The representation of phalli in Neolithic Thessaly, Greece

The view that Greek Neolithic imagery is dominated by female figures has been considered a fact for some time and is still widely circulated (Gimbutas 1982; Kokkinidou and Nikolaidou 1993.58; 1997. 91; Marangou 1991.15; 1992.163; Orphanidi 1998; Perlès 2001.256; Bailey 2005.151). Usually, this view is accepted unproblematically, either without actual statistics to back it up (e.g. Kokkinidou and Nikolaidou 1993; 1997), or with criteria that are clearly projections of current presuppositions, e.g. when gender is attributed on account of the general form of the human body, equating women with voluptuous bodies, having protruding buttocks and a thin waist (Kokkinidou and Nikolaidou 1993; 1997; Marangou 1991; Mina 2007; 2008a; 2008b). Accordingly, a reassessment of the evidence is long overdue, especially in view of recent work done in other cases where similar assertions were made for years (e.g. Çatalhöyük, see Meskell and Nakamura 2005; 2007; 2009; Nakamura and Meskell 2004; 2006; 2008; see also Kuijt and Chesson 2005; 2007 for the Near East). This new work clearly shows that certain arguments cannot stand scrutiny, and that we need a more rigorous methodology in categorizing bodies, either of clay, stone, bone or flesh. It is mandatory to put our criteria in the open and change our perspective to gain a better, that is a more reflexive understanding of the material. To do so, and in accordance with arguments in favour of blurring our accepted categories and subverting our disciplinary boundaries (Meskell 2004; 2008; Nakamura and Meskell 2004), I will refrain from tracing sexed/gendered figurines and instead attempt to follow a different path, one that could produce alternative associations within the material. I will try to map the occurrence of particular traits across types of artefact and ask how these are represented in the various media, focusing on the materiality of each representation and, accordingly, the potential effect on the people using it (see also Nanoglou 2009a; 2009b; Nanoglou and Pappa 2009).


Introduction
The view that Greek Neolithic imagery is dominated by female figures has been considered a fact for some time and is still widely circulated (Gimbutas 1982;Kokkinidou and Nikolaidou 1993.58;1997. 91;Marangou 1991.15;1992.163;Orphanidi 1998;Perlès 2001.256;Bailey 2005.151).Usually, this view is accepted unproblematically, either without actual statistics to back it up (e.g.Kokkinidou and Nikolaidou 1993;1997), or with criteria that are clearly projections of current presuppositions, e.g. when gender is attributed on account of the general form of the human body, equating women with voluptuous bodies, having protruding buttocks and a thin waist (Kokkinidou and Nikolaidou 1993;1997;Marangou 1991;Mina 2007;2008a;2008b).Accordingly, a reassessment of the evidence is long overdue, especially in view of recent work done in other cases where similar assertions were made for years (e.g.Çatalhöyük, see Meskell and Nakamura 2005;2007;2009;Nakamura and Meskell 2004;2006;2008; see also Kuijt and Chesson 2005;2007 for the Near East).This new work clearly shows that certain arguments cannot stand scrutiny, and that we need a more rigorous methodology in categorizing bodies, either of clay, stone, bone or flesh.It is mandatory to put our criteria in the open and change our perspective to gain a better, that is a more reflexive understanding of the material.To do so, and in accordance with arguments in favour of blurring our accepted categories and subverting our disciplinary boundaries (Meskell 2004;2008;Nakamura and Meskell 2004), I will refrain from tracing sexed/gendered figurines and instead attempt to follow a different path, one that could produce alternative associations within the material.I will try to map the occurrence of particular traits across types of artefact and ask how these are represented in the various media, focusing on the materiality of each representation and, accordingly, the potential effect on the people using it (see also Nanoglou 2009a;2009b;Nanoglou and Pappa 2009).
