Repressive Online Tolerance?

(Originally published: 25 October 2021)
In a provocative 1965 essay entitled “Repressive Tolerance”, the philosopher Herbert Marcuse calls for intolerance of right-wing views that favor the (unjust) status quo. Marcuse, who was forced to leave his homeland of Germany in the 1930s as the Nazi’s rose to power, was fast becoming a key figure in radical politics in the US. His views had great influence on the emerging “new left” movement of the 60s, and more recent attempts to disrupt talks by conservative speakers might be traced back to his ideas. His contribution in the essay was to blow apart the Western faith in the value of free speech.
The classic liberal defense of free speech is of course found in chapter 2 of John Stuart Mill’s On Liberty. According to Mill, we should allow uncensored views – even ones that might appear to be outlandish or repugnant – to be expressed because this is our best bet of arriving at the truth of any matter (scientific, philosophical, or political). If we stop a view from being expressed, says Mill, we risk discarding what might end up being the truth. But even if we are already in possession of the truth, we might nonetheless have reasons for allowing challenges to be voiced. Silencing these false views would deprive us of ‘the clearer perception and livelier impression of truth produced by its collision with error’. Perhaps we only really possess knowledge once we are required to defend the truth against objections.
Marcuse agrees with Mill insofar as he thinks that, at least sometimes, free and open discussion is the most promising way of getting the correct answer to any question. However, the qualification he adds is crucial. The instrumental promise of free speech only holds in some circumstances. And those circumstances in which we can expect the happy mechanics that Mill describes are, for Marcuse, certainly not the circumstances characteristic of late capitalism. Here, the role of ideological beliefs in sustaining systems of exploitation and oppression give certain views – namely, those that seek to justify the status quo – an unfair advantage in the marketplace of ideas. If people come across two opinions – one favorable to existing power structures and one opposed to them – they are more likely to discount the latter irrespective of the quality of the argument. In situations such as this, tolerating differences of opinion means extending tolerance to ‘manipulated and indoctrinated individuals who parrot, as their own, the opinions of their masters, for whom heteronomy has become autonomy’ (p.90).
What is to be done? Marcuse suggests that giving all views a fair hearing would require a re-balancing process. We need ‘information slanted in the opposite direction’ (p.99), and what this suggests in practice is ‘the withdrawal of toleration of speech and assembly from groups and movements which promote aggressive policies, armament, chauvinism, discrimination on the grounds of race and religion, or which oppose the extension of public services, social security, medical care, etc.’ (p.100). For Marcuse, achieving emancipation from injustice may require intolerance towards those who defend it, perhaps by preventing them from speaking through tactical disruptions to events.
Marcuse is often viewed as representing the very worst excesses of twentieth century Marxism. He is thought of as the stereotypical paranoid leftist, who sees imaginary enemies at every turn. But even Marcuse may not have envisaged a society in which an algorithm actively pushes right-wing opinions to the forefront of our attention in a site of public debate. Yet this is precisely the society that many of us have found ourselves in. A recent study by Twitter found that the timeline of users in six of seven countries from which data was collected promoted tweets by right-leaning news organizations and politicians at the expense of others. Who needs ideology when you can convince the masses to vote against their own interests with the help of computers?
Marcuse himself presented a worry to his proposed strategy: how are we to know which views we should withdraw tolerance to? His answer is that individuals should decide this once they have been freed of ideological beliefs, but therein lies the problem – how are we to free ourselves of ideological beliefs if the trends Marcuse describes as sustaining ideology are still ongoing. His “solution” ‘presuppose[s] that which is still to be accomplished: the reversal of the trend’ (p.101). Fortunately, a parallel problem does not haunt the solutions to biased timelines and newsfeeds of online platforms. Assuming we can quantify the propensity of a given algorithm to promote certain views over others, a re-balancing exercise should be a fairly straightforward technical task. The real problem here, though, is one of power: can we really trust private companies to take the necessary steps to ensure that public debate achieves the Millian ideal? We have now moved much of our public sphere online, under the control of big tech, and no amount of ideology critique will be enough when the real barrier to human emancipation is the result of a computer program