Harris and Sørensen's critique of the archaeological inattention to emotion and their recognition of the material mediation of affect bring to the fore perennial epistemological problems defining the broader archaeological enterprise. The immediate citation of the long-discredited Hawkesian ladder of inference challenges the assumption that past emotional states are unrecoverable from archaeological contexts, just as an earlier generation of archaeologists rejected processual theory that meaning, conceptual schemas and symbolism fell beyond the pale of scientific inference. Of course, Hawkes was not a materialist in the strict sense of the term, and he recognized that value systems transcended the epiphenomenal and played a vital role in structuring social practice and shaping historical process. It was his contention, however, that conceptual and symbolic schemes and their role in social reproduction were simply too complex to be read satisfactorily from material remains (Hawkes 1954; see Fogelin 2008, 129–30). He wrote that ‘there is nothing in North American ecology . . . to compel either Iroquois institutions . . . or the constitution of the United States’ (Hawkes 1954, 163). To be sure, Hawkes probably would not have denied that moved to move is intrinsic to the human condition and that affective dispositions were a force in individual experiences and the collective fortunes and self-representations of past communities. At the same time, he probably gave little consideration to the dialectical interdependence of the material world and emotion, a relationship that has captured the imagination of recent scholars. Hawkes would no doubt have scoffed at the notion that emotion as ontological problem, cultural construct or variable of social interaction is amenable to archaeological interpretation.

I applaud the authors’ endeavour to develop a heuristic package, an analytical ‘vocabulary’, as they cautiously refer to it, which will aid the investigation of emotion in past societies and contribute to our understanding of the inherent material conditioning of affect. Indeed, this article represents a valuable contribution to material culture studies and problems of deciphering how artefacts and the built environment were constituted by and constitutive of human emotion. Although I part company with Hawkes that emotion eludes archaeological inquiry, my main criticism of the article is centred on his acknowledgement of the complexity of behaviour and the fundamental theoretical challenges inherent in isolating affect as an independent variable. Just as ecology cannot predict Iroquoian institutions, the same could be said of emotion. Of course, Harris and Sørensen would emphatically agree with this premise, as made clear in their paper, but the thrust of their argument still tends to reify emotion in a way that elides the cultural and historical specificity of affective experience. I do commend the authors for venturing forth on the dizzying anthropological tightrope separating the chasms of extreme relativism on one side and reductive universalism on the other. In citing Rosaldo, Strathern and Lutz, Harris and Sørensen condemn notions of ‘transhistorical humanity’ and agree that emotions, like belief, are culturally constructed (see also Meskell 1999; Tarlow 2000). Nevertheless, they still recognize that embodied emotion constitutes a general condition of the human species and is a phenomenon amenable to comparative analysis and anthropological generalization – a position I fully endorse. However, the historical context of emotion in the Late Neolithic period is still sidelined in the paper, despite statements to the contrary and numerous caveats that assert the importance of just such a context. At the end of the paper, emotion reads to me as a nebulous, decontextualized, and somewhat free-floating ‘thing’ – even despite the authors’ compelling argument that emotion implicates both mind and body and is inextricably tied to practice and specific sociohistorical traditions.

What I found particularly striking about the paper is that the theoretical critique closely mirrors recent exposés on the archaeology of ritual (Bradley 2005; Fogelin 2007; Kyriakidis 2007). If emotion were substituted by ritual performance, the thrust of the analysis would have changed insignificantly and our understanding of the emotional valence of Mount Pleasant would hardly be dissimilar. To provide an example, the analytical category of attunement is reminiscent of Renfrew's (1994) argument that ritual ‘focuses attention’, or Smith's (1982) contention that ritual serves as a ‘focusing lens’ that heightens consciousness and emotional sensibilities. Returning to the anthropological tightrope, certain scholars have also questioned the validity of ritual as a general category, and several archaeologists contend that the focus on ritualism distorts interpretation of past structures of practice and the existence of plural rationalities. If ritual is inseparable from instrumental action operating within specific cultural constructions of the world, and if it is irreducible to representation and communication, then its analytical worth as a cross-cultural category is thought to be compromised (see Brück 1999; Gerholm 1988; Goody 1977). Nonetheless, despite the myriad cultural permutations of ceremonial behaviour and constructions of the sacred, a majority of social scientists would contend that highly formalized, symbolically charged and rule-governed practices, occasioning a disconnect between action and normative intention, constitute a generalizable, if not universal, human phenomenon demanding archaeological scrutiny (Bradley 2005, 33; Bell 1992; 1998; Humphrey and Laidlaw 1994). In such perspectives, ritual is cast as a particular mode of embodied practice or action, and it has been touted as the trope that will enable the transcendence of mind/body and subject/object reductionisms and even of Eurocentric notions of the sacred and profane (Bell 1992; Comaroff and Comaroff 1991; Handelman and Lindquist 2004). Harris and Sørensen's take on emotion aims for a similar surmounting of such simplifying dichotomies. Certainly, the research of Houseman and Severi (1998), as well as of Humphrey and Laidlaw (1994), on the prediscursive aspects of ritual acts has shown that religious meaning is polysemic, ambiguous and paradoxical; ritual as action invites a multitude of possible significations with diverse political repercussions – where meaning is not simply a priori and mechanically reproduced in ritual but is variably generated in the structure and flow of rite. In other words, the primacy of mind (or mythopraxis) is countered and more attention is given to phenomenological issues and to questions of affect (Thomas 2002). Ritual is action in the subjunctive mood, as Turner argued (1967), and it normally involves the focusing and intensifying of attention, public or individual, which implicates distinctive frames of practice, thought, performance and, by extension, emotion (Lewis 1980; Tambiah 1979).

