We would like to thank the five commentators for their thorough and stimulating reflections on, and criticism of, our article. The different comments raise various issues, and we appreciate their diversity of perspectives and their analysis of problems in our attempt at a rethinking of emotion in archaeology. The comments are each in their own way highly rewarding for us, and they certainly bring concerns to the fore that we have left out. Here we identify several issues that the commentators address in different voices and with varying intensities, and would like to examine these in turn. First, we consider the question of ritual at Mount Pleasant and the absence of the quotidian from our account. Second, we engage with the worry expressed over the lack of specificity of emotions in our given scenarios. Third, the phenomenological perspective in our article is given some critical thought. Fourth, we address the important point on which several of the commentators agree: that we leave out how emotions unfold in historically specific and context-dependent situations. Finally we turn back to the issue of our vocabulary to see how it stands the test of both application and critique.

Ritual and daily life

In our approach to emotions, we aimed to move beyond emotionally ‘obvious’ contexts, such as funerals, where a number of specific emotions might be expected to occur. Thus we turned to Mount Pleasant, a site at which a range of practices occurred, only a tiny minority of which have any connection to the dead. The practices at Mount Pleasant, as we see them, vary enormously to the degree to which they are structured, and, indeed, ritualized. Swenson, in his comments, urges us to consider the ritualized nature of the practices at Mount Pleasant, arguing that this would allow us to provide a more situated and contextualized understanding of the affective fields (to use our term). Whilst we acknowledge there is more to say about the temporality and intensity of the different practices at the site, we are reluctant to categorize Mount Pleasant as a purely ritual locale. The acts of building and consumption alongside the ad hoc practices of deposition, which Brück (Reference Brück2001) so accurately describes, do not have the feeling of a separated sphere of activity to us. We are deeply suspicious of attempts to segregate ritual from daily life, rather seeing the two as entwined and emergent from, rather than prefiguring, certain kinds of practice. Teasing out where ritual aspects emerged at Mount Pleasant, and specifying these, would undoubtedly have added to our account, but we would hold back from the more general reconsideration Swenson calls for. Equally an attention to more quotidian contexts (whether ritual or not, Mount Pleasant is undoubtedly out of the ordinary), something Berggren also recommends, would certainly add to our broader understandings (and see Harris Reference Harris2009). In this context we can simply plead the defence of lack of space; we had room for one case study, and chose Mount Pleasant precisely for the breadth of contexts it offered at a single site.

The specificity of emotion

Several of the commentators, including Kus, Smith and Tarlow, are critical of our failure to identify specific emotions at the site. They would like us to specify the kinds of emotion and, particularly in Tarlow's case, to allow their intensity to emerge. For us, however, attending to emotions is not simply about looking for intense or particularly rich emotions, because many or most of the feelings that people experience may be trivial, familiar or fleeting. These emotional contexts are just as important to appreciate as part of an emotional reality, even thought they may be much harder to detect or deduct from archaeological data compared to stronger emotional contexts. Here we may recall how historian Eelco Runia (Reference Runia2006) describes the trivial and the mundane, that which has become clichéd, as an inherent challenge to the sense of presence, which is another affective occurrence or atmosphere that we might add to our address of emotion. Seeking to create a feeling of presence is certainly about attuning people's attention to a given occasion, situation or place.

‘A sense of presence’ may sound like a very indistinct emotional reference or feel, but we are wary of attributing more specific emotions to a prehistoric context, which we believe runs the risk of colonizing past emotions with our own. This may also be a reason why our analysis of emotions at Mount Pleasant can appear ‘bloodless’, in Tarlow's term, or dissociated from the emotional specificities that we would have been able to construct in an ethnographic or historical context, where living informants or text could have provided clues to the particular feelings and atmospheres that people experience. This, of course, highlights some of the inherent challenges when trying to approach emotions in a purely archaeological context; that is, from the perspective of material culture (Sørensen Reference Sørensen2010). It is interesting to note here that Berggren too struggles to identify particular emotions in her case study. The potential to describe more vibrant and specific pasts is one we should all be aiming for, and to this extent we acknowledge that there is much more to be said in our account.

