24/02/2011

The Dilthey Prize


People like to make lists of the good things humanity's been getting up to. But they always jabber exclusively about natural-science: hard tech. This is probably because physical apparatus is louder, and life-saving in an obvious way. So: some quiet (Heideggerian) technology that was also massive:

Grandest findings of the human sciences
(broadly construed)
in their first century:


1931: Maths is not logic.

This doctrine, "Logicism", or "Fregeanism-with-respect-to-the-foundations-of-maths" was a highly impressive attempt to make the world make sense. It consists in the two linked theses:

1. mathematical concepts can be defined in terms of logical ones (no math-primitives)
and
2. mathematical principles can be derived from axioms (given definitions of concepts).

Why does this matter?
Why did people want it to obtain?
Why is Gödel's incompleteness theorem so renowned?

I suppose logicism matters because we live in a world where the most (academically, politically, rhetorically) credible analyses are the ones that cloak themselves in formal algebra. It's a paradigm where even fatuous formalization is preferred to other methods. We are reassured by the difficult and symbolic; it tastes of rigour even where it is unrigorous. The logical positivists were still invested in the idea long after Russell's Principia Mathematica, logicism's finest hour, went south.

People want it (particularly math-physicists like Hilbert) because aporia is fucking unpleasant. A great many philosophical problems disappear when you have appropriate rational grounding.

The Incompleteness Theorem's power is that it's an intuitionist's rebuttal of logicism using logicists' own methodology: if the Theorem is true, no finitary system can ever cover everything that arithmetic does (or seems to do).

"...the human mind is not capable of formulating all of its mathematical intuitions."
- Gödel

Quizas.





from 1898: Marx was wrong.

At least insofar as:
(It was mostly Marxists that found all this out for us.)

His other things - the concepts of social alienation and exploitation; his methods of social analysis; his theory of technical change; and his pioneering thought about market failure and ideology are right enough that pretty much everybody who thinks about society uses them.

Despite what it looked like from where he was, there probably aren't "fatal structural contradictions in the economic dynamics of capitalism". Welfare capitalism has (despite recent appearances) smoothened some of it out. Before the C20th (and depositor insurance in particular), there was a banking panic of 2008's scale every twenty years. (Note that this isn't an endorsement of the "Great Moderation" presentist nonsense: just that capitalism's structure isn't obviously essentially volatile.)

Marx was right; Capitalism does limit us in all sorts of ways. But not near so much as his followers have sadly tended to.


(Winners: Von Bortkiewicz, Marshall, Bakunin, Keynes, Kolakowski, Hayek, JK Galbraith, Okishio, maybe Fukuyama ... and the Prague Spring)



1945 & 1961:
Evil is banal.

i.e. that 'Good' people are capable of horrific things too.

(Winners: the Nuremberg courts, Arendt, Asch, Milgram, Hofling, Zimbardo, Bandura)



1951: Voting may never work properly.

In 1951, an unusually human economist, Kenneth Arrow, published a game theory / psephology / ethics paper, which, starting from a very few modest value judgments like "votes should count the same," and "the system should operate stably" goes on to prove that any putative way of organising a vote will fail to satisfy all of the values at once.

The result sometimes gets reported as "democracy sux!", but the actual conclusion should be something like: "a voting mechanism defined for all possible preference orders cannot simultaneously comply with all four of these desirable conditions."

i.e. Although we have thought up lots of ways to decide together, flaws and "manipulability" are endemic to social choice. Others have shown up the limits to our science and our reason with regards to all "interpersonal comparisons of utility" - but there is conditional hope for a partial fix.

Principle #1: Anonymity (or, "Non-dictatorship")
The social welfare function should account for the wishes of multiple voters. It cannot simply mimic the preferences of a single voter.

Principle #2: Universality
For any set of individual voter preferences, the social welfare function should yield a deterministic, unique and complete ranking of societal choices.

Principle #3: Independence of irrelevant alternatives (IIA)
The social preference between x and y should depend only on the individual preferences between x and y (Pairwise Independence). More generally, changes in individuals' rankings of irrelevant alternatives (ones outside a certain subset) should have no impact on the societal ranking of the subset.

Principle #4: Unanimity
If every individual prefers a certain option to another, then so must the resulting societal preference order. This, again, is a demand that the social welfare function will be minimally sensitive to the preference profile.

Pick your virtues.




1956: Capitalism is not natural.

Highly artificial conditions are required for a market to arise, and work is involved in keeping the bastard thing from spoiling when it does.

Let's be generous to the marketeers; let's grant them even the Efficient Markets Hypothesis. Problem is, once you depart one tiny little jot from the ideal Platonic "perfectly competitive" market, all bets are off. Lipsey and Lancaster showed, - sixty bloody years ago! - that in every actual economy, and almost every physically possible economy, satisfying lots of the criteria for efficiency will not necessarily improve efficiency. It logically requires that you catch em all. This is why, irl, perfectly rational, expert free-market advice is so often counter-productive;
they're advising that you take a flight that gets you almost to your destination, where "almost" is "the sea".

(Pure) economic libertarianism is a fatally flawed doctrine. This same set of findings helped dismantle Social Darwinism. This is one of the biggest social-scientific works ever, but even in the field it simply is not well grasped.

(Winners: Lipsey and Lancaster, Nash, Dixit, the neo-Darwinists)


stalled? : Existing society is bigoted and brutal.

