The Concept of "Religion"

In my philosophy of religion class yesterday I gave everyone in the class a chance to give their own concise answer to the following question: “What is religion?” (To be more precise, I asked them to imagine they were being interrogated by space aliens, and that the fate of the Earth depended on their answer).


Not surprisingly, there were many diverse responses—some emphasizing social and institutional phenomena, some emphasizing beliefs or ways of looking at the world, some emphasizing practices or ways of life, and some stressing inner spiritual experience. Some definitions were, I’d say, quite gilded—that is, they used language aimed at highlighting the beauty or value of the thing being defined. Other definitions were quite the opposite. For example, one student defined religion as a system for justifying the exclusion or marginalization of people from a community.

Once I had the chalkboard covered with these various accounts, I pointed out how this diversity is also represented among scholars—with understandings of religion ranging from the more private, personal, “feeling”-oriented understanding (favored by theologian Friedrich Schleiermacher and philosopher/psychologist William James), to more sociological understandings (promulgated by, for example, Emile Durkheim).

I then spent a few minutes considering the idea I advanced in my book—namely, that “religion” is what Wittgenstein calls a “family resemblance” term (see p. 15 of Is God a Delusion? for an account of this idea). Then, in the last few minutes of the class, I turned to another approach—one that, based on some further reflection I’ve done since writing my book, I’m becoming increasingly convinced is the right one. According to this approach, “religion” is best understood as what philosopher W. B. Gallie called an “essentially contested concept”—but with a twist.

Since I didn’t have time to fully explain this idea in class, I want to do so in this post. In fact, I’ve already done so on this blog—here and here. But since it’s always helpful to try to explain ideas in different ways, let me have another go at it here.

What Gallie noticed was that there are some terms whose proper use, rather than being determined by an established definition (one that sets out the necessary and sufficient conditions for something to fall within the term’s scope), is instead determined by a shared set of complex exemplars or paradigms along with a shared appraisive meaning. So for example, there isn’t a common definition of “rape.” Instead, there are a bunch of exemplars—sexual acts that we all agree count as rape—together with general agreement that when an act is labeled “rape” there’s a strongly negative appraisal that goes along with that.

Here’s the thing about “rape.” It just isn’t and never will be a neutral, purely descriptive term. To call something rape is (among other things) to condemn it in a particular way. That condemnation is part of the meaning of the term. And so it matters a lot whether or not a particular act qualifies as rape. Acts of rape are morally worse than other classes of sexual acts (such as seduction, say, or aggressive lovemaking, or adultery).

The paradigms of “rape” exist because there are a bunch of things that we all agree deserve to be condemned in this distinctive way. But these paradigms are complex. They have lots of different features. And we don't all agree on what it is about these paradigms that makes them deserving of the negative appraisal. And this means that there are controversial cases.

Consider: A guy keeps pressuring his high school girlfriend to have sex. She doesn’t want to. He threatens to break up with her. She closes in on herself. He backs off for a few minutes, then begins groping her again. She doesn’t resist. He undresses her. She remains totally passive and unresponsive. He puts on a condom and penetrates her.

Is it rape? More people would be inclined to say “yes” today than twenty years ago—but there are still many who’d say it isn’t, that the guy is being insensitive but isn’t a rapist.

The reason for the dispute is that there isn’t agreement about whether the boys behavior in this case deserves the negative implications of the “rape” label. In other words, this is a moral dispute about what warrants a certain kind of negative appraisal.

And moral disputes can’t be resolved through definitional fiat. Suppose someone says, “From now on, rape will mean an act in which someone uses physical force to overcome a woman who is actively resisting sexual penetration. As such, the case at hand isn’t rape.” Such a move isn’t going to just be accepted. Why? Because to call something “rape” is to say that there's a certain kind of “badness” to it—more precisely, the same kind of badness that the agreed paradigms of rape possess. And so, to define rape as “an act in which someone uses physical force to overcome a woman who is actively resisting sexual penetration” is to say, in effect, that only acts which meet these conditions are bad in the relevant way. Put another way, to define “rape” is to take a stand in a moral dispute.

And as long as there is moral dispute, to impose a uniform definition of “rape” on a community of speakers is to impose one disputed answer to a moral question on everyone in the community. This wouldn’t be merely an act of establishing a linguistic convention. It would be an act of using language to truncate debate and to effectively delegitimize certain moral views.

And this is why some concepts become essentially contested. Their being essentially contested is a good thing—a way to keep some voices in a moral debate from being illegitimately silenced through definitional fiat.

My claim is that this idea of essential contestability is useful for understanding religion—but not if we accept Gallie’s idea without modification. Religion, I think, is an essentially contested concept with a twist. And what’s the twist? Here’s how I explain it in a forthcoming article (“Moving the Goal Posts?” to be published in Philo: A Journal of Philosophy):

But unlike “art,” whose appraisive meaning is positive, or “terrorism,” whose appraisive meaning is negative, “religion” has come to be used such that there are two competing communities of discourse, each using the term in an essentially contested way. But whereas one community of discourse treats “religion” as a positive appraisive concept and seeks to gauge which features of the paradigms warrant the positive appraisal, the other treats it as a negative one and seeks to judge which features warrant the negative appraisal. When a concept comes to be used in this way, we might call it a “bifurcated essentially contested concept.”

Unlike essentially contested concepts as Gallie understood them, I’m not at all convinced that bifurcated essentially contested concepts serve a useful function. When an essentially contested concept becomes “bifurcated,” what happens? On the one hand, you have those who attach a positive appraisive meaning to the paradigms of religion. They will be formulating their definition of religion by looking for what it is about the paradigms of religion that justifies the positive appraisal (and so will sift out of their understanding of religion anything in the paradigms that warrants a negative appraisal). On the other hand, those who attach a negative appraisive meaning to "religion" will be doing to opposite. The result may be that you have two parties with virtually identical value systems, who therefore make the same appraisive judgments about the various features of the religious paradigms—but who appear to be utterly at odds. An analogy—again from my forthcoming article—can be helpful:

It’s as if one community of discourse attaches to the term “sex” the appraisive meaning that typically attaches to “rape,” while another attaches to it the appraisive sense of “making love.” The former group looks at the range of phenomena that go by the label “sex” (ignoring, of course, those phenomena which no one would ever call rape) and tries to identify what justifies the negative appraisal. The latter does the same (ignoring the phenomena, such as rape paradigms, which no one would ever call “making love”), in the attempt to identify the parameters within which the positive appraisal is warranted. The latter holds up its results, saying, “This is the kind of sex (by which we mean making love) that deserves label!” The former protests, “That’s not sex (by which we mean rape) at all!”

This, I think, is what’s going on in the conversation between Christopher Hitchens and Unitarian Universalist minister Marilyn Sewell, whose unusual debate inspired one of my recent Religion Dispatches articles. It may also help to explain some of the common charges leveled against my book—charges to the effect that I respond to the new atheists by coming up with this definition of religion that has nothing to do with real religion as it exists in the real world.

Of course, what I defend in my book has a great deal to do with actual religions—but when I look at those real-world phenomena, I’m trying to identify the features which might justify a positive appraisal (what I call the germ of a true religion that might be salvaged from the crud of “superstition” and “fundamentalism” and “religionism”). My critics, meanwhile, are sifting through the same phenomena in an attempt to identify what makes religion so bad. And what do they pinpoint? What, from my standpoint, is the crud from which true religion needs to be salvaged. And so they’re holding up the crud and calling it religion, while I’m holding up the gem that was buried in the crud. And they protest, “That’s not religion at all!”