21 October 2018

Fact and Value: The False Dichotomy

Image credit: The Guardian.
Posted by Thomas Scarborough
The fact-value distinction is one of the most important problems of philosophy.  The Scottish philosopher David Hume gave it its classical formulation: it is impossible to derive an ‘ought’ from an ‘is’.  That is, it is impossible to establish any value amidst an ocean of facts. 
On the surface of it, Hume would seem to be unimpeachably right.  The facts cannot tell us what to do.  But here is a problem.  Neither can value.  While one cannot derive an ‘ought’ from an ‘is’, neither can one derive an ‘ought’ from an ‘ought’, as it were.

Take, as an example, a statement of fact: ‘We are ready to hoist the spinnaker.’  Such a statement gives us no idea as to whether we should hoist the spinnaker.

Yet we find no difference with a statement of value: ‘We ought now to hoist the spinnaker.’  Why ought we to?  What gives us the authority to say so?  We find that such a statement is quite adrift, and equally unable to tell us whether we should or not.

What, then, was Hume thinking when he wrote about ‘ought’? 

It would seem to me that Hume made a lazy assumption, of the kind that philosophers fail to examine any further, on thinking that they have gained a special insight.  The assumption would be something like this: that there is a certainty which lies in value which fact does not possess.  Thus Hume equated value with a ground for our behaviour—if one should ever find it. 

To put it another way, Hume’s fact-value distinction would seem to be a false dichotomy.

What is it, then, that fact and value have in common, that neither will deliver ‘value’—in the sense of a ground for our behaviour?

I would propose that the scope of both fact and value is too limited for either to deliver universal truth.  Both statements of fact and statements of value exist in limited contexts, without being referenced to any fixed points except their own—while the question of certainty lies beyond this, in something which is far more expansive.

It may be easy to see, for instance, that I should hoist a spinnaker if I wish to win the race—and this I may state both as an ‘is’ and as an ‘ought’:

Given such and such conditions, the spinnaker will secure a win—alternatively, I ought to hoist the spinnaker to clinch it.  In both cases, it would be true and compelling that I need a spinnaker—yet not if I should expand my horizon, to ask whether I should have entered this race at all.

How then might we reference statements to something broader than simple fact and value?  An analogy might help.

I am in a boat on the ocean, to anchor a buoy.  If I reference its position to the seaweed I see underneath it, or the birds which circle overhead, I have in this case an unstable reference.  Or I may reference it to a spit of land that I see in the distance.  This would seem to be more stable, though not completely so—the wind and the waves may change it.  Or I may reference it to the stars—but even the stars will move. 

Ideally, my buoy would be referenced to everything.

This may not be as absurd as it sounds.  If the context is big enough—and if we should know just what kind of a context this should be—we may well be able to ground both fact and value.

If we reference everything to everything, there may be a way forward.  While space does not allow me to explore this further here, readers may refer to a post in which I sketched some thoughts on how this might be done: How Shall We Re-Establish Ethics in Our Time?

10 comments:

  1. I do think the question is a good one, (thank you Thomas for introducing it) but this discussion lost me around the analogy - and surely we should not end by telling readers to go and read more elsewhere!

    Ah, there we have an ought, or rather an 'ought not'. But at the same time it is also an 'is', a claim about the world. Thus "Blog posts for Pi should be completely comprehensible (modular) and not require or request that readers read a bit more on the subject somewhere else.

    Let me try to generalise this. If I say "eating horsemeat is wrong" I am also saying "you should not eat horsemeat".

    Okay, you might say "is wrong" is a special case, but if I say "eating horsemeat is cruelty to wonderful, gentle animals" then again the line between the 'is' and the 'ought' seems to be vanishingly fine.

    ReplyDelete
  2. Thank you, Thomas, for shedding interesting new dimensions on the ‘is’–‘ought’ puzzle that David Hume famously laid out.

    For what it’s worth, I’d like to offer a possibly complementary take on Hume’s position: In particular, might ‘is’, ‘ought’, and ‘should/will’ compose a sequence?

    Here’s a scenario that might make more concrete what I mean:

    An Aztec high priest, venerated for boundless wisdom, says to his people, ‘The gods demand that we sacrifice our enemy captives by cutting out their hearts, to offer the gods the highest of all honours, human blood’. (That’s the ‘is’, or statement of fact, namely, the Aztecs’ religious creed.)

    The Aztec high priest then says, ‘We therefore have a moral obligation to sacrifice our enemy captives and rejoice in human blood, to properly honour the gods’. (That’s the ‘ought’, or statement of moral belief drawn from the gods’ hallowed demand for human sacrifice.)

