Postscript: Finding the Invisible
A tough wedge must be sought for a tough log.
—Publilius Syrus, Sententiae
Past scholars have generally dismissed the history of easily perishable commodities like cloth as unreconstructable, on the ground that there was no evidence. By tracking down a great deal of evidence from unusual sources, however, we have reconstructed much about ancient textiles and the people and societies that made them.
Women’s work consisted largely of making perishables—especially food and clothing. So if we are to retrieve significant amounts of women’s history, or of the history of any evanescent occupation in particular (and I am thinking of such things as music and dance as well as food and clothing), we need better evidence than just that which falls into our laps. We need the skill to glean all surviving evidence and to wring out of it every last drop of information and useful analysis. A hypothesis, after all, is no better than the evidence that supports it, and hypotheses without evidence are mere wishful thinking. A tall order? Perhaps, but also a delightful challenge to those who, like Agatha Christie’s immortal Poirot, enjoy “exercising the little grey cells” in the chase.
Let us start by asking: What are the general directions from which the evidence for such ancient objects and activities might come? The conventional sources are archaeology and surviving texts. But the problem with archaeological remains is that what we want to study in our case may not survive directly or in a recognizable form. Cloth itself, for example, seldom makes it through the millennia except in tiny, hardly recognizable shreds. Until recently excavators tended to throw even those away, assuming they were of no value. Loom weights, on the contrary, survived in great quantities but were also assumed to hold little information, so their fate was just as bad. It didn’t occur to either diggers or scholars how much these unprepossessing blobs of clay could reveal about the development of the looms needed to weave the cloth and about the users of the looms.
The problem with texts, moreover, is twofold. They may not discuss what we wish to learn—for example, few ancient texts talk about women’s lives, partly because very few women wrote or dictated texts—and even when they do touch the subject we want, they may not tell us what we need to know. Thus the economic texts that discuss clothing use so many unknown technical terms that we learn next to nothing about clothing from them. The scribes knew what they meant. Why should they think of explaining the details to us, three or four millennia later and half a world away?
Archaeology, however, has gone through revolutions in the last century. For a long while it was little more than a branch of art history, its task being to fill up art museums and private collections with handsome curiosities. (Those whose acquaintance with archaeology comes only from films like Raiders of the Lost Ark may think it is still that way; the film may be great entertainment, but it is terrible archaeology.) Around the turn of the century, however, archaeology began to pull itself up to the level of an investigative science, through the efforts of a few individuals who wanted to learn how ancient people lived. Scholars such as the great Egyptologist Sir W. M. F. Petrie were realizing that in removing an antiquity from its context the finder (whether “scientist” or “treasure hunter”) destroyed forever any social information recoverable from the find group. The taker therefore had a duty to humanity to record everything about that context, no matter how small.
This new view leapt to international fame through the notorious act of a French specialist in the Coptic period (the early centuries A.D.), named Émile Amélineau. From 1894 to 1898 he obtained permission to excavate the incomparable tombs of the earliest Egyptian pharaohs at Abydos, who flourished around 3000 B.C. After ransacking them of their contents, keeping little record of what was found where, Amélineau deliberately burned or smashed to bits any and every object that he chose not to take back to France, so as to make those he took more valuable, because unique. The burning lasted for days. A horrified Petrie rushed in as soon as Amélineau had left, to glean “a rich harvest of history . . . from the site which was said to be exhausted,” by carefully sifting through the rubble and putting back together what he could of the five hundred years of pivotal human history that the treasure hunters had wantonly shattered. (This was, after all, the dawn of human writing and civilization.) If it did nothing else, l’affaire Amélineau (as it came to be called) finally got people’s attention, and a more responsible archaeology began to develop, characterized by the sort of exhaustive recording that Petrie specialized in.
Such changes in basic approach took time, however, and Petrie ended up putting into his own collection (now in the Petrie Museum at University College London) many types of artifacts that no one else considered of value yet. Thus it was that in the 1970s two women, an Egyptologist and a textile conservationist, found treasure in “a tumbled heap of dirty linens” among the masses of labeled items stored in the Petrie collection: the earliest complete garment that has come down to us (see fig. 5.3). It is a fine off-white linen shirt of the First Dynasty. Its seams, fringes, and elaborate pleating are intact and still show the creases at the elbows that remained when its owner last stripped it off over his head, five thousand years ago. Petrie had seen the value in this linen shirt and trusted that eventually textile history would come into its own.
Even at its blindest, when diggers of the mid-twentieth century were dutifully recording facts of no known use just because they were there, the new method of trying hard to preserve all objects and information found has proved its worth magnificently. In 1961 Emmett Bennett was able to reconstruct the filing system of the Mycenaean scribes, hence a great deal about their economic practices and even the geography of the kingdom, because the excavators of the Pylos archives decades before had meticulously recorded the exact three-dimensional findspot of every as-yet-undeciphered tablet scattered through the fill.1 We relied on some of this information in reconstructing the Mycenaean textile industry in Chapter 9.
But a second revolution was coming, too, one that radically increased the amount of information recoverable from what was dug up. After World War II a wide range of military technology and scientific knowledge that had accrued during the great struggle gradually became available for other uses, including the interpretation of archaeological data. Radiocarbon dating, infrared photography for seeing through unremovable dirt, isotope “fingerprinting” for tracing sources of raw materials like stone and ore, thin-layer chromatography for analyzing dyes, and a hundred other methods were and still are being worked out. Thus we have appealed over and over in this book to archaeological information that has been further interpreted through the natural sciences, from chemical analysis of the Lascaux string fragments to palaeobiological reconstruction of the era when sheep became woolly. Among the innovations have been a number of improved means for finding sites and artifacts. Thus magnetometers and other devices for finding metals, stone walls, and other anomalies have helped with archaeological prospecting, while new techniques of flotation and sieving have allowed excavators to find the tiny seeds, leaves, bone fragments, and so forth that tell us worlds about ancient environments. We learn what ancient climates were like, that ancient grain supplies were plagued by mice and weeds, and that hemp, a favorite fiber plant, was in use in Europe in the fifth millennium B.C., four thousand years before the narcotic subvariety (marijuana) was brought in from southern Asia. Improved methods of microscopic analysis, special photography, spectrometry, and other nondestructive techniques have added to our knowledge. For example, careful microscopic probing of an Iron Age woolen vest found in the salt mines at Hallstatt, Austria, showed that the ancient owner had been pestered by lice; the garment’s seams were full of lice eggs. Textiles and other readily perishable objects are, as a result of all this, much easier to study now than they were twenty years ago, right within the discipline of archaeology and its scientific helpers.
But we need all the help we can get. What else is out there?
In this book we have often appealed to language itself via linguistic science, in addition to looking at such texts as have come our way. For example, we have used such revealing etymologies as tunic, shirt, and to ret to throw light on the history of clothing and the processes of preparing fibers. (We even added that the English word robe comes from rob because clothing was one of the most frequent forms of plunder in the Middle Ages, as in many another time and place.) The discussion in Chapter 9 of the stages of Mycenaean textile manufacture, too, was based on a careful linguistic analysis of the names of female professions in the Linear B tablets, worked out in detail elsewhere. We also combined archaeological with linguistic arguments to interpret the layering of vocabulary found in Greek, in which only the most primitive aspects of weaving show Indo-European names, the rest having been borrowed later. Since the technology for which they borrowed these terms—weaving on the great warp-weighted loom—is known from the loom weights to have developed in central and southern Europe in the Neolithic, the prehistoric Greeks must have learned the craft fairly late, as they were moving into Greece from farther east.
