Synthetic Philosophy within the division of Labor

by Eric Schliesser on January 30, 2024

When I first published on what I call ‘synthetic philosophy’ back in 2019, I presented the two key components of the view in such a way that it caused confusion about the position I was trying to describe as a sociological phenomenon within philosophy of science. I developed the idea of ‘synthetic philosophy’ in order to give philosophers of science a better conception of what they actually do and how this might fit in the modern university (and their grant agencies). I introduced the idea with the following characterization:

‘synthetic philosophy’ [is] a style of philosophy that brings together insights, knowledge, and arguments from the special sciences with the aim to offer a coherent account of complex systems and connect these to a wider culture or other philosophical projects (or both).

It is quite natural that my readers thought that synthetic philosophy just is a kind of integrative project. In contemporary philosophy, Philip Kitcher is (recall) the spokesperson for a view like pretty much this (including the use of ‘synthetic philosophy’) in which it is part and parcel of contemporary pragmatism. My friend, Catarina Dutilh Novaes, also advocates for a version of this view (see, for example, here at DailyNous). In Bad Beliefs, Neil Levy emphasized and developed a slightly different version of this view, too. This approach is also nicely defended by Adam Smith in the context of describing philosophy’s role in the division of labor at the start of Wealth of Nations.

My unease about this program is due to the fact that what does the integration, the integrative glue, as it were, is too unconstrained or (to use one of Timothy Williamson’s favorite words) undisciplined. I also worry that it opens the door to the image of the philosopher as creative genius who has mystical powers at understanding the totality of things. I reject the anthropological (and moral) assumption on which such a heroic figure is based. In addition — and I was myself not as clear about this back in 2019 —, hyper-specialization makes the kind of integrative project Kitcher wishes to defend a glorious, fool’s errand. (To be sure Kitcher himself is quite explicit that he rejects the anti-egalitarian commitments that are inscribed in the creative genius image.)

For, as my regular readers know (recall) Elijah Millgram’s The Great Endarkenment: Philosophy for an Age of Hyperspecialization convinced me that even within traditional disciplines it is likely that nobody has an unerring eye for detecting the objective merits of theories because of the effects of hyperspecialization and the development of local modeling languages and pidgins with their abstruse vocabulary. And so much the worse across disciplines.

That is, most of what even very well educated and curious scholars think they know about fields outside their own falls in the category of authoritated belief. (I mean by this that non-expert held believer’s endorsing or adopting the beliefs held by the relevant epistemic authorities (at a suitable level of simplification).) This makes any expertise vulnerable to impostures when they engage with others. I would argue that in Republic Plato already diagnosed this problem (recalland here) and proposed to deal with it by organizing all of society’s institution toward breeding a caste that would be dedicated to developing the skilled exercise of practical knowledge. In a way Plato’s far-reaching (and authoritarian) proposed solution indicates how difficult this problem is.

To put this bluntly, and give you a sense of why you should take this seriously: modern academic and research grant administrators (as well as journalists, lawyers, and judges) are clueless about most of other people’s disciplines’ underlying expertise. This is why following procedure has become so important to officials in charge, and why most self-aware academic leaders sound so vacuous (and are willing to be led by PR folk) when they speak in public. It’s a sign of their intelligence that they understand they have no other choice. (‘Better to sound like a robot than an ignoramus,’ is the motto.)

As it happens, in my original article, and just a few pages later, I also added the following claim:

Synthetic philosophy requires a general theory such as {Darwinism or} game theory or information theory (and perhaps Bayes’ theorem) that is thin and flexible enough to be applied in different special sciences, but rich enough that, when applied, it allows for connection to be developed among them. (pp. 7-8)

That is, back in 2019, for me the ‘integrative glue’’ had to be a theory or model that is in some sense capable of some generality (say, being being applicable to multiple disciplines). This would allow a kind of methodizing and disciplined approach to the synthetic aspect of the practice. I articulated the position in the context of a review of Dennett and Godfrey-Smith (and described similar features in Darwin and Rachel Carson).

