Wednesday, July 05, 2023

Evening Note for Wednesday, July 5

 Thought for the Evening: Social Ontology

I don't think it can seriously be denied that analytic philosophy is in a crumbling state today. (There are people who would deny it; they are wrong.) Nonetheless, and also regardless of whatever criticisms one might have of analytic philosophy, there are parts of it that are currently thriving in which immensely promising work has been done. I've mentioned modal metaphysics before; that's the obvious one. But the second most important, I think, is social ontology. 

Analytic social ontology derives primarily from John Searle, who branched out to it from speech act theory. It is the theory of 'social entities'. There are things in the world that seem to be real and have real-world effects but which seem to be mind-dependent: money, corporations, nations, political offices, and so forth. Searle's best discussion of this, I think, is "Language and Social Ontology". He takes the origins of social entities to lie in language. Once we have a language, we have commitments, and insofar as these commitments bind us together with reasons, they are what Searle calls deontologies. Society is built out of deontologies that we recognize as things to which we are committed. For instance, if I tell you something in a matter of importance, we both recognize this as involving a normative expectation of truth-telling; we both, in one way or another, have a reason, independent of our particular desires, to give and to demand truth in such a matter. Searle holds that these deontologies are created by collective intentionality which imposes status functions on things. Collective intentionality is what it sounds like -- it's when we are operating together in the stance of 'We' rather than individually in different stances of 'I'. When we do so, we sometimes engage in what Searle calls Declarations, of which a common form is:

X counts as Y in C.

(Searle originally held that this was the only form, but later was persuaded that some cases require a different from, like 'Y exists in C'.) X is a person or object; C are the presupposed conditions. Y is the status function. For instance, pieces of paper count as dollars when they are printed in a certain way by the Mint acting according to statutory and regulative authorization and other conditions. A person counts as the President of the United States on condition that in an election they have received the appropriate number of Electoral College votes as counted by Congress and have not completed a term of office. And so forth. (It's worth noting that both of these examples show the layering of social entities: the Mint, the Electoral College, and Congress are all themselves social entities on which other social entities depend.) Because these Declarations are put forward by general acceptance (collective intentionality) operating within linguistic deontologies, the social entities created by them have deontic powers, which are powers to change social relations as determined by those deontologies. Dollars can get you doughnuts; Presidents can issue commands to the United States military. 

Searle's account is still the general reference point. Almost every part of it has been criticized; but no criticisms have tended to gain general acceptance. Attempts to replace it often also fail; the best attempts are those of Francesco Guala, who proposes that social entities actually are just systems of rules and incentives in equilibrium states. A lot of Guala's work is excellent -- he is the second most important person in the field, without any doubt -- and many of his criticisms of Searle are at least roughly right, but I think his general idea is doomed to fail; when you look at his examples, what he is calling 'incentives' often obviously are explained by social entities rather than explanatory of them. In any case, that's a complicated question.

I think there are a few criticisms of Searle that are particularly important, though. One is that he regularly conflates statuses (which are a matter of how we classify things) and status functions (which are a matter how we use those classifications). These are pretty clearly distinct. When we have 'X counts as Y in C' structures, the Y is often really a status, not itself a function. Second, he assumes that there are no functions in nature, which means that there is a hard division between social ontology and physical ontology. Searle's view of functions is idiosyncratic, and I think it causes problems for his understanding of what kinds of things can be status functions. One of the reasons Guala's rules-in-equilibria account works as well as it does is that it makes no such presupposition about what goes into making status functions. Related to this, Searle ties social ontology entirely to language. The problem is that there seem to be social entities that are prior to language (society itself, for instance), social entities that are created by non-language-users (animals marking out territory, for instance), deontic commitments that are prior to rather than posterior to language (like those involved in reason itself), and originations of social entities that don't involve, even implicitly, anyone declaring anything at all (like an uncrossable body of water being a boundary). Language is obviously the means by which we deliberately and artificially create social entities, and is obviously important for having extremely complicated social entities, but there are other kinds of social entities, and it isn't really clear that Searle's account, inspired by speech act theory, can wholly do justice to them. Social scientists (represented philosophically by Guala) also often complain that Searle's account is useless for the kinds of things that social scientists do; while I don't have a huge amount of sympathy for this (one might as well complain that social science theories aren't very helpful for what philosophers of language do, which is just as true), it is nonetheless the case that attempts to use an account like Searle's to describe things like money keep failing -- you get initially plausible descriptions that break down completely when you look at how money actually has to work in real life.

