polemical circles (2010) by Michael Hartl

I have never cared about the Pi vs Tau debate. (If you haven’t heard, the argument is over whether we should use 2π or the absolutely equivalent τ in maths texts. It’s a wonderful, nominally bitter feud where both sides admit that the result of the feud has absolutely no practical importance.)

But I’ve realised that your view in Pi vs Tau is an analogy for an actually deep question: it serves as an intuition pump for ‘how much should we change huge complex systems?’

Let’s get some machinery set up.

One way of slicing the confused term “liberalism” is into two great streams: the metaphorically “French” and the metaphorically “British”. 4

  • 'French' liberalism is top-down: it has a single large theory of the world, and is willing to make radical changes to make the world accord with the theory. (A stark example, maybe the inspiration for Hayek coining the distinction, is the difference between Napoleonic civil law and English common law.)

  • 'British' liberalism is bottom-up: it emphasises the limits of our knowledge, the difficulty of making large changes successfully, and the value of evolved institutions, as repositories of successful adaptations.

(It is in this sense that e.g. Marx is French - a top-down rationalist interventionist - where e.g. Taleb is British, a halting curator of what has persisted.)

Your Hayekian ‘Frenchness’ is not just a political question: it bears on how you interact with all current and possible states of affairs. e.g. geometrical symbols.

That is, you can tie in the maths feud as follows:

  • Pi people as British liberals. They hold to either traditionalism ("what's old is good because it is so invested with meaning") or cautionary pragmatism ("what's old is known and evolutionarily stable").
    Why make obsolete perfectly good textbooks and course notes? Why create a generation gap?

  • Tau people as French liberals. They hold to rationalism ("what makes sense is good") or perfectionism ("only what is best is good") or aestheticism ("what looks nice is good").
    If even mathematicians cannot get these things right, for want of courage and standards, then we are all screwed.

This is a really bad analogy

Okay, okay:

grand social reorganisations might be different from the inplace alteration of mathematical symbols in a couple ways (e.g. cost in billions, number of people affected in billions, risk to human life in megadeaths, extent of uncertainty, degrees of freedom in the policy space, tenor of the respective debates, average rationality of participants, guns brandished).

I'd guess that most political radicals are incapable of caring about Pi vs Tau, and that most Tauists are not political radicals. 3

But it's still good to think about, and in the tiny sample of people who care about both, I predict a correlation between political radicalism and Tauism of like r > 0.5.

You might wonder how ‘French’ I am. Boringly, I don’t know. 6

Traditions are often internally incoherent, just because they are a giant heap of ideas all added without editorial care. 5 Not everything which persists a long time is adaptive : bad local maxima exist all over nature and society. 1 And the status quo is monstrous and intolerable in many, many obvious ways.

But large, sudden social changes have probably been destructive more often than constructive. I am instinctively top-down, but I’d like to be more empirical.

In a world with no actually pressing problems (like malaria, adversarial geopolitics, pollution, imprudent AI research, suffering, death, or stubbed toes) then I would probably push for mass textbook revision in tau’s favour. de Condorcet on a lotus throne. 2

  1. Unless you define "adaptedness" as "extent of persistence through time", you rogue.
  2. You might think it's churlish or pious to bring down Tauism in favour of heavy moral matters. It's not a moral question, after all!

    But tauification would cost quite a lot, and since the invention of money, resources are resources; diverting them to an inconsequential aesthetic crusade is a morally loaded action.

    (...if and only if the money could be diverted to something of moral importance.)

  3. Because, overwhelmingly, most people hate or don't care about maths, and most people are not political radicals even in word.

    (Though there's amusing weak evidence that Brits are simultaneously more leftist and rightist than their politicians are.)
  4. Of course, these concepts don't cleanly divide French or British thinkers: it just names a pair of tendencies in the C18th. (Hayek, the modern populariser of the idea, knew this.)

    His exceptions include:
    • British 'French' people: Thomas Hobbes, Joseph Priestley, Richard Price, Thomas Paine.

    • French 'British' people: de Montesquieu, Benjamin Constant, Alexis de Tocqueville.

    Hayek also came up with a competing binary for liberalism: 'rationalist constructivism' vs 'skeptical individualism'. This division corresponds more closely to the familiar, more-or-less damaging left-right axis we use now.

  5. So is incoherence adaptive?
  6. God, what an insightful intuition pump this has been after all.


Post a comment (with Markdown):