5

COUNTERINTUITIVE IMAGES AND THE MECHANICAL ARTS

Now the human understanding is infected by the sight of what takes place in the mechanical arts, in which the alteration of bodies proceeds chiefly by composition or separation, and so imagines that something similar goes on in the universal nature of things.

—Francis Bacon, The New Organon
[Novum Organum 1620], Works I: 47–69

Imagine that the composite figures I have been describing were not just images of fictitious animals, but actual sentient beings. Their adaptive strategies and evolution could now be described in some detail. We would observe how they have subsisted within a wide range of environments, and that their origins probably extend back as far as our own. But for much of their early prehistory, the relationship of these creatures to humans was largely one of avoidance. They remained generally very rare and special kinds of animals, only occasionally seen, even by those few people who became experts in their patterns of behavior. All of this changed with the emergence of a new and complex type of ecology, around six thousand years ago. Urban and state-like societies offered a setting in which composites could, for the first time, thrive and multiply in significant numbers, leaving a clear and striking taphonomic imprint on the archaeological record.

Clearly not all varieties of composite animal thrived to the same extent, even within their newly discovered niche. Cases of “arrested development” are not difficult to find. One such case may be illustrated by an atypical cuneiform text, known as the Babylonian Göttertypentext. Most probably composed in the second millennium BC, it survives only in Neo-Assyrian copies.1 The text gives a detailed technical program—of just the kind recommended by Leonardo da Vinci (see chapter 2)—for the depiction of composite beings, and uses them in a very unusual way to personify nouns.² The following image was intended, for instance, to personify “Grief”:

The head is supplied with a cap and horns of an ox; he is supplied with ears of an ox; the hair hangs loose over his back; the face is that of a woman; his hands are those of a man; he is supplied with wings and his hands are stretched outwards towards his wings; the naked body is that of a woman; his feet stand in the huppu posture.³

As has often been noted, however, almost none of the composites described in the Göttertypentext are attested within the extensive record of Mesopotamian art. Another, and perhaps clearer, case of limited reproduction derives from a palace workshop on Bronze Age Crete. It concerns the miniature creations of the so-called Zakros Master—hyper-composites, combining manifold elements of human and animal bodies—which appear never to have traveled far beyond their point of inception, despite clear attempts to replicate particular images on multiple seals (figure 5.1).⁴ If composites, to paraphrase Rudolph Arnheim, “can be bred the cheap or the hard way,” then we must also acknowledge that some remained stillborn.

At the other end of the distributional spectrum, we find the contrasting phenomenon of supermobile composites, hopping cultural boundaries with astonishing bravado, and leaving behind them a trail of connections that charts the frontiers of an expanding urban economy. It is unclear how the concept of “fittingness,” introduced in the previous chapter, could account for this particular aspect of their distribution. Unlike, for example, the invention of the wheel (another highly contagious feature of Bronze Age culture), the particular form taken by composite figures is unconstrained by technological function. How, then, are we to explain the frequency with which particular composites moved, often in only slightly modified forms, across cultural frontiers? What kind of constraints were acting on their transmission? What was the source of their intercultural stability and appeal?

5.1. Hyper-composite figures impressed onto clay sealings, found at the palace town of Zakros on Crete, mid-second millennium BC (after Weingarten 1983, pls. 20–25; courtesy Judith Weingarten).

To address these questions, we need a sharper picture of the processes through which composites spread among different regions. In the following chapter, I will propose some distinct patterns of transmission that are attested across multiple chronological periods and regional settings, shedding further light on the institutional contexts of image transmission in the Bronze and Iron Ages. First, however, more needs to be said about the strange “reproductive habits” of the composite animal, and its strong association with mechanical modes of image making. It may be instructive here to step outside the realm of composites proper and consider another type of image, which occupies a middle-ground between masking and the depiction of imaginary beings: the face of Humbaba, legendary guardian of the Cedar Forest defeated by Gilgamesh in battle, exemplified here by a cast terracotta plaque of the second millennium BC from central Iraq (figure 5.2a). The face of Humbaba and its distant Greek relative the gorgoneion exhibit remarkably broad distributions in time and space.⁵ Both are to some degree “counterintuitive” images, in which the “usual conventions and typical classifications [of physiognomy] are syncopated and intermixed.” “In disrupting the features that make up a human face,” as Jean-Pierre Vernant wrote of the gorgoneion, “it produces an effect of disconcerting strangeness that expresses a form of the monstrous, oscillating between two extremes: the horror of the terrifying and the hilarity of the grotesque.”⁶ Before asserting a relationship between the limited counterintuitiveness of these images and their cultural catchiness, we should consider the technological background to their dissemination. A recent discovery at Tiryns offers a point of departure.

