I would like to thank Silvia Tomášková, Christoph Ulf and James Wright for their comments, which allow me to clarify and expand my argument. As I cannot address all the points they raise in this response, I have decided to concentrate on three themes: (i) the application of moral theory, (ii) Homeric questions and (iii) the focus on the shaft graves of Mycenae.

The application of moral theory

The commentators’ views on my use of moral theory were divided. Tomášková agrees that moral philosophy can offer refreshing new insights of agency and personhood. Wright feels that my application of moral philosophy is weak, but does not really spell out specific weaknesses. Finally, Ulf does not see at all what moral philosophy has to offer.

Let me therefore explicitly address the question: why moral theory? And why not sociology or social psychology, for example? These, according to Ulf, would be more relevant to my enquiry, and to archaeology and history in general. To start with, because most sociological studies deal only with agency. For instance, Granovetter's (Reference Granovetter1973) conclusion that individual experiences are closely bound up with membership of larger groups (adopted by Ulf) may sound similar to my notion of relational personhood. However, it is not. Granovetter's main unit of analysis is always the modern rational, autonomous agent who manipulates networks in order to achieve his/her personal goals (Granovetter Reference Granovetter1973, 1370). He discusses agency and social interaction in a modern context, while I want to discuss agency and personhood in a premodern, relational world. Here Tomášková's comments give me further ammunition: we need to confront the otherness of the past, we need to devise ways to conceive of personhood in the past as being different. As she rightly points out, this may entail that in some societies certain categories – children, old people, cripples, war captives, slaves and so on – were not full persons.

Would social psychology provide a more appropriate entry to personhood? I have serious doubts. Let me give an example. Ulf correctly points out that categorization is an essential stage of human cognitive processes. But is categorization based on universal psychological dispositions, or rather on cultural and moral traditions? According to Brewer and Miller (Reference Brewer and Miller1996, 6; a study of intergroup relations in the modern world referred to by Ulf), the first distinction in cognitive psychology is between natural categories (plants, animals) and cultural categories (a political party, an ethnic group). This categorization is not useful for a discussion of personhood in a premodern society, where persons may merge with objects, animals may be persons, or persons may transform themselves into spirits. Ulf, I fear, has misunderstood the purpose of my paper, which was not to discuss only agency, the position of the person in the social system (although this is, of course, part of my argument), but personhood, the position of the person in the cosmological universe. To put it simply: studies, and even entire fields of study, that presuppose Western rational categories and the individual as the main unit of analysis cannot be a source of inspiration to anyone trying to understand personhood in a premodern context.

It was actually my uneasiness with sociological theory that made me leave out Dornan's (Reference Dornan2002) discussion of agency theory, a study according to Wright essential for my discussion. Her discussion is, of course, useful for someone who wants to find his/her way in the labyrinth of agency theory, but suffers from a serious weakness. While she criticizes the use of the Western notion of individuality when studying other cultures (Dornan Reference Dornan2002, 315), this is more an afterthought than an integral part of her argument, which dwells throughout on the tension between the individual agent and the social structure. Interestingly, Dornan sees the future of agency studies in archaeology in studies such as Thomas (Reference Thomas1996) and Meskell (Reference Meskell1999), which I criticized for retaining incongruent, or underdeveloped, notions of agency alongside their nuanced discussion of personhood. As a reaction to these studies, I developed the notion of relational agency by using insights from moral philosophy.

I now turn to the question put by the commentators: why moral philosophy? My answer would be to reverse the question: why have we not used moral philosophy so far? (although both Michael Walzer, who is approvingly mentioned by both Ulf and Wright, and Heidegger, who has deeply influenced, among others, Julian Thomas's discussion of personhood, are moral philosophers). I suspect that this resistance to moral philosophy is symptomatic of the academic division between sociology, history and philosophy – a division which Alasdair MacIntyre vehemently deplores: ‘There ought not to be two histories, one of political and moral action and one of political and moral theorizing, because there were no two pasts, one populated only by actions, the other only by theories.’ And he continues, ‘Every action is the bearer and expression of . . . theory-laden beliefs . . . every piece of theorizing . . . is a political and moral action’ (MacIntyre Reference MacIntyre1985, 61).

Let me restate my position: if we want to discuss agency, and if we reject the notion of free will, we need moral theory in order to reflect on motivation. Moral theory tells us that we always act, whether consciously or unconsciously, with a certain purpose in mind, a set of goals defined within a moral code, a cultural tradition and a set of historical conditions. This is why I find MacIntyre's work inspiring: because he brings together morality, culture and history.

