In my recent roundup, I recommended Cate Hall’s advice to be more agentic. In this post, I offer disjointed reflections on her idea of radical agency.
Agency, more or less
Agency is a matter of degree. You can have more or less of it, and you can improve. Hall writes:
Over the years, as I’ve gradually grown dumber relative to my peers through a combination of aging and making smarter friends, one of the main ways I’ve compensated has been through dialing up my agency, which I think of as something like “manifest determination to make things happen.”
As a first pass, “determination to make things happen” is not a bad definition of agency. I’m not sure what role “manifest” is playing here, but maybe “conscious” and/or “deliberate” are what she has in mind. Either way, agents make things they want happen—they are efficacious and purposive. Human agents, in particular, do so by way of desires, preferences, intentions, planning, and coordination—these constitute what I take the “determination” part to be. Finally, such determination will often explicitly appeal to reasons to do certain things and to do them a certain way. Typically, though not always, such reasons will be “manifest” to us—we access, weigh, evaluate, choose, and endorse them through a deliberative process that results in intention and execution. And we humans do this more reflectively and consciously than other animals, at least as far as we can tell, thanks to our abilities to not only consider our reasons but to reject them, dispute them, exchange them with others, and recursively reflect on such a process, and so on. Moreover, as Michael Bratman has long explained, human agency isn’t just purposive—so is animal agency—it is distinctively organized and embedded in planning structures, which require complex, higher-order, temporally extended representations and social coordination. I’ll discuss those differences in a future post.
However, much agency depends on “subpersonal” processes—instincts, drives, emotions, habits, routines, reflexes, and biases shape our behavior in ways that we typically only appreciate in hindsight and partially, if ever and at all. Human agency often happens much less explicitly, less consciously, and much too quickly, for conscious processing and deliberate reasoning. More often than we like to admit, the whole process is quite messy and opaque. This makes our agency often a lot more similar to animal agency. Still, humans are agents when and to the extent that they intentionally move things around in the world as part of larger plans, and respond explicitly to (at least apparent) reasons to do so. The hallmarks of human agency are not pure delusions; if they are, they’re extremely compelling.
If agency is a matter of degree, I suspect almost everything in this account can be dialed down. If I alter any of these aspects of agency, holding everything else constant, I may exercise my agency to a lesser degree. Moreover, I can dial them down only to a point before I stop exercising my agency, but instead start behaving automatically, irrationally, or under coercion—my agency can also be impaired, unwittingly, by mental illness, temporary loss of control, intoxication, addiction, delusion, hallucination, hypnosis, etc.
On the other hand, if Hall is right, I can also dial it up. Up to what point? I don’t know. There is such a thing as minimal agency (and I think at least most animals have it), but there may be no such thing as maximal agency. If the sky is the limit, this is good news: we have plenty of room to cultivate our agency and make more things happen! I’m not saying it’s easy, but it’s possible. Take Hall:
As a result, I’ve done a bunch of cool stuff in different domains: I was a Supreme Court advocate and the number one female poker player in the world; started art and perfume companies; and led operations at Alvea, a pandemic medicine company I co-founded, when it set the record for the fastest startup to take a drug to clinical trials. All of these things I did in my 30s.
Pretty impressive. Notice that these are indeed things Hall has done or made, not just things she’s been or experienced. Some people care more about the being (or appearing to be) part than the doing part (cf. social media), but Hall seems to care a great deal about the doing part.
On the edge
Hall then introduces the idea of radical agency.
In my way of thinking, radical agency is about finding real edges: things you are willing to do that others aren’t, often because they’re annoying or unpleasant. These don’t always surface in awareness to the point one is actually choosing -- often they live in a cloud of aversion that strategically obscures the tradeoff.
Hall finds radical agency on the edges: “things you are willing to do that others aren’t”. Radical agency reminds me of an idea popular among economists: thinking on the margin (I sometimes hear “at the margin”). The thing about margins is that what makes them interesting is extrinsic—relative to that of which they are a margin. Because diminishing returns are a thing, the same quantity of a given good (e.g., monetary value) is not worth the same (in prudential value) to everyone. Tipping your server benefits them more than tipping Taylor Swift. For any possible choice, ask yourself what additional value it creates (in your or somebody else’s life): compare marginal benefits with marginal cost, not previous costs, which are now irrelevant. Forget about sunk costs or fidelity to your past investments. Just think: how much net good can this act do now or in the future? Marginal value can be realized by picking up overlooked low-hanging fruit or identifying a tractable way to overcome a major challenge. The former is easy but neglected; the latter is hard but the payoff is high. When you hit a point of diminishing returns, the marginal value of further investment decreases and your efforts are better spent elsewhere. Did you invest in a losing proposition? Move on. Your further investments won’t do any good. Is there a relatively unexplored space you could dig into? Go for it.
Suppose you’re training for a marathon. There is an optimal number of specific workouts and total mileage that will maximize your chances of hitting your target pace. However, the limitations of the human body are such that the marginal value of any particular workout will not be constant. If you have already run a lot, or if you’re approaching the race, you will hit a point of diminishing returns. But it is also true if you are undertrained. The marginal value of an extra workout before the race may be small or even negative, because to derive extra fitness from the workout you need to rely on compounding effects, and added fatigue shortly before a race is not worth the meager and uncertain gain, and on, and so forth. On the other hand, months of training will offer some edges, windows of time and energy where you can perform challenging workouts that are a little scary and quite painful, but which can go a long way toward improving your fitness on top of a solid base. Here the value is not diminished by the base you already have; it comes, as the payoff, from going where your body is not naturally inclined to go. (Note that, on the edges, marginal costs can also be high—here, increased risk of injury and burnout. Go there only if the expected payoff exceeds the cost.)
