What is the significance of the debate on scientific realism?

There is a certain tendency to link it to the question "why trust science?", but that science is to be trusted seems undisputed within this debate. The arch-anti-realist van Fraassen freely admits that we could act as if major theories were true - so what, do you think, is at stake?

(Perhaps I should add that this is not intended as an argument against scientific realism. I would personally subscribe to realism when asked (with some qualifications), but I am wondering about the significance of the debate itself. I mean, and this is not a rhetorical question, what difference does it make whether realists or antirealists prevail?)

Thanks for any responses!

  • 2
    Followung your approach we can add: who cares about philosophy? Commented Aug 25, 2023 at 19:01
  • 4
    It has implications for how developing future theories is approached. For instance, in the development of quantum mechanics the 'anti-realist' view (Bohr's) prevailed over the 'realist' one (Einstein's), see Bohr–Einstein debates, and quantum field theories were subsequently developed in the more abstract mathematical fashion rather than according to the realist intuitions that Einstein favored. Feynman makes some insightful remarks about this in his Nobel lecture.
    – Conifold
    Commented Aug 25, 2023 at 19:41
  • @ Mauro: That would certainly be a legitimate question, don't you think? Note that I was not implying that we should not care about the realism debate, but asking about the extent to which you think it is important. @ Conifold: Would you apply that to science policy issues as well?
    – Turtur
    Commented Aug 25, 2023 at 21:20
  • Yes, which policies get adopted, what research gets funded more, and other pragmatic issues also depend on which views of scientific methods and theories the decision-makers favor.
    – Conifold
    Commented Aug 25, 2023 at 23:03
  • 1
    @D.Halsey have you taken a look at my referenced recent paper about the possibly testable non-equivalence? Now since you accept ontic existence of complex numbers in the same way as reals, there must be some uniqueness about complex numbers where the unobservable imaginary part beyond this actualized observed world is one main feature, thus for me it directly means anti-realism thesis instead of the usual Copenhagen route without hidden variables. Due to embodied cognition most people don't like to abandon realism and believe QM should be explained entirely of observables in our world... Commented Sep 21, 2023 at 5:18

7 Answers 7


My sense of why it matters to those who oppose it, is their feeling that under scientific realism they will lack a certain intellectual 'elbow room' for the things they find precious, like a role of significance for consciousness.

Let's take a modern example. Donald Hoffman came to at least greater prominence, with his TED talk about why we can't trust the results of evolution to show us reality, Do We See Reality As It Is? He made good points there, and it's thought provoking. He then went on to develop his own solution for the Hard Problem where he makes consciousness fundamental, with a kind of monism that has aspects of Idealism and Pansychism (see eg Donald Hoffman’s Conscious Realism vs. Panpsychism and Idealism). He is I think seeking to give consciousness a central role in cosmology, out of a sense that the evolutionary and physicalist-materialist picture of the origins of consciousness make it a kind of footnote.

I think the hubris of certain scientists that sees them portray their endeavour like this:

"The grand underlying principles have been firmly established... further truths of physics are to be looked for in the sixth place of decimals."

-Albert Michelson, in a speech at the inauguration of the Ryerson Physics Laboratory in Chicago 1894, a couple of years before failing to find the luminiferous aether & so making way for Relativity

makes some people fear that science is presenting what Rorty called a 'final vocabulary'. Rorty says drawing on Harold Blooms writing, that one of the 'strong poet's greatest fears' is that he will discover that he has been operating within someone else's final vocabulary all along, and has not "self-created". This article describes cases of philisophers succeeding in that rather poetic task: Nāgārjuna, Nietzsche, and Rorty’s Strange Looping Trick.

I would draw a parallel to how religion has developed, from the era when prophets could reconfigure things to meet new social challenges, to having a 'hard coded' book, which sooner or later is likely to lose relevance to a changed society, ending the progressive revelation of that community engaging with it's past in this way. I see a fear people have, that scientific realism is portraying a future where revelation has ended, no intellectual elbow room remains to play among half-known things and find space in the cosmology for ourselves, instead we are only a curious footnote with nothing much to separate them from dissipative systems like a zephyr or whirlpool.

