Value Capture in Philosophy
“Value capture is… where your values are rich and subtle or in the process of developing that way. And you get put in a setting or near a technology or an institution that presents you with a simplified, typically quantified version, and the simplified version takes over.”
That’s C. Thi Nguyen, professor of philosophy at the University of Utah and author of the newly released The Score: How to Stop Playing Somebody Else’s Game, in an interview on the podcast of journalist and commentator Chris Hayes.
[“Slices of Time” (detail) by Emmanuelle Moureaux]
He continues: “You might’ve thought that if anybody is not vulnerable to this, it would be [philosophers].” But as it turns out, “we are vulnerable as hell.”
He then shares part of his story:
So I went into philosophy because I loved weird, interesting questions about like the meaning of life and why any of this is here and why are things beautiful. And then I went into philosophy grad school and I got enculturated. And in philosophy, there is a status ranking collated by survey of what the status is of the journals you publish in and the departments you can be at. And you don’t know any of this when you go to grad school. You’re like, philosophy is cool. I want to think about the meaning of life. And then almost everyone comes out being like, what is my highest rank publication? What is the ranking of my job? And there’s a shift.
And this is the shift that I started thinking about because when that happened, what my sense of what I was trying to do changed because as it turns out for various complex institutional reasons, a lot of the stuff that gets you published in high status journals is not big meaningful questions. It’s fairly technical work on fairly arcane issues and it’s kind of micro moves. I mean, I would say some of these micro moves are very valuable, but it’s like tiny technical logic pushing… And I got obsessed with this.
I got obsessed with moving up the ranking and I started writing philosophy that bored me. And then I wanted to die. And then at some point I was in some place where I was like, I’ve made the weirdest decision in my life, which is to abandon any possible financial reward by not being a lawyer, by leaving Silicon Valley, by being in this weird ass dumb profession, which the only possible reason you could do it is out of love. And yet somehow… I have found myself writing boring philosophy I hate in order to move up this internal ranking system. And then I was like, and I think this is, there’s this raw reaction I often have, which is kind of like, what’s going on? What are we doing here?
The full interview is below. Discussion welcome.
Hear hear! *Continues working on a third-order problem in chmess*
This has long been identified in sociology as “goal displacement.”
If the implicature of this comment is that Nguyen is dressing up an established idea in new garb, I would suggest that the indirect assertion is mistaken.
Value capture is more specifically existential. It’s about what happens to an agent’s rich, vague, hard-to-express, hard-to-measure, identity-constituting cares and values and senses of meaningfulness when the agent who has these things is situated in large-scale institutions. What happens in value capture, Nguyen argues, is (roughly) that the agent chooses to alienate himself by letting the large-scale institution clarify and define and mechanize his values for him. The large-scale institution does this by providing clear but obviously blunt proxies for what the agent really cares about and by providing clear means for meeting the goals. The large-scale institution in effect tells the agent what to care about, what’s important, what success looks like.
The problem is not just that the large-scale institution is wrong in its direction, flattening meaning — though it often is, and laughably so, because the metrics pushed by the large-scale institution are designed for something other than defining a life well-lived. And it’s not just that the institution-defined values “come from outside” the agent — for most of our values rightly do. The problem is also that the agent is incentivized to relinquish the freedom to interpret and order his values in the way that makes the best sense for him and his context. What “being a good parent” means for one may not be “cashed out” in exactly the same way for another, given the innumerable particularities of context.
But if I misread the implicature of your comment, then I just rambled on for nothing!
This is valiant, but come on, value capture is similar enough to the well studied phenomenon of goal displacement that if Nguyen doesn’t discuss it, then, well, that says something else about our discipline.
I actually agree with you that Nguyen should mention goal-displacement and clarify the similarities to and differences from his concept of value-capture.
(I’m three-quarters through the book, and have read many of his articles, and I don’t think I’ve seen it mentioned yet.)
But, again, his concept has a distinct existential flavor — a concern with the soul, if you will — that the sociological concept does not. And it’s not as if he didn’t do his sociological homework. His articles and his book are chock full of references to sociological works.
One of the great things about The Score is that Nguyen doesn’t spend much time citing people and discussing their ideas. When I read it, I realized that he must have known that there were about 100 more philosophers he could have cited, but he just doesn’t. Instead, most chapters just have one thinker who’s helped him to see something important.
This is clearly by design, and it appears to be part of his project of breaking free of the rigid, stultifying norms that dominate younger scholars in academia. If you’re saying big, important things, there is no way to cite even 10% of the prior scholars who have said similar things. I find his style extremely refreshing, and I am saddened for the poor scholars who feel forced to write whole books where 2/3rds of each chapter is taken up with various forms of homage.
The fact is that all the great works we study are not constrained in this way. To enact this norm discipline-wide is to deprive scholars of the freedom to just write. If you read The Score, you might understand why this kind of freedom matters.
