<< Back to Project Gallery

Antiquity 82 Issue 318 December 2008

A response to 'Juggling the evidence: the purported "acrobat" from Tell Brak'

Theya Molleson

The authors are to be thanked for their detailed and most useful comments on our interpretation of the skeleton from Tell Brak (see http://www.antiquity.ac.uk/ProjGall/jurmain/index.html and reproduced below). It highlights the difficulties and limitations of such work, which we fully acknowledge. Since most of the comments are directed at that part of the paper to which I contributed it behoves me to reply so that my colleagues are not tarred with the same brush. Because of the constraints of presenting the details of the bones in a short paper intended for the general reader I restricted citations of references to those salient to the specific case. This contribution was not intended to be a methodology paper.

It is true that no two people react to a given activity in the same way but that is not to say that a set or syndrome of entheses, as distinct from enthesopathies, of the ligament and muscle attachment areas on the bones of an individual cannot be interpreted as reflecting habitual activity (Robb 1994; Molleson 2007). It is the basis of clinical diagnosis of individual patients. The activities that have been inferred from archaeological contexts are very often highly specialized involving restricted movements undertaken from an early age while the bones were still growing; where several activities are involved the movements are usually too generalised to be informative.

The preservation and injuries observed on the bones of FS 1374C were disposed of first. Healed injuries are so often the outcome of a single traumatic event, which may or may not have been due to the individual’s lifestyle. The specific changes that identified highly developed ligament and muscle attachment areas I took to reflect sustained joint and muscle use, the signs of habitual activity. For example, Keith Manchester many years ago recognised that pits developed on the ventral (palmar) surface of phalanges were the consequence of flexing the fingers. He had seen it in lepers — claw-hand deformity of leprous paralysis (Manchester 1984: 168). Here we see the analogous feature in the claw-toes of the Tell Brak skeleton, in response to flexing the digits. By identifying the bone features I intended to provide the evidence for the informed reader to form an opinion; to help the general reader I gave my interpretation. Taken together with the impressions of the cruciate ligaments and muscle attachments on the knee joints the syndrome of signs on the Tell Brak skeleton seemed to be a response to sustained jumping, turning movements. Of course there could be other explanations for the pattern of morphological change seen on the skeleton. As is so often the case in clinical practice the diagnosis cannot be proved. We do know from textual evidence, however, that there were at Brak well-known people whose occupation involved ‘jumping about’ and who were not ordinary ‘dancers’.


There can be no better response to the assertion that genetics determine the appearance of entheses than Jim Tanner’s conclusion at the end of the comparative study of 137 athletes taking part in the 1960 Olympics 'athletes are born and made' (Tanner 1964: 113). He expected his study to stimulate the thoughts of his colleagues of the year 2000 (Tanner 1964: 9). The subject is fighting fit, for it is through these challenges that answers come.

The completed skeleton chart of FS 1374C in black.


  • MANCHESTER, K. 1984. Tuberculosis and leprosy in antiquity: an interpretation. Medical History 28: 162–73.
  • MOLLESON, T. 2007. A method for the study of activity related skeletal morphologies. Bioarchaeology of the Near East 1: 5–33.
  • ROBB, J. 1994. Skeletal signs of activity in the Italian Metal Ages: methodological and interpretative notes. Human Evolution 9: 215–29.
  • TANNER, J.M. 1964. Physique of the Olympic athlete. A study of 137 track and field athletes at the XVIIth Olympic Games, Rome 1960: and a comparison with weight-lifters and wrestlers. London: George Allen &Unwin Ltd, London.


  • Theya Molleson Department of Palaeontology, The Natural History Museum, London SW7 5BD, UK

Specific comments on the text

We are responding to the submission by Oates et al., 'Equids and an "acrobat": closure rituals at Tell Brak' (http://www.antiquity.ac.uk/ant/082/ant0820390.htm). The paper describes the interpretation of a partially preserved skeleton with bone changes that indicate ‘habitual energetic activity’ (p. 398), or a person who was ‘physically active, using jumping and twisting movements in a very disciplined way with the feet pointed downwards during leaps, as can be seen in modern dancers’ (p. 396). While we acknowledge that the skeleton was deposited in an unusual way, and that documentary sources at Elba and seals at Tell Brak suggest people were present who were characterised as 'always jumping about' (translation of a Semitic word which can also infer an 'entertainer, juggler or horseman'), we would like to explore further what is described as the main subject of this article, namely the 'occupational' interpretation of an incomplete skeleton as that of an 'acrobat.'

In bioarchaeology some readers will be aware that there have been numerous conference presentations and published papers focusing on how skeletal lesions are used to determine 'occupation', especially in the last 10 years since the International Journal of Osteoaearchaeology’s special issue in 1998. A recent survey has also provided a critical retrospective of research since then (Henderson et al. 2008). As bioarchaeologists with a very long experience working with archaeologically derived skeletal remains, and with expertise in the particular subject of reconstructing behavioral attributes from human skeletal remains, we are especially concerned about this paper and its misleading aspects, especially for those readers with no experience of this type of study.

