Skip to content

Free speech in the marketplace of ideas.

In a couple of weeks, the verdict in the case against Geert Wilder’s for inciting hatred will be announced. Wilders is charged under laws that have been enacted in many jurisdictions, but which are controversial. I don’t know whether these laws are good or bad. Here I just want to address one argument in favour of unfettered free speech.

The argument is a familiar one. It might be called a Millian argument, after John Stuart Mill who was an influential proponent. Mill argued in favour of a ‘market place of ideas’; in this marketplace, good ideas – that is, truth – would displace bad. “Wrong opinions and practices gradually yield to fact and argument: but facts and arguments, to produce any effect on the mind, must be brought before it”. On these and other grounds (including the claim that only if we know and can refute rival ideas are our own views justified), Mill argued for complete freedom of speech. For proponents of this idea, we ought not ban hate speech; we should allow its expression so that it can be tested and publicly refuted.

The idea is very attractive. But it seems – at least in its more optimistic form – to rest on an implicit presupposition of greater human rationality than is in fact warranted. The evidence suggests that we are less good at assessing ideas than the optimists suppose, and, moreover, that our rational assessment has less effect on our attitudes than might be thought.

We are especially bad at evaluating arguments about topics about which we already have strong current views. When we encounter apparent evidence against a view we hold strongly, we devote considerable of effort to discounting it, whereas we are quick to accept arguments for views we already hold (Ditto and Lopez 1992). In a classic study, Lord, Ross, and Lepper (1979) divided subjects into groups depending on their attitudes toward capital punishment. Each group was then presented with the results of (fictitious) studies examining the claim that capital punishment has a deterrent effect on crimes, one of which presented evidence in favor of the claim and one which gave evidence against it. For one group, the study supporting the claim employed a cross-state comparison to establish its conclusion while the study opposing it used within-state figures (ie, comparing the crime rates before and after the introduction of capital punishment). Subjects were quick to find fault with the methodology employed by the study that argued for the conclusion they reject, while finding the other study to be methodologically sound. Thus, attitude toward methodology is best predicted by the subjects’ attitude toward the conclusion offered, rather than by features of the methodology.

Even after we have evaluated claims, with whatever degree of success, we remain vulnerable to psychological mechanisms that leave us with unjustified beliefs. The sleeper effect produces a delayed increase in the persuasiveness of a claim (Pratkanis, et al. 1988). Subjects who hear negative statements about unknown individuals are left with more negative impressions of those individuals then subjects who hear neutral sentences about the same individuals , even if the subjects are later told  that they are not true (Wegner et al. 1981). This bias towards acceptance may explain people’s well-know tendency to persevere in beliefs about themselves, even after they told that the feedback upon which their self-assessment was based was systematically misleading (Ross et al. 1975). Wegner et al. (1985) told subjects that the feedback they would receive would be systematically misleading before they got it; nevertheless, they tended to accept it as valid.

We can therefore predict two things, with reasonable confidence. One: even if the intellectual case for a view – such as a racist or sexist opinion – is weak, many otherwise intelligent people will fail to be convinced of its falsity. Two (and to mind more disturbingly): the refutation of racist or sexists opinions will probably nevertheless leave those who hear them with residues; their attitudes to people who fall into the stigmatized groups will be more negative than they would have been had they not heard the claims. And this will be true regardless of whether they accept the claims.

This does not entail that we ought to ban hate speech. There are predictable costs of government intervention as well as of an unfettered marketplace of ideas. It does entail that a marketplace of ideas will likely have costs, and that these costs will likely fall more heavily on racial minorities. If we come down in favour of a marketplace of ideas, we ought to do so in full awareness of these costs.

Share on

9 Comment on this post

  1. Dr,

    I am curious to know what your reply is to Stanley Fish’s (and Robert Post’s) claim that the “Marketplace of Ideas”, being itself necessarily defined by ideological presuppositions, can only be constituted by the same political—ideological—considerations it is designed to hold at bay, and would, consequently, not be “unfettered” or “complete”—at least, not in the allegedly flawed, robust, Millian sense you invoke; it seems to me that Fish’s work presents us with a devastating exposé of Mill’s theory.

  2. I’m sure that you are right in stating that the “sleeper effect” and the subjective choice of arguments or methodologies exist. But should this lead us to the pessimistic conclusion that we cannot rely on individuals to freely form correct moral judgements, that free speech cannot be justified by the Millsian argument, or that there is a form of Gresham’s law in which in the market-place of free speech, the bad drives out the good ?

    If we look at history, we observe that racism, homophobia and sexism have existed for centuries but are, globally, on the decline. Which means that somehow we have been able to accept that they are wrong, despite cultures which widely broadcast and implemented these beliefs. We have in fact improved our moral judgement (this is not to say that there’s not much more still to do !)

