In 1937, the German philosopher Max Horkheimer, recently relocated from Germany to the United States, published an essay titled “Traditional and Critical Theory”. As much as any event in 20th Century history, this marked the end of the expansion of academia’s embrace of the Enlightenment program of rationality, objectivity, tolerance, and individuality, and a return by our supposed scholars to superstition, fable, hatred, and group stereotyping. It does not seem, from reading the text, that he wished it to be so. But as I will argue here, these are, unfortunately, the ultimate fruits of his labor, and the labor of his colleagues at the Frankfurt School.
Edit: this story used to also cover (academic, legal) Critical Race Theory, which has given us some really important insights but which, I argued, has been compromised by rather than improved by its association with critical theory. However, the article was too long and even so failed to cover Critical Legal Studies which is the essential link between Critical Theory and CRT. Most comments were also about the critical theory portion. So I have revised the content to better focus on the central theme of the article: Critical Theory itself.
The rest of this article justifies the above statements, explaining in brief the key flaws of Critical Theory, and suggesting ways to keep the insights while dropping cognitively pernicious baggage.
Part One: the Flaws of Critical Theory
“If, however, the theoretician and his specific object are seen as forming a dynamic unity with the oppressed class, so that his expression of societal contradictions is not merely an expression of the concrete historical situation but also a force within it to stimulate change, then his real function emerges.” (Horkheimer, Traditional and Critical Theory (p. 215 in link at top of article).)
Hidden within this deceptively admirable-seeming statement lies one of Critical Theory’s biggest practical flaws. (Note: in context, the “specific object” is “the proletariat”, which in this context constitutes “the oppressed class”.)
Can you see the flaw, enormous but hidden? If you can’t, you’re in good company: Horkheimer was no intellectual weakling, and neither he nor his highly intelligent compatriots saw it, at least at first.
Within philosophy, the topic of epistemology covers how we come to know things — not just feel things, believe things, but know things. This both includes the idea of the distinction between knowledge and some other sort of belief (e.g. because knowledge exhibits a manner of cognitive success which belief does not), and the methods by which we can come to know rather than merely believe. (This is in contrast to ontology, which covers what really is — never mind how we manage to figure out what really is.)
The most successful epistemology by far, bar none, is the scientific method. My favorite expression of this is that given by Carl Sagan in The Varieties of Scientific Experience (Penguin Press, 2006; p. 217): “In a complex universe, in a society undergoing unprecedented change, how can we find the truth if we are not willing to question everything and to give a fair hearing to everything?” Part of the willingness to question includes the willingness to accept doubt. As Richard Feynman said, “We have found it of paramount importance that in order to progress we must recognize the ignorance and leave room for doubt.” Or, if you’re more hard-nosed about it, you could read Karl Popper on falsification and the scientific method.
Of course, to doubt you need to leave yourself adequate psychological safety to be able to doubt. What if you can’t? Carl Sagan again (p. 216–217): “We kill each other, or threaten to kill each other, in part, I think, because we are afraid we might not ourselves know the truth, that someone else with a different doctrine might have a closer approximation to the truth. […] If I can’t convince you, I must kill you. […] You are a threat to my version of the truth, especially the truth about who I am and what my nature is. […] The thought that I might have dedicated my life to a lie […] is a very painful realization.” (Sagan clarifies that he is personifying the attitude so as not to sound like he’s accusing his audience, but this “I” is a hypothetical “I” only (maybe the science-averse Anti-Sagan?).)
Can you see, now, the diabolical flaw in Horkheimer’s vision of a Critical Theorist?
The social practice of science is intensely devoted to opening up enough space between the correctness of ideas from the value of a scientist so that the scientist can bring themselves to deeply question their own ideas rather than viewing everything as a threat. Critical Theory does the exact opposite, encouraging people to deeply enmesh their self-worth with their belief in their findings.