In fact, ritual is commonly viewed as practice that heightens or alters consciousness and induces intense emotional states (Bateson 1986; Turner 1967; 1982). It has been variably argued that the multisensory experience of ritual theatre is an effective tool in forging social solidarity, not so much by articulating shared belief among participants but rather through inculcating collective experience predicated on emotionally arresting action (Fernandez 1965; Kertzer 1988, 99–100). Kertzer (1988, 76) notes that

In truth, Sørensen and Harris's article is markedly, if unwittingly, Durkheimian in its theoretical stance, especially in its emphasis on the role of emotion in creating a sense of shared community identity at Mount Pleasant. Durkheim's notion of collective effervescence (1965) resonates directly with their statement that ‘emotions can be experienced by social groups and at times by large crowds, sometimes even by an entire people or nation, or by large parts of the world’ (p. 150). Highly pertinent to my appraisal, almost every example of emotion mobilized by Harris and Sørensen, both ancient and modern, conforms to what many would classify as ritualistic behaviour. This is apparent even despite their call to liberate emotion from the reductive framework of ritualism, as reflected in their referencing of Tarlow's critique of the distillation of death to the singular analytical framework of ritual (which ignores feelings of grief, sorrow and loss). Nevertheless, the immediate context for these emotions is still death itself, and emotions cannot be extricated from the situational, cultural and semantic framework in which mourning practices are embedded. In the same vein, the peculiar affective power of the henge monuments of Mount Pleasant is contingent on the practices expected or performed at the monument, including architectural construction, feasting, initiation, musical production, political posturing, intense self-reflection and even the possible ritual enactment of mythohistories. Harris and Sørensen analyse emotion in an exclusively ritual context, and it remains problematic that this context is explicitly downplayed and undertheorized in their analysis. They state their desire to move beyond Thomas's notion of ‘dramatization’, but I am not convinced that this is fully achieved, and it is hard to deny the dramatic and highly theatrical aspects of this impressive ‘ceremonial’ site.

This realization brings me to the heart of my critique: emotion cannot be divorced from specific historical, political and experiential contexts – ritual or otherwise, and traditional contextual–hermeneutic reconstructions must come first in archaeological interpretations of emotion (Meskell 1999; Tarlow 2000, 728–29). One might object that such a perspective relegates emotion to epiphenomenal reactions to traditional social, religious or material factors. However, stressing emotion's contingency does not necessarily imply such a simplistic or determinative chain of causation. Instead, what must be privileged is the social and phenomenal framing of past events and landscapes that jointly implicate the cultural, emotional and embodied, as Harris and Sørensen implicitly acknowledge. A particular context steeped in public ritualism or high theatre is clearly distinct from private arenas of domestic production and consumption. Whether Bourdieu's fields, Holland's figured worlds, or the ludic frame as theorized by performance theorists, emotion is dialectically propelled in specific, if changing, contexts of practice – that are at once cultural, political, economic and affective (but not strictly so in the latter case) (Bourdieu 1993; Holland et al. 1998; Köepping 1997). It is for this reason that I find the heuristic of the ‘affective field’ problematic; it could easily mire the analyst into circular reasoning and to the troubling reification of affect. Can diverse social activities, including an evening meal, a carnival, the harvesting of corn, the recitation of genealogies or the procession through a henge best be understood in terms of an affective field? Or are they more accurately culturally constructed social events of differing significance in which distinct modes and intensities of emotion are experienced, generated and performed (and, by extension, where these emotions are generative of future emotions, possible status distinctions, and changing conceptual categories)?