5.2. Furrowed faces on (a) a cast terracotta plaque from central Iraq, early second millennium BC (after M.-T. Barrelet. 1968. Figurines et reliefs en terre cuite de la Mésopotamie antique. Volume 1. Paris: Paul Geuthner, fig. 759), and (b) on the reconstructed underside of a mold-made faience cup from Tiryns, Greece, thirteenth century BC (courtesy Joseph Maran and Maria Kostoula).

THE MATERIALITY OF (MINIMALLY) COUNTERINTUITIVE FORMS: A VIEW FROM TIRYNS

There are perhaps few better settings for thinking about the cultural significance of monsters, in the broader colloquial sense, than the ruins of Tiryns, overlooking the Gulf of Argos. The Argolid was Bronze Age Greece’s face to the East, just as the Gulf of Corinth—on the other side of an isthmus dominated by the palace at Mycenae—was its face to the West. According to Pausanias, Tiryns became the kingdom of Perseus after his return from the ends of the world with the head of the Gorgon, and his subsequent shaming and exile from Argos.7 The people of Argos claimed that the lethal prize, which retained the power to kill by its stare, was buried beneath their marketplace: a shrewd insurance policy against future ruination, as well as a potent foundation myth. Pausanias’s Description of Greece credited the building of Tiryns’s walls to another race of monsters, the one-eyed giants, or Cyclopes, who undergo a marked transformation in Greek literature, from Homer’s “idle and unruly” troublemakers to the diligent artisans and builders of culture—masters of sculpture and metalworking—that are found in classical sources.8

Dating to the thirteenth century BC, the remains of a multimedia workshop are among the recent finds from a destruction layer in the Lower Citadel of Tiryns.⁹ The workshop specialized in producing composite faience and sheet metal objects: exotic techniques, imported from the east, and conceivably practiced here by foreign craft-workers. Its products included a distinctive range of head-shaped vessels, the basic forms of which were achieved by pressing successive layers of faience into an intaglio mold. The resulting container was fired to a dark glaze, painted, and ornamented with inlaid eyeballs and perhaps also gold foil attachments. Composite vessels of this kind are rare in Bronze Age Greece, but good parallels are known from urban centers on Cyprus and the Levantine coast, whose trade links to the Argolid are attested in a variety of luxury imports at Tiryns and other Mycenaean palaces.¹⁰

Having no base, the cups were designed for rapid drinking, probably of wine, in which the act of consumption—from a spectator’s point of view—replaced the face of the drinker with that of the creature depicted on the cup. As practical objects, they united the ritual techniques of drinking and masking, both of which imply submission to a public act of metamorphosis. One particular vessel from Tiryns is distinguished by a grotesquely furrowed face, and an unusual aperture in its gaping mouth which could be blocked with a finger while drinking, or used for theatrical effect to emit a gory stream of red liquid, perhaps accompanied by a gargling sound (figure 5.2b). Parallels have been drawn with a contemporary series of terracotta masks that were found on Cyprus, and these in turn echo Near Eastern depictions of the ogre Humbaba, whose head (bleeding heavily, no doubt, from the mouth) was taken in combat by Gilgamesh, and transformed into a protective image.11

Both the faience vessels and the terracotta masks they resemble were most often created by pressing soft material into a reusable mold, which allowed precise replication of a complex—and counterintuitive—version of a face.¹² A late, and apparently isolated, revival of the same technical procedure occurs in the Peloponnese during the Iron Age, when cast gorgonesque masks appear prominently among dedications from the Spartan sanctuary of Artemis Orthia.¹³ Its deep origins, however, lie in Mesopotamia, where the one-piece mold was first applied to the reproduction of terracotta images at the end of the third millennium BC and where grotesque faces were produced by this method no later than the beginning of the second, together with a range of composite figures.¹⁴ In considering the popularity of this particular type of image—its ability to become, as Pascal Boyer puts it, “both relatively stable within a group and recurrent among different groups”—we can hardly afford to ignore the material procedures through which it was replicated.

A FURTHER PARADOX IN THE EPIDEMIOLOGY OF CULTURE

From the standpoint of an “epidemiological” approach to culture (see chapter 2), the close association between mechanical techniques of image production and the spread of (minimally) counterintuitive forms is puzzling. It implies a strong element of redundancy—a kind of superfluous cultural prosthesis to cognitive predispositions that are already biased towards the reception of such images. The paradox becomes more acute when we take into account that Mesopotamia, the heartland of composite animals, is also the region where mechanical methods were first widely applied to the reproduction of images, via stamp and cylinder seals, initially, and then also via the terracotta mold (see chapter 4).¹⁵ Moreover, the subsequent proliferation of composites during the Bronze Age follows closely the spread, to neighboring parts of the Old World, of mechanical modes of image production, and of the palace and temple institutions where they were used to promulgate officially sanctioned signs.16