However, a moral tradition is only an abstract scheme which constrains, but does not fully determine, our actions. For this reason I move from MacIntyre's perhaps static construct to Walzer's (Reference Walzer1994) more nuanced position, which allows for different moral choices (in this respect, I am rather surprised by Wright's remark that I ignored Walzer's argument, since I made precisely this point in my paper). In this way, I can insert a relational element in discussions of motivation, and thereby conceptualize forms of agency different from our own and compatible with the idea of a relational personhood. My choice, however, is also dictated by my observations on the shaft graves, the diversity in mortuary practices and the different rate of innovation, but also the shared features and parallel developments among the burials. Adherence to norms is seen alongside their imperceptible redefinition and their dramatic transformation; discord alongside agreement (here, I am puzzled by Ulf's remark that I downplay discord in social life).

I used moral theory for another reason: in order to understand imagery. If we read MacIntyre's brilliant discussion of contest and cooperation among warriors, only a cursory look at the Lion Hunt dagger will suffice to demonstrate that free agents did not inhabit the social imaginary of the emerging elites. Incidentally, a recent study of the Homeric texts reaches independently a similar conclusion (Graziosi and Haubold Reference Graziosi and Haubold2005, 104–10). As Tomášková rightly remarks, MacIntyre's discussion about the relationship between narrative and selfhood is extremely useful when trying to understand the explosion of figurative art in this period. I agree with her that images constitute, visualize and fix a moral discourse about status, excellence and beauty. Questions immediately arise: why is the moral discourse about men so coherent, why is the representation of women inchoate and underdeveloped, and why are children totally absent? I intend to explore these problems elsewhere, but here I can expand on Tomášková's other interesting question: why the sudden explosion of images at that particular moment? I believe that narrative, imagery and mortuary ostentation are three strategies, or rather three facets of the same strategy, aimed at stabilizing this emerging moral code in a period of change, fluidity and discord. It is, of course, not a coincidence that these three facets converge on the shaft graves of Mycenae. I will return to this point in the last part of my response.

To conclude: moral theory (rather than sociology or social psychology) is a necessary addition to our theoretical apparatus, if we want to redefine agency and make it commensurable to a relational definition of personhood.

Homeric questions

Opinion on the relevance of the Homeric epics for the transition to the Mycenaean period was equally divided. While Wright agrees that Homer can help us explore the moral code in the early Mycenaean period, Ulf recommends that the reader skip this section altogether. According to him, no discussion of Homeric parallels is possible, unless we first address all the questions arising from the long and passionate debates surrounding the epics. I am afraid I disagree with this position. A certain consensus has been reached on some of the questions Ulf lists. Though not everyone agrees, the ideas that the poems were written down between 750 and 650 B.C. (Ulf Reference Ulf, Raaflaub and van Wees2009b, 81), that they contain elements of different periods, and that they do not faithfully describe a specific historical stage are now gaining ground. Whether the epics arose out of oral traditions or from written sources, whether they received ‘oriental’ influences or not, and so on, is not really relevant to my argument. The question that matters is this: does Homer express the moral code of this heroic world (a code which, needless to say, allows for different views, oscillations and dilemmas, as the protagonists shift their position and reflect on their actions – Williams Reference Williams1993; Graziosi and Haubold Reference Graziosi and Haubold2005)? Here the answer is clearly yes. The next question is crucial: does this code bear any relation to the values prevailing during the transition to the Late Bronze Age? Ulf does not address this question directly, but tacitly adopts the now prevailing position: the Homeric poems may contain some archaizing elements to create ‘epic distance’, but for the rest they reflect upon the 8th century, and bear no relation to the Mycenaean world (see e.g. Finley Reference Finley1954; Morris Reference Morris and Cairns2001; Morris and Powell Reference Morris and Powell1997). However, there is an important flaw in this conclusion (reached usually by ancient historians): it equates ‘the Mycenaean world’ with the last phase of the palatial system, the 13th century. But this is only a brief interlude in a long process (see table 1) which includes the formative stages, starting with the recovery of the mainland societies during MH I–MH II, and continuing with the intensification of changes during MH III–LH I (the focus of this paper), the rise of regional centres in LH I–LH II, the appearance and peak of the palatial system in the later Mycenaean period, and finally the slow and uneven process of disintegration which stretches across the ‘Dark Ages’ until the 8th century.

It is impossible to deny that this Heroic Age contains elements of the early Mycenaean period: the arguments about boar's tusk helmets, for example, or pre-Linear B linguistic elements in the Homeric language (Horrocks Reference Horrocks, Morris and Powell1997), have been mentioned all too often. Some scholars have argued that the kind of stories immortalized in the Homeric epics were circulating in the early Mycenaean period (see e.g. the ‘Siege Rhyton’, figure 5e, though other examples exist; Morris Reference Morris1989; Hiller Reference Hiller, Hardy, Doumas, Sakellarakis and Warren1990; Negbi Reference Negbi1994). I agree with this view; I cannot believe that heroic achievements were for the first time depicted during this period, but were not narrated, or avidly listened to. Indeed Sherratt (Reference Sherratt1990, 817) has explicitly argued that the early Mycenaean period saw the initial creation of the bardic tradition.

But we need to go one step further: do the values expressed in the Homeric epics (not just the stories) characterize also the onset of the Mycenaean period? Let me be clear about this: I am not assuming that they were; I am trying to demonstrate exactly this point. I have explained body modification and adornment and the emphasis on weaponry in male burials of the shaft graves as materialization, or visualization, of a new social and moral prototype: the young warrior. And I have detected the values of beauty and strength, of cooperation and competition in the narrative of the Lion Hunt dagger or the golden rings.

But my point is not only that the early Mycenaean era saw the earliest expression of a tradition which became transformed and crystallized into the Homeric epos. I want to stress that there are important parallels between the transition to the Mycenaean period (the 18th–17th centuries, or MH III–LH I periods) and the 8th-century ‘renaissance’: both societies are characterized by a simple agricultural economy, with small-scale trade alongside exchange of gifts and luxuries displayed in feasts and ostentatious funerals; age and gender divisions were important in both, and status was based on descent and personal (especially military) achievement; in both societies men dominated, though elite women had some form of authority (the features describing the 8th century taken from Ulf Reference Ulf2009b). During both phases the mainland recovers from a deep and prolonged crisis, and rising elites seek to define their image and fix their status by means of lavish burial and visual (and verbal?) representations of military prowess (Sherratt Reference Sherratt1990, 821). Both are effervescent periods characterized by rapid internal change, discord and fluidity, but also increasing integration to a wider world.

To conclude: I do not use the Homeric epics as an explanatory model, but as an appropriate, i.e. structural rather than formal, analogy.

The focus on the shaft graves of Mycenae

Wright criticizes the narrow scope of my paper by pointing out first the missing comparative dimension. I am not sure I could cover the moral and cosmic codes of the entire eastern Mediterranean in one paper, but I also doubt whether I should. I believe (though cannot discuss this further here; see Voutsaki Reference Voutsaki and Kilian-Dirlmeier1999) that the societies of the southern mainland were exposed to formative influences by the Minoan palatial civilization and the Aegean maritime polities in MH II–LH I, but did not come into direct contact with the complex societies of the eastern Mediterranean.

Further, Wright expresses reservations about my focus on the shaft graves of Mycenae. He is absolutely right that we are dealing with numerically few graves, which are anomalous in terms of complexity, rate of innovation and wealth. I hope my discussion above explains why I concentrate on the graves of these ‘trendsetters’ (to use Wright's expression) in a discussion about agency and personhood. Further, he suggests that I should have included the MH II period, the graves surrounding the shaft graves (the ‘Prehistoric Cemetery’), and the other cemeteries in the Argolid. He is absolutely right that the shaft graves should not be treated in isolation – but we are doing all the things he suggests (and a lot more) in the Middle Helladic Argolid Project. There we analyse the evidence from all cemeteries and settlements; we undertake a detailed comparison of mortuary practices and household organization across the entire region; we re-examine all skeletons and study pathology and health status; we carry out an extensive programme of radiocarbon, stable isotopes and ancient DNA analyses (see http://www.MHArgolid.nl). I could not possibly fit seven years of work by a team of 12 people in one paper. I should stress, however, that all these analyses of different types of data are only meaningful because of this one initial step, of building up a theoretical framework around personhood and agency. This allows us to study broad trends and individual variation, and to understand how different persons, groups and communities responded to, benefited from or became excluded within the wider processes of change that took place during the end of the MBA. To finish, I would like to thank James Wright for extending the discussion into the Mycenaean era, and thereby presenting the outcome of these processes of change which engulfed, and indelibly transformed, the entire southern mainland. I am grateful also to Christoph Ulf and Silvia Tomášková for opening up the discussion and making it relevant beyond the Mycenaean world.