Facts vs Opinions

The American educational system teaches children to distinguish between "facts" and "opinions". A recent paper in Misinformation Review has even made mastery of this distinction a marker of civic political competence. Per this paper facts are those statements that "can be proved or disproved with objective evidence" whereas opinions are those statements that "depend on personal values and preferences". I think this is a bogus distinction and should not have any role as a marker of political competence or as part of children's education. 

Now, this is a topic other better philosophers have handled and I agree with their critiques. Corvino outlines the trouble with trying to draw the distinction in any coherent fashion. The NYT piece by McBrayer linked above correctly points out there is no way this could sensibly divide all claims in the manner sometimes suggested. Jenkins-Ichikawa points out the kind of category error involved, how the distinction's contrast ironically encourages a confusion between reality and our perspective thereon. Douglas and Hawley rescued British youth from their cousin's miseducated fate by explaining to the government how this distinction does not actually help people sort lies from error as it was purported to do. (As, indeed, the Misinformation Review piece seems to be hoping it will do.) These critiques all seem right to me and part of the point of this blog post is just to have them linked in one place. Thus it is so.

But I want to add a more high level perspective to them. Because I think they do not address what it is that the fact/opinion dichotomy monger is trying to achieve by proffering their distinction. I think that in their best moments they hope ultimately to help us avoid certain kinds of pointless arguments - pointless argument, in particular, which bad faith actors may try and exploit. Where the bad faith actor is clearly wrong they will fool us into arguing about matters which cannot reasonably be resolved, and encourage us to shrug our shoulders and conclude with an indifferent whomst can say. Where the bad faith actor wishes to push an arbitrary perspective upon us they will act as if their mere preferences reflect deep truths rationality compels us to accept -- mind you I find that this way round is less often worried about in misinformation studies -- more on that towards the end. By educating people on the fact/opinion dichotomy, I believe, the hope is we can avoid this sort of civic manipulation. So in what remains I will explain why I think this a bad idea.

People should offer reasons in favour of the claims they want you to accept, but have the good grace to accept that not all disagreements need to be resolved. If for some reason it is important for us to accept a claim as the basis of action, but I doubt the pertinent claim, then you really ought try to persuade me peaceably and rationally. We may also argue for fun (who's a better captain, Kirk or Picard? Who was the better striker, Bergkamp or Henry? etc etc) and if we do that then the game will typically be more fun if you commit to the bit and actually offer decent reasons in favour of your preferred option. All good. But also you should accept that we don't always need to do this until agreement is reached. It's ok if we just don't agree re what is the best Marvel film, civic peace requires we allow disagreements about the precise nature of the afterlife to go unresolved, the evidence simply doesn't settle whether or not there is sentient extraterrestrial life so at some point we gotta just agree to disagree. Also all good.

If I am right, this is the distinction the fact/opinion distinction is trying to capture. Facts are meant to play the role of: things where, if we disagree, we should reasonably expect to come to agreement if all players respond to evidence rationally. Whereas opinions play the role of: things where we may faultlessly disagree, or at least where we might have to accept that rational persuasion may fall short of securing consensus. As per the above I am not opposed to that distinction per se - it's a good and important one and it's a kind of civic virtue of finesse to recognise when one is dealing with each type of claim. My problem is just that I think the fact-opinion distinction as so understood is a terrible way of characterising this.