This is my suspicion, anyway. I think it could be tested, by presenting gloomy theories about the role and nature of life and consciousness to people who identify with scientific realism or not, and seeing if how they react corresponds. And by asking people about what scope they see for radical new understanding to come from science - I suspect the scientufic realists will be far more optimistic.

The status of mathematics, and whether A(G)I can be conscious, are also diagnostic tools that can help understand the motivations of anti-realists.

  • Downvoter/s do please give critiques...
    – CriglCragl
    Commented Sep 20, 2023 at 14:13
  • i downvoted cos i was pissed of with my downvotes (without explanation). you give reasons for why people think 'realism' matters, less than why realism matters. does that make sense? there is a difference.
    – user67675
    Commented Sep 20, 2023 at 14:20
  • 2
    Interesting Nāgārjuna makes the cut! He's been admitted to the clan. Bet he'd scare the living daylights out of Nietszche.
    – Wayfarer
    Commented Sep 21, 2023 at 7:45
  • 2
    Interesting TED talk. Hoffman arrives at phenomenology via science! : Evolutionarily, the human mind sees the 'tomato' as a nutritious fruit, (not a thing-in-itself). However, his terminology re. 'reality' differs from Heidegger's who refers to Hoffman's 'reality' as Aristotle's φύσις (phusis/nature/origin): being (Principle of Reason, p 64), and to reality as networks of involvements "ontologically grounded in the Being of Dasein" SEP; Being & Time, p 255. Otherwise a great move to establish phenomenology scientifically. Commented Sep 22, 2023 at 9:21
  • 1
    +1 Ontological elbow room. :D I like it! I'm going to add it the toolbox next to ontological commitment and ontological shock.
    – J D
    Commented Sep 24, 2023 at 17:16

"What is the significance of the debate on scientific realism?" I don't claim to be able to identify "the" significance of such a debate, but would like nonetheless to point out one of its aspects.  The OP mentioned the anti-realist van Fraassen.  Often van Fraassen's opposite number is considered to be the mathematician/philosopher Putnam.  Regardless of whether mathematics is considered to be one of the sciences, the debate in question certainly includes the issue of realists vs anti-realists with regard to mathematical entities, such as for example the real numbers R or the integers N. Internalizing a realist viewpoint towards numbers, while it may have its advantages in research, can also be an impediment to the reception of novel approaches to R, N, etc.  Thus, axiomatic approaches to infinitesimal analysis find unlimited integers within N itself, and infinitesimals within R itself. 

Realists tend to have more difficulty in the reception of such ideas: if R and N were finally identified (and set in stone) by Weierstrass, Dedekind, and others over 150 years ago, how could the above be possible?  In a recent exchange at MSE, a student, who is a priori interested in nonstandard analysis, reacted as follows to the idea that there might be unlimited numbers in N: "if that's what you claim then this disqualifies you from making competent contributions at this site" (the comment was addressed to another user, not myself, and has been subsequently deleted).  Such reactions come not only from students but also from seasoned specialists in set theory, one of whom (a self-declared realist) referred to such an approach as "idiosyncratic" in a recent comment at MO (the last I checked, the comment was still there).

In my experience, mathematicians working in axiomatic nonstandard analysis tend to take some distance with regard to realist views. In conclusion, whether or not one adopts realist views can arguably color one's reactions to novel developments (as was the case for Einstein, as mentioned in the comments).  To put it another way, the progress of science may well be dependent on the prevailing metaphysics.  The development of analysis in the second half of the 19th century may be a case in point.