    The Aztec high priest then sacrifices a succession of enemy captives over a sacrificial stone for their blood, which the high priest dedicates to the gods — the conviction being that the ritual will result in the gods delivering a better harvest of crops, good weather, and increased strength in his own warriors. (That’s the ‘should/will’, or acting out of the consequences of ‘is’ and ‘ought’.)

    Moving beyond the Aztec model, such a sequence might conceivably end at ‘is’ (a fact, which need not lead to anything more); or it might end instead at ‘is’–‘ought’ (fact segueing to moral agency); or it might conceivably proceed to ‘is’–‘ought’–‘should/will’ (fact segueing to moral agency segueing to action taken).

    So, with deference to Hume, might there in fact be situations where it’s possible to derive an ‘ought’ from an ‘is’ and a ‘should/will’ from an ‘ought’?

    ReplyDelete
  3. Thank you Martin and Keith. Point taken, about posts being self-contained. Hume's view may be represented like this (top), followed by my own (bottom):

    • fact vs. value and ground
    • fact and value vs. ground

    The core issue in my post is that neither fact nor value can yield what Hume called 'ought'. If this is true, then secondary issues arise: why are both fact and value incapable of providing a ground for our behaviour, and how might one find such a ground.

    There has been, I think, a lot of promising yet unconvincing debate around possible similarities between fact and value. The analogy in my post is intended to show that we use fact and value in limited contexts, and that their ground vanishes when we think beyond those contexts.

    As for what fact and value might have in common, take physical science as an example. Newton told us: the world is thus and thus. What he really believed was: the world ought to be thus and thus. We know now that his 'is' was an 'ought'. This is however a limited example which will not meet all objections.

    Anyway, if 'the question is a good one' as Martin suggests, then we may have redefined one of the biggest problems of philosophy.

    ReplyDelete
  4. First, on Keith's Aztec example,

    "‘The gods demand that we sacrifice our enemy captives by cutting out their hearts, to offer the gods the highest of all honours, human blood’. (That’s the ‘is’, or statement of fact, namely, the Aztecs’ religious creed.)"

    - but it looks to me exactly as an 'ought' statement. "The Gods say we ought to sacrifce..." etc

    That leads me to my second query, this time of Thomas. Now I may be 'missing the point' but then perhaps the point needs re-explaining. Thomas asks: "why are both fact and value incapable of providing a ground for our behaviour" - but isn't an ought statement EXACTLY the reason. Isn't that waht we mean by 'ought'. "You ought not to write complicated posts" means at the same time "it is bad to write complicated posts" and "the correct, virtuous path is to not write compliacted posts".

    ReplyDelete
  5. Ah! But where do you put the idea of complicated, on which ground? Why are both fact and value incapable of providing a ground for our behavior, and how might one find such ground? Our lives evolve within limited contexts. Though many different limited contexts do not adhere to similar facts and values. There is no common ground for an ought.

    ReplyDelete
  6. I get your point, Martin. Thanks for the observation. What I was trying to say in the first of the three propositions was really little more than ‘To the Aztecs, the gods exist’. In other words, an ‘is’ or fact. It seems that by my having elaborated on the proposition — perhaps inelegantly — the nature of the proposition shifted, as you point out. If, however, I were to pare the proposition down to my original intent — that is, to some simple variation on ‘To the Aztecs, the gods exist’— it seems to me that I’m back to the statement of fact (‘is’) I was really searching for. In which case, the pared-down first proposition might better support my ultimate conclusion regarding ‘is–ought–should/will’: that is, it’s possible to derive an ‘ought’ from an ‘is’ and a ‘should/will’ from an ‘ought’.

    ReplyDelete
  7. Dear editor, I am interested in publishing a short article (1500 words circa) regarding the Common Good and Nash equilibrium. Would that be of any interest to you and/or your readers? Thanking you in advance. Alan

    ReplyDelete
    Replies
    1. Actually www.the-philosopher.co.uk does longer pieces, and this sounds a bit too 'formal' for the blog anyway.. You could also send it to us full length to see if it might be right for that purpose.

      Delete
  8. Sounds interesting, yes. However we ordinarily limit pieces to 700-800 words. Use a 'conversational' style, something that an 'educated public' would find comfortable reading. You can send it to me, Thomas Scarborough and I shall copy it to my co-editor Martin Cohen. We acknowledge receipt within a few days.

    ReplyDelete