Within large-scale social formations (where the origin of commodities was increasingly mysterious and remote), the use of image-bearing seals to mark and secure the containers of goods, taken into the body and household, provided evidence of elite claims to secure the materials of life and wellbeing. For millennia, in the ancient Near East and Mediterranean, sealing devices formed the “canvas” on which an ever-evolving imagery of power and protection unfolded: the comings and goings of gods and goddesses, fabulous animals and kings, intricate scenes of industry and travel, feasting and worship, conquest and defeat—all the dazzling visual innovations of the urban state, harnessed to the most elementary social activity of them all, the sharing and distribution of food and drink. Sealing practices created extended webs of accountability between producers and consumers, while the highly specialized techniques of seal manufacture (using intaglio carving and, from the second millennium BC, wheel-cutting of miniature designs) ensured centralized control over the circulation and modification of designs.¹⁷

Such practices had a dual effect on the dissemination of images, comparable in some respects to that of the printing press in late medieval and early modern Europe.¹⁸ They provided a method for accelerating the replication of powerful visual formulae, simultaneously restricting their production to a small group of artisans and their elite patrons. This use of mechanical image production to fix ritual values within stable media of transmission, extending their dissemination beyond ephemeral performances into the spaces of everyday transactions, is a distinctive and neglected feature of early state formation in the western Old World. It can be counted among those cultural strategies through which elite groups made “legible” their cosmological and political roles in society,¹⁹ and it has a direct bearing on the distribution of composite figures in the visual record. Cognition may render us peculiarly susceptible to such images, but viewed in this light their dissemination seems more closely reliant upon technological and political forces.

Might the argument for nurture over nature be pushed further still? Ernst Gombrich, in The Sense of Order, appears to suggest just such a possibility, claiming that pictures of imaginary, composite animals—far from being cognitively infectious—confront unusual obstacles to transmission. Of medieval European grotesques he wrote:

There are no names in our language, no categories in our thought, to come to grips with this elusive dream-imagery in which “all things are mixed.” It outrages both our “sense of order” and our search for meaning . . . [those] tendencies which dominate our perception. They alone enable us to form those expectations which can be subsequently confirmed or refuted and which thus permit us to “make sense” of the information we pick up with our eyes. . . . Not only do the limbs of these composite creatures defy our classifications, often we cannot even tell where they begin or end—they are not individuals, because their bodies merge or join. . . . There is nothing to hold on to, nothing fixed, the deformitas is hard to “code” and harder still to remember, for everything is in flux.20

We are on the verge of a seductive hypothesis, which must be considered if only to be immediately rejected: is mechanical replication in some way a necessary supplement to our innate capacities for recalling and faithfully reproducing counterfactual images on an extended spatial and temporal scale? Let us admit that the idea has its attractions. It provides an economical, if only partial,²¹ explanation for the sparseness of composites in the prehistoric visual record, and it tallies reasonably well with their observable spread in later periods. Gombrich, however, offered no evidence—experimental or otherwise—to support the notion that pictures of imaginary creatures might be more difficult to remember and replicate than pictures of living species, and no such evidence has to my knowledge been supplied before or since.22 Furthermore, any attempt to formulate a causal or universal equation between the use of mechanical reproduction and the dissemination of fantastic imagery confronts an important counterexample, within the ambit of early Old World cultures.²³

THE CASE OF CHINESE BRONZES

For Mikhail Rostovtzeff, with whom this study began, China marked the end of a journey on the monster’s tail, and this chapter will follow suit. An association has long been recognized between the earliest development of cities and urban elites in the valley of the Yellow River and the inception of new modes of display that made great use of fantastic imagery.²⁴ A by-now-familiar contrast can be drawn, in this respect, with preurban societies in the same region. The Neolithic Yangshao culture (ca. 5000–3000 BC) possessed a rich tradition of painted ceramics, the decoration of which comprises, for the most part, geometric motifs.²⁵ On those rare occasions when figurative images are present, they depict recognizable animals (such as fish and birds) or schematic human forms.²⁶ The medium in which imaginary creatures first appear on a significant scale is that of cast bronze vessels (figure 5.3a,b), coinciding with the territorial expansion of Shang power in the latter part of the second millennium BC.²⁷