  • ...To sexual minorities (fabulous LGBT activism; no monolith occurs)
  • ...To the poor (Fabians, Marx)
  • ...and in fact everyone; itself (MARX, Weber, Debord, Habermas, Barthes, Foucault, Chomsky, Deleuze-Guattari, Derrida, Zizek)


1973: Money can make us crap.

Running counter to the suggested "efficiency wage" effect and our generally bottom-line-crazed culture, another psychology bit: "overjustification".



perennial: Existing society is endangering itself (and much, much more).


("Winners" - people who realized we were losing -

Activism

Aldo Leopold
Rachel Carson,
James Lovelock,
the Sierra Club,
eventually, guilty voters

special thanks to: Science

1957: Revelle & Keeling (the carbon penny drops),
1974: Molina & Rowland (o no ozone)


forever: Mind is hard.

The issue of what the fuck the mind is might have looked easy at the start of the century, but it lingers, and by no means just because of ideology and self-service. Functionalism and property physicalism were only recently born, though they grow quickly and evenly. But the issue is hardly solved, nor is it clear that cognitive science will breakthrough (for it may be chiselling at the wrong wall).


(Winners: Semon, Ryle, McCullough and Pitts, Turing, Chomsky, Newell and Simon, Putnam, Fodor, Dawkins, Dennett
vs
vs



Recently: Atheists are not less moral.

We can grant the religions that si dieu n'existe pas, tout est permis might have seemed likely. But only from the decrepit metaphysics of human nature they all tote.

(Winners: Epicurus [...], Russell, Hauser, de Waal, this just in)


*********************************************************

And now the section that explains why this kind of award doesn't happen:

Disagreement happens

Calling things objective "achievements" as I have done above is more or less taboo in good, rigorous philosophical work: all things flow; the infinite cornucopia is not yet eaten out, and there may anyway be no fact of the matter.

Individuals are sometimes allowed achievements; for entire fields we tend to speak instead of "developments", a neutral term which does not invite applause.

(This wiki page is excellent in tracing the agonies of the topic.)

Further, the humanities and half of the social sciences deal with values: entities which get called "arbitrary" from objectivist positions (which miss the fucken point) and of which it's never possible to speak of with finality.


1. The lure of logicism remains; the project recently underwent a resurrection (so-called "neo-logicism") in work by "The Scottish School" (Crispin Wright et al) and "The Stanford School" (Bernard Linsky et al). Parts of Gödel's proof and certainly von Neumann's have come under suspicion.

2. Under a very broad definition of "Marxist",the movement's themes are very much alive; the "Critical Theory" approach (straight outta Frankfurt) can be thought of as cultural neo-Marxism.

3. Not even the world faiths contest this one, I think.

4. Workarounds in "social choice theory" and "public choice theory" have been suggested, so that partial success is usually assured. (e.g. difference between resolute and nonresolute situations)

5. All that is true does not glitter: there aren't many serious pure libertarians anymore (even Nozick was made to bite bullets). Those that come close to the doctrine are roundly mocked even by other frothing types.

6. There's danger here; feminism, always attacked and denigrated, has come to be seen as "finished" by too many. Some overenthusiastic people declared Obama's coronation to be "the end of race politics" in America. The animal rights thing rolls on, neither revolutionizing nor stagnating.

7. Unorthodox, without much replication. Yet.

8. The meat of the concern has long reached 99% consensus, but at the fringes there is considerable distortion and pressing questions...


Mind you, disagreement is endemic in the natural sciences too. But they're orders of magnitude better at enforcing method, and thus a PR image of conformist progress.

6 comments:

  1. Hey man, only started to read the first part of this!

    It's not just an old fashioned thing to describe numbers with logic. Well, perhaps it is, but it is also a recreational form of mathematics particularly among computer scientists. For instance, the natural numbers can be defined as sets, and subtraction can be defined as logical predicates as follows:

    It isn't required for the logic, but:

    Let Zero be the empty set.
    And in this way, 0 = {}. 1 = {{}}. 2 = {{{}}}.

    Subtraction is performed as follows.
    For some set X, for some set Y, Subtract(X, Y) implies:
    Either X or Y being empty implies X.
    otherwise,
    for some set A, such that A is a child of X,
    for some set B such that B is a child of Y,
    then Subtract(A, B) is implied.

    Example, subtracting one from two
    Subtract({{{}}}, {{}}) implies
    Subtract({{}}, {}) implies
    {{}}. Or 1 as we like to call it.

    ReplyDelete
  2. Oh, I ken it's useful, and highly homologous: that's why the result was so surprising. Flick down and you'll see that folk are still trying...

    ReplyDelete
  3. The title of the post could have been "Nobel Prize for the Mardy Sciences"

    ReplyDelete
  4. Hehe, I'll have to read the rest later. It is long. I was happy my example provided something useless! Hehe!

    ReplyDelete
  5. Ahhh, I see what you were getting at now. To be honest, it's not a question that interests me really. Unless you have a specific problem that requires one to be treated as the other, it's a bit mad to force one to be a subset of the other.

    ReplyDelete
  6. Grrr, posting things when I am angry is silly! Well really it is quite interesting but I'm curious as to what degree the proof would be required for it to be accepted. I mean, maths can be embedded within the lambda calculus (or at least all the maths we know now...) and the lambda calculus can be described using logic. So you've got to wonder nowadays what the value is in such theories; when instead you could be could be compiling either logic or math into turing machines.

    ReplyDelete