The fact-opinion dichotomy gets things wrong in both directions. There are things which it calls facts which fall on the "let it be" side of the seek-reasoned-consensus/let-it-be dichotomy, and there are things which fall on the seek-reasoned-consensus side of the seek-reasoned-consensus/let-it-be dichotomy which it calls opinions. It is easier to give examples of the latter, since the former requires resolving an ambiguity which I shall get to in a moment. But re the former one can think of cases wherein a group decision depends on personal preferences -- should we buy pasta to cook this weekend or ingredients for a curry? My partner and I are faced with this choice reasonably often given our cooking repertoire. Clearly "what's best for dinner" is meant to fall on the opinion side of the dichotomy, but also clearly given our shared life and the fact we need to eat some kinda decision has to be made. We gotta buy something, and even "buy enough for us each to cook our own meals" counts as a decision we need to make together. And this will go at higher levels of seriousness too - I do not think we can agree to disagree re the permissibility of murder, we have to make collective decisions about what the best form of governance is, our shared public spaces have to look some way or another so I cannot be entirely indifferent to ideals of beauty. So on and so forth. Plenty of things on the opinion side are things which we require reasonable argumentation on.

Sometimes people here retreat to a kind of "ah but ultimately there is no disputing taste" on these points in a way that I think somewhat bad faith. So sure they will say I want you to think murder is bad, but ultimately that is... somehow... not the sort of thing I can rationally argue about if we just disagree on core moral claims, whereas facts are the sort of thing we can really demonstrate who is correct about. For one thing, whatever they say in these circumstances they do not actually act in their life as if moral or political or even what-we-shall-have-for-dinner argumentation is impossible. If you are reading this blog post I guarantee you are the kind of person who quasi-regularly engages in all those sorts of arguments, exchanging reasons in as rational a fashion as you can manage, on the regular. No different from other arguments.

 Of course in these arguments someone may always be stubborn or find that when core premises just aren't shared they will not budge or... etc... but the same goes for arguments about matters of fact. I can no more prove to someone that Peano's axioms entail that 2 + 2 = 4 if they will not accept modus ponens than I can prove to someone that torturing puppies is bad if they will not accept that inflicting needless suffering morally counts against an activity. And do you really have a better argument in favour of such basic core principles in the factual case than you do the moral? The distinction between fact and opinion is certainly not that unlike with opinions shared premises are not needed to argue about facts, or premises in arguments about facts simply cannot be denied.

How about then the factual side of things? Here I think we must resolve an ambiguity in the idea of something being "(dis)provable by objective evidence". This might mean something like -- within our present capacities to (dis)confirm, given technical limits and degree of social cooperation etc. But then that is obviously going to misclassify a bunch of claims: it is not, in the intended sense of the distinction, a matter of opinion whether there is a teapot orbiting Alpha Centauri. But clearly we are not in a position to decisively (dis)prove that such exists, it is simply beyond the capacity of our best telescopes. (If you're about to take the teapot thing too seriously just imagine the claim is about an asteroid within certain dimensional parameters that plausibly may or may not in such an orbit.) And nor do I want to put all such claims on the side of let-it-be. We sometimes are forced to act before we can (dis)prove certain claims upon which the decision must turn. In the time available we will have to seek reasoned consensus as best we can.

Instead, therefore, perhaps the claim is that those claims (dis)provable by objective evidence are those which we could, in principle, (dis)prove if our scientific and technical capacities were far superior. And I think that while this will probably classify more claims in the way the fact/opinion dichotomy promoters intended, it clearly does not capture the claims we ought to argue about. Because, of course, our capacities are not ideal, so there will be plenty which we would fruitlessly argue about if we were to focus on the facts so understood. And nor is that merely hypothetical! As a sort of inverse to the above point, it would be terrible to waste our time trying to settle whether future technologies could, in principle, resolve all worries with climate change -- if there is no practical path from here to those technologies that is simply a waste of time when action is demanded now. What in principle can be settled does not align, at all, with what in fact can be settled, and thus certainly not with what it is fruitful to spend our time debating.

The reason this dichotomy gets things wrong is, I think, because it is based on a folk epistemology and metaphysics that is ultimately ill thought through. According to this poor man's verificationism the world is fully independent of our ideas, devoid of evaluative properties, and as such knowable. Were we shorn of delusions and bias we'd all quickly agree as to what is the case, within the limits of our present scientific capacity at least. In addition to claims about this knowable world we express other opinions too; these may be more or less benign but they are ultimately just refined expressions of taste. Irritatingly enough though our biases, ultimately rooted in our emotions, often lead to these tastes interfering with our understanding of the world. The fact/opinion dichotomy is meant to help us stick to the knowable non-evaluative claims wherein we can reach rational agreement, and recognise the expressions of mere taste for what they are.