The authors argue that the affective field is the substratum in which variable and even ‘incongruent’ atmospheres are ‘felt’ and conceptualized by different groups of actors. This recognition of different emotional states parallels Gerholm's ‘postmodern’ analysis of a Hindu funeral ritual in Trinidad, wherein specific individuals were found to hold disparate understandings of the event despite a generally shared awareness of its emotional gravity (Gerholm 1988). Harris and Sørensen's ‘atmosphere’ also intends to overcome the common critique that phenomenological approaches in archaeology impose a singular and ahistorical sensual framework onto past subjects. The traditional phenomenological stance implies that built forms elicit shared affective responses, thus forging homogenous subjectivities and world views regardless of political and cultural differences (see critiques by Brück 2001; Johnson 2006). Indeed, the influence of the British phenomenological school of archaeology is strongly evident in the authors’ explicit linking of movement to the etymology and experience of emotion (‘to be moved’). In the end, the identification of differently constructed atmospheres at Mount Pleasant is far from revelatory and once again consists more of an exercise in semantic substitution. Making private offerings in the visually obscured ditch is contrasted with public feasting and the affective power and visually disorienting configuration of the henge rings, but the authors do not take into account that ritual events may have been sequentially integrated and that both emotion and meanings could have been embedded in a singular temporal unfolding of highly structured activities (both dramatic and mundane in nature).

As stated previously, context is all-important, and an inferential ladder will inevitably guide analysis (Tarlow 2000, 729). The henge, associated artefacts, feature emplacements, the perimeter stockade and so on must be interpreted as a contextual tableau in conjunction with the broader interrelationship of Mount Pleasant with neighbouring residential settlements and similar ceremonial sites (Hodder and Hutson 2003; Swenson 2008). In other words, emotion can only be inferred after meaningful social, historical and spatial contexts have been approximately delineated. Once again, this may reaffirm an inferential ladder, but it does not imply that emotion is interpretively more inaccessible or that it is socially epiphenomenal. Changes in the function and architectural configuration of Mount Pleasant, implying shifts in social memory and ecology, and possible transformations in religious beliefs and political affiliation, also must be considered before emotion can be effectively interpreted. Harris and Sørensen may maintain that they did grapple with contextual meanings in speculating on the possible symbolism of the house and the significance of the different strategies employed in dismantling the later palisade. However, this component of their analysis is underdeveloped and should have been given primacy in their exploration of the cultural and possible cosmological framing of emotion in the Late Neolithic and Early Bronze Age context (see Treherne 1995; Van Dyke 2009).

How would Harris and Sørensen approach the intersection of emotion and materiality in a household setting (as opposed to an awe-inspiring monument and ritual context such as Mount Pleasant – a ‘controlled environment’, as Smith (1982) famously described built environments of a sacred nature)? In reflecting on quotidian spaces, Bourdieu's (1973) celebrated ethnographic study of the Kabyle house demonstrated that the spatio-temporal and material framing of the everyday plays a key role in socialization and the reproduction of unquestioned power relations. In contrast, ritual practice and its material expressions have often been differentiated from the material parameters of the domestic and quotidian (see Barrett 2001, 158; Inomata and Coben 2006; Van Dyke 2009, 238–39). Religious experience is equated with calculated ideological production, ‘discursive consciousness’ and active or contested subject formation, while the common household is identified with the taken-for-granted, ‘practical consciousness’ and the apolitical (although Bourdieu's notions of doxa and habitus are concerned first and foremost with the reproduction of misrecognized structures of power). The authors write that ‘on occasions material things can stand out, demand to be noticed and defy attempts to ignore them; here they are “present-at-hand”’ (p. 151) (as opposed to Heidegger's ‘ready-to-hand’). In an archaeological context, the authors seem to imply that aesthetically and religiously charged items represent the default ‘ready-to-hand’ emotional stimulators. Of course, the dichotomization of consciousness and emotion within the framework of the mundane and ceremonial (and simplified archaeological–spatial domains of house and monument) is problematic; ritually enacted liturgies can be horrifically dull and emotionally understated, while social dramas involving heightened feelings (joy, conflict, satisfaction, fear) often unfold within everyday domestic contexts. At the same time, quotidian space can be highly ritualized and imbued with meaning wherein ceremony is incorporated into daily, practical routines. Discursive consciousness and heightened ideological awareness might normally find material expression in highly performative domains and public ritual spectacles, but this is far from universal; nor does it occur in a prescribed or ahistorical fashion.

