Not so recently now, Jason Brennan wrote on a popular “gotcha” in internet debates, and while his take on it took a beating in the comments, it got me thinking.
Is there a justification for science needing philosophy beyond “but whenever you try to argue that you don’t, that is itself an act of philosophy”?
That particular gotcha seems reminiscent of proof by contradiction. But if we formalise it, it falls apart:
Assumption: Science doesn’t need philosophy
Implication 1: No scientist would ever engage in philosophical discourse
Premise 1: The arguer is a scientist
Premise 2: This argument itself is a part of philosophy
Conclusion 1: A scientist is doing philosophy whenever they engage in this argument
Conclusion 2: Since conclusion 1 contradicts Implication 1, the Assumption must be incorrect.
Written out like this, it seems much more obvious that the ‘gotcha’ isn’t even an argument. It’s a signal for philosophers and philosophy-fans in internet comments to snicker and pat themselves on the back for being so much cleverer than those stupid self-contradicting scientismists. This is why Brennan’s counter-gotcha is likewise a signal for scientists and science-fans on the internet to giggle and raise their glasses in commemoration of how much more productive they are than those angel-counting philosophismists.
Or, to put it explicitly: Implication 1 doesn’t follow from the Assumption. Not even close. Premise 1 might not be true in any given use of the argument, since it’s mostly deployed against science-fans. Premise 2 is deeply dubious and seems to rely on mixing meta-level “discussion about philosophy” into “philosophy.” Yes, if we say “philosophy” is a field of study that includes all studies of itself, studies of studies of itself, and so on ad infinitum, that’s permissible, but that’s not what we were saying at the beginning.
Trying to prove that science doesn’t need philosophy by getting science to engage in a meta-discussion of philosophy is like trying to prove that you need to know group theory to solve a Rubik’s cube. Actually, it’s not even that good. It’s like arguing that you need to know Zermelo-Fraenkel Set Theory in order to write a diary, since every argument you make that that’s a stupid idea can be expressed using ZFC. Or that you need to carefully study human languages in order to do meaningful heart surgery, since any argument against the notion will be expressed using language.
I worry that the above is not an adequate expression of the counter-position. Let me elaborate.
When someone argues “scientists need to study philosophy” they are referring to a particular body of thought and knowledge. The most common case being ethics, but also frequently a demand that scientists should (for reasons explored below) understand the epistemological framework that underlies the scientific method, and the justifications of the idea of Knowledge that underlie that etc etc. But then when someone objects that actually that isn’t the case, their objection is not part of that same body of philosophy, but rather a broader “meta-philosophy” of ideas about philosophy itself.
How does one argue from “all objections to studying philosophy are meta-philosophical in nature” to “and therefore you should study philosophy”, without equivocating between the philosophy that was originally meant by the injunction to study philosophy, and the meta-philosophy that was introduced as part of the gotcha? Isn’t this a case of constructing a false referent class, “philosophy plus meta-philosophy (plus m2-philosophy etc)” that – while not unnatural – would not be a grouping that would be used in this discussion except to make people think they’ve already accepted that they are doing one thing when they’re actually doing the other? The original discussion, remember, was about ethics or the nature of truth or something. Proving that you can’t avoid arguments about what you can and can’t argue about doesn’t prove that you have to do any of those things. It’s a classic semantic bait-and-switch, and people need to stop using it.
To conclude this section: this gotcha stands only as long as it’s not thought about, at which point it collapses into a word-game and some shady inferences.
* – For example, you could argue that while scientists do engage in philosophical discourse, that they should never need to engage in philosophical discourse about science (which this argument is). Really, though, trying to salvage a solid argument from this rhetorical trap is a mug’s game.
Let’s examine further the question of why it’s considered necessary for scientists to justify their basic epistemological framework. One answer springs immediately to mind. I do try to resist the urge to be cynical, and fair warning: this is very cynical. But nonetheless, isn’t it possible that the reason philosophismists love to demand that scientists provide a rigorous defense of their underpinning assumptions (which usually aren’t too far from naïve falsificationism) is that the philosophy-fan in question has a really smack-bang-whiz-pop takedown of the expected reply? Something that they’re just aching to unleash on someone who never read a bunch of impenetrable books about how truth is dead? This is definitely what the scientism followers are afraid of when they’re wary, or even outright dismissive, of engaging in such discussions. Eventually (actually, very quickly) you get tired of hearing that knowledge is merely opinion and therefore you can’t know nuffin’, and of cleverer arguments to the same effect.
