Harris and Sørensen's paper is a welcome attempt to address the question of ‘finding’ and interpreting emotion in the deep past. Their contribution to this emerging area of debate is particularly valuable in that it treats upon a prehistoric context, and is thus unaided by the rich contextual information provided by historical sources; and it moves the debate away from the mortuary context towards other areas of lived experience – here the processes of constructing, inhabiting, engaging with and ultimately destroying an architecturally defined space. These are more ambitious and less obvious contexts for constructing emotional pasts. Where the authors are most successful is in the identification of cultural loci where emotions are developed and are involved in the construction of experience – such as in the production and reproduction of ‘sense of community’. Less convincing, for me, were the places where they follow the (mainly British prehistoric) ‘phenomenological’ tradition. My own view is that shared and expressed emotional values are more amenable to archaeological identification and analysis than personal emotional experience, and I shall try to explain why. I think by separating the social meanings of emotion from the physical experience of emotion the authors of this article might be able to pursue more fruitful kinds of enquiry.