Now, interestingly enough, ‘synthetic philosophy’ was coined by Herbert Spencer, without him (I think) ever defining it back in 1864. For him, it seems to cover his whole (systematic) philosophy. The Britannica usefully suggests that in Spencer philosophy itself is the synthesis of (and articulates) the fundamental principles of the special sciences. Evolution (but through acquired characteristics) plays a huge role in this project. In 2019, I adopted the phrase as a homage to Spencer, but in the safe knowledge that nobody — this is actually a standing joke about him when philosophers mention him in our age — actually reads him now. (That is a bit of a shame because (i) it means we miss what was common ground for many in the late nineteenth century and early twentieth centuries; (ii) and while there is a kind of dull relentlessness about his project, it’s full of creative and interesting arguments.) An homage is not an endorsement of his ethical and political views.

Post Darwin, Spencer’s project clearly inspired Huxley (recall) and Peirce (recall) to use Darwinian evolution as a kind of universal acid that could explain and integrate many different phenomena not the least cosmic evolution. (This is also in Spencer, but Spencer is happier to also include arguments not out of place among seventeenth century rationalists.) The Darwinian version also constrains the arguments and modeling a bit more than it does in Spencer.

While Kitcher mentions Dewey, on my reading of his project it is, in fact, continuous with the post-Spencerian sensibility we find in Huxley and Peirce. Everyone that knows my own writing knows I am drawn to such speculative philosophers as subjects of scholarship. But such enterprises also cannot shake the aura of speculative dilettantism. In response to an earlier version of the present argument, Helen de Cruz offered a spirited defense of the philosophical dabbler like herself (see here). My view is that this stance works well for individuals, but is not a good stance for a field or a discipline to embrace.

By contrast, however, the way I understand synthetic philosophy, involves an expertise first and foremost rooted in skilled understanding or practice with theories, models, and particular (formal, computational, conceptual, etc.) methods that have a broad generality. (They need not be wholly topic neutral or universal.) The emphasis is more on the expertise and less on integration. My restatement of the position is thus:

  • Synthetic philosophy is style of philosophy that presupposes or develops expertise in a general theory (or a model, a certain method/technique, etc.) that is thin and flexible enough to be applied in/to different special sciences, but rich enough that, when applied, it allows for connections to be developed among them with the aim to offer a coherent account of complex systems and connect these to a wider culture, the sciences, or other philosophical projects (or both).

That is to say, through the expertise with a modelling practice (a theory, a set of techniques) the synthetic philosopher can speak in several disciplinary languages and perhaps even generate a philosophical pidgin. Her habitat is to be found in (what Peter Gallison calls) trading zones, or arbitrage opportunities, among disciplines as well as, potentially, on the scientific research frontier, or public facing philosophy or science communication.* I have in mind philosophers who work on foundations of some formal apparatus or have the freedom to develop a theory or (say) computing into novel areas.

Now, as this account suggests I, thus, view synthetic philosopher, properly understood, as a means toward managing or coping with the problems that are the effect of hyper-specialization. Because synthetic philosophers have (ahh) feet in multiple disciplines or specializations. That is, in fact, part of their specialism. However, synthetic philosophers, while capacious in sensibility, are not to be thought of as generalists. (I am myself no synthetic philosopher.) Yes, there will always be polymaths among us. But even they are constrained by time to understand only understand a fraction.

Let me wrap up. Of course, sometimes the presence of the same formal apparatus in different fields is not illuminating. [Insert your favorite example!] I myself have a known aversion to Bayesians who use Bayes theory to hammer away at everything. Even so, I would emphasize that conceiving philosophy of science and ‘philosophy in science’ (see here Pradeu et. al.) as ‘synthetic philosophy’ is a means of disciplining professional philosophy; and of keeping philosophy salient in the modern university and to grant agencies in virtue of its potential contributions to the sciences, practical affairs, and by constituting and contributing to the manifest image. Of course, there is a risk that synthetic philosophy may not even be seen as or be housed in philosophy departments. The upside is that it’s a way of constituting the interdisciplinary potential of philosophy without just being wishy washy or tending toward escapism and that’s because permanent learning from (particular) sciences is internalized in the practice.



Bill Benzon 01.30.24 at 9:42 pm

Ah, you’re that Eric Schliesser. I hadn’t realized.