(If I were American I would also be inclined to argue that it is a subtle violation of the establishment clause, a kind of creeping metaphysical secularism wherein liberal neutrality becomes a metaphysical certitude. Certain religious traditions would have it that a great many things this dichotomy calls "opinions" can indeed be known through objective evidence, namely reading the plain words written down in certain holy texts. By teaching kids to think in terms of the fact/opinion dichotomy it seems the state is coming out against this perspective. This habit of converting political neutrality into a metaphysics wherein rationality demands neutrality is, I think, surprisingly common, and another reason I am suspicious of liberalism.)

Now I just think this is a bad picture of ourselves and our reasoning. The world may not be so knowable, facts are not so neatly independent of our will, the "emotions are biasing" thing is overplayed. But even all these philosophical disagreements aside, I think that outside of some moments of high minded epistemic idealism all this is in the end just a technocrat's delusion. They do not really worry about "murder is wrong" going in the opinion category because they are not worried about serious disagreement on that point. What they most want is a clear realm wherein their expertise is recognised and deferred to and taken as the common basis of political action. This is why they are less worried about false impositions of opinion than they are about relativising the things which ought to be taken as facts. Sometimes the political nature of the distinction is laid bare (witness this from the Misinformation Review paper: "Objective evidence is often quantifiable... and comes from verifiable sources and methods such as official government records" -- the American state as arbiter of what the facts are!) but more usually it is presented as simply a philosophical given, an obvious part of the conceptual landscape. And it is this underhandedness which bothers me. Here we have a grab for status crudely disguised by a toy philosophical distinction. Ultimately my objection to this distinction and the will to power it barely masks is that it is tawdry.

The ordered operation of opinion contrasted with the kaleidoscopic chaos of reality.


  1. I'd say that fact are beliefs that are grounded in "common premises." Common premises are those that if rejected entitle us to dismiss out of hand all arguments seemingly derived from that rejection. In this way the assertion of "fact" justifies the ultimatum that the asserter put-up or shut-up.

    In this way we can say about the asteroid both that it is your opinion that it is in orbit around Alpha Centari and that we believe there is a fact of the matter about whether it is there. The former indicates that your belief is not, in fact, grounded by common premises but that the belief that someone could ground such a belief in common premises is itself grounded in common premises. An assertion of fact could be justified but not by you, here and now.

    Opinions are assertions which are grounded neither in common premises nor in any premises that contradict common premises.

  2. I am mostly convinced by the argument, but I resist the suggestion that it tells against liberalism. I think that the misinformation discourse looks for technocratic means of delegitimizing, de-platforming, or even suppressing ideas that its practitioners dislike while maintaining their political neutrality. The facts/opinions distinctions aids in this project b/c it suggests that incorrect factual assertions (say, that the COVID-19 virus was created in a lab) can be suppressed w/o really suppressing anyone's democratic freedom to express their political opinions.

    If the above is what someone means by liberal neutrality, yeah it's bogus, but I don't think that's what it has to mean. Liberals just need to allow the debate to be as open as possible while having a decision mechanism (voting?) when collective action is required. To me, the need for collective action is what transforms an argument from a game (what is the best Marvel Movie) into something that requires agreement (what will we buy for dinner). Of course, we don't have to agree that in fact curry is better than pasta, we just need a means of deciding what we will "do" before we go hungry.

    1. Ah to be clear I do agree with you! I don't think liberal neutrality just necessarily gets construed this way. But its a worrying tendency that it in fact occurs imo.


Post a Comment

Popular posts from this blog

How I Am A Marxist

On Not Believing In One's Work

Arguments in Philosophy