From Natural (or Naturalized) to Social Epistemology

post by HalMorris · 2014-08-06T00:48:44.425Z · LW · GW · Legacy · 4 comments

Contents

4 comments

   I've been reading an anthology called Naturalizing Epistemology (1986) edited by Hilary Kornblith.

   "Naturalizing" epistemology has been heavily identified with W.V.O. Quine (author of the 2 first articles in _Naturalizing Epistemology).

   Others draw parallels between naturalized epistemology and the much earlier philosophy of pragmatism, or John Dewey in particular, as in Transactions of the Charles S. Peirce Society Vol. 32, No. 4, Fall, 1996, "Dewey, Quine, and Pragmatic Naturalized Epistemology".  Or see Stich 1993 "Naturalizing Epistemology: Quine, Simon and the Prospects for Pragmatism".  The title alludes to Herb Simon, who is no doubt better known to lesswrong.com than most philosophical epistemologists.  [For some good articles and quite a few broken links, see: http://www.ucs.louisiana.edu/~kak7409/EpistPapersBySubject.html] Naturalized epistemology, like many other intellectual approaches (to whatever) has a strong and a weak program, or position.  The strong might be represented by Quine's "Why not settle for psychology".

   Basically, I think general naturalized epistemology aims to ground epistemology in something solid and material like people, and their scientific study -- as opposed to reasoning with purely mental constructs.  Another tendency that claims to be "naturalizing" epistemology is to study how "good reasoners" arrive at what they think is the truth, and this may mean trying to rigorously define how scientists think.

   "Why not settle for psychology" is to pass the buck or forward all questions to another department (e.g. psychology, sociology, history of science), as if the disciplinary traditions of philosophy have nothing to offer.  Do they really have nothing to offer?

   I think one way to not pass the buck is to focus on certain habits that seem to affect, or afflict, virtually all of philosophy when it deals with thoughts, truth, etc.  Namely to talk as if our subject is some "canonical knower", talking of what "is known" without reference to any particular knower, seeming to forget about the fact that I am in my mind and you are in yours.  Descartes' "Cogito ergo sum" might be more naturalistic if he had written "So you think maybe you don't exist?  But then aren't you experiencing something, an awareness of words and/or pictures that seem to be inside your body.  Call that your 'self' and you can't avoid thinking it exists in some form, whether as a disembodied spirit cohabiting with a body, or a biological process, or a computer simulation.  This only works for your existence.  It won't convince you that I (Descartes) -- in fact I might be long dead as you read this."

   The main move in "naturalizing" epistemology without passing the buck is, in my opinion to keep grounded in the realization that there is only me knowing, or you knowing, or either of us believing mistakenly, and there are the processes by which we came to know or believe.  And the canonical knower is a fiction, and declarations like "it is known" or "it is knowable" are just to unclear.  This grounds me in the realization that the vast bulk of what I think I know is due to having gotten it from some trusted source.  It used to be trendy to say that would make me an "authoritarianism", but if "authoritarianism" is a real thing to be avoided, it can't just be something we all do because there is no alternative.

   There may be a "right" way to establish a scientific fact, but in almost all cases, hardly one person in a million has actually witnessed it being established.  The vast majority "know" it because they read it in a book.

   So we are left mostly with the sources we have chosen to trust, and the question of what can justify that trust.  I expect most readers of LW believe they do a good job or determining who to trust, and we all know people who we thing don't do such a good job.  Alvin Goldman, the only other author besides Quine allotted 2 articles in Naturalizing Epistemology is now (some years after the book was published) the best known proponent of one of two conflicting schools of Social Epistemology.

   I want to suggest if you approach naturalized epistemology right, then social epistemology is a natural outcome. Goldman treats the question of "Who to Trust" seriously in "Experts: Which Ones should you Trust?" in the journal Philosophy and Phenomenological Research, Vol 63, No. 2 (7/2001)

   [NOTE: many online papers by Goldman are generously provided at
        http://fas-philosophy.rutgers.edu/goldman/Papers.htm]

   If one were to teach general principals of knowing who to trust in K-12, what would these principles look like?  The following examples involve social reasoning as well as the intentional stance beloved of evolutionists like Daniel Dennett.

   One principle might be "look in the back of a magazine to see if its ads are directed to really gullible people -- if so, suspect the rest of the magazine is directed to really gullible people.  I suspect a great deal of our ability to discriminate trustworthy sources is based on somewhat similar rules of thumb.  So if the magazine that advertises the "X-ray glasses" says we will all be flying around with jet packs in a couple of decades (a typical example from the 1960s), enable bullshit detector.

