Daniel Little’s blog post ‘Moral progress and critical realism‘ raises some important issues for critical realists and indeed social scientists more generally. I’m sympathetic to the general orientation of his piece, and have made similar arguments elsewhere (summarised in my previous post on Materially Social). I thought it would be useful, though, to add some further discussion of how Daniel’s argument relates to critical realism itself, and in particular to the status of Roy Bhaskar’s theory of explanatory critique
[this post is a republication of a guest post that Daniel kindly allowed me to post on his own blog Understanding Society].
While critical realists agree that there is a real world that exists independently of what we think about it, they need not – and do not – agree on exactly which classes of things exist within that world. Moral realism is a case in point. Roy Bhaskar explicitly identified himself as a moral realist, and offered several different justifications for this in the course of his work. Some critical realists accept all of those justifications, some are ambivalent or selective about which they accept, and others like Andrew Sayer and myself, for example, reject moral realism outright.
I’d like to focus here on one of Bhaskar’s arguments: the theory of explanatory critique. Technically this is an argument for ethical naturalism rather than moral realism (I’ll come back to that), although it is sometimes regarded as supporting both. The classic statement of the theory can be found in his book Scientific Realism and Human Emancipation (p. 177):
“Let a belief P, which has some object O, have a source (causal explanation) S. I am going to contend that if we possess: (i) adequate grounds for supposing P is false; and (ii) adequate grounds for supposing that S co-explains P, then we may, and must, pass immediately to (iii) a negative evaluation of S (CP); and (iv) a positive evaluation of action rationally directed at the removal of S (CP)”
This argument can be read and/or employed in a number of different ways. Let me discuss three. First, it is a variation on the classic Marxist critique of ideology – it suggests that there are social institutions (S) which generate false beliefs (P) about other (or the same) social institutions (O) and that we ought to get rid of them. For example: if capitalist-controlled media sources mislead us about the nature of capitalism then we should replace those media sources (note that Bhaskar is careful to qualify the argument with a ceteris paribus clause (CP), and thus acknowledges that other factors must also be taken into consideration). As a critical ethical claim this seems reasonable and attractive, and it gains some of its appeal by being rather more direct than most versions of ideology critique.
But this is not the point of the theory of explanatory critique, which brings us to the second reading. On this reading, the purpose of Bhaskar’s statement is to support his advocacy of ethical naturalism: the claim that we can derive ethical conclusions from purely factual premises. Bhaskar maintains that the premises (i) and (ii) are purely factual, and lead logically (“we may, and must, pass immediately to…”) to the ethical conclusions (iii) and (iv). But as a number of people have pointed out, there is a flaw in this argument. The premises are indeed purely factual, and the conclusions are indeed ethical, but the premises are not sufficient to entail the conclusions. To arrive at these conclusions, we need a further premise: we must also believe that it is wrong to generate, advocate, or support false beliefs. Of course, most of us DO believe that, and if so we may well be happy to accept the conclusion in reading one. But that doesn’t mean that Bhaskar has shown us how to derive an ethical conclusion from purely factual premises: his argument for ethical naturalism is false.
One also finds critical realists who think that the theory of explanatory critique provides a justification for moral realism: the claim that there are moral facts that are objectively right, good, or true regardless of what people may think about them. As far as I am aware Bhaskar himself does not claim that the theory of explanatory critique entails moral realism, and when he does advocate moral realism explicitly in his later work he offers other arguments to support it. But most critical realists are uncomfortable with those later arguments, and so it is important to establish whether or not the theory of explanatory critique does support moral realism. Let me call this a third reading of the argument, although it also depends on the second. On this reading, the argument for ethical naturalism establishes that we can indeed derive ethical claims from non-ethical facts, and this further implies that those ethical claims must therefore be objectively true. The logic is pretty straightforward: if it is objectively true that there are social institutions (S) which generate false beliefs (P) about other (or the same) social institutions (O), and if we can logically derive an ethical claim from these objective facts, then it would seem to be objectively true that we ought to get rid of those social institutions, irrespective of what any person or social group might believe about the issues. But it is quite clear that this is NOT a tenable conclusion, because reading two is itself false: the ethical conclusions of Bhaskar’s explanatory critique depend on ethical as well as factual premises, so even if the factual premises are objectively true there is no basis to conclude that the ethical conclusions are also objectively true.
While this argument may have been a little technical for a blog post, I think it is important to clarify these distinctions. I regularly encounter (and read) fellow critical realists who cite Bhaskar’s theory of explanatory critique as support for ethical naturalism and moral realism. I suspect that some of them have been seduced by the appeal of the first reading above into believing that this justifies the second and third readings as well. It does not!