Another tool we have used in a particularly novel way is the comparative method of reconstruction. This method was gradually worked out in the nineteenth century by linguists interested in determining the historical relationships between languages. They began to realize that many languages were changed later forms of a common ancestor language (like French, Spanish, and Italian, all of which are simply later forms of Latin, each having changed gradually in its own way during many centuries) and that the parent could be reconstructed to a fair extent by meticulous comparison of the structures—especially the sound structures—of the daughter languages. The method works best when the structures used are both arbitrary cultural conventions (as language is) and so habitual that people don’t think about them much. When we talk, we are worrying principally about framing our sentences well enough to get our message through to the other person, not whether our tongue is hitting this or that little spot in the mouth. What gymnastics the tongue has to go through to produce the needed words is its problem, as it were; the tongue is on automatic. The sounds of language are thus excellent fodder for the comparative method. But we have the same sort of dichotomy in the cultural conventions of clothing. On the conscious level we worry about fashion and momentary social messages, but we take for granted and scarcely, if ever, think about the basic notions of what constitutes dress within our culture, including (depending on the culture) what is appropriate for certain social classes, sexes, etc. These “automatic” aspects of clothing yield to comparative reconstruction back into prehistory. Very conservative forms of decoration (see Chapter 6) turn out to be partly reconstructable in this way also, and ancient music and dance may, too. One of the checks we have on such work comes in collecting the most archaic vocabulary connected with these fields and comparing its reconstruction to the reconstructions of the costumes, decoration, dances, etc.2
Other sources of evidence include mythology—a difficult field not widely understood as a potential helpmate for archaeological problems since, having been roundly abused by some, it has been rejected by most.3 Then there is ethnology. Its particular virtue is to suggest possible solutions to archaeological problems by showing parallel behavior in other human cultures. That is, ethnographic studies can help determine the range of possibility and likelihood for what people did, especially if the researchers will take the time and courage to get a firsthand knowledge of whatever they are trying to study. Besides, it could even be fun!
For a century or more, pottery has been the mainstay of archaeological chronologies. It is central to the field. But few archaeologists have ever made a pot—ever kneaded the clay (let alone found it and dug it out), built up the vessel, dried it, decorated it, fired it . . . and watched it come out of the kiln in shattered pieces the first time because as novices they hadn’t gotten all the air out of the clay during the kneading. One learns just what the ancients faced by trying to do what they did, and overnight one’s theories become a great deal more realistic. The same, of course, is true of spinning, weaving, cooking, woodworking, and any other craft.
Avigail Sheffer tells of worrying whether the soft, crumbly, doughnut-shaped weights she was digging up by the dozens at Iron Age sites in Israel could really function as loom weights, as everyone assumed. She thought they would break as they swung around on the loom. So she and her colleagues made up a stack of unbaked clay doughnuts, strung up a makeshift warp-weighted loom with wool hand-spun for them by some local Bedouin women, and began to weave. She reports: “The weaving was very easy and quick, taking less than an hour to produce a piece of material one metre long. No damage occurred to the weights even when the loom had to be moved from place to place.” She learned a lot by trying it out—even more than she had set out to learn—such as how quickly one could weave on such a loom.
Pottery “experts” for a long time could get away with knowing little about how pottery was made, because there was so much else to study among the wealth of potsherds. (Today some scholars are also learning to make pots.) But with the study of perishable objects, we do well to start by learning the craft firsthand and to keep experimenting at every turn as we go along. We need all the practical help we can get because if we take a wrong turn somewhere in the logic of our theories, we don’t have piles of other evidence lying around to warn us of our mistake. I have found that the considerable time it takes to replicate ancient practices is always amply rewarded, as when I rewove the Hallstatt plaid (figs. 0.1 and 0.2). Theories are kept on a sounder footing and new information gathered about the problems and limitations people faced in those days. And there is the pleasure of doing something different, something so old that it is new again.4
Yet another way to get evidence—and most difficult of all—is to develop entirely new sources by looking carefully at the nature of the problem to be studied. For example, archaeologists have long been perplexed by evidence that new groups of people have infiltrated or overrun an area. How to tell whether whole families have arrived or just bands of men—traders, warriors, or the like? The problem diminishes, however, once we have discovered what women were doing that men weren’t and then ask what traces these activities may have left. We have seen that women spent most of their time raising young children and preparing the daily food and household cloth and clothing. We can follow the evidence for those to find the women.