    One of, or a combination of, two things thus appears to have happened.
    Either :
    1. The changes in our beliefs have happened intellectually through ethical reflection, and this despite the previous moral majorities, in which case it seems that at least part of this change is in fact due to free speech.
    2. These beliefs have changed due to political or economic pressure from victims of discrimination, and part of their ability to change things has been due to free speech (as well perhaps as a host of other freedoms).

    In either case, the words of Mill seem still appropriate : “Wrong opinions and practices gradually yield to fact and argument: but facts and arguments, to produce any effect on the mind, must be brought before it”

  3. Anthony, you are right that we need to explain these shifts. We need to explain the success of science, too – including the very science which I’m citing here. Part of the explanation for the success of science is its social nature – peer review and publicity of methods and data, and (perhaps especially) formal methods makes it harder to hold on to a failed hypothesis. I think part of the success of moral argument is due to similar factors: collective deliberation can outperform individual. Even for the individual, the effects I note do not by any means always swamp the power of argument. But they can be shown to swamp it all too often. I suspect that it is actually a combination of rational and arational factors that lead to actual moral progress. Many people can’t be convinced, and many of the ones who are convinced don’t change their implicit attitudes despite changing their explicit beliefs. True progress occurs only when a new generation is socialized into new attitudes.
    Sebastian, I don’t know the work you’re referring to. Just like all markets, a marketplace of ideas is going to be distorted by monopolies, collusion, asymmetries in information, and do forth. I believe that these facts ensure that markets need to be regulated, more tightly with regard to some goods than others (and it may be appropriate to distribute some goods by other mechanisms). But markets have their benefits as well, in terms of efficiencies and allocation of resources. I suspect there is more room for genuine markets in the production of knowledge than in knowledge itself; markets presuppose information rather than distributing it. I don’t know how similar these concerns are to Fish’s.

  4. There are serious costs to allowing erroneous information to spread, whether that information comes from malice or stupidity or some other source of error. But the question of what is erroneous is often contested and error sometimes becomes a matter of faith or ideology. How are these costs to be avoided or contained? I don’t think the market in ideas is sufficiently robust and active to take care of the problem, especially around election time.

    Would anyone really accept some sort of authority deciding what should be held to be error and what true? We have had experience with such authority, and most of us realize that there is no such thing as an efficient impartial authority. Courts of law woun’t work either because lawsuits are costly and that cost could deter persons from contesting error. Moreover, erroneous decisions can yield punishing damage awards that harm seriously persons seeking to contest what they feel is error. See Lee Kwan Yew’s use of the courts in Singapore.

    In general, a noisy marketplace is a good thing, whether it is of merchandise or news or ideas. Whether it is a newspaper, a university or the internet. The cost of error is serious, but that of preventing error is greater in its potential for harm to the marketplace. So, suck it up; live with it. Teach your children skepticism and critical thinking and hope someone else is doing that for their children. Have a drink.

  5. Dennis, I appreciate these concerns, and I am far from sure that I don’t agree with you that the risks of restrictions are greater than the benefits. But I think the risks of unrestricted speech are very great. First, I deny that we can’t confidently identify claims that it is mischievous to assert (an example: there was recently an article in the Age newspaper on the Armenian genocide. The Turkish genocide wrote a letter denying there was any such genocide. That’s a lie, of course). Second, the risks can be seen in current inaction on climate change. You say we have to live with it, and just teach our children skepticism. I am afraid that many people’s children will die with it, rather than living with it.

  6. The case of climate change denial is an interesting one since (i) it concerns denial of a risk, rather than of a historical event (such as genocide denial), and (ii) it is more an incitement to *inaction* rather than incitement to (undesirable) action (as in the case of hate speech). I’m not immediately aware of any case in liberal / Western-style democracies where this kind of “misinformation” would actually be banned, and I’m not really aware of political discussion in this direction either. Should there be?

  7. Peter, I would be very uncomfortable urging the banning of climate delusionism (as John Quiggin has appropriately dubbed it). But there may be actions short of bans that could help to reduce the contamination of public discourse. We have social norms that penalize holocaust denialism; perhaps we ought to have social norms that penalize this kind of speech too (instead we do the opposite: the media pursues a policy of so-called balance on climate, but not holocaust denialism).

  8. Thanks Neil. I share your discomfort at the idea of urging banning of “climate delusionism” (nice term!) per se, not least because it would certainly play into the hands of the deniers. But there is also a wider question, namely whether either (i) denial of a risk for which the evidence is overwhelming or (ii) incitement to dangerous inaction could ever be a grounds for restricting free speech. The reason I find this interesting is precisely that I’m not aware of any existing legal precedent for this in liberal democracies, where there clearly are examples in relation to historical events (e.g. holocaust denial in Germany) and in relation to incitement to violence etc. Is this something we should argue for? If not, why not, and if so, how would we go about it?

  9. There is no marketplace of ideas. There is a pisspot of ideas. The media may promote some idea more than others but it has nothing to do with correctness. It is probably just a matter of whose interests are served and who can pay for the promotion.

Comments are closed.