Clearly, if someone wished to map all the characteristics usually associated with gender, a significantly longer treatise than a mere paper would be necessary.Accordingly, and in a further attempt to bring out aspects of Neolithic imagery that have been largely neglected (but see Marangou 1991), I will focus on the representation of phalli in Neolithic Thessaly, following recent reappraisals of similar material in other studies (Meskell and Nakamura 2005;Meskell 2007; see also Nakamura and Meskell 2009;Mithen et al. 2005;Nergis 2008;Özdogan 2003).One could argue that this is an insignificant aspect of Neolithic imagery in Greece, but a review of the available numbers shows that it is as good as any, at least as good as vulvae or even breasts (see below).The specificities of the representation of phalli will show that there is an important array of unexplored issues regarding the production of bodies in Neolithic Greece, and will allow us to consider the possibility of ontological questions quite different from our own.

The sample
The discussion offered here is based on an analysis of a sample of more than 1250 figurines from sites all over Thessaly (Nanoglou 2005.appendix;2006. 158).I have examined most of them personally, but for some I had to rely on published reports.In the latter case, I have not used items with no published photos of the relevant anatomical parts.Of course, not all of the objects preserve these parts, so the total number of artefacts on which the research is based is limited to less half of the total (Tab.1).Before embarking on the analysis per se, a caveat is in order: the sample does not entirely come from controlled excavations or surveys.Some objects do, but many were collected randomly over the years, and many are the product of excavations conducted long ago with questionable retrieval strategies.So, the sample is statistically problematic -one might say, biased -since it was likely that figurines with more pronounced traits would have been favourably treated.I suggest that this, in fact, strengthens one of the arguments of this paper, which is that the occurrence of genitals has been overestimated in the re-presentational landscape of Neolithic Greece (see below).
There are many issues concerning the identification of a trait as a phallus or a vulva, issues that bear upon the quantification of the sample and the results offered.I will not go over them now, but I need to say that, based on an extensive study of figures from Greece, it seems to me that in most cases the incision usually identified as a vulva is no more (or less) than an incision that separates the legs of the figures, something that is more visible in some examples, less in others, where most probably the representation of anatomical parts is more condensed (Nanoglou 2005.144, Figs. 3-4).Accordingly, I will include here only examples where such ambiguity can be reasonably rejected. 1Considering all that, plus the fact that most researchers have talked about the overwhelmingly female character of representational material in Neolithic Greece, but unfortunately with no actual statistics to verify their claim 2 , it is significant to see the occurrence of genitals on figurines from Greece, to get an idea about this unquestionable fact.
The figures are divided according to their preservation: items preserving the upper torso (at least from neck to waist) are used to tabulate the occurrence of breasts, and items preserving the lower torso (at least from waist to hips) are used to tabulate genitals.The two sets are not straightforwardly comparable and thus any discussion of the occurrence of both breasts and genitals must be limited to figures preserving both the upper and the lower torso (at least from neck to hips).As shown in Table 1, this last category includes somewhat more than half the sample, whereas the number of fragments preserv-1 I realize that this strategy may result in underestimating, rather than overestimating, the occurrence of vulvae, but, given the focus of the paper, I will refrain from a detailed discussion of the issue here.2 Both Marangou (1992) and Mina (2008b) provide numbers and percentages of their categories, but it is not possible to see the characteristics they recognize on each figurine, since there is no catalogue with the traits recognized on each object and therefore our databases cannot be juxtaposed.Tab. 1.The sample ing the upper torso is more or less twice the number of those preserving the lower torso.

Discussion
A first analysis of the sample is presented in Tables 2-3.It is immediately apparent that genitals and pubic triangles are indeed scarce.In Thessaly as a whole, figures with genitals amount to 9.45% of the total.Breasts, however, are more common, but not very common, either -the percentage is 21.25%3 .If we limit the sample to material from excavations rather than unsystematic collection, the figures are almost always even lower (Tabs.4 and 5)4 .There are important issues concerning distribution and dating to which I will return presently, but it is necessary to state the obvious: a) the traits on which gender identification are usually based are not that common on the whole, and b) nothing suggests the overtly feminine iconographic landscape, which is often presented (Kokkinidou and Nikolaidou 1997;Marangou 1991;1992;Mina 2007;2008a;2008b;Orphanidi 1998).In fact, phalli and vulvae are more or less equally represented, making need to focus on the neglected phalli more critical than ever.