One thing that women typically impart to their children is the first elements of language, including the vocabulary of highest frequency, and we have seen some of the value in analyzing the linguistic inheritance. Second, spinning and weaving were almost always women’s work (except in a few cases of urban specialization; see Chapter 11). In this case we can trace the women by tracking their tools. The job is made easier because the textile tools were mostly very humble—a few sticks, clay spindle whorls, and perhaps clay weights, with no intrinsic value for trade. So when we find Avigail Sheffer’s crumbly clay doughnut weights, well known in Europe and western Anatolia, suddenly turning up in Israel in the Iron Age, far outside the homeland of the warp-weighted loom, we have every right to suspect that a group of women had moved in, along with their families, from rather far to the northwest. One may even suspect a connection with the biblical Philistines, whose pottery (as we said in Chapter 10) is related somehow to that of the Mycenaean Greeks and whom the Egyptians show arriving at the gates of Egypt around 1200 B.C., with husbands, wives, children, and baggage piled on oxcarts and in boats, looking for a home. Fought off and turned back by the Egyptians (who wrote the attackers’ name, consonants only, as P-l-s-t), they settled around Gaza and gave their name to the whole area—Palestine.
On the other hand (as an apparent counterexample that serves to prove the rule), we have no right to assume migration of women when we find fancy little Syrian spindles carved of local ivory turning up occasionally in Mycenaean settlements in Greece in the Late Bronze Age. Themselves handsome, these objects must have sold for a good price—a wonderful trinket for a merchant or sailor to take home to his sweetheart.
Once we have located good sources of evidence, we need to sharpen our ability to make the most of what is there.
The first step, in my experience, is to trick oneself into focusing on every part of the data. Draw it, count it, map it, chart it, and if necessary (or possible) re-create it.
For example—a personal one—I inspected photographs of the Venus of Lespugue a dozen times, but it was not until I made my own tracing (fig. 2.1) that I noticed the marks showing that the strings of her string skirt were fraying out at the bottom, telling me that the sculptor knew of string made from twisted fibers twenty thousand years ago. The act of drawing forced me to pay minute attention to every tiny detail of the statuette for the first time. Similarly, it was not until I decided to color by hand my photocopies of all the known Mycenaean frescoes showing clothing that I began to appreciate how frequently a particular border pattern occurred on the frescoes as well as on the clothes and how easy it would be to weave it. That, in turn, prompted me to try to weave it, and during the relatively slow, step-by-step process of doing so I realized that I was making just the sort of band that I had seen—but not thought about—in black-and-white photos of rural women of this century starting their cloth on a warp-weighted loom in the traditional way. Except for color, the designs were the same between the modern Norwegians and the ancient Mycenaeans.
Mapping and charting help, too. Even after I had worked for ten years collecting the descriptions of every fragment of prehistoric cloth and textile tools that I could find in the archaeological literature of Europe and the Near East, I had no idea that these data separated into three main zones of development—until in desperation I sat down to map the evidence. To my astonishment, the distribution of fiber use and loom types exactly correlated with the types of pattern weaving for which I had data (see Chapter 11).
The acquisition of facts in a tough subject seldom goes in a straight line—that is, it seldom goes where you think you want to go with it. I often have data I don’t know how to use, on the one hand, and on the other hand, I am always missing lots of data I’d like to have. Faced with a pile of opaque evidence, one can usefully ask, “What are all the individual things I can deduce here?” Thus we deduced from the bare bones of a lawsuit against a New Kingdom Egyptian woman (accused of using another woman’s linens to help pay for two slaves; see Chapter 11) that linen could be used as a sort of currency, that women could themselves make such commercial deals as buying slaves, and that women were directly responsible in a court of law. We could also have learned about the price of slaves and the workings of the legal system had we wished. Once the details are systematically chipped loose from the matrix, it is often easier to figure out where they can be usefully filed—these under textile studies, those under women’s rights, and so forth.