Take another example: recordings of Beethoven’s piano sonatas. I’m not talking about playing them; just listening to them. Yes, there is such a thing as aesthetic agency—something we do together, partly by sharing music we love. Beethoven published 32 of them (starting with No.1, op. 2, excluding the ‘juvenile’ ones). They have been performed and recorded ad nauseam, and for good reason: most of them are wonderful. But suppose you spend time browsing albums more or less randomly on Spotify, Apple Music, YouTube, or your local music store. In that case, you’ll notice that some well-known sonatas tend to be more often recorded than others—Moonlight, Appassionata, Pathétique, Waldstein, and the late sonatas (No. 28-32). Yes, like Haydn’s symphonies, the named sonatas are more frequently played than the unnamed ones (the names were rarely Haydn’s or Beethoven’s). You’ve almost certainly heard some of them. I suppose there is a good reason for professional pianists to want to record a complete cycle of Beethoven’s sonatas. There is great value in such sets, even though we already have many of them. On the other hand, suppose you want to listen to one of Beethoven’s sonatas tomorrow morning. You’re unsure which. There’s comfort in choosing a familiar one. And some sonatas you just can’t get weary of. Still, I want to suggest that the marginal value of choosing a sonata you’ve never listened to is much greater. Go ahead, tomorrow find a recording (any) of any unnamed sonata you are not already familiar with. The average value of a well-known sonata might be higher than the average value of the unknown ones (maybe), but this doesn’t affect the relative marginal values of those two possible experiences. Another way to put the point is: go for the underrated.
This one, No. 5 in C minor, is not exactly obscure, but the early sonatas tend to be glossed over more easily. They’re lots of fun. (Granted, very few if any of Beethoven’s sonatas qualify as unknown or underrated. It’s all relative.)
Kinds of value
The value you realize through your agency can be purely instrumental—the expected marginal utility, or impact, of your action relative to what ultimately matters, whether it be economic growth, pleasure, welfare, or status points. The value could also be intrinsic. Suppose happiness is an intrinsic good. How much extra happiness can I expect to create by making yet another dad joke in class (a lot, of course)? Finally, the value could be constitutive—I value running for its own sake (so intrinsically or as a final good), but there are many constitutive goods I also care about, such as a certain balance of sensations, weather, scenery, challenge, and achievement. I don’t value these things either as an end or as a means; they’re part of a complex package, all aspects of which are inextricably valuable. But I can finetune each aspect to increase the constitutive value of the activity. For each kind of value—instrumental, intrinsic, constitutive—certain choices will give me the option to increase it by a margin, and that margin can be tiny or huge. Importantly, what lends my choice significance is not previous investment (sunk costs), or total value, or average value, but the marginal value resulting from that choice. The reason why the marginal value of yet another middling recording the Appassionata says little about the total (or average) value already realized by performances of the sonata, which is enormous other than that such a recording would have to be monumental to add anything of significance relative to the total value.
The edges are where you can create marginal value. The cost may be high (effort required, unpleasantness, risk of failure), but so may the payoff. Notice that the structure of agency is here built into the nature of the task. It’s precisely in virtue of its difficulty that a task requires dialed-up agency, and it’s no coincidence that such tasks typically generate value. I say typically because we can all think of arduous work of little value and some easy ways to create a lot of value. But there is a non-fortuitous relationship between difficulty, effort, and value, and I think agency is what articulates this relationship. Exercising your agency on the edge can be painful. There will be obstacles to overcome and risks to take on. Do I think this makes it any less valuable? No. Au contraire.
If that’s right, then dialed-up agency provides us with a potentially generative route to value that does not depend on hedonic experience, or at least not in the transparent, straightforward way you would expect. Radical agency generates value that may come apart from the hedonic value implied by the activity. Maybe it is meaningful, and maybe there is value in being on the edge. The process may be painful overall; the outcome may not be pleasurable. Because that’s not the point. If it were, there would be easier ways to generate pleasure, which don’t require radical agency.
Work is overrated
However, the edges don’t necessarily imply more work. It’s where most people go that you have to increase the work you put in to increase your chance of reaping benefits. The edges, rather, are neglected or scary or require specialized skills. Not more work. Hall contrasts “the idea of finding real edges” with “eking out wins by grinding harder than everyone.” Speaking of pokey, she mentions “a massive edge [] available in the form of physical reads, but almost entirely ignored.”
All of my agency hacks are kind of like this, in my opinion -- big, glaring edges that people might rather ignore.
So, more difficult in a sense, but not necessarily more work. Just better work.
Hall makes it abundantly clear in the last section of her essay, aptly titled “Don’t work too hard”:
This might be the most important item on the list. It took me almost 40 years to learn it, because my instinct is to think more hours mean more productivity as long as you’re really trying to be productive -- that’s just multiplication, right? No. The reality is that grinding, even if it temporarily increases output, kills creativity and big picture thinking.
Burnout is the ultimate agency-killer. This is so true that I’ve learned to identify a reduction in agency as one of the first signs of burnout, one that shows up even before I consciously realize what’s happening.
Duly noted. Don’t work too hard. Be more agentic.
Next time, I’ll explore what it means for animals to be agentic and what it might mean for them to exercise radical agency.
