> On Mar 29, 2015, at 2:40 PM, Benjamin Udell <[email protected] > <mailto:[email protected]>> wrote: > 1. Peircean abductive inference is not Bayesian at any level. Insofar as you > involve Bayesian subjective priors at all (even if not on all levels), your > idea of abduction (or at least of how abduction normatively ought to work) > differs from Peirce's. You're allowed to disagree with Peirce, of course, I'm > just pointing out where you diverge from Peirce.
I’ve missed most of the discussion, but we should note that Peirce was a frequentist and not a bayesian in terms of statistics. I’m not sure that entails we can’t find bayesian inference useful in certain contexts. I definitely use it. But I’m not sure it has an inherent meaning different from any other computational tool for prediction such as doing large inner products to find a shortest distance in some multidimensional space. (This is the typical approach to find relevance in search queries) It’s interesting to me that Peirce doesn’t attempt to really “solve” induction or abduction but leaves them fairly open beyond a general technique. Others here might disagree but I suspect we could arrive at a pseudo-abduction using training of various sorts in neural networks or other complex treatments. I’m not sure this would be the same as abduction but would seem to my eyes to be closer to induction given the relationship to past inputs. Honestly computational induction and abduction are interesting but I don’t find the taxonomical debate too helpful ultimately. To me it seems like abduction in Peirce inherently appeals to non-computational practices and judgments like aesthetic value and human practices such that true abduction intrinsically isn’t computational. Beauty and simplicity (which you mention below) apply to our adductive judgments but I don’t think can be treated computationally except via a simplified representation. > 2. You discuss the need for abduction as arising in changes of situation or > in lack of understanding of how to solve a problem with which one is > confronted. I guess that that is close enough to Peirce's idea that abduction > is the first step of inquiry and arises in situations of surprise or extreme > complication. > 3. You have abduction guessing particular circumstances and general rules in > connection with particular circumstances. Close enough to Peirce. > 4a. You don't discuss abduction as inferring through an icon or likeness. > Can't evaluate your treatment of abduction in that respect. > 4b. You don't discuss intrinsic value of an abduction - in particular, > natural simplicity, plausibility. Can't evaluate your treatment of abduction > in that respect. > 4c. You discuss the testing of abduced hypotheses. That's Peircean enough in > a vague way. You don't discuss the selection of hypotheses for testing, > either for the tests' cheapness or for their strategic value (breadth, > caution, incomplexity) in expediting inquiry. You don't discuss abduced > hypotheses in terms of the (pragmatistic) emphasis on conceivable > (imaginable) practical implications, although more generally you do see > cognitions as implicit in 'action' modules, which I take to be like > dispositions for conduct. Can't evaluate your treatment of abduction in those > respects. > > Now, I've already discussed 4b and 4c in previous messages, and have thought > that perhaps you've regarded them as implicit in your discussion of abduction > or elsewhere in the paper, so I was figuring that you would reply by > explaining how you have them covered. Maybe I should have explicitly asked > you! > > If you have not dealt with 4a, 4b, and 4c in your treatment of abduction, > then I'd say that your idea of abduction is vague enough in some important > ways with regard to Peirce's, that one could say that it's partly like > Peirce's, and possibly like and possibly unlike Peirce's idea in ways that > you have not made explicit or that you haven't addressed in your paper. > > Best, Ben > > On 3/29/2015 1:25 PM, Danko Nikolic wrote: > >> Dear Ben, >> >> On 29/03/15 17:42, Benjamin Udell wrote: >>> In reading your links, I didn't seen an emphasis on abduction's >>> plausibility as naturalness and simplicity, though you did seem to accept >>> such ideas in one of your messages to peirce-l. >> >> I discuss abduction in section 3.4 entitled "Anapoiesis and abduction -- the >> mind" of the following paper: >> >> http://www.sciencedirect.com/science/article/pii/S002251931500106X >> <http://www.sciencedirect.com/science/article/pii/S002251931500106X> >> >> I am seeking feedback on that.
----------------------------- PEIRCE-L subscribers: Click on "Reply List" or "Reply All" to REPLY ON PEIRCE-L to this message. PEIRCE-L posts should go to [email protected] . To UNSUBSCRIBE, send a message not to PEIRCE-L but to [email protected] with the line "UNSubscribe PEIRCE-L" in the BODY of the message. More at http://www.cspeirce.com/peirce-l/peirce-l.htm .
