Interpreting Iron Age finds requires the use of archaeological method. Two main types of evidence provide our data: excavated finds and literary sources from the classical Mediterranean. Both classes of evidence are incomplete, subject to the vagaries of preservation and transmission, fraught with possible errors and must be interpreted critically.
Nearly all the objects studied here were incorporated into funerary assemblages. A tomb assemblage presents the opportunity to try to explain the associations of objects and the interred, the choice of objects and the stylistic range of grave goods. Mortuary analysis teaches us that sex, gender, age, cultural or ethnic identity, and various roles in the society help determine the type of burial, its associated ritual and the selection of objects buried with the dead. Age can be determined with a good degree of certainty -- perhaps eighty percent -- if a skeleton is preserved. The skull may often be identified as having predominantly the characteristics of a general racial type. DNA analysis, when undertaken, may reveal genetic groupings and anomalies. Physical anthropology may be extremely informative about an individual's health, diet, growth pattern, cause of death, etc. However, bones cannot reveal how the individual felt or thought about any of those biological factors or how he or she was regarded by his/her social milieu. Race is only very generally identifiable on the basis of human remains; ethnic and cultural identity not at all, since those are matters of variable self-definition based on a complex of factors not represented in the archaeological record. In analogous fashion, sex can be read from bones, but gender, being a matter of social, cultural and individual subjectivity, requires examination of the the entire find complex, comparative study of other burials and consultation of the non-Celtic literary sources.
It is striking that a field of archaeology in which a large number of the most opulent and significant sites are female burials has concerned itself so little with the issues of sex and gender. To understand the choice, function and style of a work of art, we wish to know for and by whom it was created, what its original functional and symbolic purposes were, and why it came to rest in its final find spot. Together with age, sex and gender represent aspects of identity that are fundamental both to the individual's social, political and religious roles while living and to how that individual is treated in death and thus enters the archaeological record. It is clear that our interpretations of a find complex as "elite," "warrior," "princely," "priestly," etc., will be strongly colored by the addition of "female" to any of these terms. The scenario envisioned surrounding the manufacture, use and deposition of the individual artifacts must in turn be influenced. Why, then, is so little attention paid to this crucial issue in "Celtic" archaeology?
This short treatment of the subject cannot resolve the problems of sex and gender in studies in the European Iron Age. The questions are laid out in an urgent plea for a reevaluation of the material and our approaches to it in respect to sex and gender analysis.
Biological sex is obvious -- a person is either male or female. This common truism is called into question by phenomena of our modern existence. The popular press abounds with stories of transsexuals successfully living as members of the "opposite" sex, including cases in which sexual partners were thoroughly fooled, sometimes for years. DNA testing is considered necessary to determine sex in Olympic athletes. Like race, sex turns out to be much less obvious and easily determined than appears at first blush.
The history of the field shows further similarities to the history of the study of ethnicity and race. The sexual identity of "Celtic" archaeological remains has traditionally been established by male scholars or those working within the male-dominated anthropological and archaeological establishments. The questions posed by these researchers and their presuppositions will necessarily differ from those that would have informed their female counterparts, and the evidence would unquestionably have been interpreted quite differently; Aristotle's wife no doubt knew how many teeth she had.
Human remains are very seldom preserved from antiquity with enough soft tissues intact to make a visual determination of sex. Sexing skeletons involves the comparison of pelvic and skull measurements. In the best-case scenario, where a skeleton is well preserved and there is a broad base of comparative data available, skeletal sexing in adults is accurate to around eighty percent. It is essential to stress that sexual differentiation varies markedly between populations -- comparative measurements that would be quite useful in sexing modern Australian Aborigines may be entirely off the mark when applied to Iron Age Europeans (Renfrew and Bahn 1996, 404-406). That said, within a circumscribed population sample, female pelvises and skulls do reveal morphological differences from those of males that allow an individual adult specimen to be sexed relative to the population as a whole. This recognition of physical features as neither specifically male or female, but rather lying along a continuum of more-male to more-female, astonishes the scholar steeped in the Western tradition of an absolute male-female dichotomy. It also disconcerts the student of ancient art to finds that the confident assertions of sex as determined by archaeological means with which our literature is rife are by no means as infallible as we like to believe.