This work by Dave Elder-Vass is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
This (and the post before it) in my view spreads more darkness than light. Your critique of Bhaskar’s position is transcendent, not immanent, i.e. based on a travesty of his views.
First, your notion that Bhaskar’s moral realism holds that human values are ‘right, good, and true independently of any human or social reality’, with ‘absolute justifications’, is absurd (fake news?). What kind of nutter would nowadays think that? When an important thinker appears to be uttering an absurdity, the probability is always that the fault is in yourself and not the thinker – especially when, as in this case, the thinker runs a critique of epistemic absolutism throughout his oeuvre. All Bhaskar’s arguments are conditional and from geo-historically relative premises; nothing is absolute. Your own position that people as a matter of fact have certain basic rights and needs that arguably should be met is actually a species of Bhaskar’s moral realism, but you don’t notice the contradiction. This is because all your talk of ‘automatically justifying’ or ‘overriding’, ‘set in stone’, etc. completely ignores the conditionality of Bhaskar’s arguments and his deployment of the CP clause. Dan Little, to whose blog you refer, does the same.
Second, you foist onto Bhaskar your own, arguably false, positivist understanding of what a ‘fact’ is, namely, that it is completely free of value-commitments. In doing this you forget that social reality is suffused with values and that in your own everyday and professional practice you regularly criticize beliefs as inadequate, misleading, etc. and on that basis give up those beliefs (CP) and accept that practices informed by those beliefs should be changed (CP). People move from facts to values in this way all the time. Hume notices this in the Treatise, but says he can’t find any grounds for it. This is because, in thinking of facts as value-free, he extrudes himself and his beliefs from the (social) world – exactly what you’re doing. Hume says he prefers the destruction of the whole world to the scratching of his finger, forgetting that, if the world goes, so does his finger, together with his philosophy. It’s not an objection to Bhaskar’s argument to say, as you do, that his premises include a value-commitment to truth. Commitment to truth is a condition of possibility of factual discourse (‘an aspect, as it were, of the logical geography of the concept of a belief’). You fail to notice that Bhaskar completely recasts the positivist understanding of what a fact is: a fact incorporates a commitment to truth. ‘The positivist account of a fact was always in fact false.’
The sentences in quotes come from Bhaskar, Enlightened Common Sense. This gives an accessible and clear account of Bhaskar’s views and arguments on this subject (esp. pp. 95-99, 139). It’s important to bracket your own views and try to understand it from the inside before launching into critique.
Mervyn, Thank you for your thought-provoking comments and your pointers to Bhaskar’s discussion of these issues in Enlightened Common Sense. My original posts were based on quotes from his earlier work but those pages in ECS seem to confirm my argument rather than undermine it.
This is particularly clear on the question of moral realism: Bhaskar writes in ECS (p. 139) that “Moral realism holds that morality is an objective (intransitive) property of the world. Ethical naturalism grounds a distinction, within moral realism, between the domains of actually existing human morality… and the moral real (moral alethia or object/ive) of the human species… which explanatory critical philosophy and science may discover”. So it is clear that he is committed to a moral real that is independent of “actually existing human morality” and that can somehow be discovered, and the word alethia, as I understand it, implies that he thinks that the moral real is also true in some sense. There doesn’t seem to be anything conditional or geo-historically relevant about that other than our capacity to discover it.
[my full reply is over the limit for comment length so I will break it here]
On facts and values, the issues are more complex. As you say, Hume thinks of facts as value-free, but this means that Hume’s Law – that we cannot move directly from facts to values – can therefore only be read as meaning “we cannot move from facts that are value-free to values”. As you also point out, however, Bhaskar’s innovation is that he claims that there are no value-free facts in the first place. For Bhaskar to claim that he has refuted Hume’s Law would be incoherent if what he means by that is that he has shown we can move from facts, understood as already including values, to values. No-one (including Hume, as far as I am aware) disputes that we can move from claims that already have a value component to other value claims. Still, if there were no value-free facts in the first place then Hume’s Law would be irrelevant and so we need to look at Bhaskar’s claim in a different light.
He bases the claim that there are no value-free facts on the argument (which extends Habermas’s argument about ideal speech situations) that “the value that truth is a good and falsity is an ill… is a condition of factual discourse (an aspect, as it were, of the logical geography of the concept of a belief) and so it does not involve anything other than considerations intrinsic to factual discourse to legitimate the deduction of values” (ECS p. 98).