It’s a popular book, rather than a strictly academic one. Remember: it’s published by Penguin-Random House.
For an audience of academic philosophers, we can check out his OUP book, Games: Agency as Art.
I like this. Thanks.
I’d still be a little surprised were he not to have stumbled onto the concept of goal displacement, given how much reading in sociology he’s obviously done for the last 5-10 years.
But, as you suggest, I also think it’s okay to have simply discovered and articulated a concept independently and to have done something really cool and interesting with it.
Which leads me to posit: I don’t think this sub-thread would exist — I don’t think there’d be sniping at his neglect of this or that reference — if Nguyen weren’t getting such well-deserved public praise for his work.
Makes sense. And this seems related to Goodhart’s Law: “When a measure becomes a target, it ceases to be a good measure.”
E.g., in social media: let’s say what’s valuable is engagement, and a measure of engagement can be the number of followers or likes on a post. But if you fixate on those measures — crafting posts to maximize your numbers — then that’s counterproductive to engagement, which implies authenticity. You’re just producing clickbait for ego and/or money.
Whoops…listening to this podcast (or video interview disguised as a podcast), I see that Thi mentioned Goodhart’s Law at 18:14. 😉
while there’s plenty of production for ego/money alone, i think that those cynical motives often aren’t the only ones. many feel compelled to produce more-of-x if x was very well-received. sure, that’s partly capitalizing on x’s success but it’s also a good feeling for many to make people happy/be appreciated/accrue goodwill. we then begin to associate our value with the value of producing more-of-x rather than with interests, questions, ideas, skills that brought us to x in the first place.
As much as I appreciate what Thi Nguyen is doing, I think value capture is the wrong framework to make sense of modern academic problems. I get that this is one of the things that he’s best known for but I think this is an instance of everything looking like a nail when all you have is a hammer. We definitely have institutional problems in academia and many of those are about the overproduction of bad work but that’s been true for decades and decades.
The phrase “publish or perish” has its own wikipedia page which notes that the “pressure to publish has been strongly criticized on the basis that over-emphasis on publishing may decrease the value of resulting scholarship, as scholars must spend more time scrambling to publish whatever they can get into print, rather than spending time developing significant research agendas.” It also says that the first use of the phrase was in 1928 to describe academia.
My preferred framework is the Kuhnian one. Academia has been in a long period of “normal science” where scholars work within their cloisters and institutional rewards are structured around publications, grants, and impact. Technologies have threatened to take academia into crisis (most recently LLMs but I recall the panic over MOOCs, paper mills, fiverr, Yik Yak, etc etc). As of yet, normal science prevails. I do think AI/LLMs have the potential to really bring us to crisis, especially as the ability to access these technologies becomes more and more discrete (at my university we are not allowed to ask students with the new Meta Raybans to remove them).
I also think that some of the changes to academic philosophy make sense. Much as in other mature disciplines going through normal science (I’m looking at you physics), our problem spaces are well known and the questions have largely become so complex and technical that individuals likely can’t solve them. It takes teams of people working through really technical issues and sub-problems to advance things. And I don’t think this is a bad thing either.
When scholars like Thi Nguyen lament the fact that they feel like they can’t write about “weird, interesting questions about like the meaning of life and why any of this is here and why are things beautiful’ I struggle to make sense of what they mean. These are well-trodden topics. Unless you’re going to just re-invent the wheel, anything we have to say about “the meaning of life” needs to make contact with a long and sophisticated history on that very question. Anyone who fails to make contact with that literature will almost certainly be writing bad philosophy (we have all had that student who believes that their genius solution to the problem of evil finally settles it once and for all but it’s really just absolutely ignorant of the collosal history behind the problem). Philosophers today are working on weird interesting problems but they’re doing so while recognizing that this is an old problem about which much has already been said.
To me, the problems with modern philosophy are less about “value capture” and more about the failure to appreciate the value of academic work. It’s because journal editors, reviewers, and writers are not paid that our publication process is so bad. In a sane universe, it shouldn’t take 9-18 months to publish a peer reviewed article. We shouldn’t be teaching hundreds of students of year. We should have more time to work more closely with a smaller number of students. It’s neoliberalism that’s killing academia, not the fact that we value publication counts (though I admit that someone who values the mere fact that something was published is doing something wrong).
“[A]t my university we are not allowed to ask students with the new Meta Raybans to remove them.” Jeez… I named my oldest son Jean-Luc because I was hoping he would battle the Borg, not become one… We are in trouble, folks.
Sounds like he might become Locutus
Wow can you share more about your school’s Meta glasses policy? I need to get ahead of anything like that here at my own.
Dee,
University lawyers are, in my view, being overzealous about potential ADA based lawsuits. If the glasses are prescription then we are not allowed to ask them to remove them or to bring in a non-recording pair of prescription glasses.
