I have been thinking lately about identity politics. More precisely, why so many people deeply committed to nondiscrimination nevertheless are uncomfortable with identity politics.

An initial difficulty is to define identity politics. Here’s the definition in the Google dictionary: Identity politics is “a tendency for people of a particular religion, race, social background, etc., to form exclusive political alliances, moving away from traditional broad-based party politics.” (For expository convenience, I will group these identity criteria under the terms “ethnicity” and “ethnic status”). This definition focuses on political alliances. An obvious criticism of such tendency is to observe that people can pursue worthy goals more effectively by forming broad-based alliances (some have criticized Hillary Clinton’s campaign on that score). Such approach is entirely pragmatic: political goals are more likely to be achieved if more people participate in the movement.

A second definition of identity politics is the idea that ethnic groups are entitled to benefits (usually coercively obtained) by virtue of their ethnic status. At first blush, this does not seem very appealing: to say that I’m entitled to a benefit B by virtue of my ethnic status sounds, well, not nice. But, of course, identity politics understood in this way, to be persuasive, must be linked to some other narrative. One is a history of past injustice. If so, my claim that I am entitled to a benefit B by virtue of my being a Latino is a short-hand for the claim that I am entitled to B as compensation or redress for the wrongs inflicted on Latinos now or in the past. Being Latino, by itself, does not suffice to generate the claim. A common criticism of this position is that these claims are hard to sort out, and often result in benefitting or punishing undeserving people.

A very different narrative is diversity, the current dominant rationale for ethnic preferences. I, a Latino, am entitled to a benefit B because possessing that benefit will promote a public good, diversity in the workplace or the classroom, or diversity among those who will get the benefit B. A common criticism of this position is that, in the real world, it applies only to groups that have political clout. It does not apply to everyone who can contribute to diversity, such as conservatives or libertarians, for example, or even to ethnic groups out of favor. If the claim is amended to embrace all diversity, then it ceases to be identity politics (it still may be open to objections, but those are beyond my concern here).

There is another dimension of identity politics. It is the idea that who states a proposition is relevant to the truth of that proposition. This is mistaken. “As a Latino, I support affirmative action!” is a non-sequitur (or a tautology simply saying that I support affirmative action). The validity of an argument depends on the facts and sound reasoning about those facts. They do not (cannot) depend on who the speaker is. Sure, someone may be prone to identify a new argument because of who she is. As an example, feminists did a great service when they showed why the common law of rape was unjust. But that speaks to what Popper calls the context of discovery of a proposition. It doesn’t speak to the context of justification of that proposition. So, one who says, “as a Latino, I believe X, Y, and Z” is invoking an irrelevant reason in favor of the truth of X,Y, and Z. The correct statement would be: “I believe X, Y, and Z are true for reasons A, B, and C. And I was motivated to think about X, Y, and Z because I am a a Latino.”

These are preliminary thoughts. A more definitive assessment of identity politics requires further research into each of these versions of the concept and the available objections.