On the other hand, the problem of missing data may respond to the query “If I can’t get at it by the straight path, how else can I get at it?” This takes us back to the question of creative use of sources, already discussed. For example, when I began working on textiles, I wanted to know about those specifically from Greece. But textiles don’t survive in Greece. Eventually it became apparent that one could deduce information about Greek cloth from the textile tools left there, from the vocabulary, and from mapping the development of cloth and other related artifacts in neighboring areas.
Finally, we need the best possible methods of analysis for our hard-won evidence. Methods of working from the internal structure of the data and those based on logical deduction are well known. Thus, for example, allover designs can be classed not just randomly by motif but exhaustively according to symmetry types.
Studies also show that our sense of symmetry operates on a far less conscious level than our awareness of motifs and use of color. So when we see that the early Minoans use only Symmetry Type A but presently add to it the use of Symmetry Type B, which happens to be the Mycenaeans’ favorite kind, we can deduce that a much more intimate contact than casual trade has started to occur between the two cultures. In fact, the use of the new symmetry type begins just about when we get evidence for the Mycenaeans’ taking over part of Crete, and the decorative change can be used to help date that take-over.
The pursuit of the perishable, however, requires careful attention to a variety of invisible factors as well, factors that most people forget to consider because in some sense they “aren’t”: The factors aren’t said, or aren’t conscious, or aren’t seen. It is their absence that is the problem, precisely because they don’t obtrude on our attention.
For example, one difficulty in working with texts is that ancient scribes didn’t think to tell us much of what.we most need to know. They already knew all about it. (We moderns are no wiser. When I first went to Greece, no one thought to explain to me that in that country when you nod your head down, it means “yes,” but when you nod your head up, it means “no.” I got on several wrong buses before I began to suspect that Greek head nodding was even a problem to me.) Among our textile discussions in Chapter 11 we investigated the dress that the Athenian women made for Athena’s festival. But working out the problem has been greatly hampered by the fact that Athenian writers never bother to mention which of Athena’s statues on the Acropolis, the big or the small one, was to be clothed in the new peplos. We are stranded to deduce it as best we can.
On the other hand, silent assumptions come quite as often from ourselves. (I had assumed I knew what head nodding meant, just as my Greek friends had assumed that I knew what head nodding meant.) In making their deductions about Athena’s dress, several scholars assumed that, since it took nine months to make, the dress must have been very big and therefore must have adorned the huge statue. The assumption that large size is the only reasonable cause for a long manufacturing time is unwarranted, however, as we see when we explore the technology thoroughly. Statistics of various sorts about weaving show that exactly that kind of time would have been needed to make a storytelling cloth big enough for the small statue. Furthermore, we are told of the setting up of a single weighted warp nine months before the festival. A Greek warp-weighted loom was not equipped to make a cloth much larger than that needed for the small statue.
Finding one’s own unwarranted assumptions is one of the most difficult things to wrestle with, precisely because they are so hard to recognize. Trying simply to state all one’s assumptions explicitly is the first major step. Thus, in the example of Athena’s dress, it helps to say, “I am assuming that the long time needed to make the cloth is due to its size. Is that the only possible cause for lengthy time in weaving?” Put that way, one can begin to see some other possibilities: A small Persian carpet may take years to weave because the method is slow.
Another type of “missing data” can be found by systematic comparison. I had been working for years with the Egyptian material on spinning before I realized that nowhere was there a picture of a distaff in use, and nowhere among the thousands of textile artifacts a surviving distaff. So how were they draft-spinning a fiber that is typically longer than one’s arm? (A distaff holds the unspun fibers and acts basically as an arm extender when the fibers are very long, so that a little group of fibers can be pulled or “drafted” most of the way past the next group, and so forth, to make a continuously long, thin thread.) The answer, I gradually discovered, was that they were not draft-spinning (unwarranted assumption discovered). Instead of pulling the fibers past each other, the Egyptians were separating them entirely and splicing them end to end (see Chapter 8). All the phases of the work were represented in the tomb paintings, but I was not ready to understand the details of what I saw until I had been forced to discard my wrong assumptions. This occurred when I noticed that a crucial element for my theory was missing, as the result of careful comparison to well-known examples of how people spin in most other parts of the world.