If you want to induce cognitive bias in people, threatening something fundamental to their being is a spectacularly effective strategy. Some particularly stark examples are given by Jonathan Haidt in The Righteous Mind (Vintage Books, 2012; pp. 61–63). Upon subconsciously provoking a moral judgment about innocuous actions, he observes that the test subjects “made up absurd reasons to justify judgments that they had made on the basis of gut feelings — feelings Thalia had implanted with hypnosis”. He then goes on to call out himself for doing exactly this while in the middle of doing an experiment showing that this thing happens.
On February 3, 2007, shortly before lunch, I discovered that I was a chronic liar. … My wife, Jayne…asked me not to leave dirty dishes on the counter where she prepared our baby’s food. Her request was polite, but its tone added a postscript: “As I have asked you a hundred times before.” My mouth started moving before hers had stopped. […] something about the baby having woken up at the same time as the elderly dog barked to ask for a walk and I’m sorry but I just put my breakfast dishes down wherever I could. [… (later) …] [I was] writing about how people automatically fabricate justifications of their gut feelings, when suddenly I realized I had just done the same thing with my wife. […] It’s true that I’d eaten breakfast, given Max his first bottle, and let Andy out for his first walk, but these events had all happened at separate times. Only when my wife criticized me did I merge them into a composite image of a harried father with too few hands, and I created this fabrication by the time she had completed her one-sentence criticism. … I then lied so quickly and convincingly that my wife and I both believed me.
Here we have a scholar, a scientist, rigorously trained, devising careful experiments, trying to remember to keep doubt in mind, still fabricating subconsciously all kinds of fanciful and counterfactual explanations in response to a mild implicit criticism, and doing so so deftly that it almost escaped his notice!
What chance, then, does the Critical Theorist stand whose loyalty is not first to truth, not first to understanding the complexities of the universe and our society within it as it undergoes unprecedented change, but to dynamic unity with the oppressed. Without this, the Critical Theorist loses their real function.
If the alarm bells are not now going off full-blast in your head, they should be. Could you really, truly, embed yourself so passionately into an area that you believe to be a critical part of oppression, bind up your entire identity with solving it and unifying with those oppressed, and yet manage to accurately perceive reality? Or might you exaggerate every threat, fabricate every excuse to conform to whatever prejudices and biases might be expressed by the oppressed group you’ve decided to conform to?
Now, we should be tentative in our conclusions — maybe in practice it’s actually not a problem. Maybe critical theory is just a different angle on Sagan’s exhortation to question everything, and it all works out fine.
Maybe you won’t excuse your own failures in advance. (Horkheimer, p. 220: “Knowledge in [the] traditional sense, including every type of experience, is preserved in critical theory and in practice. But in regard to the essential kind of change at which the critical theory aims, there can be no concrete perception of it until it actually comes about.” Note — later practitioners of critical theory who embraced postmodern ideas then dropped the pretense of preserving knowledge in the traditional sense.)
Maybe you will embrace testability, with clear ways to determine whether the project is succeeding or failing. (Horkheimer, p. 242: “There are no general criteria for judging the critical theory as a whole, for it is always based on the recurrence of events and thus on a self-reproducing totality. Nor is there a social class by whose acceptance of the theory one could be guided. It is possible for the consciousness of every social stratum today to be limited and corrupted by ideology.”)
Or maybe Critical Theory, for all its good intentions, is an absolute disaster of pragmatic epistemology.
And this is just early Horkheimer, who was still moderately enamored of the scientific method and of rational thought!
Habermas (in Knowledge and Human Interests): “The methodological framework that determines the meaning of the validity of critical propositions of this category is established by the concept of self-reflection. The latter releases the subject from dependence on hypostatized powers. Self-reflection is determined by an emancipatory cognitive interest.”
So, you decide you want to be free (emancipated cognitively, at least, or maybe you’re part of or thinking about the oppressed class). You self-reflect, which means you don’t have to listen to any existing wisdom or pay any attention to inconvenient realities (dismissed as “hypostatized powers” — which means models of reality being confused with actual reality). And you end up not rationalizing and fabricating everything because…??
