Blood, War, and Ritual: Religious Ecology, ‘Strong’ Culture, and Human Sacrifice in the Premodern World

1 We would like to thank the editors and anonymous reviewers at Anthropological Forum for the helpful comments during the review and revision process. We would also like to thank Jeffrey C. Alexander, Philip Smith, Derral Cheatwood, and Andrea L. Ruiz for comments on an early draft of this paper. A version of this paper was presented at the 2008 Society for the Scientific Study of Religion Annual Conference, Lexington, Kentucky.

2 The introductory section of our review of the academic literature on this topic is based on Winkelman (Citation1998) and Ernandes et al. (2002).

3 We need to reiterate here that we apply aspects of the strong program that emphasise the idea of culture as an autonomous social force. We are aware that a central tenet of Alexander and Smith's work is the ‘reading’ of culture as text, in which comparative analysis would be a less suitable analytical strategy than a case study approach. The strong program would dictate a closer and more detailed analysis of a single culture with emphasis on the binary codes that make up the meaning systems of that culture. Our aim here is not so much, as in a pure strong program approach, to textually analyse a premodern culture and to render an interpretive analysis of meaning systems that may shed light on the process of human sacrifice as a whole. We are interested in the effect of cultural forces as they are evident in empirical data and how those forces stand up in the face of structural variables. A classic example of a more purist structural account that predates the strong program is found in Hubert and Mauss (Citation1961).

4 Contemporary language and region codes are formulated and available in Burton (Citation1999).

5 Although highly regarded by many scholars, cross-cultural research using SCCS data has its detractors and has been criticised on several counts (for a thorough discussion, see Ember and Ember Citation2000). A common criticism is that the inherent uniqueness of individual cultures makes the comparison of cultural traits between societies unfeasible. In other words, because each society is an independent entity that is made up of distinct cultural and structural characteristics, comparison between societies is not possible. This criticism is contested, however, by pointing out that cross-cultural comparison does not deny that societies may share similar traits. Instead, researchers should be cautious not to conflate ethnographic findings with cross-cultural analysis of the sort we report below. Ethnographies tend to focus on what makes a particular society distinct, while cross-cultural research reveals commonalities in human behaviour across social systems, and consequently, highlights the general traits that exist and vary between societies. So in other words, ‘[e]thnography reports what is distinctive about a particular culture. Cross-cultural comparison relates what is general, what is true for some or many or all human cultures’ (Ember and Ember Citation2000, 353). A second criticism is that the sampling in cross-cultural research is biased, since it is essentially impossible to identify the sampling ‘universe’ of all possible cases, as is standard in random probability sampling that occurs when individuals are the unit of analysis. To address this issue fully, however, we would need a list of all world societies from which a random sample is drawn, a task that is clearly unrealistic. However, as Ember and Ember note, it is common in even highly structured survey research and population studies to assume a numerical universe of observations will include large numbers of missing cases (Ember and Ember Citation2000, 356–357). Even the most advanced population studies using random probability sampling assume that a sampling universe does not include individuals that are highly mobile, homeless, or simply lack permanent addresses.