The tables show that most of the artefacts with genitals are from the earlier part of the Neolithic.This difference pertains not only to the absolute numbers of artefacts, but also to the relative percentage of figures with genitals within the total number.From 8.41% in the earlier Neolithic (6500-5300 BC) the percentage declines to 3.39% in the later Neolithic (5300-4000 BC).There is also a decline in the percentage of figures with breasts, but the fall is at first sight less significant (17.24% to 14.46%).It should be stressed, however, that these numbers pertain only to clay figurines; stone figurines, which proliferate during the later Neolithic (Nanoglou 2008a), show neither genitals nor breasts, something that lowers even further the percentage of figures with genitals and/or breasts in this period (to 11.43% for figures with breasts and 2.60% for figures with genitals).This trend resonates with the diminished desire for detail that is observable in the later Neolithic, when figures focused more on the generic human image (Nanoglou 2005; see also below).Although there is great variety during the earlier Neolithic, some types can be teased out of the total population.Generally, phalli are associated with these types of artefact (starting with figurines and moving on to other media): Human figures with a phallus (Figure 1 and vignette a in Figure 4), Clay vessels with a handle in the shape of a phallus (Fig. 2),

Tab. 3. Occurrence of breasts on Neolithic clay figurines from Thessaly.
Stone pendants in the shape of a phallus (Fig. 3).

I should stress the point already made by
Hourmouziadis that all examples of purportedly autonomous phalli (e.g.Marangou 1991.Fig. 4;Theocharis 1973.Fig. 242) are actually broken at the base (Hourmouziadis [1974]  A first significant point is that certain representations of genitals show a rather specific geographical distribution (see also Nanoglou 2006).For example, a group of figures with a phallus as their only distinctive body part that date to the earlier Neolithic (see Hourmouziadis [1974]1994.88,Fig. 9 for some examples) was found over the same area, in the southwestern part of the plain of Thessaly (the modern Karditsa-Farsala region).A different subtype of figures with a phallus, again of earlier Neolithic date, is well known from various publications: figures seated on a stool which often, although not always, have a phallus.Their distribution is wide, ranging from the southeastern to the northwestern part of Thessaly (Fig. 4).This kind of information is not very helpful, but the study of some examples which have gained a lot of attention lately (Bailey 2005.163-4;Hamilton 2000;Talalay 2000) can be more informative.This sub-subtype combines the seated posture and phallus with the presence of breasts, and a crescentlike artefact in the right hand, which rests on the shoulder or leg (vignette a in Figure 4).There is also a standing figure with these features (vignette b in Figure 4).A significant point is that all the figures in this group come from nearby settlements -shown with triangles and a diamond on the map -in the middle of the plain of Thessaly (Fig. 4).In this restricted region, then, there is a group of figures that make sense as being seated (at least most of the time), with a phallus and breasts, and holding an object.These are figures that invoke the genitals in particular combinations (namely, in relation to a pair of breasts and a crescent-like object) and, I would suggest, in particular contexts.It would not be out of bounds to say that genitals would have no meaning if they stood on their own.
It is, perhaps, pertinent that the third type of artefact, vessels (ladles) with handles in the form of a phallus, comprises objects that can be carried in one hand, and are vaguely reminiscent of the crescent-like objects in the hands of the figures mentioned above (or vice-versa).In any case, even if the object depicted on the figures is not a ladle, we still have a correlation between the representation of a phallus and a very specific artefact.Significantly, the vessels are

Tab. 4. Figures with genitals from selected excavations.
contemporary with the figures, but do not share the same distribution, which should warn us against an easy association of thematically similar representations (although there are two seated figurines with a phallus and protruding breasts from Sesklo, but with no artefact in the hand).On the other hand, it might suggest that similar requirements (e.g. the presence of a vessel with a phallus) materialized through different avenues (either as an actual vessel, or as a figure with such a vessel).
