Abstract
Far from being a universal feature of culture, the concept of religion has distinctly western origins. What, then, is religion, and how shall it be empirically studied? I suggest, as one of many possible alternatives, an etymologically-based approach to religion, understood as trust in sources of epistemic and moral authority. Four authorities are considered, including institutional religion, science, nature, and the state. I present results of a survey-based empirical inquiry of U.S. adults, enriched by means of follow-up interviews exploring their trust or distrust in these domains of authority. Based on this inquiry, two hybrid forms are at the forefront of religious debate among Americans: theocracy, a linking of trust in institutional religion and government, and ecology, a combined trust in nature and science. These results are regionally variable in the United States, and cross-national data clarify the exceptionalist position of the United States with respect to European countries. Trust in authority emerges as a fruitful means to link seemingly disparate realms of social life, and offers an important basis for geographic comparison. Yet whether understood broadly as trust in authority or along other lines, the geography of religion will benefit from greater theoretical precision and methodological pluralism as suggested in this study.
Acknowledgments
The project was supported by National Science Foundation research grant BCS-0082009. I appreciate the major contribution of two University of California–Santa Barbara graduate students, Evan Berry and Tricia Mein, who worked alongside me.
Notes
aFactor extraction: principal components analysis. Rotation: varimax (Kaiser Normalization). Factor loadings under 0.1 omitted.
1. For a much fuller account of natural kinds, see CitationPutnam (1975).
2. Two methods for factor retention were employed: all factors were retained with eigenvalues (representing the amount of explained variance) greater than 1, and/or those that preceded an abrupt change in slope on an ordered eigenvalue scree plot (CitationCattell 1978; CitationKim and Mueller 1978).
3. The Religion II module is stored in the University of Cologne's Central Archive for Empirical Social Research as ZA3190 (www.gesis.org/ZA); variables V20 and V22 were included in this analysis.
4. Religion II variable V31.
5. The Environment II module is stored in the University of Cologne's Central Archive for Empirical Social Research as ZA3440 (www.gesis.org/ZA); variables V10 and V18 were included in this analysis.