In early China, bronzes were used in the performance of sacrificial offerings, through which descent groups—the basic units of political organization throughout the Bronze Age—sought to maintain favorable relationships with a host of spirit beings.²⁸ Ceremonial bronze vessels, cast with a sophisticated piece-mold technology, were first produced in quantity with the onset of urban life. They remained central components of dynastic culture for more than a thousand years, spanning the period from around 1500 to 400 BC.²⁹ Writing before the stylistic development of ritual bronzes had been established,³⁰ Rostovtzeff proposed a westerly source for the fantastic creatures that appear on Shang and Zhou vessels—in particular, the taotie (figure 5.3c), which he confidently referred to as a “Chinese ogre mask”; another design that he called an “eagle-griffin”; and a third that he designated a “dragon.” “The main problem of the Chinese animal style,” he wrote in 1929, “is to solve the question of the origin of these three fantastic monsters. Is it possible that the same figures, with the same peculiarities, were invented independently both in the Near East and the Far East?.” “Have the Mesopotamian fantastic animals and those of China a common origin?’³¹

5.3. Bronze vessels from (a) Yueyang, Hunan, and (b) Sanxingdui (Shang period, thirteenth to twelfth century BC, height 50 cm; after drawing by Li Xiating in R. Bagley. 1999. Shang archaeology. In The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C., ed. M. Loewe and E. L. Shaughnessy, pp. 124–231. Cambridge, UK: Cambridge University Press), and (c) rubbing of taotie motif (after Rawson 1987: 27, fig. 9.2).

Chronological impossibilities aside, it has since been suggested that the fantastic creatures of early Chinese art are constructed on quite different principles to their counterparts in the ancient Near East and Mediterranean. Robert Bagley notes a long-standing practice in Chinese art history of viewing prehistoric and Bronze Age images through the retrospective lens of much later, Han-period designs, with their iconic depictions of composite animals.32 Even the enigmatic carvings on Neolithic jades, popular as grave goods on the east coast of China from the fifth to the third millennia BC, have sometimes been confidently identified as dragon motifs.³³ The dragons of Han art, as Bagley points out, “are not materially different from the ones we see today on placemats in Chinese restaurants.” But their remote ancestors, he suggests, “were very different indeed, different in the imaginative processes that produced them and different in their elusive, utterly unfamiliar relationship to animals of the real world.”³⁴

Accordingly to Bagley, figural designs on Shang and Western Zhou bronzes should be clearly distinguished from the composites of ancient Near Eastern art.³⁵ The examples of imaginary animals that he considers do not result from clearly articulated variations on animal anatomy. Instead they comprise tantalizing suggestions of organic form that, on closer scrutiny, dissolve back into larger patterns of interlaced lines and projections. In other words, they exemplify just the kind of visual elusiveness that so irritated Gombrich in The Sense of Order.³⁶ Other historians of Chinese art have taken a different approach. Ledderose describes the taotie in terms much closer to those I have used for the design of imaginary composites in the western Old World. “The anatomical parts of the taotie,” he proposes, “may be regarded as modules in a decorative system . . . they are composite, interchangeable parts combined into units.” The taotie, he suggests, can be constructed from up to ten standard bodily elements (such as body, eyes, nose, horns, and claws), which retain their individual shapes even when detached from the body and left “floating in a sea of spirals.”37

From a comparative perspective, and within the wider terms of my discussion, what stands out is the systematic avoidance by Chinese bronze casters of mechanical techniques for the replication of complex visual designs, despite the availability of such methods in other media of lower status.³⁸ Decorative images were painstakingly carved by hand either onto the core from which casting molds were formed³⁹ or onto individual mold sections.⁴⁰ The overall system of manufacture was modular but not mechanical, and reflects the wider principles on which urban crafts were organized:

For decoration the bronze makers developed a modular system that allowed them to assemble countless combinations from a limited repertoire of motifs and compartments. They also devised a modular technical system for casting their vessels. These systems offered the best solution to the task that the ancient Chinese had set for themselves: to produce high-quality bronzes in sets. . . . To have religious and political life function properly, the Shang aristocracy may have needed in, say, the twelfth century BC, sets of ritual bronzes totaling several thousand units. The bronze makers met the demand by devising modular systems. Modular products lend themselves to a division of labor. They are most smoothly and efficiently fabricated in a production system in which the work is compartmentalized.⁴¹

Only from around the eighth century BC onward were clay pattern blocks, such as those discovered at the Houma foundry in Shanxi Province, used to impress designs mechanically onto the mold sections from which bronzes were cast.⁴² The adoption of this mechanical technique coincides broadly with the absorption, into an existing visual repertory, of restricted elements from a foreign pictorial style that derived ultimately from urban centers of the Near

East.43 Carried onto the Central Chinese Plain through interaction with nomadic populations, these foreign elements included new ways of constructing composite animals and arranging them in confrontational poses that were formerly absent from Chinese art.⁴⁴ In no way, however, can the earlier avoidance of mechanical replication techniques in bronze casting be said to have impeded the dissemination of standard templates for the visualization of imaginary beings, and the overall impact of mechanical reproduction upon this longe-stablished tradition appears to have been quite minimal.