Very often, in the case of Iron Age Europe, it is not the skeleton at all that is sexed. The associated goods are still considered diagnostic, and even in cases where a skeleton is preserved, it is often not studied by physical or forensic anthropologists. The most famous of all ancient Celtic skeletons, the "lady " from Vix, on the other hand, has been subjected to exhaustive anthropological study. The results are anything but reassuring. The skull's cranial capacity of 1425 cc approaches the modern average European male minimum of 1450 much more closely than the female maximum of 1300. The long bones are less well preserved than the skull, but reconstruction of the left femur suggests a possible body height of between 1.58 and 1.67m (Sauter 1980, 99). Langlois's 1987 report reevaluates the findings that led Sauter, among others, to question the determinability of the "lady's" sex, and concludes that they indicate that she was in fact female (212-214). A stumbling block has been the incontestably robust character of the "lady's" skeleton; more recent study of the Hochdorf chieftain's remains suggests that the elite among the ancient "Celts" were larger than has been thought in the past. I can see no reason other than modern prejudice that would lead us to expect "Celtic" women to be small and delicate; the classical sources certainly suggest quite otherwise.
It is painful to acknowledge that Hochdorf and Vix are the only properly studied and published elite "Celtic" skeletons of around 500 BCE of which I am aware. This is not an adequate database on which the build a model of "Celtic" dimorphism. Even if more skeletons were preserved, it is particularly disheartening to read in Brothwell that "there is a constant danger of incorrect sexing, and indeed ... there is a 12 per cent bias in favour of males" (1981, 59). We may add that, in the case of Iron Age Europe, there is an almost overwhelming bias in favor of the particular scholar's list artifactually-based criteria. This leads to such bizarre phenomena as the resexing of skeletons from female to male based solely on the presence of weapons in the tomb. Grave 116 in the non-elite cemetery at the Dürrnberg, for example, contains projectile points and other weapons; although the skeleton is clearly female according to anthropological criteria, it is considered male because of the weapons and counted as such in the demographic analyses of the site (Schwidetzky 1978, 562).
The biological sex of an individual is an essential, although not the only factor, in the determination of that person's gender identification. Since the anthropological basis for the latter is demonstrably anything but solid in "Celtic" studies, we should not be astonished that some truly bizarre reconstructions have been proposed.
Gender is a cultural construct, part of the way an individual defines him/herself and is defined by others in the society. Gender is connected only in part to biological sex and issues of sexuality and reproduction; to a great extent, we study gender in the context of economic and power relations, of production and class, of ritual, belief and ideology (e.g., Wright 1996).
Iron Age Europe exposes, more clearly than perhaps any other field of archaeological inquiry, to what degree modern interpretations of ancient gender are the products of our modern-day constructs of the male and female. A simple example is the distress caused archaeologists by the inclusion of drinking vessels in apparently female burials. In 1934, Jacobsthal was horrified at the suggestion that the Kleinaspergle burial might be female, thus exposing the ancient women of Swabia as lushes the equals of their Etruscan counterparts (1934, 19). It is in the same tone of horror that young scholars and excavators react today when asked about the possibility that a burial containing weapons might be female (oral communications, Spring 1995). Since the mere presence of weapons has led to the statistical resexing of anthropologically female skeletons as male [see I.], we should not be surprised by the attitude toward gender revealed in the preliminary publications of the ongoing Glauberg excavation. The first report, in 1995, was entitled "Celtic Princess with Rich Dowry" (note 2). When further X-ray examination revealed the presence of spear points, however, the very next report simply changed the identity of the occupant of the tomb to male (Herrmann 1995, 47-48). It will be extremely interesting to see what the skeleton will reveal, once the excavation has reached that point.
On the point of "Celtic" gender, archaeologists are notably influenced by neither the classical sources nor parallels in the archaeological record, such as the numerous Sauro-Sarmatian "warrior-women" tomb complexes (e.g., Davis-Kimball 1997). Instead, their constructs of what a "Celtic" woman could or could not be overrides all other considerations. Thus, "Celtic" women "couldn't possibly " be buried with weapons; ergo, even burials in which the skeleton is clearly anthropologically female are declared male [see I.]. What, then, to make of interpretations of burials with no preserved or sexed skeletons? Pauli (1972) and others have articulated the current list of excluded and included gender markers (see Arnold 1991, 368 ff.). When a burial appears to contain a female assemblage according to those criteria, and yet the very scholars who establish the categories find it impossible to declare the occupant female, how are we to account for this behavior, if not as "reluctance to accord women significant social status" (Arnold 1991, 372)? Both Pauli (1972, ) and Spindler (1983, 108) found their way out of the dilemma by attributing the problem assemblages to a hypothetical practice of ritual transvestism! The simple suggestion that burials containing typically female assemblages might in fact be female has refreshingly been made by Arnold (1991). How long it will take the prehistoric establishment to come to terms with this revolutionary possibility is yet to be seen.
The prospect of any productive study of early "Celtic" gender issues thus remains frustratingly remote. Until large-scale comparative anthropological analyses of skeletal remains and comparative evaluations of known male and female burial complexes can be undertaken, conjecture and personal preference rule the day.