  • +1 On the perspective that realism has psychological implications.
    – J D
    Commented Sep 24, 2023 at 17:10
  • Or to put it another way, lack of progress may be a result of inflexibility about metaphysics..! You seem to bring up a case of someone who describes infinity as not literally real. To say there are no infinite sets is just objectively obviously wrong. Could you have got the two views confounded..? At least in regard to the 'expert'?
    – CriglCragl
    Commented Sep 25, 2023 at 20:10
  • @CriglCragl : I elaborated in an earlier answer on the idea that "lack of progress may be a result of inflexibility about metaphysics" (see this answer). Professional mathematicians working in ZF today all operate with infinite sets, so that's not the issue. Commented Sep 26, 2023 at 12:33

The issue here stems from what I might call the 'occurs/is' distinction. Science describes or models events (what occurs), not objects (what is/are). In fact, science often abstracts objects away to mere tokens in order to isolate and examine the events themselves. For instance, if I want to analyze the dynamics of a pool shot — where a cue ball is used to knock a numbered ball into a pocket — the first thing I would do is forget about the cue ball, numbered ball, and the felt table-top. I'd talk about abstract point masses with certain velocities, spins, coefficients of friction, angles of contact, etc. The science problem is a question of the exchange of the energy in linear and rotational moments of inertia; the objects involved are (in an odd way) incidental.

Scientific realism (effectively) asserts the events we observe and measure are functions of properties of real (ontological) objects. Thus in our pool-shot example, the cue ball and numbered ball are objects that 'have' inertia, and contact between them transfers energy so that each ball then 'has' a different inertia. That is a naïve, intuitionist explanation that fits well with our sensory experience, but it is oddly metaphysical. Why should we insist that these 'objects' exist (or at least exist as we imagine them) when all we actually experience are events of interaction. I mean, if we think about it, we can simply change the inertial frame and make either the cue ball or numbered ball appear at rest, or to have any inertia we like as long as the system balances. It makes no difference to the laws of motion, because the laws of motion are only concerned with the interaction. If we have a pool ball that never, ever interacts with anything else, what does it even mean to say it 'has' inertia? Practically speaking, inertia only 'appears' at a moment of contact: i.e., the experience of an 'outside' force. Projecting it as an intrinsic property of the object is supposition or unverifiable inference.

Consider a thought experiment… Say I claim to have a special cue ball (call it an e-cue, 'e' for esoteric). I reach into my pocket, pull out what appears to be absolutely nothing, make a gesture of placing that nothing on the table, go through the motions of taking a shot with the cue stick, and then a moment or two later a numbered ball silently accelerates and rolls into a pocket. It looks exactly like a pool shot, except the e-cue cannot be seen, felt, or heard. Is it correct (if counter-intuitive) to assert this e-cue is a 'real' object, because the events we see play out exactly as we expect of a real object? Or if not, why should we assert that a normal cue ball is a 'real' object? And before you answer, think about how much of modern physics relies on exactly these types of indirect events.

  • 1
    Re your "The science problem is a question of the exchange of linear and rotational moments of inertia", are you sure you literally mean classic physics deals with exchange of ' linear and rotational moments of inertia', not their respective energy? (energy is the core concept in classic physics) Commented Sep 20, 2023 at 23:24
  • @DoubleKnot: I thought I said that. Oops! Commented Sep 20, 2023 at 23:33

I've taken time to review the responses of the venerable contributors here, before I attempted to answer, so don't judge too harshly if I claim my answer is indebted to the other answers by way of abstraction.

Scientific realism matters because it appeals to our psychological needs.

Let's go through a partial laundry list you've been proffered. (My apologies at the violence my simplification does to the articulate ideas of other contributors.) Scientific realism matters because it:

  • Affects one's research politics and funding.
  • Grants ontological elbow room to accommodate diverse worldviews.
  • Allows for the application of theoretical parsimony.
  • Permits scientific consensus.
  • Carries with it overtones of being right and wrong.
  • Is about the broader discourse of events rather than a mere tool of specifying the appropriate role of objectification.
  • Validates our experience.

One of the most difficult ideas that we seem to resist when studying the Western Canon is the idea that the truth is not as objective as we are led to believe ironically by scientific realism (SEP) itself. The implication is there is one description of reality that must prevail and we may squabble, but ultimately, we are unified in getting to know the one, true physical existence. Scientific realism assures us that we may not have the right answers, but that there are right answers to be found. This is in stark contrast to strongly anti-realist interpretations.