It may well be, however, that Kus's contribution indicates the way forward on this issue. Her use of psychological anthropology points out that a further development of vocabulary may be required in order to begin to talk about more defined and specific emotion-words. This offers the potential to identify areas of ‘overlap’ between different contexts, and to build from that overlap to recognize alterity both between and within emotion-worlds. Additionally, the potential of metaphor that both Tarlow and Kus stress seems to us a very promising way of developing further understandings of the way things and people constitute emotions within particular affective fields. Indeed, it may well be that affective fields are themselves constituted through material metaphors, as much as through practice, and this area of our thought demands further attention.

Phenomenology

If we are wary of essentialism in this regard, it is something we also take very seriously when it comes to the body. Thus we would like to take issue with the critique of the phenomenological disposition offered by Tarlow. More specifically, we would like to distance ourselves from the version of archaeological phenomenology she associates us with. Let us be explicit: we are in total agreement with critiques (most notably Brück Reference Brück1998; Reference Brück2005) that have demonstrated that landscape phenomenology often relies upon an essentialist version of a transcendent human body, that is the same in all times and places.

Often, this version of phenomenology has been taken to be synonymous with ‘experience’, and ‘subjectivity’ has been reduced to ‘personal experience’. We believe that such readings have only limited usefulness, and omit the cornerstone of phenomenology that can be summed up as the epoché or ‘bracketing’ of enculturated expectations to experience and – in our case – emotions. If we bracket – or, in other words, suspend – our personal expectations of emotions in our analysis, then we may be at risk of creating more abstract or even pale reflections of past worlds, but on the other hand, we may be less liable to project our own cultural context onto the other. In this connection, ritual may be a point of reference that deserves more critical scrutiny in order to avoid being framed by modern associations with what constitutes or characterizes a ritual situation or context.

Furthermore, our reading of Heidegger does not require us to hold an essentialist view of the body, because the body, like other aspects of being-in-the-world, is disclosed against particular and historically specific backgrounds. Thus the kind of body disclosed to Dasein varies. Similarly, kinds of emotions (or moods, to be Heideggerian about it) vary through the worlds into which a person is thrown. This version of phenomenology may privilege experience (as opposed to a truly symmetrical reading of things and people; see Latour Reference Latour1993; Webmoor and Witmore Reference Webmoor and Witmore2008), but it does not demand a singular eternal body.

Historical context

To return to our article: the abstract rendering of past worlds that may be the consequence of a bracketed analysis could be at risk of overlooking the historical particularity of certain emotional contexts, as proposed by Tarlow, Swenson and Smith. We believe, however, that starting out on the basis of the archaeological material means that we are always historically anchored and consequently need to make culturally, historically and locally grounded interpretations of emotion-worlds. Being contextually grounded also means, in turn, that the proposed vocabulary can be refined, adapted to specific requirements and adjusted according the nature of one's analytical context.

However, part of our aim of setting up a vocabulary was also to allow the analysis of emotions to begin at the unprejudiced rather than the assumptive, whereby the predetermination of emotional implications of the analyst's expectations of the context may be suspended. This would offer the analysis a potential transparency, as our vocabulary does not subscribe to any notion of natural or universally recognizable emotions. In other words, analytically it might be necessary to free oneself of the prefiguration that, for example, a funeral is necessarily about a certain range of feelings (e.g. sorrow, sadness, tension). Analytically, we cannot sustain such a prejudice and maintain our academic credibility, which forces us to identify specific contexts before making interpretations of the emotions associated with them. So, in other words, the particularity of emotions in given historical contexts will remain a challenge to the researcher's gaze, and we believe that the starting point must be the archaeological material from which emotions should be extrapolated.

As the commentators point out, there is undoubtedly more to say about the specific historical setting of Mount Pleasant, and the world in which this and other henges were constructed in the third millennium B.C. An analysis of the broader context would have allowed us to discuss the development of this landscape in detail, to tease out the different kinds of material practice that constituted differing scales of community, alternative forms of identity, and the kinds of emotional regime implicated in this (cf. Harris Reference Harris2006). We cannot, however, agree with Smith that we need to begin with institutions and the social setting, and allow emotions to emerge from that. As archaeologists, whilst we encounter the material first and foremost, our interpretive task is to recognize how different worlds are possible within these material conditions (Barrett Reference Barrett and Hodder2001). Or, better yet, to explore how the social and material are never in fact separate from one another, but rather emerge together, in the worlds we explore. Therefore emotions should neither precede nor follow from the emergence of particular locales, but are rather central to this process of emergence, precisely because this is the very means by which places come to be recognized (Harris Reference Harris2009; in press).

