In this essay the nature of crime and death in opera is dealt with, focussing on a select number of operas, such as Verdi’s MacBeth, Wagner’s Die Walküre and Strauss’s Salome and Elektra. It appears that crime in opera is pervasive and appears in many different forms, as well as in a wide variety of contexts. An attempt is made to analyse the way composers manipulate the circumstances of killings – lenghtening or shortening or replacing scenes – in order to get maximum results in terms of emotional content.

This article reflects on some of the qualities and strengths, as well as some potential weaknesses, of a research methodology used to study ‘hard-to-reach’ groups, such as the undocumented. This approach, known as the PEER (Participatory Ethnographic Evaluation and Research) approach, is introduced in terms of its key elements of trust, anonymity, in-depth data and flexibility. Its suitability for sensitive, or ‘liminal’ research issues, involving groups of vulnerable informants, is explained. The method is based on relations of trust, which are maintained through anonymity in data collection, and extend from social researchers to informants, through the intermediation of trained community-based peer researchers. It is they who interview others in their own social networks; since trust is the key ingredient in making this ethically-informed methodology work well, trust must be invested also in the peer researchers, who form part of the research team.