Norm channels John Stuart Mill:

It is a commonplace of political liberalism that discussion and debate are good; we learn through considering different points of view, including those to which we are opposed. Commonplace because true.

Really? At least one famous study (pdf) has found that we don't learn from considering different points of view. Quite the opposite. Back in 1979 Lee Ross and colleagues got supporters and opponents of capital punishment to read arguments against their beliefs.Far from causing people to learn, this had the opposite effect:

The result of exposing contending factions in a social dispute to an identical body of relevant empirical evidence may be not a narrowing of disagreement but rather an increase in polarization...Social scientists can not expect rationality, enlightenment, and consensus about policy to emerge from their attempts to furnish "objective" data about burning social issues.

What's going on here is the confirmation bias (pdf). People give great weight to evidence which supports their priors whilst giving contrary evidence critical scrutiny. And that's if they even see such evidence. It's widely thought that the increased freedom of speech which the internet has given us exacerbates the tendency towards filter bubbles and echo chambers in which people create a "Daily Me" which corroborates their prejudices.

Discussion and debate lead not so much to learning as the mere exchange of prejudice. Free speech gives us not a rational pursuit of truth but rather the mindless and often dishonest ventings of "Sam Bacile", George Galloway, Kelvin Mackenzie and countless other fanatics. Mill's defence of free speech seems to have been a rationalist Victorian optimism which isn't supported by the evidence.

Or is it? Two things suggest otherwise:

- In academia, free expression is a necessary condition for scientific progress, though not a sufficient one; it must be supplemented by peer review and benchmarking theories against evidence.

- Western societies which have enjoyed long periods of free expression have enjoyed what Norm and I would regard as progress towards truth, in the sense of increased equality for homosexuals, women and ethnic minorities. There might be a causal link here.

On balance, I suspect Mill's consequentialist argument for free speech is weak. If it is to rescued, it's by pointing to the tiny fraction of intelligent expression it permits, rather than to the vast majority of worthless out-pourings.

This, though, is not to argue against free speech. Rather, I mean to suggest that the argument for it should be a non-consequentialist one. Quite simply, nothing - not majority opinion, national security, decency or the will of God - can be sufficient licence to suppress it.