This emerging picture of high specificity is still evident in the last of the types I mentioned: stone pendants in the form of a phallus (unfortunately, most have no clear dating information, see Kyparissi-Apostolika 2001, but perhaps of later Neolithic date, Nanoglou 2005.152).The general trend during the entire Greek Neolithic was to represent the whole body, rather than just parts of it.The rare finds that represented only a part of the body almost always had a suspension hole and were made of stone.In this case, the stone phallus becomes a detachable part of the body that wears it -if indeed we agree that this was their use -enabling someone to have a phallus, irrespective of its materiality, whether of flesh, marble or any other material.It seems that if someone required a phallus for a specific occasion, then one could as well bring one along (more on this below).
As in other cases, contextual information on the figurines is scanty (Nanoglou 2008b), and for the particular artefacts, non-existent.So any inference about their use cannot be strongly supported, but it is possible to suggest that figurines in the earlier Neolithic were used in groups, or at least were comprehensible in terms of groups (Nanoglou 2008b).Whatever the occasion of their use, their presence would have affected the frame of reference for whatever actions were taken, and for whatever subjects were constituted.Figurines would have conditioned the way people understood themselves and the world, being material referents for the action taken (see the concept of 'citation precedents' coined by R. Joyce in Joyce 1998;2000a;2000b;2008).So, particular bodies would have been produced in relation to these figures, which framed the field of intelligibility, providing and maintaining models of sanctioned and possibly desirable bodies with which people could identify generically or specifically.
So, how should we approach these artefacts?What kind of bodies were they produced within a world populated by these figures?Were they articulated as part of a discourse that highlighted reproduction?It seems significant that, from the thousands of figures we know, only a Late Neolithic one from Sesklo seems to depict an adult and child (Theocharis 1973.

Reproduced by permission).
figures interpreted as a family (Gallis 1985;2001) is the well-known building-model from Platia Magoula Zarkou which contained eight figurines.These figurines have no genitals, although some have breasts.They were deposited as a group, following an earlier Neolithic tradition, and they depict some kind of action (either posture or gesture), again following an earlier Neolithic tradition.In fact, the contradistinction between the actions depicted by the figures may well have been a significant feature of the assemblage, whereas genitals are absent from the scene.It is possible then that these figures suggest a categorization of bodies according to their actions, rather than their 'gender role' (see Meskell and Nakamura 2005;2007;Nakamura and Meskell 2004;2006;2009 for similar observations on the material from Çatalhöyük).So, again: what kind of bodies were produced when these phalli were invoked?Their occurrence is highly particular, with specific combinations, suggesting that their meaning resides in these combinations.Phalli were brought to bear upon the lives of people in articulations that were anything but common, and furthermore, seem to have defined a specific context for them to be 'useful' or appropriate.As stated, we know little about their context of use, but if we presume that they were at least visible in some gatherings, of whatever sort, then they would have affected the whole process of addressing others and being addressed by others.A person would have been able, and to a certain extent obliged, to find a place within this gathering, but also beyond it, in reference to the figure and the articulation of genitals along with the other traits.In one distinctive case, what mattered was the presence of a body, preferably seated, with breasts, a phallus and a crescent-like object.Someone, or many, had to perform this combination, either in clay, in flesh or any other possible material, in order to be able to represent themselves in an intelligible manner.They may or may not have had to carry a ladle with a handle in the shape of a phallus; they may have had to carry a pendant in the form of a phallus; but the point is that they would have to describe themselves by referring to these fi-gures along with other things, and accordingly produce themselves in relation to these bodies.We could suggest that this was a special role reserved for someone in the long term, or for everyone at a specific time.The