And then there’s the small detail that basically everything that critical theorists predicted was impossible without critical theory (but was good) ended up either coming true or making huge leaps forward using standard Enlightenment-era approaches and proper scientific methodology, save for the radical societal transformation through revolution. Marcuse apparently longed for radical societal transformation so much that he literally invented “radical subjectivity” as a most desirable state to induce in people, where people would find the existing society intolerable and would revolt against it.
Our hypothesis, then, shouldn’t be that critical theory generates rubbish that is transparently so. Rather, we should suspect that it will be used with rationalizations and fabrications to support an unreasonably extreme and revolutionary perspective that isn’t truly grounded in reality, but at most inspired by it. Transparent rubbish is easy to handle: you see that it is rubbish, and you throw it in the bin. But strongly-motivated rationalization is dangerous. Haidt didn’t just fool himself; he fooled his wife, too.
Critical theory is not the kind of thing you should let near anything important. Used as a spice or garnish, to remind you to look at the big picture, ask if things are fair, ask if you are making assumptions that are wrong but convenient for the powers that be, it perhaps still has some merit. But there are far better alternatives (as we’ll see later), so the safer bet is to discard it entirely.
A brief example: critical theory thinking in action in Critical Race Theory
In Critical Race Theory: An Introduction (3rd ed; Delgado and Stefancic; NYU Press, 2017), the authors express what they view as key components of CRT, while acknowledging that critical race theorists are a diverse group with many different perspectives.
Delgado and Stefancic appear very proud of the emphasis on narrative (pp. 45–46): “Although some writers criticize CRT for excessive negativity and failure to develop a positive program, legal storytelling and narrative analysis are clear-cut advances that the movement can claim.”
They then proceed to give an example of why it’s a complete disaster epistemologically and socially. They contrast two narratives: the narrative of the U.S. as the land of rights and freedom (told by whites), and the narrative of the U.S. as the land of gory brutality and injustice (told by blacks, Latinos, and Middle-Easterners). Page 48:
How can there be such divergent stories? Why do they not reconcile? To the first question, critical race theory answers, “experience”. People of different races have radically different experiences as they go through life. […] To the second, it answers that empathy is in short supply. […] Literary and narrative theory holds that we each occupy a normative universe or “nomos” (or perhaps many of them), from which we are not easily dislodged.
Um…and…recognizing these problems, do we want to try to build anything important on top of it? Of course a story can help humanize an otherwise dry description of people, but from this, the take-home message should be of the danger of narrative, both established and alternative. The take-home message shouldn’t be “they didn’t reconcile” but real things actually happened, and because real things can be objectively determined, we can gain agreement despite initially different perspectives if we focus on objective reality rather than narration. Narration was the problem.
Nonetheless, Delgado and Stefancic revel in the effectiveness of narrative to sway the legal process. This is only slightly less sensible than the idea that taking the One Ring and using it against the Enemy only to preserve Gondor and the Free Peoples is a good idea or will even work. How could this have come out so backwards? (If you guess “maybe Critical Theory?”, I would be inclined to agree.)
Part Two: Beyond Critical Theory
Humans have a superpower. We can fly. We can swim to the deepest ocean. We can unbridle the fury of the sun. We have telepathy. We are practically immune to pathogens.
Humans have a superpower. We can think.
We can fly because we understand fluid mechanics and combustion and heat transfer. We can swim to the deepest ocean because we understand metallurgy and pressure and the metabolic requirement for oxygen. We can unbridle the fury of the sun because we understand the difference in rest mass between nuclei of different elements and how nuclear decay can be accelerated by neutron capture. We have telepathy because we understand the electromagnetic spectrum and the semiconducting properties of silicon and how to encode voice with mathematics and mathematics with logic and logic with bits of silicon. We are practically immune to pathogens because we understood optics and genetics and cellular biology and identified the mechanisms of disease. (We don’t feel like we’re almost immune, especially in the era of Covid, because we are so used to the status quo. Compared to history, however, childhood mortality especially is vanishingly low.)