There are several omissions in the osteological analyses, particularly notable given the central importance this skeleton is accorded. Firstly, there is no diagram or complete listing of preserved elements, but only a brief summary indicating the skeleton is quite fragmentary (p. 393), and in Table 1 a list of measurements taken. When attempting to infer meaning from a skeleton from an archaeological site, firstly a complete skeleton is desirable so that distribution patterns of lesions can be seen and a most likely 'diagnosis' presented [Note 1]; secondly, it is also preferable to have a large sample of skeletons at a site to determine the range of variation in lesion expression, especially for those lesions that may be related to increasing age (this is to see if indeed a particular skeleton is unique) [Note 2]. Figure 3 also states that the skeleton in question is the ‘most complete of the three specimens’ but the image is not very clear and an arrow would have been helpful on the figure [Note 3]. There is also a comment at the foot of p. 391 that the upper part of the skeleton, has already been reported elsewhere, but it is impossible to tell if any of those observations are included in this current paper. For example, one humerus is reported to have osteoarthritis, but there is no metrical data for that humerus in Table 1 [Note 4]. Of note is the attribution of ‘mature adult’ to the remains (p. 393), but no citations relevant to methods used are provided; such documentation is both standard and relevant to supporting those particular age-related changes to the bones of the skeleton evaluated by the authors [Note 5].

In terms of methodological concerns, specifically, there is only superficial description of lesions in the feet and no differential diagnosis is offered [Note 6]; Figure 4b does not give clear enough detail to support the description [Note 7]. Nor is there any consideration of a differential diagnosis which might help explain the unusually short stature of this individual [Note 8]. Lastly, apparently no radiographs were taken, and certainly no radiological information is reported; this may have helped in interpretation [Note 9].

The weakest part of the article is also its core, that is, the interpretation of presumed behavior based on skeletal/morphological changes which are mainly lesions associated with muscle and ligament insertion sites on the bones preserved. Other than some determination of acute trauma in the lower limb (which we do not have adequate grounds to dispute), this analysis focuses almost entirely on those changes in muscle and ligamentous insertions, or what bioarchaeologists commonly refer to as 'entheses', 'enthesopathies', or one of a suite of 'musculoskeletal markers of stress/activity'. There is a surprising lack of reference to published literature on the methods used to analyse this skeleton and how data may be interpreted; this would normally provide readers a context to assess the validity of this approach [Note 10]. Without providing this crucial groundwork, most readers will be left completely in the dark without any possibility of assessing the very basis of the claims made by the authors. Indeed, only three citations that concern relevant skeletal methods/interpretations are provided, with just one from the last decade; furthermore, one of the earlier ones was authored by one of the authors, where a similarly unfounded behavioral interpretation from this same site was made [Note 11].

As might be expected, there is indeed a considerable literature on this subject with numerous older publications supporting the notion that enthesopathies provide a useful indicator of, at least, general levels of activity (e.g. Dutour 1986; Kennedy 1989; Hawkey & Merbs 1995). Nevertheless, over the last several years there have also been critiques emphasising the multifactorial etiology, oversimplification, and resulting over-interpretation of these types of bone changes and their ostensible relationship to activity (Waldron 1994; Jurmain 1999). Ageing, biological sex, genetics, work/activity, obesity, diet, and other disease can all lead to the development of MSMs together or separately (Resnick & Niwayama 1983; Rogers et al. 1997 [Note 12]).

More recent attempts have controlled for the confounding factors of age, sex and body size (Molnar 2006; Weiss 2007; Cardoso 2008a & b) and, in so doing, the correlations with activity have become yet weaker. Cardoso’s research (Cardoso 2008a) has also indicated that, even with historical documentation about 'activity' associated with nineteenth and early twentieth-century Portuguese skeletal remains, there was no correlation between 'activity' with either enthesophytes or with osteoarthritis (Cardoso 2008b; Henderson &Cardoso 2008). In fact, entesopathies and osteoarthritis were not necessarily associated with demanding activities. Furthermore, recent research has indicated that the anatomical basis for enthesophyte development is poorly understood in the bioarchaeological field and that bioarchaeologists would be wise to develop a better understanding of the anatomy of the soft tissues in which the lesions develop (Henderson 2008; Henderson & Cardoso 2008) [Note 13].

Finally, recent experimental research using rigorous controls on nonhuman animals, has shown virtually no relationship of activity and any identifiable enthesial bone change (Zumwalt 2006). From these results the researcher concluded, ‘In spite of decades of assumption otherwise, there appears to be no direct causal relationship between muscle size or activity and attachment site morphology, and reconstructions of behavior based on these features should be viewed with caution’ (p. 444).

