The critical importance of Gary Goertz and James Mahoney's well-written depiction of two specific research cultures within quantitative and qualitative social science becomes clearest against the backdrop of the book's central interlocutor: Gary King, Robert Keohane, and Sidney Verba's Designing Social Inquiry (1994). Colloquially known by political science graduate students the world over as “KKV,” this highly influential methodological treatise contains at its core two basic arguments: first, that causal inference is the highest, most noble aim of the social sciences and, second, that valid causal inference can be established only through a single, unified logic of inquiry that takes its starting premises and evaluative standards from additive, linear statistics.
Goertz and Mahoney endorse KKV's first argument but make it the central task of their book to decisively dispute its second. They do so by demonstrating that alongside the dominant additive, linear statistical approach characterized by a search for the average effects of independent variables (effects-of-causes), there exists a vibrant and irreducibly distinct qualitative tradition of establishing causal inference through a set-theoretic logic marked by a search for the necessary and sufficient causes of a dependent variable of interest (causes-of-effects). For Goertz and Mahoney, “Good science is concerned with both kinds of questions” (p. 41), and, consequently, “there is no set of principles that unifies all social scientific work” (p. 220). This is a clarion counter to KKV's performative utterance (John Austin, How to Do Things with Words, 1962) about a single, unified logic of inquiry, and it would be a laudable leap into better if not “the best of times” in the discipline if every graduate methods class featuring their treatise were henceforth to also assign A Tale of Two Cultures as a persuasive rejoinder.
While catchy, however, the book's Dickensian title risks misleading by overstating its scope. A more accurate, though fatally awkward, title might read: A Tale of Two Subcultures: Set-Theoretic and Additive, Linear Approaches to Causal Inference in the Social Sciences. Goertz and Mahoney are certainly aware of the existence of other, important subcultures in both the qualitative and quantitative social sciences (p. 4), but generic references to “qualitative and quantitative research” in the book's subtitle and throughout the text create the danger of an unjustified conflation of “set-theoretic” with “qualitative” and “additive, linear” with “quantitative” to the detriment of distinct alternatives such as interpretive and Bayesian approaches.
Following a brief “mathematical prelude,” which serves as a useful refresher on set theory and basic logic, the bulk of Goertz and Mahoney's 17 short chapters demonstrates the consequences that follow from their critical distinction between an additive, linear effects-of-causes quantitative approach and a set-theoretic, causes-of-effects qualitative approach. They show in clear prose how these consequences impact nearly every domain of a typical research process oriented toward establishing causal inference, including research design, case selection, concept development, measurement, hypothesis testing, and generalization. The book's final chapter offers a series of helpful tables summarizing these domains.
Throughout the text, concrete examples helpfully illuminate the implications of working within each research tradition. Most dramatic among them is the stark contrast in the answers that each tradition gave to the question concerning how many votes were lost for George W. Bush in Florida as a result of an early media call of a victory for Al Gore in the 2000 U.S. presidential election. Employing an additive, linear statistical approach, John Lott (“Gore Might Lose a Second Round: Media Suppressed the Bush Vote,” Philadelphia Inquirer, 14 November 2000) found that at least 10,000 votes were lost for Bush. In contrast, utilizing a set-theoretic qualitative approach, Henry Brady (“Data-Set Observations Versus Causal-Process Observations: The 2000 U.S. Presidential Election,” in Henry Brady and David Collier, eds., Rethinking Social Inquiry: Diverse Tools, Shared Standards, 2010) argued that no more than 224 votes could possibly have been lost. Ten thousand versus 224: consequential differences indeed!
Although Goertz and Mahoney largely maintain a discursive and illustrative tone, they also occasionally formulate general rules (p. 210), principles (p. 165), and, presumably for matters they feel especially insistent about, fundamental principles (p. 153). These formalized sprinklings are not accidental: A large part of their contribution lies in excavating and verbalizing research practices that have remained implicit, undertheorized, and walked rather than talked in the set-theoretic qualitative tradition. Indeed, this work of explicating the implicit creates moments in the text when the authors encourage their fellow set-theoretic qualitative practitioners to abandon the dominant language deployed by their additive, linear quantitative counterparts altogether and to adopt instead a vocabulary more in keeping with their tradition's own values (see, for example, the discussion of variable-indicator versus concept-data, p. 140).
Such moments are what enable a generative reading of this project as a kind of interpretive anthropology in which the authors, positioned as native informants who self-identify with the qualitative side of the two traditions, seek to describe, interpret, and codify the differences between additive, linear quantitative and set-theoretic qualitative research by examining both existing practices, as well as the metapragmatic speech produced by each tradition about its own practices. Indeed, Goertz and Mahoney explicitly acknowledge the interpretive character of their enterprise, stating that “our two cultures argument is, broadly speaking, an exercise is description and interpretation” (p. 5, n. 2).
Curiously, however, the very interpretive approaches utilized by the authors are intentionally excluded from the scope of the book itself: “[I]nterpretive approaches are not featured in our two cultures argument…. Such a book would bring to light fundamental clashes over epistemology and ontology that exist within parts of the social sciences. In this book, however, we focus on scholars who agree on many basic issues of epistemology and ontology, including the centrality of causal analysis for understanding the social world” (pp. 4–5).
I can sympathize with Goertz and Mahoney's self-aware decision to exclude interpretive approaches; scope conditions, after all, are an essential part of the very qualitative tradition they describe (p. 210), and the exclusion of interpretive approaches undoubtedly makes for a more parsimonious book unmarred by the messy “fundamental clashes” to which they allude. However, by sidestepping these clashes in favor of neatness, the authors forgo a valuable opportunity to make an otherwise superb book even more significant. Just as contrasting set-theoretic qualitative and additive, linear quantitative traditions allow the unstated assumptions and implicit practices of each to materialize, so too would the inclusion of interpretive approaches have elucidated some of the core taken-for-granted assumptions that underlie both set-theoretic and additive, linear approaches.
Instead, by framing their overall project as pertaining to “qualitative and quantitative research in the social sciences” while nonetheless excluding interpretive traditions entirely, the authors run the risk that their otherwise important argument against a single, hegemonic logic of inquiry within the social sciences might, despite itself, reproduce its own “worst of times” version of hegemony. (Indeed, it is worth noting that the value and understanding of causality within the interpretive tradition remains actively contested. For two recent discussions, see Peregrine Schwartz-Shea and Dvora Yanow, Interpretive Research Design: Concepts and Processes [2012], especially pp. 49–54, and Patrick Thaddeus Jackson, The Conduct of Inquiry in International Relations: Philosophy of Science and Its Implications for the Study of World Politics [2011]).
The pluralistic multiculturalism advocated by Goertz and Mahoney is thus an uneasy one. Not only does it leave unresolved fundamental tensions between the two traditions they do describe (the distinction between 10,000 votes and 224 votes cannot, at day's end, be settled by politically correct pleas that we all just get along), it also leaves unmapped the key portions of the terrain that constitute contemporary social science. The shifting and contested borderlands of this uneasy pluralism are exactly where scholars seeking to expand on the authors' impressive book should take up their own cartographic instruments in order to carry this important effort forward.