Vocabulary

Part of the reason that the historical contingency of emotion is not as foregrounded as it could be in our article is undoubtedly due to the dual task we set ourselves, and the limitations on space this created. We were certainly interested in the emotional and affective valences of a site like Mount Pleasant, but we also aimed to construct and test a vocabulary that would help us to access these aspects.

Within this vocabulary, Swenson is critical of our notion of affective fields. He suggests that the concept could easily be mired in circular reason and is wary that multiple different activities could be subsumed within a single affective field, reducing them all to the outcome of a reified set of relationships. Of course, he is right to be suspicious of this. But our point was never that these different activities (eating, processing, harvesting and so on) could be placed within one well-defined affective field. Rather, affective fields are always multiple and are characterized by fuzzy contours: affective fields are socially shared; hence their significance may differ from person to person. Thus whilst there are links across affective fields, particularly through the way they are materially instantiated, they cannot be reduced to the singular.

Metaphor here is crucial again, though certainly underexplored in our article. How do things move across and between affective fields? How might acts of harvesting, for example, resonate emotionally with gathering up the belongings of the dead, or the remnants of a feast? Without reducing these to singular affective fields the power of an approach rooted in metaphor and poetics, as Tarlow and Kus suggest, may well be to allow us to see how these fields allow emotions to resonate across contexts. If things can be sticky with emotion (sensu Ahmed Reference Ahmed2004), why not metaphors as well? In fact, were we to collapse the divide here between concept and thing (cf. Henare, Holbraad and Wastell Reference Henare, Holbraad, Wastell, Henare, Holbraad and Wastell2007; Holbraad Reference Holbraad, Henare, Holbraad and Wastell2007), we might find further ground for exploring the multiplicity of emotional responses in multiple forms of practice.

Similarly, Smith is critical of our assimilation of the difference between affect and emotion. Our aim in doing so (following Simonsen Reference Simonsen2007) was to avoid the simplistic separation of affect being equivalent to body and emotion to mind, as Smith points out. His criticism that this potentially leaves the roles of sublimation and dissimulation underrepresented is well taken, however. Whilst this is an issue that requires further thought, we are reluctant to reinstate the gap simply to meet this lacuna. Instead we suggest this may be an area (alongside the specifics of emotion, as discussed by Kus) that requires us to refine our terminology, perhaps offering a fifth term to go alongside those we have already designed. This is not the place to outline such a term, but we are happy to acknowledge the importance of Smith's point here as we move forwards.

As a final point we were excited by Berggren's use of our terms. We are very hopeful that as her account shows, this vocabulary – augmented by other tools, no doubt – does offer us a way to begin tackling the questions of emotion and material culture in the deep past. We believe these questions are important. Swenson queries whether it is wise to separate out emotion from other variables. This, he suggests, runs the risk of reifying our approach and ignores the complexity of human behaviour. There is much to be said for this. However, like any area of study that has gone undervalued a necessary first step is to focus on it explicitly to move the debate forward. Much in the way in which memory, personhood, identity and so forth are not in fact separate areas of human lives, but rather part of the tapestry of life's rich pageant, emotion can benefit, temporarily, from our undivided attention. Once our theories and methodologies are sufficiently developed (as they are now with memory, we suggest) the time will come to reincorporate our approaches back into the mainstream.

There are many other important insights the commentators raise that we do not have space to do justice to here in our reply, but will continue to provide food for thought as we develop our perspectives on these issues. Once again, we would like to thank our interlocutors, the two anonymous peer reviewers and the editorial board of Archaeological dialogues for this opportunity. We hope that our proposal for an analytical vocabulary may help others thinking about emotion in purely material contexts. Despite work on the theme over the past two decades, we still believe that emotion needs to be integrated more intimately in studies of prehistory, and we suggest that a rethinking along the lines of our analytical tools for understanding emotions through material cultural analysis may help with this. Furthermore, notions of both material and affective ‘turns’ currently abound in the social sciences (e.g. Bennett and Joyce Reference Bennett and Joyce2010; Clough and Halley Reference Clough and Halley2007). By taking an approach that unites these two perspectives, archaeology has the potential to make a significant contribution to wider debates.