So, then, where does this leave us in the archaeological study of emotion? Can inferences on ‘emotively significant’ landscapes only be made for monumental, ritual spaces (where distinct theoretical varnishes – say, a ritual bias or an analysis foregrounding intense emotion – precede comparable methodologies and interpretations)? Or are all landscapes emotively significant to the point that such an analysis is at once futile and pointless? As stated earlier, both ritual and emotion are amenable to analytical demarcation in archaeology, and the cynical question posed above is for the most part rhetorical. However, a hermeneutic analysis and attention to context will provide the only means to solve these very complex and important problems. For instance, future studies on the covariation of household design with shifts in the construction and use of henge monuments should prove illuminating in understanding the dialectic of place-making and the role played by human emotions at Mount Pleasant and neighbouring sites. Is there a notable continuity in residential architecture during periods of henge destruction or renovation? Are there any meaningful correlations between domestic structures and Mount Pleasant which might affirm that Site IV actually represented a ‘communal house’, thus allowing more robust inferences on the affective resonance of the household as symbol and setting of everyday practices? Were house lots also fastidiously curated spaces, and does a comparable trajectory of abandonment, burning or purposeful decay distinguish different types of site and structure? Or are residential villages and henges characterized rather consistently by distinct spatial biographies? I agree with the authors that the lack of houses at Mount Pleasant is likely relevant to deciphering the interpenetration of affect, meaning, built environments and possibly plural subjectivities in Late Neolithic and Early Bronze Age Wessex; clearly more research is necessary to examine the contextual disjunctures or interrelationships between archaeologically distinctive landscapes (as undertaken by Thomas (2007) at Durrington Walls). In a similar vein, did certain decorated wares mentioned by the authors both ‘figure’ worlds and promote emotional attunement given their particular and changing contexts (where emotion and consciousness are mutually constituted)? Does the deposition of ceramics in ditches express the emotional draw of communal events of festive consumption, or, rather, does it point to the consumptive destruction of the vessels themselves, as understood within a narrative of cosmic process and the emotions that the possible ritual re-enactment of cosmogonic myth might have invoked? Harris and Sørensen recognize the importance of archaeological context in discussing the differential deposition of grooved ware and decorated ceramics – but the affective field and its incongruous atmospheres are never satisfactorily connected to an underlying social and discursive matrix, beyond vague references to communal houses and moderated invocations of community solidarity in the spirit of Durkheim. I say ‘moderated’ because a sense of shared identity is inferred, but a sense of community is argued by Harris and Sørensen to have been ‘felt’ differently – as purportedly exemplified by the three distinct strategies of palisade destruction. Of course, Site IV, its concentric henge circle, and the placement and careful removal of the colossal wooden beams of the later palisade point to considerable planning, deep-seated beliefs and protracted labour investment, and not simply to ‘embodied emotions’.

To conclude, Harris and Sørensen's article has forced me to rethink certain assumptions about the role of emotion in ritual experience, as briefly outlined above, and I feel obliged to follow up on their request to critically evaluate the heuristic vocabulary introduced in their article. Although I have little concrete to offer in suggesting how the four constructs relate, it is significant that an inferential hierarchy (ladder) is implied in their analysis despite claims to the contrary; the primacy of affective fields in generating atmospheres through processes of individual attunement is made explicit in their analysis. As alluded to above, I recommend the embedding or even sublimation of ‘affective fields’ within specific cultural-historical and performative ‘frames’, which will no doubt demand a delineation of distinct but possibly overlapping contexts – including public ritual arenas, living spaces, agricultural fields, symbolically bounded landscapes, gendered places, exchange locales and so on. A more explicit demarcation of the broader context will prevent the undesired reification of emotion as an autonomous and ahistorical agent. Finally, sensitivity to these more encompassing frames will demand a reconsideration of the argument that ‘contestation’ at Mount Pleasant was not a question of power, but of ‘conflicting emotional understandings’ of the settlement's monumental sense of place (p. 161). I am not sure what is meant by ‘emotional understandings’, a fascinating juxtaposition that encapsulates many of the unresolved tensions of Harris and Sørensen's article. Does emotional understanding simultaneously implicate meaning, discursive consciousness, cognitive and sensual attunement (‘attending to the world’), as well as Heidegger's ‘present-to-hand’ or Lutz and White's (1986) ‘embodied thoughts’? The idea that ‘disparate affective relations’ can account exclusively for the differential destruction of the palisade is thus rendered suspect; timber removal, protracted decay, and the immediacy of burning were likely motivated by different religious or political convictions (of varied emotional weight) or even by a more integrated cultural understanding of death and regeneration. Harris and Sørensen's notion of affective relations is reminiscent of Reddy's argument that ‘neither the concept of discourse nor that of practice is adequate to capture the active power of emotion’ (cited in Tarlow 2000, 717). However, the power of affect defies interpretation if its discursive, material and structural contexts are ignored (Meskell 1999; Tarlow 1999; 2000; Treherne 1995); I agree with Tarlow (2000, 725–28) that the search for emotion in past societies is impossible if this search does not include the exploration of past meanings and how they are jointly materialized. Although involving intense emotions (as does any explicit performative negotiation of power), contestation is irreducible to ‘emotion’, whether simply felt or fully understood.