Intersectional Alienation

Anna Cooper - ``Liam is a fan of how
generally disappointed with everyone &
everything I seem to be in this photo.''

A long time ago me and some comrades -- the magnificent Dan Malinsky and the terrific Morgan Thompson -- got together to discuss intersectionality theory. Intersectionality theory is a set of ideas that has recently gained prominence after influential work by Kimberlé Crenshaw and the origins of which is often traced back to work by Anna Julia Cooper. It stresses the various ways in which the fact that we simultaneously occupy multiple demographic categories can complicate or undermine social theories that attempt to focus on the consequences of just one demographic category at the time. Examples of allegedly insufficiently intersectional approaches to studying social life are feminist theories of patriarchy, or Marxist theories of class oppression -- the charge of the intersectionality theorist is that something very important is missed out by neglecting the fact that we are not just workers, or women. We are, say, African-American working class women, or Japanese upper-bourgeois immigrant transmen, or.... etc. (In fact the charge is often more specifically that: what happens in non-intersectional theorising is that the most privileged sub groups within the relevant category end up having their perspective and needs taken as representative of the whole: so the slogan goes that when theorising about race and gender, we end up acting like ``all the women are white and and all the blacks are men''.)  What is more, there is usually an eye to policy or activism (often especially as it pertains to the oppression of black women) in light of our knowledge; it is a branch of theorising with a distinctly pragmatic or political bent -- the point is not just to understand the world but to change it.

Malinsky - ``Liam is very very confident
that I am not going to be happy with the fact
that Liam described me as `terrific' above.
But if I don't like it I should probably just
start my own blog or maybe stop always
being so generally terrific, so it's not like I
really have grounds upon which to complain.''

Well, my comrades and I are all broadly interested in methodology in the human sciences, and are interested in various matters relating to social or political organisation. Morgan Thompson brought to our attention the fact that for certain intersectional claims, ones that she was interested in testing in spheres of mutual interest, there was no agreed upon methodology for testing them on large data sets. What is worse, as we looked around the sociological literature, it became apparent that intersectionality theorists have often been charged with failing to offer a coherent methodological perspective, and there are even claims out there to the effect that this methodological slipperiness is effectively letting intersectionality theorists dodge falsification of some of their characteristic claims. So, we thought, here was a chance for us to advance our own research and also use our skills in a way that might be beneficial, and we produced a paper outlining means of adapting certain statistical tools so that they are appropriate for represent and test characteristic claims of intersectionality theory. The paper was eventually published earlier this year, it's called `Causally Interpreting Intersectionality Theory' (CIIT), and it's available here if anyone wants to give it a read.

Morgan Thompson - ``OK, but how would
it even help if we had our own blogs
 though? How would that stop you writing
awkward effervescent praise? That doesn't
even begin to make any sense, Liam.''

A couple of citations to CIIT have recently caught my attention and made me think about our relationship to how our work is received. The first of these is in a paper in a journal that focusses on gender relations in the sociology of work and organisations. A section of the paper is dedicated to looking at ways of `operationalizing' intersectionality theory. They set up the problem of the section we feature in as follows:

For scholars in work and organizations, this challenge can be daunting as we need to address the two thorny issues inherent in all intersectional research design. First, to translate intersectionality theory into concrete methodologies (Christensen and Jensen, 2012) and, second, to develop analyses that interrogate intersectional paradoxes insightfully while capturing the simultaneous interrelations between the subjective and the structural. In addition, as scholars in work and organizations, we also need to engage with the reality that our discipline is dominated by a functionalist epistemology and positivist methods... 

And go on to discuss a number of methods of dealing with it. Our work features in the following way:

Recently, some scholars (e.g. Martinez Dy et al., 2014; Woodhams et al., 2015a, 2015b) have drawn on a critical realist positioning to expand both the methodological understanding as well as the empirical study of intersectionality in work and organizations. Empirical works following this approach have used quantitative analyses of large data sets to measure identities as variables, determining their interrelationships and ultimate impact on different material realities (e.g. employment outcomes). They argue that quantitative methods allow scholars to test empirical hypotheses and relationships among variables, have the potential to offer definitive evidence of causal relations, and account for non-additive relationships (Bright et al., 2016). For example, Bright et al (2016) argue that interventionism and causal graphical modelling using Bayesian statistics may provide a means for testing claims based on the intersection of certain variables. The argument for positivist, quantitative approaches is bolstered by the legitimacy and authority afforded to them in what counts as rigorous and legitimate knowledge production in the field of work and organizations.

Liam Kofi Bright - ``Yo dawg
I heard you like recursion...''