When you say, though, “it’s neo-liberalism that’s killing academia”, isn’t it the case that the means by which it does this is largely through techno-bureaucratic processes which, if they dominate, would necessarily displace goals with measurements of goals ala Goodhart’s law?
What is neo-liberalism without an overabundance of bureaucracy? (I don’t get the difference, i.e. I don’t get what more is concerning about neoliberalism besides this very thing about value capture – I’m worried that I’m missing something more you have in mind.) And what is a more efficient bureaucracy than one that offloads activities humans have found ways to simplify onto software or machine learning systems?
Maybe one can be reassured about the ability of philosophers to recognize meaningful work by the fact that C. Thi Nguyen has published the work for which he’s widely praised in Phil Review, Phil Imprint, Mind, Ergo, JESP, and other leading journals, not to mention the multiple prizes he’s received. I’d say professional philosophy has been quite receptive to his work, and not the boring kind. Deservedly so! So maybe value capture need not dominate the publishing landscape. That’s good news.
I think an important part of this is that, underneath the numerical metrics of citation counts and journal status, is a set of human judgments in the form of peer review. Despite all the problems of peer review, it does at least keep the weird and wacky human judgment in it.
Some people like the idea that philosophical work can get attention through the viral processes of social media, as a way to bypass the gatekeepers of traditional peer-reviewed journals – but I worry that these viral processes of social media are more likely sites of value capture (optimizing for the superficial appeal of an idea, rather than the slightly deeper process of reading a paper for a few hours as in traditional peer review, or the longer processes of intellectual engagement that go into most citations of new work).
To turn it back to our perennial topic of AI, I think it’s an important fact that people close to AI and bullish on it in many other ways keep saying that “good taste” is one of the major comparative advantages humans are likely to maintain for a long time. Substituting referee judgments of “is this interesting” with more easily optimizable judgments of “can this argument be refuted” is a way to lose this value of the peer review system.
I completely agree, especially about peer review vs social media channels.
Read books by Mark Hamilton! You’ll be glad you Did!
The value of philosophy captured me, and I thrived/suffered only because of that.
On the one hand, philosophy isn’t like cartography: There probably are many huge philosophy-continents left for us to discover, and some of the philosophy-continents that we already know about are still poorly or incorrectly mapped out. On the other hand, philosophy’s progression over time from the big-picture to the nitty-gritty is probably an overall good thing. Even before you’ve found all the continents, if you do have pretty good maps of some of the continents, it makes sense to map those ones out as precisely as you can. In fact, at some stages of disciplinary progression, that might be the only productive thing you can do. On a third hand, given that some philosophy-continents remain undiscovered or are poorly/incorrectly mapped out, it’s important for there to be some philosophers who are working on voyages to those ill-understood lands.
Here’s an alternative partial explanation:
You start out with a vague still-developing idea about some big philosophical question like the meaning of life or the nature of beauty and you want to develop it further, maybe formulate it more clearly or figure out what deeper ideas underlie it. And you discover that your initial idea actually has different parts or aspects that need to be thought about separately. And then that each of them has different parts, so that to progress with clarifying your initial big idea you need to address a lot of smaller questions. If “God dwells in the details” you’re now seeking God.
I have no doubt that value capture happens, but sometimes the move from bigger to smaller questions is an intellectual one, based on a recognition that the big question that initially attracted you is actually more complex than you then thought and that seriously engaging with it requires acknowledging that complexity.
It’s worth defining value capture and goal displacement in a way that’s neutral on whether the rewarded behaviour is conducive to the displaced goal or not.
Some think that philosophical work should always engage directly with big ideas, and others think progress is best made by normal science type work (“micro-moves”). If there’s a system that rewards micro-moves, then often that’s what both groups will end up trying to produce. And both groups will exhibit value capture (if defined neutrally) because they’re pursuing the reward, not merely the lofty goal they started out with. But the former group will, of course, be much more conflicted about this.
A corollary is that, even if there’s value capture in philosophy (and there absolutely is!), it doesn’t follow that the reward structures that are in place must be reformed or abolished. It may be that, on the contrary, these structures do good, if imperfect, work by making some philosophers better pursue the original goals than they would otherwise have done.
If only Frederick Winslow Taylor had appended a note at the end of “The Principles of Scientific Management” to the effect of “This stuff was mostly developed in the context of metal die machining, for God’s sake don’t try to apply it to philosophy.”
This is a problem from the perspective of the institution as well. They don’t ultimately want people with high h-indices, they want deep, original thinkers who commit to their claims. I’m fairly sure this awful situation is partly explained by Goodhart’s Law.
“They don’t ultimately want people with high h-indices, they want deep, original thinkers who commit to their claims.”
You obviously don’t work in Australia! (We’re told that all measures need to be “objective”, and h-indices are “objective” while of course being a “deep, original thinker” isn’t. My impression is that this is something we’ve imported from the UK, but I’m not completely sure about that.)