discovery of similar examples in different locales suggests an iterated occasion and not a unique event.So these scenes where people, figures and/or vessels converged would probably have been sufficiently institutionalized to produce multiple moments of articulation.However, within the wider context of representational practices, these scenes, or in any case the roles reserved for figures with genitals, were only small elements.The first point is that these figures would have informed the actions and concomitant identity of these people in a way that does not resemble our prioritization of gender.Anthropomorphic figures generally helped produce genderless bodies, by which I do not mean ambiguous bodies (Talalay and Cullen 2001), but bodies that did not refer to a categorization according to their genitals.For the most part, the bodies that anthropomorphic figures in earlier Neolithic Thessaly facilitated or imposed cited a varied gamut of actions (gestures and postures,) and only occasionally genitals, again in relation to specific actions.If the common denominator of these representations was the action performed, and genitals were invoked only in the context of a certain group of actions, then we should seriously consider whether genitals were only a secondary attribute of a particular role in specific scenes where people engaged with these objects and each other.The reiteration of these combinations would have naturalized a body with a general requirement to act (at least, on the particular occasions) and only sporadically to bring along breasts, a phallus and/or a phallus-shaped object.Even then, the body was obliged to do something, not merely have breasts and a phallus.The invocation of genitals was in excess of the general requirements a body had to fulfill in order to be acknowledged as such.Figurines and other representations in earlier Neolithic Thessaly enabled and, inflicted a categorization of bodies according to their actions, and reserved the invocation of genitals only for specific contexts where they were needed.
It seems significant that, on a certain level, in most of the cases where phalli are present, they seem to be something additional to the body, at least from our point of view, either in terms of 'superfluous' genitals and/or breasts, or pendants that hang from the body, or even vessels carried around.Whether all these were actually considered additional or not is impossible to know, perhaps; but the point is that at least some of them present the possibility of an ambiguous status, due to their materiality: they can be brought along or left aside; they are not fixed, and their character enables people to invoke them at will.This does not mean that their invocation would not have been sanctioned by certain rules, which would have conditioned any intentionality.We should not see intentionality as an opportunistic endeavour.Despite their transient character, no performance would have been totally unrestrained, and their use in such performances would have been regulated in some way.Besides the theoretical necessity of such rules, the specified articulation of the traits discussed bears witness to them.Nevertheless, on a first level, what matters is that a body that was materialized by drawing on these resources probably bears little resemblance to a body that is produced as either having or not having certain genitals, either having or not having a specific gender.If the sample on which the above picture is based is considered viable, then we can hardly argue that the representational landscape populated by these figures inflicts a categorization of bodies according to their gender.Instead, it is possible that the invocation of genitals was relevant only for certain people, those who could have had a particular role to play within the community that used the figures.Perhaps the prerequisites for such a role were not confined to a certain body type outside the context of this specific performance; that is, the kind of body someone might have had in the rest of their life was not relevant.What mattered was the body that entered the specific scene, where figurines and/or vessels with phalli were used.It is interesting, and perhaps mandatory, to consider the ontological status of this body: part flesh and bone (and vessels, and organs etc.), part clay or stone -a true hybrid (Latour 1993;Meskell 2004;2008) or cyborg (Haraway 1991).Even if, today, we can only comprehend this as a body with add-ons, it is still significant that, in order to perform in the particular context (or perhaps various contexts), it required these add-ons, which redrew the contours of the body, its very boundaries and, of course, its very ontological status.