So we don't quite advocate Bayesian statistics but this isn't bad. We do indeed provide arguments for (a particular set of) quantitative approaches to studying intersectionality. And I am happy to hear that an additional benefit thereof will be helping people gain legitimacy in their field when they engage in such studies. Among the coauthors of CIIT there would probably be differences in opinion as to how we would want to relate to positivism -- but we ourselves describe our project as one of explication, so it's not entirely off. Also I strongly suspect that in the relevant disciplines ``positivist'' just means ``uses statistics'' or something close to this, so this was less an acknowledgement of our shout out to Carnap and more just a way of saying we do stats. In any case, upon reading this I was largely happy that our work had been understood, and even virtues that I had not fully realised it possessed were appreciated.

Nothing so nice, alas, can be said regarding our place in the second paper of interest. Here we play the villain. The author sets up a contrast as such:

....intersectionality has been used either radically, when it acknowledges lived experiences and context to advance transformative politics against domination, or ornamentally, to accommodate other theoretical frameworks, subsequently depoliticising and limiting its transformative scope.... This distinction is akin to debates within intersectionality over its methodology: we can identify the contrast between an additive model of political inclusion (‘adding up’ identities, differences, and experiences to ‘include’ them into a schema -1), and a politics of radical change, which dismisses boundaries and mere counting.

The `-1' in that is a footnote, which leads to CIIT's citation: ``Intersectionality has even been coupled with ‘graphical causal modelling’ – see Bright et. al. (2016)''. So here, it seems, our use of intersectionality is an especially striking example of the ornamental, merely additive, depoliticised, limited, maybe some kind of boundary enforcing, but in any case `mere counting', use of intersectionality. Not so good.

Naturally, of course, I don't agree! We actually expend a great deal of effort in the paper ensuring that the picture of intersectionality one gets is not one of merely additive difference made by considering various demographic categories; we have an extended discussion (section 5 of CIIT) about the advantages our model confers in planning policy or action designed to change the world rather than just study it; we are explicit at multiple points that we are not arguing against more qualitative methods, we are not involved in boundary policing them away. Indeed, who is boundary policing who, given that the complaint seems to amount to that we are using quantitive methods (we are among the `mere counters') in a domain in which the author does not approve of them, coupled with an objection to accommodating various theoretical frameworks?

Such, at least, was my first reaction. But when I thought about it, I came to see it in a different way. The second citation is, in some sense, the pessimistic mirror image of the first. The first cites us thankfully, thinking that by showing that intersectional theorising can be done with the kind of `positivist' methods folk in their part of gender studies value we shall therefore boost the esteem of intersectionality theory, encouraging more people to work on it and take it seriously. The second cites us scornfully, apparently taking us as just an especially outlandish warning sign of the gentrification to come, wherein intersectionality will suffer a kind of death by kindness. Sure people increasingly will pay lipservice to intersectionality, but this is at the cost of losing sight of the original theoretical goals and values that underlay it. Both papers, then, predict that intersectionality theory's fortunes-in-terms-of-popularity are waxing, and both papers cite us as exemplary of (and maybe even causally relevant to) this turn of events. But the difference between them is what they think shall result from increased popularity; pessimistically, will intersectionality theory be hollowed out, become ornamental, a mere buzzword? Or optimistically, will it be strengthened, renewed, developed to new heights?

It is much closer to our intent, of course, that our paper should bolster or advance intersectional theorising, and we ourselves do try and maintain solidarity with its original spirit by offering pragmatic defences of our explication, in terms of the benefits that may be accrued to helping guide political action. We came to praise intersectionality, not to bury it. But what these two responses to CIIT really drove home to me is just how powerless we are in this regard. The content of our paper, the specifics of the arguments we gave, seem to me just not the kind of thing that will make a difference as to whether the world goes in either of those directions. If the optimistic scenario comes about, I really doubt it will be even a little bit because people were convinced by our arguments. Whereas if the pessimistic scenario comes about, I don't think anything we have done or said in the paper will prevent our work being used in a way that renders the intersectional theorising a mere ornament. Now the paper is out there, how it is received, which (if either) of these futures shall be realised, is to some very significant extent beyond our control. Perhaps to more experienced scholars this is old hat, and in any case if I reflect on the fact-value distinction, or inductive risk, or the difference between the context of discovery and the context of justification then I can probably make this salient to myself by such a priori means. But I am a junior scholar, and this was striking to me. The alienation of labour under capitalism extends even into the rarified world of academic intellectual production.

Before closing, I thought I'd just note a nice coda to all this. In that first paper, wherein we are discussed in a manner that is mostly coincident with our intent, I am listed in the bibliography as L. F. Bright. My middle name is in fact `Kofi'. The author is well and truly dead.