Me and Viliam On: Intersectionality and Insight

Below I reproduce a short exchange I had with frequent LessWrong commenter Viliam last week on this thread. 

Some thoughts on the thread in retrospect: I come from a position in stark opposition to 'intersectionality' and similar notions arising from critical theory, and would ordinarily be content to dismiss it out of hand. However, in the recent past I've had to reconsider a similar position regarding the whole body of postmodern philosophy and related studies, owing to the writings of David Chapman (see this article and this comment), which find part of their base in postmodern and critical figures, most prominently Heidegger and Foucault, and make the case that those thinkers were, in fact, very careful and profound, but that their insights were diluted both by the fashionable nonsense styles they wrote in and by their successors, who were not trained in the Western rational corpus before attempting to reform (or destroy) it. 

This by no means invalidates criticisms like the Sokal Hoax or its sequel, nor does it ignore the absurdity of contemporary critical scholarship, but it does eliminate the easy condemnation of the whole field. One must pay close and effortful attention in order to dismiss this scholarship, though it does answer to a straightforward heuristic, "was it published before or after 1985?"

Still, I could no longer be so glib with pomo or early critical theory, so I had to give intersectionality a bit of room. In attempting to steelman the ideology, I promote this passage from James C Scott, which I believe exemplifies a weaker form of the intersectional proposition that I can get behind. 

It is also worth noting that this conversation regards not any specific scholars or texts, but a general milieu which both of us understand (or think we do). 

Me

I found a passage in James C Scott's Seeing Like a State that shifted me a little closer towards agreeing with intersectionality.

I think that a "woman's eye," for lack of a better term, was essential to Jacobs's frame of reference. A good many men, to be sure, were insightful critics of high-modernist urban planning, and Jacobs refers to many of their writings. Nevertheless, it is difficult to imagine her argument being made in quite the same way by a man ... The eyes with which she sees the street are, by turns, those of shoppers running errands, mothers pushing baby carriages, children playing, friends having coffee or a bite to eat, lovers strolling, people looking from their windows, shopkeepers dealing with customers, old people sitting on park benches ... A concern with public space puts both the interior of the home and the office as factory outside her purview.  The activities that she observes so carefully, from taking a walk to window-shopping, are largely activities that do not have a single purpose or that have no conscious purpose in the narrow sense.

-Seeing Like A State (p. 138)

This passage, and the book in general, provided some concrete examples of 'different ways of knowing' that I could wrap my head around. Going forward, I will take claims to the importance of researcher identity a bit more seriously in certain fields (ie urban planning and agriculture, but not in mathematics and CS). 

This seems to point to a weaker form of intersectionality/identity-focused theories that makes a lot more sense: 'researcher identity has distinct effects on research in complex and/or culturally charged fields' vs the orthodox strong form, 'researcher identity is the primary lens through which research must be judged in all fields.' 

I experienced a similar nudge reading David Chapman, who insists on the importance of early postmodern philosophers such as Foucault. I previously dismissed both groups out of hand, now I take a moment to assess individual claims and expect that some have validity. 

Has anyone experienced similar nudges, or have particular comments on this subject?

----------------

Viliam

To me, intersectionality feels like one of those "motte and bailey" things. Yes, seeing things from other people's perspective can be very useful, and yes, it is often better to invite an X to participate in the project than trying to guess how things might seem from X's perspective. This is definitely worth paying attention to! 

But it also comes with the political baggage, the official list of "groups that matter" which suggests that it is useful to consider people's gender or sexual orientation, but not that useful to consider e.g. people's social class or degree of autism. Unless you discuss the latter from the perspective of the former. (You don't discuss poverty, unless you discuss e.g. poverty of women, or poverty of LGBT; and the correct solution is always a support of given group, not a support of poor people in general. The other causes exist only to serve the important causes.) 

It also considers people within the groups replaceable. If Alice and Betty have the same gender, same race, same sexual orientation, then Alice is assumed to also speak in Betty's name, especially if Alice is politically on "the right side of history". (On the other hand, if Betty is conservative, she is not even allowed to speak for herself. Alice knows better what is best for her. You should listen to Alice and ignore Betty, otherwise you are "against women".) There is a somewhat related concept of "token" person, which kinda admits that having one representative of a group is not enough, but the main concern is about choosing Betty as that representative. 

Plus the usual "critical" bias, where one person being unhappy about something is more important than ten people being happy and opposing the change. (Because the true reason for the fight is promoting social change towards "the right side of history", not solving specific problems of specific people. Individuals only exist to serve the cause.) 

So I am in favor of listening to different perspectives, but if you do it with an open mind, chance is that you may ask ten X's and receive ten different answers. The answers will probably have something in common, so hopefully you can work with that, but can be frustratingly different at the parts where you hoped to get one unified "X perspective" as an answer. (You can't make everyone happy, and you can't even make all X happy.) Therefore, you can get better data by asking people different from you for a perspective, but at the end of the day, you have to apply your own judgment to the somewhat contradictory answers you have received.

----------------

Me

Good points. Perhaps 'intersectionality' isn't the right term. I also considered 'positionality,' trying to refer to ' ideology that emphasizes identity over reasoning.' Or maybe I'm thinking of the 'motte' form, so that [whatever the Scott quote represents] is a weaker form of motte!intersectionality is a weaker form of bailey!intersectionality. 

Though I think the Scott quote represents something stronger than 'paying attention to identity X's perspective'. It looks more like 'identity X may provide information and insights in unpredictable ways.' 

This is not compatible with reflexively applying a narrative to an identity group, as so often happens. If identity X's insights line up perfectly with your preexisting beliefs, there's something else going on. 

Perhaps more specifically, I newly endorse the proposal, "Identity has distinct and unpredictable effects on research," but not the more extreme proposals: 

"Identity group members are replaceable." 

"Identity groups have a 'correct' position." 

"Problems must be examined first in relation to identity groups." 

I'm personally coming into this with a heavy bias against intersectionality and critical theory, so I'm trying to steelman where possible.

----------------

Viliam

Considering that anyone can provide an unpredictable insight, and you can't invite everyone to the debate, so you need to use some heuristic to get maximum insight per number of people invited... the social justice heuristic (focusing on gender, race, sexual orientation) is actually quite good. 

It can be further improved by also considering social class and religion/politics.

----------------

Me

I hadn't considered that angle. Still, that heuristic assumes 

a) that the field is one where those differences are salient (I maintain mathematics at least is exempt) and 

b) that the people you're inviting have sufficient background to make meaningful contributions, contra the orthodox intersectional considerations you mentioned before. 

I'm tempted say that this heuristic (diversity of identity) is strictly less effective than diversity of thought/ideology, but that seems to be what Scott runs against. It would indicate that there are insights not available just through ideology but through (to use an abused phrase) lived experience. 

As to how these cross over and whether they're intersectional, that's another can of worms I'm not going to open.