There are a number of ways to read this, but let me focus on what seems to be the most plausible: the argument that for factual discourse to flourish it is necessary that most factual discourse is truthful, and that we need factual discourse, hence it is instrumentally necessary that we adopt the value that truth is good. Now this is an argument about what is involved in the practice of making factual statements. So the content of a typical factual statement would not imply that truth is good (and so in terms of their content there could still be value-free facts), but the act of making this or any other statement would do. This might look like an example of a derivation of a value (truth is good) from factual premises, and of course we only need one good example to refute Hume’s dictum. But, like all instrumental arguments to value conclusions, this deduction only works if it has a further value premise – the fact that we need factual discourse only entails value conclusions if we take it as a further value premise that our needs should be met. So Hume’s dictum is not refuted. And it seems to me that value-free facts, in the sense necessary for Hume’s argument to work, have not been ruled out either.
As in the main blog, it’s important to distinguish the interesting kernel of Bhaskar’s argument here from the illegitimate use that he makes of it. The argument that we need factual discourse and that for factual discourse to work it must generally be truthful is a good reason for societies to adopt the value that truth is a good thing. It is also interesting as a speculation about the social evolutionary causal factors that have led us to think that truth is a good thing. But it is not an argument that provides moral certainty (“deduction”), and it is not an argument in which a value is deduced from a purely factual premise.
As you say, social reality is suffused with values and we do criticise beliefs as inadequate or misleading and we do argue that they should be given up. But those criticisms depend on value assumptions that cannot be derived from purely factual claims. Bhaskar clearly argues that we can deduce values (at the end of the quote from p. 98 above) and this appears to be what he means when he says that we can discover the moral real by philosophical methods. But the claim that facts necessarily entail certain values can’t be sustained, and so we have to give up the scholastic practice of trying to deduce values. Instead we must go back to the far more human and social practice of debating our values, offering good reasons to support them, but also taking account of other points of view if we are to establish sustainable and workable values.
On the issue of ‘deduction’, ‘certainty’, ‘scholastic practice’ (!) etc., you badly need to bone up on the nature of Bhaskar’s transcendental arguments – see McWherter, Morgan, etc. The conclusion of a TA follows with logical necessity, but the premises may be challenged, change, etc. So the conclusion is conditional and relative, and the position argued for makes no claim to be the only posssible one but the best currently available. There’s nothing ‘scholastic’ about Bhaskar’s TAs whatsoever; they’re adapted from Kant, who’s hardly a scholastic philosopher, and Bhaskar clearly and abundantly distinguishes his metaphysics from first philosophy and runs a powerful critique of foundationalism and infallibililism thoughout his oeuvre. Its a bit much – and poor scholarship – when a major thinker persistenetly stresses that his ideas are relative and fallible and then a maestro like you comes along and says in ignorance, ‘Oh, but he’s an absolutist.’ ‘Transcendental’ isn’t even in your lexicon, is it? You don’t seem to be able to recognize a TA when you see one.
You’re still going on at the end about ‘purely factual’ in Hume’s sense. But as you concede at the outset, there’s no such thing – like phlogiston – so it’s irrelevant. So we need to figure out what a fact actually is. You get Bhaskar completely wrong on this (as often). He’s not at all trying to derive the value that truth is a good from facts, and there are not ‘a number of ways to read’ what he says on the topic (the muddle is in you, not him) – he clearly indicates that commitment to truth is a [transcendentally] necessary condition of possibility of a fact. You can’t discover facts unless you are committed to finding out truth, making a real discovery. That can surely be accepted fallibly. So a fact contains a value commitment to truth as part of what it is. But that’s the only value commitment it *necessarily* contains, and we accept this in our daily practice: when we take penicillon for a bacterial infection, we don’t suppose that the science that discovered it was necessarily contaminated by a desire to attract research funds. From facts understood in this more adequate way we can indeed derive values that we should act on, other things being equal. And Hume, who thought that you couldn’t logically derive values from facts was wrong because his concept of a fact was false.
