Pat Gunn (dachte) wrote,
Pat Gunn

  • Location:
  • Mood:
  • Music:

Models and Society

Reflection on a recent set of conversations, possibly of the broken-record sort, but I want to work on getting the phrasing right so I'll say it again even if it's not quite new:I generally don't find mathematical models very convincing when they talk about shoulds in society, from economics to voting. To really have an argument that is good *and* convincing, these things should have at least the following:

  • Good internal consistency (the math-y stuff)
  • Good basis (which includes how well the model lines up with society and what it's trying to model)
  • Value-dependent content more-or-less meets my values (meaning what one is trying to optimise, what one's notion of the public good is, what one's trying to do, other goal-type stuff) - of the three, this is the "convincing" part, the other two are broad paths to a "good" argument.
By and large, arguments about society that are very model-centric fail by my standards in at least the third criterion, and quite often in the second as well. When talking with people who love these models to death, the less careful of them deny that the third is a concern ("there are no values here except the objective values" or "there are no values here until it's applied - it's just math" paired with a failure to ever address that lacking when they start to argue their models as "should"s - both infuriating) and minimise the importance of the second. These types of understanding are seductive in the sense that it might be really cool for these often-beautiful equations to have a deep and undeniable meaning for the real world, seductive for those of us for whom math is a language, hobby, and delight, but this is at best a Potemkin Village, serving more to obscure the subjects with hidden dogma among the bases/axioms/definitions than to enlighten. I would love to hear people actually recite the values and bases of their arguments before making (or alluding to) them, so as to acknowledge that these things are not the final word (or even particularly large contributions) in the "problem spaces" of human affairs. Careful thinking and speaking would help people keep the humility they need to approach these areas carefully.

The above is largely unrelated to last night's conversation. With there, the most important thing to remember is that engineering for a field sometimes is ahead of the science of a field, progressing trial-and-error (or via intuition or weaker but faster variants on the scientific enterprise) towards its goals (note: empirical attempts at verifying induction versus deduction?). If something works, even if one's theory behind it is incorrect or absent, its noncompliance with (or lack of firm roots in) whatever the dominant scientific framework is should not be considered a flaw in it as a method. It is important to know that it works in a real sense (e.g. if in therapudic psychology, that it is not a placebo nor that the gains from it are better explained by chance, the passage of time, or other factors), but if it does, its lack of solid ties to established science is more a challenge for science than a disqualifier for it as a method. (I realise that this is pretty abstract - if you need an example to solidify it, consider psychoanalysis in this light, allowing for the possible criticisms I just outlined, although note that it is not tailored nor unique to that example).

Food pondering: Would making a paste out of black-eyed peas be good for anything (perhaps a salad? a salsa? Hmm)

Amusement: I stopped shaving last thursday as a whim, and as I grow facial hair very quickly, I now have a fairly full beard again (which I'll shave when whim strikes me - it tends to irritate my skin). While walking to the party last night with the (kosher) Moscato d'Asti, a number of Lubavichers, seeing me with hat, beard, and kosher wine, commented (in Yiddish and English) that it was a good choice of bottle. The thing I don't quite get is why each of them said it almost under their breath just as I was passing so I could barely hear them. I wonder if they didn't want to be seen talking to me (obvious to them at least not to be a Lubavicher), didn't want to start a conversation, or if there's more going on than that. People are interesting - as much as the subtle details/layers of meaning in everything we say/do give us such a potential for misunderstanding or lack of communication, they're also beautiful in that if one is paying attention and clever, one can get so much more, and one can say so many things that one can't (or doesn't want to) say directly by using turns-within-turns. I find it equally lovable somehow when people insist that they're only communicating at the most direct level as when they listen and speak using all these subtleties of layers intentionally (even in the knowledge that we don't always get all possible information out of these things and sometimes get false information as well on the further-from-primary layers). Maybe it's that in both cases, the person is aware of the potential subtlety and has decided to acknowledge/react to it, even if the manners of doing so are opposite each other.


  • Triplets

    I was pointed at two articles, and decided to add a third. The topics should be familiar (activism, feminism); the post would be a bit long to host…

  • CMU, the First Amendment, and Indecent Exposure

    Earlier on my G+ stream, I commented on the matter of a CMU student who protested the Catholic church's coverup of sexual abuse by dressing as the…

  • Diversity in Tech

    In the technical meetups I've been part of in NYC, I've been noticing a demographics shift from them being practically all-male (e.g. 1-2 women in a…

  • Post a new comment


    Anonymous comments are disabled in this journal

    default userpic

    Your reply will be screened

    Your IP address will be recorded 

  • 1 comment