The lack of correlation of activity with distinct morphological bone change is, in reality, not that surprising. What determines the timing, degree of remodeling, and ultimately development of an enthesial reaction is ultimately regulation of bone turnover physiology. In turn, bone turnover is governed by gene action, including identified specific loci that are variable between individuals (Spector & McGregor 2004). Recent experimental results have identified specific genes influencing enthesial bone formation (Chen et al. 2007) [Note 14]. In most prehistoric societies the majority of individuals were likely highly physically active, but not all show similar bone changes; in fact, many show very little bone change, even at those anatomical sites where they should be most expected.

Why is this the case? Simply put, most of the variance in physical expression at enthesial sites is not the result of activity. Most of the variance is likely due to genetic influence (including sex difference in hormonal activity), with further substantial contributions due to age and body size. The portion of variance explained by activity is, regrettably, quite small. These factors explain why more controlled research has yielded ever-more disappointing results.

In fact, while in clinical literature there are indeed reports of people developing bone lesions such as osteoarthritis in specific occupations, not all people with the same occupation will develop lesions or even the same lesions! Moreover, if a particular 'activity' is being inferred from a skeleton, one should always be referring to clinical literature to understand how specific 'activities' affect the bones of the skeleton. For example, how does whiplash affect the skeleton and do acrobats, dancers, gymnasts or any other presumably identifiable 'at risk' contemporary group of people develop bone changes such as those described in this paper?

In recent years the scientific literature has included a number of such now-unsupportable studies that attempted to reconstruct general levels of activity in past societies. However, endeavors to divine specific occupations have been much rarer. Even at the peak of enthusiasm for using skeletal 'markers of activity' (such as osteoarthritis and enthesopathies), most researchers understood the monumental obstacles of attempting such suppositions from analysis solely of skeletal remains (Jurmain 1999; Waldron 2006). When definitive artefactual associations are present, the possibilities of concluding something reasonable about 'occupation' may be better, but 'jumping' to conclusions about an individual skeleton is dangerous.

If the data, particularly those controlled most rigorously, are so lacking in support, what has led these researchers (as well as many others) to be so compelled to make hasty, superficial conclusions? In following what Tony Waldron (1994) has termed the 'alluring prospect', every bioarchaeologist at some point has been highly tempted to make similar suppositions. We, of course, want to know interesting details about the lives of those people whose skeletons we study, and we can imagine all sorts of fascinating things they may have done. Nevertheless, ultimately imagination must yield to the inescapable constraints of our science.

The general public naturally also wants to hear good stories about the past. This sort of popular demand can exert further pressure on researchers. Speculation is healthy and, oftentimes, amusing. What we choose to publish is quite another matter. The expectations of the general public regarding what can be learned from skeletal analysis often are exaggerated and may not be based in sound research. Much of this unrealistic view has been magnified by the media’s misrepresentation of forensic science research. We feel that unsupportable claims that would be eagerly received in some circles must not be made, merely because they are 'fashionable'. More generally, respectable scholarly publications do not serve the public interest by furthering such a 'CSI' mentality.


  1. However, a fractured leg-bone, say femur, can be ‘diagnosed’ from the femur alone whether or not the arm was also broken. Back to Text
  2. In clinical medicine case studies are usually described on the basis of the individual case. There is no suggestion that skeleton FS1374C is necessarily unique. The lesions are not usually age dependent although they may be cumulative and enlarge with increasing age. Back to Text
  3. Agreed. Back to Text
  4. Only the humerus head could be measured 46.4mm (measurement as defined in Brothwell 1981, given in Oates et al. 2001: 352). Back to Text
  5. All epiphyses that could be observed were fused; usually taken to indicate adult status. Back to Text
  6. Such detailed description was not considered appropriate for the journal Antiquity but will be given in a forthcoming publication. Back to Text
  7. The photographs were taken in the field in difficult circumstances. It seemed a shame to deprive the public of a very interesting and challenging case for this reason. Perhaps we were wrong. Back to Text
  8. It could have been selection of the appropriate build; we have no way of knowing, although this article may encourage the discovery of references in the contemporary literature. Having just seen the opera Aida the acrobats in the performance were notably diminutive and less stressful to support. Back to Text
  9. The lesions were commented on by an eminent radiologist, (see the acknowledgements), although the conclusions are ours. Back to Text
  10. This was not intended to be a methodology paper. Back to Text
  11. No, from Ur not Tell Brak. Back to Text
  12. The foregoing is an excellent and valid critique of the difficulties and limitations of interpreting skeletal remains. Back to Text
  13. The problem with much of the foregoing is that pathological enthesopathies have been conflated with the normal bone response during growth to sustained muscle activity. Back to Text
  14. That is to ignore the manifest effects of exercise in the production of hormones and the development of physique, the limitation of osteoporosis. Back to Text

Back to Top