Log in

No account? Create an account

Revisiting Fish

To briefly revisit a limited issue with postmodernism (spoiler: I think postmodernism is rubbish and the wrong direction to go from the fundamentally right challenges posed by relativism) and "X studies" fields (where X is gender, feminist, black, latino, or a few other things), I have a better criticism of Stanley Fish's criticism of Alan Sokal's prank.

To revisit it, Sokal decided to demonstrate that these fields are full of shit, that they play games with language, have low academic standards, and are rotten parts of academia. Sokal wrote a long paper full of dubious arguments, tailored towards flattering perspectives he thought the field liked flattered, and it made it (with some difficulty) through the review process and was published, at the same time publishing his experiences in another journal, attempting to show that as a social institution, these areas of academic study were intellectually bankrupt. It made a number of people very angry, and a number of other people very amused. Stanley Fish (whom I otherwise respect for some of his books) was among those angered by this, arguing that it was misconduct, rude, and destructive of university culture.

I confess delight at damage to postmodernism - as noted above, I think it is the wrong response to the challenges to premodern thought during the 20th century - we came to understand that morality is a social construct, that "a priori" is deceit, that all thought depends on pragmatic and otherwise unprincipled roots, and that the universe is probably not rational in any reasonable sense of the word. Postmodernism, as a response, is an attempt to dissolve these things. As a stage and in moderation, it is reasonable, as an end state, it is unacceptable. To the extent that we're influenced by Nietzsche, we should note that he was not a nihilist - he may have killed god (less metaphorically, killed the idea of god in western culture - "god is dead and we have killed him" is meant to celebrate this, not to claim there ever really were deities), but he suggested tearing down traditional morality and building something new in it's stead. Being a classicist, N's concept of traditional virtue was that of an idealised early Roman citizen (before the taint of Christianity) - while I don't much like that ideal, the basic enterprise is sound - traditionalism is destroyed to make way for new values, and relativism is the light that allows us to understand that old values were constructed just as much as the values we would construct - that they are mortal and can be slain. I don't think we should be as radical as Nietzsche nor adopt his particular aim - we should rather examine old values and traditions, learn from them, and then discard them if we can make something better that still serves humanity. Many things are worth keeping unless we can really be certain we have a better alternative (traditional academia, for example - if this opens us to the criticism that our notions of truth are oppressive and embarassing to deeply traditionalist/primitive worldview, we should be delighted to be oppressors in that sense). Postmodernism (and some other anti-foundationalist schools of thought) insist on parking and abandoning the yana (edit: wordplay - see mahayana) of civilisation in this transitional, destructive period - it is a death-urge operating under the banner of "freedom". The best conclusion to the questions posed by relativism is that of a traditional academic facing a shake-up of a field - study carefully, adapt as needed, be smart.

What Sokal's demonstration meant is akin to another great debunking - that of 「Clever Hans」. Clever Hans was a horse who was reportedly good at math, famous throughout germany. At some point, he was visited by an academe, who conducted experiments and eventually discovered that Hans could read body language of humans very well, his tapping of his hoof as answer to questions ceased when those around him made subtle (unconscious) shifts in their posture as he hit the right answer. Hans was not in fact capable of math, but he had an uncommon knack of a strong form of domestication - the horse's genius was mistaken. Debunked, the horse remained popular, because truth and popular perception can often be frustratingly distant. So it is with the Sokal affair - a fraud is discovered, and there is criticism by vested interests in the enterprise about the uncovering and lack of trust. Most modern fields of academia are not bullshit, but the fields named above, tainted with a rotten intellectual movement, either should be tossed entirely or have a cleaning of house to do, for they have little more intellectual integrity than the orthodox marxists. They invoke concepts without meaning, construct papers that are little more than patting each other on the back, and lack the history of doubt and debate for their conclusions that lies behind proper academics. Stanley Fish may be somewhat less guilty than most in this regard; his wors are wonderful at illustrating some of the deeper meaning of relativism, but they blur the line between advocacy and academics to unnecessary and harmful degrees, justifying the postmodernism he claims not to be a part of.