Radical post-modernism tries to hold relativity and lack of objectivity as a virtue, and it faces vitriol. (And rightfully so to disclose my biases.) Einstein himself struggled with non-determinism and relativity and the implications to his picture of Laplacian determination and God's refusal to play with dice. Hidden variables in response to the Quantum Revolution still have proponents despite Bell's theorem being rather open and closed. And of course, science itself has a host of enemies, particularly those who gravitate towards supernaturalism as an explanatory tool oddly dividing theologians into natural and supernatural camps. Scientific realism isn't just another scientific theory, it's a metaphysical position, one that reassures us of our place in the universe.

Intuition (SEP) is a powerful force in philosophy, and whether or not you agree with Quine in his Word and Object on semantic ascent as a tool for negotiating the language of reality, it is indisputable that each of us struggles day in and day out on how to best to use the word 'real'. Is an electron real? What about God? Aliens: are they real? Is the love I feel for others real? Philosophy seems little more than an exercise in coming to terms with what it means to be real, and scientific realism appeals to our intuition that there are certain things that are real, and certain things that are not real.

So, as a constructivist, as other constructivists who take a middle ground between reactionary realism and radical anti-realism, scientific realism matters because it carries within it an ecumenical doctrine, the doctrine of metaphysical realism which survives challenge (SEP). And it survives the challenge, not because it is definitive, but because it appeals directly to how our mind, in a Kantian sense, is constructed with primitives like space, time, and altruism. Therefore, scientific realism matters because it appeals to our psychological needs and builds on our intuitions. Anti-realist ideas are good thought tools, but they lack the certainty and coherence our minds thrive on.


Whether there is, or isn't a "reality" to our world, and how we can discover what is real, if there is one, is pretty much central to the philosophic enterprise.

And sceptics, such as Socrates, taught us to question the level of understanding we have of even our must fundamental ideas and definitions, which brings our ability to even define or characterize what "reality" is, into question, which puts the burden of justification on any claims made by advocates of scientific realism.

That rationalism is an unreliable tool to discover any possible reality, has become increasingly apparent, as Kant's Critique of Pure Reason has gotten even more reinforcement in the centuries since. See for example the demonstration that Euclidian Geometry cannot simply be assumed to be true and necessary, and Godel's demonstration that logic systems cannot be reliably used to verify themselves, and the Munchausen Trilemma showing that all justifications of our knowledge are themselves unjustifiable in principle, and the discovery that there are an infinity of different of logics.

This leaves empiricism, and its refinement, science, as the remaining favored potential option to discover "reality" if there is one. And this debate over scientific realism is therefore the byproduct of the increasingly apparent failure of realist rationalism.

This debate has two major aspects to it. One is motivational. Many of the advocates of scientific realism have been advocating for a very limited ontology/metaphysics, that presumes that things like values, and agency, are not included. The objection to such a limited/valueless worldview is behind much of the objection to scientific realism. The other part of the debate is the extension of skeptics to science, not just to rationalism. This skepticism can be applied even without the motivations of questioning a valueless worldview, but many of the anti-realists lean toward skepticism about scientific realism BECAUSE of its valueless inclinations.

Note, elaborating on the skepticism argument, IF science cannot even be defined, as the 20th century debates over the boundary question for science suggests (Verification Principle was refuted, THEN Refutability was refuted, THEN the assumptions behind science, of continually improving verisimilitude, was shown to be logically invalid, and THEN the definition of progressive and regressive Research Programmes was also shown to be logically invalid, etc., etc.) then skeptics often ask, 'how then can science be a plausible candidate for a method to arrive at "reality"'?

Note that not all scientific realists reject values, selfhood, and agency from their ontological worldview. Karl Popper is a premier example of a scientific realist who embraced a more complex ontology that leaves explicit room for agency, and values. Examples of value-based scientific realists show that the dichotomy of values vs realism, that motivates much of this debate, is not itself a necessary dichotomy.


Scientific realism is about distinguishing between what something "is" and how something behaves.

What "is" a photon? Some fuzzy thing. How a photon behaves is the subject of physics.

Since I have evidence and theory supporting my views on the behavior of a photon, I can assume it exists even though I have no idea what it is.

The existence of black holes is a product of scientific realism.