All of this is interesting (and since all of this has to do with ens rationis, the scholastic in me can't help but be amused at analytic philosophers yet again winding their way back to where scholastic philosophers were by the seventeenth century). I think a recurring problem in the field is the attempt to be 'naturalistic', which trips up much discussion in weird ways. Searle thinks his account is naturalistic (it is not, apparently even by his own understanding of naturalism); Guala thinks his account is more naturalistic (it is not, and appeals to more mysterious things than Searle's); everybody thinks the point is to explain how you can have social reality given that only fundamental physical reality is real. But this is absurd; we don't start with fundamental physical reality, but with social reality, and we have to take the latter already to be real in a very definite sense in order to get to fundamental physical reality at all, because experiments and scientific observations and statistical models are social entities. It would make more sense to ask how, in starting with social ontology, we get to physical ontology. I was recently reading a passage by Lavoisier on simple substances -- the starting-point of modern chemistry, which we now usually call 'elements'. Simple substances, he noted, are not necessarily actually simple; they count as simple substances relative to our experimental observational actions. What makes hydrogen a simple substance is not that it is simply noncomposite but that it is so relative to various chemical operations that we ourselves do. Elements are social entities. Of course, they are not merely social entities -- but few, and arguably no, social entities are merely social entities. When you recognize this, we see it is common; after all, all sciences have experiments and models, and even physicists sometimes have to declare something to be a clock under such-and-such conditions. All the sciences involve exploring physical ontology with a social-ontology framework. There is no hard and fast distinction between the physical and the social here, nor do we have reason to treat the social as merely a secondary and derivative way of talking about the physical. Nor, at this stage, do we have to worry about the question of reduction at all; we just need a good account of social entities themselves. 

Various Links of Interest

* Andrew Chignell, Demoralization and Hope: A Psychological Reading of Kant's Moral Argument (PDF)

* Kevin C. Klement, Frege's Changing Conception of Number (PDF)

* Elizabeth Jackson, Faithfully Taking Pascal's Wager (PDF)

* David A. Simon, Copyright, Moral Rights, and the Social Self (PDF)

* An interesting (and data-focused) exploration of a recent case of academic fraud, at "Data Colada":
Data Falsificada (Part 1): "Clusterfake"
Data Falsificada (Part 2): "My Class Year Is Harvard!"
Data Falsificada (Part 3): "The Cheaters Are Out of Order"
Data Falsificada (Part 4): "Forgetting the Words"

* Ragnar van der Merwe, A Pragmatist Reboot of William Whewell's Theory of Scientific Progress (PDF) -- I think this oversimplifies Whewell's theory of scientific progress; notably missing, for instance, is Whewell's influential idea that one of the elements in scientific progress is the dying off of previous generations, and, contrary to what is said at one point, Whewell doesn't think any scientific theories are really complete, even Newtonian physics -- while Newtonian physics, like every highly successful scientific theory, has uncovered necessary truths that we had previously missed, it is partly based on experimental elimination of possibilities which, however, good, may always in the future be qualified by better or newer experiments. It's also an error, I think, to overemphasize how much relativity and quantum mechanics have changed the field -- it is still the case that any adequate physics has to approximate Newtonian physics 'as a limit case', i.e., under relevantly idealized moderate assumptions, just as it is still the case that any acceptable account of planetary movement has to be one that can be modeled with circles and epicycles. A scientific theory that cannot do these things, cannot be adequate to the phenomena; there are real reasons why Ptolemy and Newton did so well. Nonetheless, this is paper is a very good exploration of the issues involved in a good theory of scientific progress.

* Jason Tosi and Brandon Warmke, The use and abuse of 'moral talk': How moral language can impede moral improvement

* Rob Alspaugh, Honor and Glory, at "Teaching Boys Badly"

* Todd DeRose, Semantic compositionality and Berkeley's divine language argument

Currently Reading

Knut Hamsun, Growth of the Soil
Pope Leo I, Sermons
J. R. R. Tolkien (Sibley, ed.), The Fall of Numenor
Nicholas J. J. Smith, Logic: The Laws of Truth