It seems to me that one reason specialized scholars write more inclusive books for general audiences is that they sense the need for a coherent account covering a wide range of phenomena than seems possible in the current era of hyperspecialization. The book may or may not have sufficiently rigorous integrative glue, but the impulse is there. And a journal like Behavioral and Brain Sciences sometimes serves as such a platform; here I’m thinking particularly of Herbet Gintis’s 2007 article, A framework for the unification of the behavioral sciences, calling on evolutionary theory incorporating game theory and decision theory.

It seems to me that we really do need such wide-ranging accounts, and ultimately they need to be developed in full technical detail and rigor. Such things aren’t going to appear miraculously out of nowhere or, for that matter, from some descendent of ChatGPT, though such descendents surely have a role to play in such projects. The only way we’re going get there is to take a crack at it, now, with what we’ve got. We can then take a look at such accounts as they appear, and work on improving them.

(On Dennett, alas, I fear he has devoted enormous effort to developing meme theory in a way that trivializes cultural evolution.)


Limericky Dicky 01.30.24 at 10:26 pm

Timothy Williamson
Had a recipe he made millions on.
Perverse stances wrapped in obfuscatory prose and logico-mathematical eristic,
While unproductive, are great for that all-important ‘widely cited’ statistic!


Maja Sidzi?ska 01.31.24 at 3:19 pm

I see Joseph Rouse as a ‘synthetic’ philosopher. His new book, Social Practice as Biological Niche Construction, is a paradigmatic example.


Eric Schliesser 01.31.24 at 4:53 pm

I have long admired Joe’s work. We were colleagues when I visited Wesleyan over twenty years ago. A very classy guy, too.
Thank you for alerting me to this work.


John Q 01.31.24 at 8:18 pm

(Reposted from Digressions)

A big problem with hyperspecialisation is that whole subfields can develop to the extent that no-one outside the subfield can assess progress or the lack of it. Within economics, I’d say this is true of DSGE macro, the refinements literature in game theory and more. Synthetic philosophy might help to bring more discipline (sic) to bear.


engels 01.31.24 at 11:08 pm

‘synthetic philosophy’ [is] a style of philosophy that brings together insights, knowledge, and arguments from the special sciences with the aim to offer a coherent account of complex systems and connect these to a wider culture or other philosophical projects (or both)

Isn’t this quite similar to the traditional “queen of sciences” idea? Apologies if I’m missing the point.


Eric Schliesser 01.31.24 at 11:09 pm

Yes, John, exactly the disciplining can also shape the special science (or a sub-field in it).


Eric Schliesser 02.01.24 at 12:29 am

No, my version of it is the rejection of that approach.


Muir Douglas 02.01.24 at 8:59 pm

I’m not an astrophysicist, but some years back I first encountered the following theory: “Many heavy elements were largely or almost entirely created by neutron star collisions.”

My initial response: “That seems ridiculous! Neutron stars are rare, so their collisions must be vanishingly rare. They couldn’t possibly account for the observed abundance of heavy elements in the universe.”

— Note from an epistemological POV, my response included one true fact (neutron stars are rare) and one plausible assumption (the universe is vast, stars are tiny and move slowly, therefore stellar collisions are rare). Neither the fact nor the assumption required specialized knowledge; both were readily available to anyone with a layman’s knowledge of astronomy. So (I would say) this was a reasonable first response.

Someone a bit more knowledgeable replied:

“Neutron stars are indeed rare, but they form from giant stars, and giant stars are much more likely to be in binary or multiple star systems. So it’s not unusual for neutron stars to occur in pairs. And when those pairs are orbiting reasonably close to each other, over astronomical time their orbits will decay, because they’re radiating orbital energy in the form of gravity waves. So neutron star collisions, while still rare, are much more common than you might think.”

So this response included facts that I did not know (neutron stars are often binary, etc.) but that were easily checkable with a few moments online. Having checked those facts, I could then amend my opinion of the theory from “seems ridiculous” to “maybe plausible”.

So there’s a family of theories that I’m incompetent to judge, but where my competence could easily be raised by asking a few checkable questions. I mention the neutron star theory specifically because it’s one that /wouldn’t/ respond well to a “thin and flexible” general theory; knowing Bayesian analysis or Darwinian theory wouldn’t really help here.

TBC I am not claiming this is usually or even often true. But perhaps it’s true enough to be its own epistemological category?

Doug M.

Comments on this entry are closed.