In fact, I would go further than this and say that there is no currently-available reply to the scientist’s philosophical justifications that is in any way worth pursuing in depth. The whole line of inquiry just leads into either an endless rabbit-hole of one or another interminable semantic and/or axiomatic argument over which assumptions are more basic**, or the inquirer says “oh okay I guess you did your research and gave exactly the answer I knew you should give in your position as a scientist, never mind.” I’ve never actually seen the latter happen, but it could. Maybe.
But it would be better to assume that the desire to have science be “grounded in (good) philosophy” is an honest one. Thus, we come to the question of what it actually means to “ground” science in good philosophy. An excellent answer to this was given by Daniel Kokotajlo, also no longer recently:
I’m not trying to say we can’t do science until we do philosophy. What I’m saying is that if we don’t do philosophy, we’re opening ourselves up to the possibility of being permanently and dramatically mistaken in our science. For example: If we think that Laws of Nature are literally Laws set down by God, as the original people who came up with the concept seem to have thought, and we don’t ever do any philosophy to question that, then the existence of God will be an unexamined assumption of science, and worse yet, people will eventually get confused and think science has actually proven it. My worry is that something similar might happen with consciousness.
This is a very good point, but I think it relies on a rather philosophy-oriented view of history. Let’s take an example. It was previously held that the universe would stand eternal, on the grounds of what seemed like rock-solid philosophical consideration. However, advances in science and engineering of steam power led to the notion of entropy, and the conclusion that everything is doomed to eventual dark and cold. The notion of an unlimited future was taken apart by this. Were there philosophers arguing for finite future extent of the world before this? Well, of course. Did some of them venture the notion of inescapably rising disorder as an a priori fact? Possibly, but not that I’ve found mention of. And this is an easy case, since you can derive the notion of entropy from any number of thought experiments once you know to think that way. But critically, even if it was voiced as an idea, it did not become widely known and lead scientists to wonder. The process seems to have been mostly the exact opposite of that, with feedback from the more abstract levels to the more concrete only occurring after reality called attention to the area and practical investigation had been done.
Even if we take the most realistic view, of side-by-side evolution – we should recall, after all, that there was no sharp distinction between ‘scientist’ and ‘philosopher’ at the time – it still seems unrealistic to say that philosophy was what took apart the eternal worldview leaving those same natural philosophers free to theorise about entropy, thereby enabling them to, in their engineering-time, design better steam engines.
Consider also the concern that people will think science has “settled” philosophy in the wrong direction. This, too, has already happened, with many thinkers of the 20th century deciding that quantum mechanics had disproved realism***. And yet we now have more interpretations of quantum mechanics than ever, including multiple realist interpretations such as Everett’s many-worlds and de Broglie/Bohm’s pilot wave. The important point here is that science didn’t become stuck after “proving” a wrong philosophical position.
Granted, it would have been better if Bohr hadn’t pushed his anti-realist interpretation so hard, and the whole philosophical question of what quantum mechanics means had just been left fallow until a broad range of ideas had been formulated. But Popper’s case wasn’t “let’s wait and see what else people think of,” but much more along the lines of “down with the anti-realist tribe!” So philosophy wasn’t exactly helping things. Also of note, Bohr was more of a philosopher than Everett or de Broglie (though a bit less than Bohm).
And consider the opposite case: people thinking science hasn’t settled philosophy when it has. The nature of time is the big one here. General Relativity might not prove Eternalism true, but it certainly wreaks a heavy toll on all non-B-theories of time. Continuing to hold to such theories seems at least as bad as the mistake that Bohr made. Especially when the people who do so don’t seem to have actually studied relativity and gained the deep understanding of it that makes it clear just how much damage the notion of a universal-now has sustained. It may be of interest to consider also William Lane Craig and his total failure to grasp the core concepts of transfinite cardinality, decades after the foundations were laid down by Cantor. But this would carry us off into mathematics, which is too useful a bridge between disciplines to trample over in this essay.
The important thing to take home is that having a state of philosophical debate on an issue isn’t sufficient to prevent scientists from thinking a result proves their own preferred philosophy to be true, but conversely scientific investigation does eventually overcome philosophical misguidedness.
