infrequent representation. Being in constant motion, they have no fixed point of regard in the visual field and perhaps in human memory. When arm and hands are crossed over the breasts, they present their narrowest aspect to the eye in an edge-on view, which suggests a rational origin for even the unusual thin "filiform" or threadlike arms of the well-known pieces from Lespugue and Willendorf. 16 

3. Inferior anterior or lower frontal surface of body.  A correctly foreshortened representation of the lower body seen from above would shrink or narrow toward the feet as if its true height had been compressed. Only the autogenous hypothesis renders sensible the compressed stature (or atrophy) of the lower body, including the diminutive feet, preserved in some PKG-style figurines. The lower body and feet are optically correct for the point of view employed in their representation.17 

It is also a fact that for a pregnant woman, inspection of the upper "half" of the body terminates at the navel with the curving outline of the distending abdomen. She must bend at the waist to bring her lower "half" into view. Thus the gravid female's direct visual experience of her full-length body involves combining two discrete views which meet at the abdomen near the level of the navel-which also, contrary to anatomical fact, appears to be the widest part of the body. When she looks down over the intervening mass of her growing abdomen, she does not see that the vertical midpoint and greatest physical width of her body really intersect at the level of the hip joint. The apparent misrepresentation of height and width routinely seen in PKG-style images is actually a sensible symmetrical combination of these otherwise discontinuous views. The necessity of uniting the two views from above and below the intervening mass of the woman's pregnant abdomen apparently produced the recurrent "lozenge composition" and the apparently incorrect proportions on which it is based (see fig 7). 

FIG. 7. Leroi-Gourhan's "lozenge composition," a product of the mental combination necessary to create a full-length image from the separate views required by female self-inspection of front body surfaces.

4. Inferior lateral or lower side of body. When one rotates at the hips and raises the arm to look down obliquely in front of the shoulder, one sees the side of the body as expanding from the lower torso toward the buttocks before contracting as the eye encounters the more distant rectus femoris and vastus lateralis muscles of the thigh and the bulging gastrocnemius of the calf. The feet may or may not be visible, often being occluded by the intervening body, particularly the more rearward the angle of regard. The apparent cantilevering of the rectus femoris in front of the lower gastrocnemius is identical with the "bent-knee" posture seen in numerous otherwise erect Upper Paleolithic images of the human figure (see fig. 8). This oblique outline of the lower side not only coincides with the arrangement of muscles seen in this region for PKG-style images, but its content is identical with the information contained in the so-called buttocks or profile image which dominates the Magdalenian (Rosenfeld 1977:90; Bosinski and Fischer 1974; Bosinski 1991). The typical absence of the upper body, shoulders, arms, and head from the visual field when one looks down upon the inferior lateral surface of the body is congruent with their conspicuous absence in this later category of image. 

FIG. 8. Autogenous visual information of lower side of body. Top, photographic simulation of modern woman's view; bottom, same view of Willendorf no. 1 (cast).
5. Inferior posterior surface of body. There are only two ways to bring the remaining dorsal surfaces of the body into direct vision-either by continuing to rotate the line of sight under the arm, thus bringing the caudal aspects of the back into sight, or to crane one's neck to look back over the shoulder. It is the autogenous form and content of these two approaches which renders comprehensible two categories of supposed anatomical distortions previously recognized in PKG-style female images (see fig. 9): the rarely encountered rearward or posterior fatty enlargement of the buttocks properly called steatopygia and the far more commonly encountered lateral deposits of adipose tissue resembling fat thighs or riding jodhpurs known as steatotrochanteria or steatomeria (Duhard 1988, 1991; Regnault 1924). 
FIG. 9. Tracings of photographs of PKG-style figurines seen from above, showing lateral displacement of posterior masses (a, c, d) and rearward projection (b). From Grimaldi, a, "yellow steatite statuette"; b, "punchinello"; c, '1ozenge"; from Brassempony' d,"dagger handle.
5a. Under-the-arm views. Depending on the effort expended in rotating and looking under the arm, the view will either be limited to a lateral segment of the lower back above the sacral triangle (tailbone) or, with greater exertion, may also include a foreshortened outline of the upper buttock below the tailbone. With or without maximum rotation, the view of this region will be dominated by the lateral bulge of the glutei medii, while the more distal glutei maximi) are either occluded entirely (with minimal rotational effort) or seen only as a foreshortened fragment (with greater rotational exertion). Thus, judging by the position of the sacral triangle, what have often been seen as unnaturally large, elevated buttocks are in fact realistic renderings of the glutei medii, properly positioned above instead of below the tailbone in the self-viewing visual field. 