If it's truly a question of acting as if major scientific theories were true, or actually believing in them (and acting the same), then it comes down to personal preference. But that does not mean we can say no more. We can still talk about the reasons a person might hold one or another preference, and consider which of these reasons are persuasive to us. Our initial preference, before thinking about this, could be different from our final preference, after considering all the ideas on both sides in full thoroughness.

Here are some reasons one might prefer to actually believe scientific realism, instead of just acting in all respects as if one believes it.

  1. It's easier, cognitively, to act in accordance with a belief you actually hold. Less effort is involved. Your brain physically needs less energy. When you lie, the frontal and parietal cortex of the brain show increased activity and energy usage on an fMRI. Plausibly, this is because you have to run two separate "programs" in parallel: what you actually believe, and what you're pretending to believe. Any computer will use more energy running two programs at once, compared to just running the one program.
  2. Why would you act in all respects as if you believe scientific realism, if you actually don't? Isn't that an inconsistency between your behavior and your beliefs? Wouldn't most people prefer to eliminate such inconsistencies, when made aware of them? If you truly believe the world is a dream (or whatever), why don't you let that belief fully guide your behavior?
  3. Occam's razor: "The world revealed by science is unreal, but it behaves as if it were real" is a more complex statement than "the world revealed by science is real." The first statement requires us to hold in our minds two things: the scientific model of the world, labeled as unreal but matching observations, and the non-scientific model of the world, labeled as real. The second statement requires us to hold only one thing: the scientific model of the world, labeled as real. The added complexity of the first statement demands commensurate justification, which doesn't seem present.
  4. Are you really acting in all respects as if you believe scientific realism, or are you making exceptions? For example, perhaps you are a creationist, and therefore you want to defund science related to life's origins. This would be a difference in your behavior, even if you act as if other science is true. I doubt there is a single person who acts in every respect as if they believed the scientific consensus, but in fact does not. Even if there were such a person, it would be impossible for us as outside observers to determine this, because they would speak and act exactly as if they held all the beliefs they don't.

On the other side:

  1. Major theories of physics are in fact wrong, in the sense of "not perfectly predictive." General relativity is wrong, and quantum mechanics is wrong. They are mutually inconsistent with each other, so they cannot both be right, and each of them yields incorrect results when applied to the domain of the other. However, we can say that major theories of physics are very good, and the real laws of physics are probably "something like that."
  2. Possibly the reason we do not have a Theory of Everything is that the universe does not obey simple laws at all. Perhaps there is no lowest level of reality at which physicists might one day declare, "we have solved it." Maybe it's turtles all the way down. It sounds like a joke, but it's not impossible.
  • This entire answer is conflating realism with truth. Commented Aug 25, 2023 at 17:54
  • @DavidGudeman "Truth" and "realism" are somewhat ambiguous terms that have been interpreted in different ways by different philosophers. In this context I'd consider them equivalent. What distinction do you draw between the two?
    – causative
    Commented Aug 25, 2023 at 18:05
  • 1
    re: Other-Side 1, that's not quite true. Effective theories mashed together out of QM, GR, and StatMech work, they just don't seem to point to a description of how the universe is, just a good way to predict what the universe does.
    – g s
    Commented Aug 25, 2023 at 18:11
  • @gs Well, combinations of QM and GR (the Standard Model) work on most domains we have access to, but it's not taken seriously that they would work everywhere. They aren't a theory of everything. Very high-energy particle collisions are one area where they leave questions. We just don't have easy access to the domains where the answers are dubious. Even on the domains where they do work, they are expected to be slightly wrong, beneath the level of precision we are able to detect. For instance, gravity is expected to have some tiny tiny unaccounted-for effect on "purely" QM phenomena.
    – causative
    Commented Aug 25, 2023 at 18:20
  • 1
    The point of anti-realism with respect to an ontological category is not that the commonly accepted theories of that ontological category are false, it is that those theories can be reduced to more fundamental theories. Most physicists today are anti-realists with respect to classical thermodynamics. That doesn't mean they consider classical thermodynamics false. Commented Aug 25, 2023 at 20:02

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .