On Comparison, Part 3

Best Practices

(avg. read time: 4–7 mins.)

As Stausberg observes, there are many challenges that the comparativist must address, in that comparative analysis: 1) requires extensive preparations; 2) can require various sorts of specialist expertise; 3) is time-consuming; 4) can create difficulty in finding matching sets of source materials; 5) is prone to mistakes because of complexity; 6) can sacrifice depth for breadth in wide-ranging comparisons; 7) may be insensitive to contexts; 8) is prone to confusion because of surface similarities or differences; 9) can be poorly received by specialists; 10) is potentially static and essentializing.¹ To meet these challenges and avoid the pitfalls of earlier analyses, it is important to outline the best practices of comparison.²

Jonathan Z. Smith identifies four parts, or what he calls “moments,” of the process of comparison.³ The first moment is description, which assumes that a step has already been taken in selecting the given comparanda, and which involves the definition or representation of each comparandum within its context (be it literary/textual, social, historical, cultural, or so on) that provides it with its significance, as well as a reception-history interacting with how others have described each comparandum and its significance. Our comparisons are unlikely to be entirely new (unless, say, it concerns a new intellectual property) and we are typically not going to be among the first to examine and describe a given comparandum. It is part of due diligence to interact with the insights of others who have done similar work before. This is also the moment at which the selection of items to be compared should be justified.

The second moment is comparison proper after the analyst has described two or more comparanda, wherein the analyst relates what he/she regards as significant using a given analytic framework/“third term.” This moment can also be described as generalization/abstraction in light of that “third term” that defines the respect in which the comparanda are being compared. Establishing a clear “third term” to organize the comparative analysis is critical for avoiding the weaknesses of superficial comparison and for enabling the possibility of illumination in comparison.

The third moment is redescription, wherein the analyst combines the work from these moments in redescribing comparanda in relation to each other. That is, one must articulate the combined insights of looking at the comparanda from multiple perspectives to establish a more holistic view. This is also the point at which similarities and differences observed previously can be cast in a new light.

The fourth moment is the rectification of the academic categories used for comparison, whether by offering a new framework or otherwise correcting an established framework.⁴ As noted in the previous two entries, there is always a possibility that using new means of comparison between two or more comparanda can challenge common understandings of previous comparisons or of the given comparanda. The established academic categories cannot and should not simply be taken for granted but must be examined themselves.

With these basic steps outlined, it is also important to know what kind of comparison one is pursuing. Bruce Lincoln suggests favoring “weak comparisons” as opposed to “strong comparisons.” The latter are wide-ranging comparisons that tend to make strong claims about universality or genealogical connection. The former are, “inquiries that are modest in scope, but intensive in scrutiny, treating a small number of examples in depth and detail, setting each in its full and proper context,” which thus produce conclusions that are, “more probative, reliable, and surprising.”⁵

In this same vein, one notion that has regularly appeared in comparative literature of the last few decades as illustrating best practices is the notion of “thick description.” Clifford Geertz, who borrowed the term from Gilbert Ryle, first popularized this notion in the realm of ethnography.⁶ According to Geertz, while a “thin description” consists of simply describing a thing (in the case of his analysis, behavior), a “thick description” of the same places the thing in, “a stratified hierarchy of meaningful structures in terms of which … [it is] produced, perceived, and interpreted,” thereby making the thing meaningful to its target audience.⁷ Such a description also functions as redescription, since the deeper understanding it produces of a thing enables the analyst to translate it into an analytic framework used for comparison with something from another context. Comparativists thus see thick description as a thoroughly contextual remedy to the problem observed in many earlier comparisons of simply extracting data from their contexts in order to place them in a context of the scholar’s own choosing.

The stress on thick description raises an additional consideration of the use of “emic” (from inside a system) and “etic” (from outside a system) categories in the comparative process. This terminological distinction between emic and etic was coined by Kenneth L. Pike in his Language in Relation to a Unified Theory of the Structure of Human Behavior (originally published in 1954), deriving from the linguistic concepts of phonemic and phonetic.⁸ Etic classifications are available apart from the analysis of a specific subject, but emic classifications emerge from the analysis of that subject.⁹ The etic perspective presents a perspective external to the system in question and uses criteria for analysis determined for relevance outside the system, but the emic perspective presents a perspective internal to the system in question and uses criteria for analysis determined for relevance from within the system for the purpose of functioning within that system.¹⁰ Etic perspectives tend to describe similarity and difference according to external criteria, but emic perspectives tend to describe similarity and difference by reference to responses from actors within the system.¹¹ To give an example, “to an external observer the physical act of swatting a fly does not appear much different in India or in the Western world. Yet, for the devout Hindu it would have profound religious implications, whereas for the Westerner it is an action to get rid of a nuisance and can even be viewed as a hygienic necessity.”¹² Ideally, etic perspectives should provide a means of entry into the system being analyzed, which are then refined by encounter with the emic.¹³ This last point makes especially clear that Pike sought not for preference of one over the other, but for each to complement the other.

As properly accounting for both similarities and differences—and thus giving one’s explanation better explanatory scope—is essential to good comparison, one also must attend to both emic and etic perspectives. To collapse the distinction in favor of the latter can muddle the distinction between “attempting to understand a given religion on the terms of those who believe and practice it, and attempting to explain that religion (or religion in general) in terms that reduce it entirely to something else.”¹⁴ To collapse the distinction in favor of the former can also lead to a distorted perspective on what the insider/outsider distinction means. All it really signifies is that there is an unequal distribution of knowledge between the insider and the outsider. If it were otherwise, the implication would be that the perspectives are somehow incommensurable and untranslatable, despites appearances and instinctive human practices to the contrary. But Jeppe Sinding Jensen observes:

Shared representations and discourses are social facts and they are only social facts because of their fundamentally linguistic nature. There is a base semantic level, a linguistic “substratum,” which permits the semantic comparison of them…. Holistic philosophical and linguistic reflections tell us that the multiple semantics of different “frameworks” rest upon a largely shared knowledge about the world and that although there may be marked interpretive differences among frameworks, discourses and meta-languages such differences do not entail that they are epistemically incommensurable or mutually untranslatable.¹⁵

Likewise, as Kevin Schilbrack is keen to remind us, an etic term can refer to a reality that is present in a context, even when the terminology the analyst uses to refer to it is not present.¹⁶

A truly deep comparative analysis must rigorously pursue the emic perspective in order to place ideas in their proper contexts (including linguistic, literary, historical/cultural, theological/philosophical, religious, and overall worldview), wherein they have their meaning in the world behind the text and the world of the text. However, comparative analysis must also rely for its organization and methodology on etic categories of thought that translate the emic ones for the world in front of the text.

My particular comparative approach in my dissertation used a form of worldview analysis that is textually focused. Thus, our next series will provide an orientation to worldview analysis.

1. Stausberg, “Comparison,” 29.

2. Also note Stausberg’s helpful checklist for comparative work in “Comparison,” 31.

3. Jonathan Z. Smith, “The ‘End’ of Comparison: Redescription and Rectification,” in A Magic Still Dwells: Comparative Religion in the Postmodern Age, ed. Kimberley C. Patton and Benjamin C. Ray (Berkeley: University of California Press, 2000), 239.

4. Cf. Barbara A. Holdrege, “What’s Beyond the Post? Comparative Analysis as Critical Method,” in A Magic Still Dwells: Comparative Religion in the Postmodern Age, ed. Kimberley C. Patton and Benjamin C. Ray (Berkeley: University of California Press, 2000), 79–84.

5. Bruce Lincoln, Apples and Oranges: Explorations in, on, and with Comparison (Chicago: University of Chicago Press, 2018), 11. Cf. Hughes, Comparison, 79.

6. Clifford Geertz, The Interpretation of Cultures: Selected Essays (New York: Basic, 1973), 3–30.

7. Geertz, Interpretation, 7.

8. Kenneth L. Pike, Language in Relation to a Unified Theory of the Structure of Human Behavior, 2nd ed., (The Hague; Paris: Mouton, 1967), 37. “Phonemic” concerns the sounds (or signs in the case of sign languages) that distinguish one word from another within a particular language. “Phonetic” concerns the sounds (or signs) that humans use and perceive in language in general.

9. Pike, Language, 37–38.

10. Pike, Language, 38.

11. Pike, Language, 38.

12. Till Mostowlansky and Andrea Rota, “A Matter of Perspective? Disentangling the Emic-Etic Debate in the Scientific Study of Religion\s,” MTSR 28 (2016): 323.

13. Pike, Language, 38–39.

14. Brad S. Gregory, “The Other Confessional History: On Secular Bias in the Study of Religion,” HistTh 45.4 (December 2006): 134 (emphases original).

15. Jensen, “Revisiting,” 40.

16. Kevin Schilbrack, “A Realist Social Ontology of Religion,” Religion 47 (2017): 161–78.