Thursday, September 10, 2020

The Argument from Agreement

The most common argument against moral realism is the argument from disagreement -- moral realism can't be true because there is too much disagreement about morals. Hanno Sauer's "The argument from agreement: How universal values undermine moral realism", in Ratio (27 Feb 2019), attempts to flip this and arguing against moral realism from agreement -- moral realism can't be true because there is too much agreement about morals.

(A1) If moral realism is true, then we would expect a lot of moral disagreement.

(A2) We do not see a lot of moral disagreement.


(A3) Moral realism is false.

The argument from disagreement has a number of problems, not least of which is that moral realism, as such, does not predict anything about moral agreement or disagreement; moral realism is by definition a position on which moral objects/entities/values/facts exist independently of our agreement or disagreement about them. Agreement-convergence or widespread agreement is sometimes used as an evidence for moral realism, but this is due to assumptions about agreement, not about moral realism. The argument from agreement suffers exactly the same problem for exactly the same reason. Sauer just flips the script here, as well, arguing for opposite assumptions about agreement. Sauer also tries to drive a wedge between convergence over time and widespread convergence, but this is somewhat useless for the argument, since both are used as evidence for moral realism, and the purported lack of both are often used as evidence against it. While there is no doubt some difference in the structure of the assumptions used, Sauer really needs to deny both; if convergence over time establishes moral realism, moral realism becomes the best explanation for widespread agreement precisely because it is independently established. It's an interesting attempt to turn common views upside down, but it's more of a clever exercise than anything else. It's nice to have, I suppose, if only to bring up when the argument from disagreement is put forward.

I was interested, though, in this:

Most scientific truths are deeply counterintuitive, and fairly recalcitrantly so. Counterintuitive claims are, by their very nature, unlikely to be believed by many people who haven’t received some sort of training (or indeed indoctrination, as with the counterintuitive teachings of many cults and religions). Consider physics: there is nothing intuitive about the idea of inertia, the relativity of simultaneity, or the mysteries of quantum mechanics. Folk physics, on the other hand, is intuitively compelling but gets it all wrong (McCloskey, Washburn, & Felch, 1983). Consider biology: even today, the idea that natural selection (and other evolutionary pressures) instead of the vastly more viscerally appealing ideas of intelligent design or Lamarckianism remains deeply counterintuitive (Medin & Atran, 2004). Finally, consider economics: economists routinely complain about the fact that the public as well as elected officials fail to grasp the basic workings of the price mechanism, comparative advantage or the nature of public goods. That is because prices and global trade are strange, and difficult to comprehend (Kahneman, Knetsch, & Thaler, 1986).

This is the sort of thing that people often say, but it is not in any obvious sense true. This is an important point that is often not grasped in the case of folk physics: folk physics is ineliminable, not merely because it's 'built-in' in some way, although it may be, but because doing scientific physics presupposes it. When a physicist looks at a dial, he could break it down into a problem in optics, but he rarely does so -- he just looks, and by folk physics assumes things like 'The light is not crossing so as to reverse the reading' (and by folk biology assumes that his eyes show the world, etc.). When people use an apparatus for an experiment, they can and sometimes do analyze it into the underlying physics -- the abstract model of the apparatus as a calculating device -- but a lot of times, they just assume folk physics things about it -- that it will not defy gravity, that it will continue to have object permanence, that volume and mass, conceived in our folkish everyday way, will be conserved. Some folk physics assumptions require correction to get the right answer for particular problems; others are essentially qualitative versions of very general assumptions for which scientific physics has quantitative versions, at least for particular kinds of situations; others are approximately, but only approximately, in the neighborhood of more scientific principles; others are simply wrong because they involve confusion across different domains, or some such. Nor is folk physics the whole story about intuitiveness in matters of physics -- many things that are categorized as folk physics are largely crude and hasty extrapolations from experience and don't rule out their opposites as counterintuitive, and there are things that are counterintuitive to experts but not to non-experts. But most scientific truths are not "deeply counterintuitive", in any case. It's not "deeply counterintuitive" that my mug won't fall through the desk; it's not "deeply counterintuitive" that green light is being reflected from grass blades; etc. And they no more become counterintuitive by translating them into more precise mathematical equations than they would if you translated them into Classical Chinese. There are endlessly many of these truths. To be sure, sometimes the further explanations may introduce something counterintuitive; but that's a long way from getting us to "most scientific truths are counterintuitive".

It's more boring to say, but a vast ocean of scientific truths are intuitive and pretty much what everyone would expect. Sometimes the scientific truths are just more precise, sometimes they are just a little different but in the vicinity of what people would have assumed. Likewise, while scientists often have intense disagreements on the many points on which they disagree, they tend toward agreement on a vast array of things -- they'd have to, because otherwise they would never be able to agree even on the correct description of an experiment or observation.