Intergroup variation in the rotational effort expended in self-inspection could thus explain not only the general lateral displacement of mass that has been called steatotrochanteria or steatomeria but the observed continuum of regional variation in this "condition" as well. Many Russian pieces appear to have unnaturally long loins, flanks, or glutei medii above the sacral triangle and atrophied or disproportionately short buttocks below (Leroi-Gourhan 1968a:520), as would be consistent with considerable rotational effort. Less effort produces the complete occlusion of the buttocks below the tailbone, and this is the key to understanding an even more enigmatic distortion found farther west -- the representation of supposedly "upside down" buttocks (Luquet 1934:434-35). In the well-known ivory figurine from Lespugue, the figurine in yellow steatite from Grimaldi, the shattered ivory torso from Brassempouy known as the "dagger handle," and a fragment of fired clay found by Klima at Pavlov, a bar or bridge of material presumably representing the tailbone lies below the apparent gluteal cleavage separating the buttocks rather than above as would be anatomically correct (see figs. 3a, c). From a self-viewing perspective, what has been seen as the gluteal cleavage between the buttocks emerges instead as the furrow of the lower spine separating the lateral glutei medii. The actual gluteal groove and the buttocks proper, which objectively extend below the tailbone, have not been represented at all, since they are in fact completely occluded in anything less than the maximum possible rotation of the head and eyes to look under the arm. Figurines with what appear to be "upside down" buttocks actually correctly represent what can be seen in an under-the-arm view. As with pieces without faces and with forearms which disappear underneath the breasts, the general principle seems to be that what cannot be seen tends not to be represented. 

An intermediate regional variation in self-inspection routines of the posterior is perhaps preserved in the arbitrary horizontal notch located immediately above the bottom edge of the atrophied "buttocks" of the Venus of Willendorf. This blunt geometric feature, which makes no anatomical sense from any point of view other than the self-viewing18, is optically transformed into a highly naturalistic foreshortened image of the lower back above a properly positioned tailbone carried above an oblique sliver of foreshortened buttocks (see fig. 10). 

FIG.  1O.  Autogenous visual information of buttocks as seen under the arm. Top, photographic simulation of modern woman's view; bottom, same view of Willendorf no. I (cast).
5 b. Over-the-shoulder view. Finally, a more difficult and presumably less frequent route of dorsal self-inspection involves sharply rotating the head, thrusting the chin over the shoulder and peering obliquely downward out of the corner of the eye. It is this view which accounts for the steatopygous form of fatty enlargement. In an over-the-shoulder view the dual masses of the glutei maximi project rearward from the body into the field as in steatopygia, complete with the deep gluteal cleavage separating the buttocks, seen in works from Savignano and Grimaldi ("the punchinello") and Monpazier (see fig. 9, b). Again what had been puzzling extremes of human anatomy become surprisingly realistic when considered from the probable point of view employed by their creators (see fig. 11). Thus, PKG-style images show the most consistent realism or organic verisimilitude when conscientiously examined from a retinal angle and distance that mimics those required for inspecting one's own body. What have been seen as gross corpulence, puzzling anatomical omissions, and exaggerated distortions become instead orderly conventions for representing the foreshortened configuration of subjective optical reality. 
FIG.  11.  Photographic simulation of modern woman’s view of buttocks as seen over the shoulder.