** – I believe this is called “doing philosophy.”
*** – The latter of these two links is a surprisingly readable account; consider it generally recommended.
If all this has given you the impression that I am a Harris-following antiphilosophe about to propose we bulldoze the humanities department to make space for a statue of Newton standing on Descartes’ neck, then allow me to reassure you: that would be entirely too much of a good thing.
Alright, maybe that’s not reassuring.
We’re left with a few more arguments for adding philosophy to science courses. Scientists do seem to have a habit of making philosophical proclamations and getting much more respect and attention from the public than actual philosophers, which must be quite upsetting. This is bad, because…? Well, it gives the public a mistaken impression of the field. But you can imagine what would happen if we poured public scrutiny onto professional philosophy: the whole field would be seen as a bunch of overpaid bearded fellows generating an endless supply of pure hot air. Politics without the fun easy-access tribal bickering.
I’m not quite willing to endorse the further statement that this is bad because the scientists who do this are never right about anything to do with philosophy. Their contention that the existence of angels can be dismissed without first studying the supposed physiology of six-limbed endoskeletes is maybe not sufficiently elaborately phrased and carefully constructed, but fundamentally comes from the right impulse. No one can afford to make a perfect study of every field they decide is nonsense, and it’s better to be too aggressive in cutting out the useless than too merciful.
And of course, there’s the argument that studying multiple disciplines makes you a well-rounded person. Sadly, this argument is pure wishful thinking. Not to mention, it’s not really clear what “well-rounded” means. We’d all like to be wide-ranging masters of many skills, warrior-poets and New Age Renaissance People. But we can’t, and pretending that learning to regurgitate some pre-thought Deep Wisdom about Hegel is the same thing is just sad.
Or, to make a more forthright claim, people become “well-rounded” by pursuing the full breadth of their own interests. You can’t teach being an interesting person, and you certainly can’t do it by the methods employed to teach philosophy or other humanities.
Which leaves an argument about ethics. I want to say, “well, tell me about all these evil scientists who destroyed the world because they didn’t learn that Human Life Is Valuable in a classroom at some point. Oh, they’re all fictional? And the stories they take part in were pretty much universally written by humanities-fans? How convenient.” But of course, if science actually had destroyed the world, we wouldn’t know about it. And the power of science grows over time: we might not be able to generalise from a history of not-destroying-the-world if the capability only recently became available. Genetic engineering is the standout example here.
However, I think there’s an important point in the above. The conception of scientists as being inherently ethically untrustworthy is from popular perception, not any actual lack of ethics. For instance, it’s not uncommon to hear that science was willing to run the Trinity test without knowing if it would ignite self-sustaining nitrogen fusion in the atmosphere and end the world right there. Of course, they knew that couldn’t happen. The same story may be repeated nearly word-for-word with the Large Hadron Collider. The fact that the popular perception lambastes scientists for taking the risk rather than boggles at the sheer level of caution required to carefully consider such far-out possibilities in the first place says a lot.
Now, I’m confident that most people reading this don’t need to be told that it was well-known that neither of these experiments would destroy the world. But is it possible that the perceptual effect, the thinking that scientists must have been very foolish or very unethical to try, still lingers even after dismissing the original urban legend? I can’t think of any good reason to suppose that scientists are insufficiently ethical, and would be quite interested to hear one.
There are several object-level reasons to insist that scientists learn some philosophy. But to suppose that certain professions engender a lack of basic human ethics that must be corrected by education, or a lack of basic human interests that must be corrected by education, or a lack of basic human respect for others’ expertise that must be corrected by education: these are claims in desperate want of evidence.
I will use the conclusion to this piece to say that it is good for a scientist to also study philosophy. It’s good for anyone to study anything, in fact. The interested-amateur – the science-fan on the internet – is in a particularly good position to gain something, since they’re not exactly busy publishing papers about neuroscience or quantum mechanics.
And I don’t think the fields should be entirely separate and never talk except through a hole in the bedsheets. There is something quite sad about the loss of the old “Natural Philosophy” discipline. I wanted to avoid making this about who more commonly tramples whose territory, because I think that concedes too much to the idea that either discipline has any territory that is in some way sacrosanct.
The only conclusion here that I’m confident of is that people should stop using the gotcha at the top of this post. It’s trivially incorrect as well as deeply flawed, and inculcates a really poor attitude that reduces a difficult and interesting question to a moment of circlejerkery. Do not do this.
I’ll go back to Python and extended Tolkien metaphors soon, I promise.