Science is a social as well as a philosophical activity, and it is not well-understood as a solitary action nor as a simple extension of the scientific method. Without the large community of academics, their traditions, arguments, and debates, it would not have the ability to effectively function. It is full of real people with values, hunches, and varying levels of methodological and philosophical sophistication, and it is fallable, but it is the best thing we have for determining truth. Academia is not entirely about science, but other fields share many of its methods, its notions of intellectual integrity, its notions of sin and virtue. When parts of academia show themselves to have such low standards as to be suceptible to this kind of prank, it becomes necessary and useful to demand things of them, and to consider those like Sokal to simply be whistleblowers. Loyalty should be to the ideals of academia, not to its institutions - if a person, discipline, or institution is rotten, we must have the ability to reform or eject them, setting loyalty aside. Any proposed ethics that could not consider and then eject astrology from academia is a bad ethics for us.

On a personal note, whenever we see papers, books, and similar that connect radically disparate ideas, we should be very skeptical, even if these ideas are fascinating or the authors brilliant/funny/nice. Yes, perhaps marxist theory and queer theory can tell us something about media consolidation, but more likely we have a few tame observations that cultural creation is becoming more open dressed up in very fancy clothes (if so, Lawrence Lessig says it better and with less questionability). Yes, perhaps cellular automata are teh sekkrit path to understanding all of reality, but more likely we have a bright guy with a few neat ideas who doesn't have enough people around him telling him when he's full of shit or unoriginal.

A further personal note - it's important to tell people when they're full of shit (or unoriginal, or both). It's too easy for people to decide "I will be original and brilliant", write their manifestos/books/etc, and not realise that their foundations lead them to reinvent the wheel for the 57th time, badly, and that there are also some rather well established criticisms for what they're trying to do if only they would look. I've had this happen a few times in my life - where I had to say "dude, go look up anarchoprimitivism", "libertarianism", "christian socialism", "turtle logo", "lisp", to someone who had essentially synthesised it themself (I've also had a number of "original" ideas that turned out to be not so original).

(note that saying "I disagree", "I disagree and find your ideas disturbing", "you're only half-right", "your criticisms are fine but your conclusions are poor", "you're full of shit", and "you reinvented the wheel" are all different things you might say - be sure it's clear in your mind which of these you're actually saying)

Largely unrelated, I am happy to see that Howard Dean is making a comeback. Alongside Russ Feingold, he's one of my favourite American political figures, and I was very disappointed that his political career did not go further. I may be happy that Barack Obama seems to have good judgement, a push for calmness in politics, and an inclusive style, but Obama isn't nearly liberal enough for me to be at ease with him. Dean may be significantly more conservative than I am (so is Bernie Sanders, for that matter), but the value-distance is at least a bit less. I approve of Howard Dean's strong criticism of the disemboweling of the health care reforms (and as a doctor, I presume he has a lot of relevant experience - I certainly trust him more than that other doctor). For a lot of (mostly personal) reasons, I also like Joe Biden, but I'd much rather see a Dean/Feingold (or Dean/Sanders) presidency than an Obama/Biden one (better yet, if some proper socialists could be found who are actually sane, I'd like to see them in office, but such beasts are as rare as centaurs and probably not electable in the US)


I've had this happen a few times in my life - where I had to say "dude, go look up anarchoprimitivism", "libertarianism", "christian socialism", "turtle logo", "lisp", to someone who had essentially synthesised it themself (I've also had a number of "original" ideas that turned out to be not so original).

Yes!!! To all of the above.

Why do I keep on feeling compelled to defend pomo when it is attacked, but to assault it when it is utilized?
My answer is that you're really defending relativism, and the broader sphere of perspectives based on relativism, but object to it itself.

I suggest as analogous people defending Atheism but attacking Objectivism (or Secular Humanism, or any other specific philosophy in the field opened up by Atheism).
Yet, pomo is so useful for "constructing" my garden! See http://vorpalbla.livejournal.com/556120.html
I have mixed feelings on this - I think there is a certain amount of confusion in PL circles - ideas that "the syntax problem is solved and uninteresting" - as an expressive medium and one on the human interface side, I reject this claim, and think it's an area of interest that will be open for a very very long time (even if, being at least partly aesthetic, it might not exactly be a hotbed of research compared to work on provability, optimisations, and related topics). Programming is kind of a wed field, because it's a form of human expression, it's (potentially) an engineering discipline, and it's an area with all sorts of research problems. So long as nobody is proprietary enough to claim the whole field (only my problems are interesting, your interests are not - claims like that), I'm generally cool with things. The normal engineering attitude of "everything is a bunch of tradeoffs, what you choose is based on what you choose to optimise/concern yourself with" is I think the best attitude towards it. Anything else is kinda broken, and it's bound to (rightly) piss people off and get them to (wrongly) say stupid things in angry response because their interests are marginalised.