We understand so much because we have discerned that — as it appears superficially — there is an objective reality. We are not all lost inside our own minds; with a few exceptions (mental illness), we agree that a woman is a woman and a hat is a hat, and they have very different properties. Nobody argues “Well, in some cultures, hats and women are treated as interchangeable; when a baby loses its mother, it may be adopted by a breastfeeding woman, or it may be adopted by a hat.” Not only does no such culture exist, it’s difficult to conceive that such a culture could exist. It’s patently absurd.
Because objectivity really is a universal experience of how reality works and everyone has been able to find it in part, peoples throughout the world found bits and pieces of the scientific method: in Egypt, in Greece, and China, and during the Islamic Golden Age. The objective perceptiveness of other cultures back deep into history is evidenced by, for instance, structures built to align with celestial phenomena.
But it was not until the Enlightenment — though his methodology was not ultimately adopted, Francis Bacon played a larger role than anyone (Descartes deserves honorable mention), and the Royal Society played a larger role than any other group — that we really started to hit on how to use our powers of observation and thought to extend our objective reach from things like the distinction between women and hats to the entire cosmos.
Now we have the almost unfathomable privilege of complaining that we have to sit, kind of cramped, for half a day to be transported with virtual certainty to the opposite side of the world, whereas our ancient ancestors who even knew there was an opposite side of the world could not have reached it at all, or if they had, would have been celebrated throughout history forever after.
The reason why objectivity is so important is twofold. First, any time we’re confused, we can just go back and ask reality, “Hey, um, did we get this right?” And, if we manage to ask carefully enough, reality will go, “Yeah, okay, not bad,” or maybe “What garbage is this even?!?!?” But secondly, and just as important, it means we can work together without fighting over who has dominance. Everyone who accepts objectivity can play the same game: learn what is known and add to it. Add, and add, and add, standing on the shoulders of giants until Zeus himself could not approach even a minuscule fraction of the destructive power wielded (but hopefully never again applied) of, for example, the President of the United States.
This is our superpower.
Any time we can make a problem look like one we can solve with this superpower, we are in so much better shape than when we’re not.
Critical theories have a propensity to throw this away — just discard it casually on the side of the road without a second thought. CRT: An Introduction is wonderfully thoughtful, reflective, evidential, and reasoned in places — admirably so. This makes the book better. It gives the conclusions a better chance of capturing important parts of what is objectively the case as opposed to being a collection of scholars’ unconstrained opinions. These are fruits of the tree of Enlightenment, and yet they’re poised to cut down the tree, or simply knock it over by accident. Despite all its power, it’s remarkably fragile. Science is at once deeply intuitive — young children are constantly improving their models of reality by testing them — and deeply unintuitive: you mean I have to doubt my own ideas? Find the most damaging ways to attack them?! Constantly and forever?!?! What madness is this!
Narration is so much easier.
And yet there is a reason we don’t have a narrative approach to microprocessor design: it doesn’t work.
The problem, of course, is that the scientific method is incredibly difficult to apply to social situations. But all hope is not lost — just because it’s difficult doesn’t mean that we can’t strive for as much objectivity as we can get in every area we can. Even small gains bring benefits as long as we remember to doubt. We’re not sure of something because we did a well-meaning but necessarily somewhat inconclusive scientifically-minded study. But we have a better chance of being more right.
In most of science, progress is made through reductionism. You take a system apart, figure out how the pieces work. Too hard? Take the piece apart, figure out how that works. Too hard? Okay, take that apart. Repeat until you get something you can understand; then start reassembling.
Systems biology takes a different approach than the standard reductionistic one. It says: let’s put the pieces together, even the top level ones, even if we don’t totally understand how it works. Let’s model the whole system. Yes, we won’t have all the details right, we must retain even more doubt about the conclusions, or the conclusions may be less specific than in reductionistic affairs. But because we know a lot about the pieces, we can learn about the whole by studying it.