The principle is powerful: the hardest thing to notice is what isn’t there—yet it may be every bit as important as what is there, and it takes the most careful of methods to ferret it out. Sherlock Holmes, master of methodical deduction, solves the mystery of the missing racehorse, in “The Adventure of Silver Blaze,” precisely when he realizes that the dog didn’t bark the night the wicked deed was done, hence the villain must have been very well known to the dog.
Finally, none of these methods will be of use unless the researcher is willing to learn what the subject has to say about itself instead of trying to make the topic come out in some predetermined way.
One of the most remarkable scholars I have studied with was Professor Albrecht Goetze of Yale University. Before I met him, I had heard that when his comprehensive book on the archaeology of ancient Anatolia was updated and republished some twenty years after it was first written the revisions consisted mostly of adding references to newly excavated material that further demonstrated the conclusions he had drawn the first time. I wanted to find out how a scholar could become so sensitive to the data as that. At first I felt very frustrated in this because he didn’t talk much about method. But gradually I began to realize that the key was on the wall of his office, in a little hand-lettered sign to which he would often refer, laughing uproariously. It said, in German, “What do I care about my garbage from yesterday?” Each new fact discovered made the picture necessarily look a little different, and he was quite happy to let go of old, outmoded views—the garbage—and move on to a new vision with a joyful laugh of discovery. He never let his ego get in the way of learning, by hanging on to an idea simply because it was his.
We women do not need to conjure a history for ourselves. Facts about women, their work, and their place in society in early times have survived in considerable quantity, if we know how to look for them. Far from being dull and in need of fanciful paint to make it more interesting, this truth is sometimes (as the saying goes) stranger than fiction, a fascinating tale in itself.
1As it happened, the baskets of clay tablets, filed by subject, had fallen off their wooden shelves one by one as the palace, and the shelves, burned. Each basketful was thus scattered in a characteristic fan across the accruing rubble as the destruction progressed, back on that terrible day around 1200 B.C.
The excavations took place in 1939 and the early 1950s, mostly before the decipherment of Linear B, which was worked out largely between 1952 and 1956. Until the tablets were readable, and until Bennett realized that different scribes could be recognized by their handwriting and that each scribe typically handled only one or a few types of accounts, no one had any inkling how useful the exact findspots would turn out to be. Recording them all was just “good excavation method.” What greater triumph can one ask of it?
2A word of warning: This is not an easy task that one can whip off in a few days or weeks. Linguistic reconstruction of sounds and of the words that ride on them, although very simple in principle, is extremely complex in practice because there are so many intricately interlocking details to consider. Similarly, to do these other types of comparative reconstruction well requires enormous amounts of time, patience, and accuracy.
3But for at least one pair of archaeologists who grasped the subject at its core, and wrote about it splendidly, see Henri and Mrs. H. A. Frankfort, “Myth and Reality,” in Before Philosophy, ed. H. and H. A. Frankfort, John A. Wilson, and Thorkild Jacobsen (Harmondsworth, 1949), 11–36. More recently, see also Paul Barber, Vampires, Burial, and Death: Folklore and Reality (New Haven, 1988). Mythology, too, is susceptible in part to the comparative method of reconstruction. See, e.g., C. Scott Littleton, The New Comparative Mythology, 2d. ed. (Berkeley/Los Angeles/London, 1973).
4Another type of perishable commodity where the experimental approach has been invaluable is in the very important industry of making salt from brine. Beatrice Hopkinson, who has done much of the work, has found that many of the most intractable questions about the artifacts received answers as soon as she began trying to make salt in the ancient way.