We do not need to interpret the figures as depictions of a certain costume (although this is still a valid possibility; see Alberti 2001 for a relevant example).The issue is that the figures (and all other objects) entered the scene as articulations that had an effect on people, giving them the opportunity and at the same time the obligation to address others in the presence of the figures, prompting a comparison among bodies that had to be understood against each other.In the scene(s) where these objects were used, there was a need for a body that could and would bring along a phallus, breasts and a ladle.It is not possible to say whether this body need have been of flesh, clay or another material, or perhaps of all of these.The bodies of the figurines were constituted largely in the absence of genitals, which had a role only in specific instances.Given the dearth of contextual evidence, it is difficult to speculate on what these instances were about.The use of a ladle could suggest the use of some kind of fluid, perhaps a bodily fluid (semen?), in an act of transaction between people that utilizes an object as a mediating agent and symbol of the transaction, once again negotiating the boundaries of the body (Herdt 1987;1999 andpapers in Gregor andTuzin 2001; see also Lingis 1994.Ch. 8;2005.127-143;Meskell 1999. 46-50;Meskell and Joyce 2003).The representation of breasts alongside phalli could suggest the parallel invocation of milk.The correspondence of milk with semen, and of breasts with penis in Sambia, come to mind here (Herdt 1982).Could it be that the figure in question, collecting all these mediatory objects and bodily fluids, synthesizes the various versions of the same substance?And could it be, to follow the reinterpretation offered by Strathern (1988.208-212),that it is not the penis or breasts that make someone male of female, but vice-versa?In Strathern's reading, one's performance activates the relevant aspects of each object, which are by definition both male and female: "what distinguishes males from females is not their appendages and orifices as such, but the social relations in whose contexts they are activated" (Strathern 1988.211).Taking this further, we could ask: what if what was activated had little or nothing to do with a pervasive categorization of bodies into male and female, but primarily related to other aspects of one's life, like age (as already suggested by Hamilton 2006;Voigt 2007 andespecially Nakamura andMeskell 2008;2009), or kin relations that marginally related to genitalia?This must remain speculation for the present.What is perhaps more important is to follow Strathern's proposal on another level and, instead of assuming that genitals, breasts or the form of the body, reveal one's identity, start to situate these objects in some kind of context that will enable us to see what kind of associations were possible and desirable in their world.

Conclusions
There are two brief points I would like to highlight as a conclusion.First, I hope this paper makes clear that the presence of genitals should not be taken to characterize a priori a 'person as a member of a subspecies of humanity' (Meskell 2000.255).Gender may not have been a primary axis of identification in the past (as was argued some time ago : Hodder 1997;Meskell 2000;Nordbladh and Yates 1990;Ha-milton 2000;Whitehouse 1998), and so we should attune our analytical methods accordingly.Relevant assertions should be based on more detailed and rigorous work, which could well have surprising results (Mithen et al. 2005).The present paper aspires to be such a move forward, although I am fully aware of the weaknesses presented by such poor contextual evidence.The second point has to do with the materiality of the objects and the bodies that the paper is concerned with.We can see how an object can provide persons with what they need, so as to be considered appropriate for the role bestowed on them, yet the object is not transparent or self-explanatory.The case offered here concerns the entanglement of persons with objects, of bodies of different material that are perfectly capable of converging to become intelligible, acceptable and even desirable, albeit on certain occasions.It is a matter of a certain approach that gives precedence to the material world at hand, a world that is populated by certain people in certain historical conditions and therefore inhabited and experienced in particular ways that exceed aspirations for cross-cultural and trans-historical givens.
I would like to thank Professor M. Budja for the invitation to a very stimulating conference.I would also like to thank N. Kyparissi-Apostolika for permission to reproduce Figure 3 and the Archaeological Society at Athens for permission to reproduce Figure 2. My deepest gratitude goes to Lynn Meskell for invaluable comments on an earlier draft.As always, all faults remain my own.ACKNOWLEDGEMENTS
Fig 56, but see two recent finds from Promachon-Topolnica, Koukouli-Chrysanthaki et al. 2007, again of Late Neolithic date).The figure has no genitals or even breasts.Apart from this, the only other case of

Fig. 4 .
Fig. 4. Map of Thessaly and part of central Macedonia, Greece, with sites discussed in the text.Dots in light grey: sites from where the sample comes.Black dots: sites with figures seated on a stool.Triangles: sites with seated figures with a phallus, breasts and holding an artefact.Diamond: site with standing figure with a phallus and holding an artefact.Stars: sites with seated figures with a phallus and breasts.NK: Nees Karyes; MK: Magoula Koutsouro or Koutouki (Zappeio 2); MP: Magoula Panagou (Chara 1); S1: Sitochoro 1.