A simple example of the transition from facts to values at the macro level. ‘Racisim is false (scientific fact). ‘Formal structures of segregation causally generate racism’ (scientific fact). ‘Therefore segregationist structures should be dismantled, CP’ (value conclusion). They have in fact been dismantled across the planet in our lifetime. The causation was complex, as it always is at the macro level, but the new scientific understanding of racism as false undoubtedly played an indispensable role. Why are you so keen to downplay or eliminate that role? Contra Elder-Vass, nobody’s claiming that scientifically derived values override everything, that other points of view shouldn’t count, etc. – that’s why the CP clause is there, as it is in all cases of application of science in an open world (e.g. ‘this building is good for withstanding hurricanes’). Today natural scientists, see e.g. the science of climate change, are far better at understanding the logic of moving from facts to values than social scientists, who by and large play an obfuscatory role. Heaven help us. We’re heading for catastrophe on a planetary scale, and you say science is useless, we can’t draw value conclusions from facts.
Let me focus on your central point. You are right that transcendental arguments are fallible, but they are not only fallible because the premises may be false. They are also fallible because the 'necessary' connection posited between the premises and the conclusion may be flawed. In particular, there may be other conditions in which the premises may come about.
For example, you say "commitment to truth is a [transcendentally] necessary condition of possibility of a fact. You can’t discover facts unless you are committed to finding out truth, making a real discovery". But that argument fails if it is possible to discover facts on the basis of different epistemological commitments. For example, a pragmatist could discover facts on the basis of a commitment to finding out knowledge that works, regarding it as irrelevant and perhaps even meaningless to consider whether it is 'true'. We don't have to agree with pragmatist epistemology to see that this means that commitment to truth is NOT a [transcendentally] necessary condition of the possibility of a fact. So if this claim is the basis of Bhaskar's theory of explanatory critique the whole argument falls apart.
Incidentally, there are several points on which you misrepresent me in your reply. I do not 'concede at the outset' that there is no such thing as a value free fact. Nor do I say that science is useless, and my argument is perfectly consistent with making critical use of science – we just have to acknowledge that the value elements of a critical argument cannot be produced by science.
Are you a pragmatist or a critical realist? It’s a condition of possibility of a fact as understood and argued for by critical realist philosophy, as I made clear. Why do you want to cut off the branch on which you’re sitting?
You’re obsessed about that ‘necessary connection’. But it’s only logical necessity. Surely you’re not questioning that there is such a thing? Perhaps you’re worried that it might grab people by the throat and make them do things? The conclusion of any valid deductive argument (the least interesting part of transcendental analysis) follows with logical necessity from the premises. The only question then is whether the premises are true. If you don’t accept them – in particular, in this case, that science makes partial but real discoveries – you won’t be bound by the conclusion.
Let me make the meaning of my final sentence, which you misconstrued, clearer: ‘We’re heading for catastrophe on a planetary scale, and you say science is useless [in this context], [in that] we can’t draw value conclusions from [scientific] facts.’ Real scientists do it all the time.
Wow, Mervyn, that's remarkable. Someone blows a massive logical hole in your argument and your response is to lecture them about logical necessity!
Let me just deal with the question of logical necessity. Transcendental arguments are often expressed as if they have a single premise, and a deduction made from that premise. My comments in my previous reply apply to that way of framing them. But you are right about logical necessity if we frame a TA like this:
Premise A (factual premise): X is the case
Premise B (condition premise): Y is a necessary condition of X
Conclusion: Y is the case
If A and B are true then the conclusion follows with logical necessity. But that doesn't undermine the substantive content of my point: transcendental arguments are fallible because either A or B could be false. My argument about pragmatism demonstrates a case where B is false. As I said, we don't have to agree with pragmatism to see that, and I don't!
Incidentally, you appear to be trying to shift the ground of your argument by introducing the point that "science makes partial but real discoveries". That's premise A in Bhaskar's central transcendental argument in the Realist Theory of Science but it's not the one that you invoked in your earlier argument for explanatory critique.
Wow, Dave, I'll leave it there thanks. You really do need to bone up on Bhaskar's transcendental arguments. When you do you'll perhaps see that it's illicit to drag a pragmatist notion of a fact into a CR transcendental analysis of the fact form (why not bring in the notion of a fact as whatever you believe?), and much else besides. Here's an up-to-date place to start: McWherter, Dustin. 2017. ‘Revisiting metaphilosophical naturalism and naturalized transcendentalism: response to Kaidesoja’. Journal of Critical Realism, DOI:
Your position is one big contradiction: you’re denying in theory what you affirm in practice all the time. Why would you criticize my belief that values can be inferred fallibly from scientific facts if you didn't think I should therefore change it, CP, and any action it informs, CP? You're not following arguments where they lead, you prefer to muddle along observing the positivist taboo – sort of. Oorevoo.
Dave, I know this is in response to a conversation that is four years old, but you are (or were 4 years ago) a very patient man.