   You might move to a new location, and at a block party, ask around about who is a good plumber or mechanic (on in some areas, where is it safe/unsafe to walk at night).  Somehow, I think most of us can do a reasonable job of deciding who to take most seriously and who is perhaps a blowhard.  Could that be taught in school?  There are few more important life skills.

   If on some momentous controversial issue, an advocate of some position sends me article after article that makes me ask "Is that the best they can do?  Is an  85 year old retired atomic scientist the best they can do to impress me with the case against Global Warming?" and similar questions depending on the article, this leads me to conclude that the supposed case they have against Global Warming is ginned up, and until I start hearing more impressive arguments, I will continue to think so based on analysis of what they have to say for themselves -- not because somebody else tells me they're full of shit.

  My knowledge of the literature is uneven, and acquired all on my own motivated by a sense that something is breaking down in terms of people's common sense about what venues to trust, and wondering what has brought this about and what to do about it.  It is for the same reason that I'm interested in Lesswrong.com

4 comments

Comments sorted by top scores.

comment by Sarunas · 2014-08-12T11:09:23.124Z · LW(p) · GW(p)

It is an intriguing topic, thank you for posting about it. I haven't read papers by Goldman yet, but I'll try to take a look at them. Some Lesswrong posts that might be somewhat related: Some Heuristics for Evaluating the Soundness of the Academic Mainstream in Unfamiliar Fields , Scholarship: how to tell good advice from bad advice? , maybe tangentially even some posts like this. In addition, potentially some posts on Overcomingbias might be relevant, as well as some posts on Slatestarcodex, e.g. Searching For One-Sided Tradeoffs. However, as you can see, all these posts are scattered here and there and do not seem to form any coherent body of ideas (well, there are some recurring themes on Overcomingbias).

It seems to be such a broad topic, it would require dozens (if not hundreds) Lesswrong posts to do some justice to it. E.g. one could probably expect different answers when analyzing what would be the best strategies for a given individual to follow versus what strategies would lead to "average wrongness in a society" being as small as possible? (well, I'll have to think about it, at this moment this is just a guess). It seems to me that natural epistemology is less likely to contain such tensions (again, a much deeper analysis is required to say whether this is actually true).

Moreover, it seems to me that it is much less likely that a single unifying framework (similar, to e.g. Bayesian inference, that was a recurring theme on Lesswrong) even exists. Again, I should probably read Goldman's papers before saying anything more.

Replies from: HalMorris, HalMorris
comment by HalMorris · 2014-08-13T03:24:40.951Z · LW(p) · GW(p)

Having looked at "Some Heuristics for Evaluating the Soundness of the Academic Mainstream in Unfamiliar Fields", I'm not convinced. I am grappling towards my own ideas for heuristics in this essay: http://ontologicalcomedian.blogspot.com/2010/07/what-is-machine.html Alternatively, I would ask the question whether the blind men are really feeling up an elephant or not. Perhaps one really is caressing a huge floppy leaf and another is hugging a tree trunk, etc. In some disciplines, I would say there really is an elephant, ergo explorations and comparing of notes will tend to converge on some solid picture. In other fields, such as literary criticism, I doubt that there is an elephant at all; people are just grabbing this and that leaf, vine, branch, tree trunk or rock and, and out of them imagining mythological animals. A lot more should be said about why academic disciplines work when they work. I'd say peer review and the other academic machinery do a better job than most any other arrangement (such as think tanks funded by people whose real interest is in promoting ideological points) -- as long as there really is an elephant. Where there is no elephant, or nobody has really found it yet, all the peer review in the world and/or attempts to mimic physics won't prevent it becoming a factory for turning out "fashionable nonsense".

comment by HalMorris · 2014-08-12T18:35:06.402Z · LW(p) · GW(p)

Thanks. I'll have a look at the links you provided. I haven't found any work in Social Epistemology that was up to my hopes and expectations, but not treating at least half of epistemology in a social context seems like utter blindness. Goldman is good at laying out what should be included in SE, including a "systems oriented" branch, and perhaps less good at realizing the program.

comment by solipsist · 2014-08-06T02:34:34.515Z · LW(p) · GW(p)

I expect most readers of LW believe they do a good job or determining who to trust...

Possibly. Then again, some readers are so self-skeptical that they question the veridicality of their senses multiple times each day.