That said, there are a *lot* of people who have, as we both would acknowledge, reinvented the wheel in pl design, aiming for nothing new in any of the three realms (pl-interestingness, expressively interesting, or having good traits for software engineering types).
I would not demean the advances that come out of PL research - I am in fact very enthusiastic about them. Advances in provability are awesome stuff, and are pretty closely tied to research in optimisation (also awesome). I am not as knowledgable about this stuff as I might like to be, but I think the problems are quite interesting. (I do have a view that you might find mildly contentious - that proofs are not hard in the deepest sense, because empiricism forbids solid belief in anything, but I believe that empirically, within their framework they have been shown to be very useful tools and have a high degree of confidence in them even if it cannot in principle be absolute in a radically empiricist framework).

I may have been unclear with what I meant on engineering. In design of languages, one might focus on traits that are interesting from a provability/optimisability perspective, as a pl-person would tend to do. One might instead focus on traits that are interesting from an expressivity perspective, as a programmer-as-artist type would tend to do. One might instead focus on traits that are interesting from a software engineering perspective, as a software engineer might do. Any of these foci are valid for language design (and I in fact appreciate them all in theory - more on that in a sec). As a PL person, you might want to heavily steer language design towards things where you could prove things about the language, or possibly do all sorts of awesome optimisations on it. As a programmer-as-artist person Larry Wall might design a language based on linguistics and psychology to make programs "roll off the mind" more easily and be very fluid. Someone at the SEI might design a language heavily focused on design by contract, easy integration with formal standards practices, and software/component reusability (languages like RESOLVE). There is a natural tension between these interests, even as there are also natural co-interests.

Personally, when it comes to programming, expressivity means a lot to me - my theoretical interests in these other things take a bit of a back seat in languages I want to use. It's not all-consuming though - while I really like fluid languages, I also want great optimisation, potentially the ability to prove things about the program (maybe even decompose it into something parallelisable), design-by-contract as an option, and all sorts of other goodies.

On occasion I have been a bit grumbly when people make statements that my preferences and interests are invalid, and I may (not sure - don't recall) have said some dumb things to let off steam when people flat out asserted that I should sacrifice more of what I love in order to get these other things I care about less - I will set my own priorities in what I like. If I did say dumb things that contradict this, consider those to be stupid angry things that should be superceded by my actual calm opinion, stated here.
There's a kind of two-way hinge - in general-purpose languages, people working on compilers will either attempt to find regularities that are specific to the code (and not guaranteed by the language) or allow flags for the programmer to make promises to use a more regular subset of the language. Knowing more about what the program does is the general task that allows the either detection of problems or ways which the program can be sensibly restructured for better performance. In language design, more restricted languages with easier formal reasoning more naturally see good optimisation - Fortran, for example, has long been considered faster than C because it is more restrictive (canonical example is its aliasing rules).

If we were to imagine a very clever and slow compiler that attempted to automatically prove things about the possible inputs and outputs to/from functions and either demonstrate side-effects to be nonexistent or get sufficient regularities about them, that compiler would be able to safely parallelise them, prove defensive programming unnecessary, reorder statements in them for various optimisations (VLIW architectures in particular), etc. To a certain extent, work on this already exists - theorem-proving in (sometimes clean subsets of) messy (from a pl perspective) languages is an interestingly difficult task. In conversations with people working on these things, I remember people citing pl-research of the sort you're interested in as being relevant. I regret that I can't provide specific examples of this though.
Those arguments are generally a waste of time anyhow. It's like people arguing about what the best ice cream flavour is, the best text editor, etc. There's a very human tendency to project one's values onto others - it's how value consensus is forged in culture, but it's often an irrational process, drawing heavily on aesthetic judgements. If vim or emacs really fit how we work well and we really like it, it's very likely that we'd be tempted to spend a lot of time trying to spread the faith, and find it frustrating if we're talking with someone who wants something different out of their editor. It's probably emergent from our genes. I generally suggest people sit those arguments out, even if they are a programmer. Usually the people who seem to have footing are just people who are good at dancing about in arguments.