When we try to think seriously about what should be done, we have a tendency or temptation to think in generic terms–about categories rather than cases.
- In social science, quantitative research evidently requires categorization; it is the search for relationships among classes of things.
- In applied philosophy/ethics, most of the discussion is about categories that can be defined by necessary and sufficient conditions, e.g., abortion, war, marriage. Thinking about categories allows what Jonathan Dancy calls “switching arguments.” For instance, you decide what is good about heterosexual marriages, and if the same reasons apply to gay marriages, you should favor them as well. By thinking categorically, you can switch from one case to another.
- In policy analysis, lots of research is about generic policies: vouchers, foreign aid payments, prison sentences. I should, however, note the important exception that some scholars study major individual policies, such as the decision to invade Iraq or the No Child Left Behind Act.
- In ideological politics, the underlying values are strong general principles, e.g., “markets are good” or “there should be more equality.” Categories of policies are then used as wedges for advancing an ideology. For example, libertarians promote school choice in order to demonstrate that markets work better (in general) than governments.
I have a gut-level preference for particularism: the idea that, in each situation, general categories are “marinaded with others to give some holistic moral gestalt” (Simon Blackburn’s phrase). That implies that applying general categories will distort one’s judgment, which should rather be based on close attention to the case as a whole.
I will back off claims that I made early in my career that we should all be thorough-going particularists, concerned mainly with individual cases and reluctant to generalize at all. My view nowadays is that there are almost always several valid levels of analysis. You can think about choice in general, about choice in schooling, about charters as a form of choice, or about whether an individual school should become a charter. All are reasonable topics. But the links among them are complex and often loose. For instance, your views about “choice” (in general) may have very limited relevance to the question of whether your neighborhood school should become a charter. Maybe the key issue there is how best to retain a fine incumbent principal. Would she leave if the school turned into a charter? That might be a more important question than whether “choice” is good.
The tendency to generalize is enhanced by certain organizational imperatives. For instance, if you work for a national political party, you need to have generic policy ideas that reinforce even more generic ideological ideas. The situation is different if you are active in a PTA. Likewise, if you are paid to do professional policy research, you are likely to have more impact if your findings can generalize–even if your theory explains only a small proportion of the variance in the world–than if you concentrate on some idiosyncratic case. On the other hand, if you are paid to write nonfictional narratives (for instance, as a historian or reporter), you can focus on a particular case.
I’m inclined to think that we devote too much attention (research money, training efforts, press coverage) to generic thinking, and not enough to particular reasoning about complex situations and institutions in their immediate contexts. There is a populist undercurrent to my complaint, since generic reasoning seems to come with expertise and power, whereas lay citizens tend to think about concrete situations. But that’s not always true. Martha Nussbaum once noted that folk morality is composed of general rules, which academic philosophers love to complicate. Some humanists and ethnographers are experts who think in concrete, particularistic terms. Nevertheless, I think we should do more to celebrate, support, and enhance laypeople’s reasoning about particular situations as a counterweight to experts’ thinking about generic issues.