Critical theory does this too — it advocates looking at the whole system, at the assumptions and models that go into it. But unlike critical theory, systems biology and other systems-level approaches to science are at least in theory supposed to be full sciences: impartial, falsifiable, and with psychological safety around concluding, “you know, actually, we got that part really wrong”.
Social psychology takes a slightly different tack by — when done carefully, which admittedly is really hard, and it’s not hard to find egregious messups — accepting the embedding within society and trying to find sufficiently independent pieces of how society works that you can take a quasi-reductive approach to study, that then reveals insights that are relevant to society. If social psychology missteps at times, at least it (so far, mostly) doesn’t do so in the epistemologically bankrupt way that Critical Theory does.
We have the tools we need to have a fighting chance to understand what is happening in society and how to use it. There’s nothing wrong with wanting a better society! Practically everyone wants that, at least nominally. By applying the best tools we have as humans, we’re less blind while working towards our wants.
So what we need is a reinvisioning of the study of society at the highest levels, a niche filled currently by critical theory, which wholeheartedly embraces the advances of the Enlightenment and the methods of science — including, absolutely, yes, the exhortation of Sagan to doubt everything, even those things it is built upon. Doubt does not mean we discard it; it means we constantly keep an open mind and ask, “What if this is wrong? What if it is incomplete?” As Neil deGrasse Tyson summarizes it, we must do whatever it takes to avoid fooling ourselves.
We can capture every worthwhile insight from critical theories while providing a framework to make the conclusions — the correct ones — increasingly hard to deny. Of course, critical theories are happy enough to embrace evidence when it aligns with the theorist’s purpose. But that’s easy mode. Hard mode — the only mode that fairly-reliably works — is embracing the limitations, too. Embracing what you see even though it’s telling you you haven’t gotten it all figured out yet, or something you thought you know was wrong.
For instance, instead of idolizing narrative, we can recognize that we already understand how to use individual accounts to open our minds to possibilities we weren’t adequately considering, but in a way that won’t cause our minds to fall out (if we’re careful): the case study. We can accept narrative itself as a perspective on a situation — again, maybe a very important perspective that hadn’t been properly appreciated! — but recognize that cognitive biases and limited experience necessarily limit and distort narratives, too. We need to take that into account if we’re to build a shared understanding. And, aside from matters of how people feel, the primary use of narrative for the purposes of building a better society is to help us gained the same, shared understanding of what actually happened.
When we propose some societal change, we can take insight from systems biology and the like and recognize that we need evaluate as best we can the likely range of impacts on society as a whole, while keeping in mind as many of the dynamic processes as we can.
As always with a scientific or science-inspired approach, leaving room for doubt must be central. A key part of this is avoiding orthodoxy: the idea that there is one right answer and that to consider anything else is inherently blasphemous. You don’t just throw up your hands and declare harebrained ideas to be good as well-supported ones. You don’t accept that very very thoroughly debunked ideas repeatedly and loudly promoted are as good as reasonably well-tested ideas. But when someone raises nonsense to compete with a decent well-founded model, it should be a teachable moment in most cases, not a time for excommunication and banishment (and punching in the face — they’re probably a Nazi, after all). We need psychological safety around speaking a contrary view (when the view is shared in good faith with as much support from reason or evidence as can be managed). We need not orthodoxy but heterodoxy.
In particular, we cannot assume that there is an oppressed class, that we must achieve dynamic unity with the oppressed class or we have no function, that racism is ordinary and everywhere, and so on. We mustn’t assume; we must ask. We have to ask: are we right? Are we wrong? To what extent? How do we know? Does it have bad or good consequences? How big are the consequences? Can we do anything about it? What manipulations cause what kind of effect? How does the part that we are studying and may wish to change interact with the whole? Can we measure that?
As Richard Feynman said, “If it disagrees with experiment it is wrong. In that simple statement is the key to science.”. If we can capture that insight, our prospects for understanding what is the case, and then using that understanding to improve things, are greatly improved.
Critical Theory is not the framework we need. Fortunately, we already have all the tools we need to do better.