Reflection V

The term “virtue-signalling” first appeared as a way of mocking the perceived insincerity of those who vocally advocate for social justice causes without making any real effort to effect them in practice, or doing so only insofar as they complement their own lifestyle choices. The anti-leftist sentiment baked into the concept is no doubt one reason why it was not initially taken seriously as a critique by its targets. Another might be that those whose sincerity is in question simply do not experience themselves as insincere. There are, after all, reasons other than self-interest to vocally espouse a cause—for example to persuade, to compel, or to mobilise. If the goal is collective action then mobilisation through discourse is its precondition, not its afterthought, and to criticise individual espousers on their lack of current individual action would seem to miss the issue.

Nevertheless, the suspicion that there is a cynical disarticulation of reasons from actions at work has not gone away, and while virtue-signalling may never have become a familiar expression in left-wing discourse terms like “performative ally” have arisen within it to describe something similar. So-called “woke capital” is guilty of this as well, with brands often producing the signs of diversity as little more than a marketing strategy. The underlying concern is that virtue-signalling does not merely mask a pre-existing absence of virtue, but actively corrodes virtue by replacing it with its performance; not just a deflection from the absence of resistance to power, but an active functional component in the auto-regulation of power relations.

This point about motive insincerity is one I’ve been chewing on a lot recently. On the one hand it seems pretty clear that virtue-signalling names a real phenomenon, and not a marginal one. On the other hand the typically left-critical explanation of it—that it stems from insincerity at the level of individuals, i.e. ostensibly selfless political or ethical expressions are actually in the service of cynical and calculated self-interest—just does not seem right at all. It seems equally clear that most people advocating for social justice are doing so from broadly sincere motives, at least in the first instance. If virtue-signalling is to be understood without appeal to motive insincerity on the part of individuals, then the puzzle is how to understand it as something more like a network effect.

An idea I’ve found useful in thinking about this is Wilfrid Sellars’ distinction between I-intentions and we-intentions. I-intentions represent self-interested goals underwritten by individual desires (e.g. I shall open the umbrella because I want to stay dry) while we-intentions represent collective goals, which may conflict with the desires of the individuals that have them (moral and political goals would fall in this category). Put in these terms, the left-critical account understands virtue-signalling as the false representation of I-intentions as we-intentions, where the expression of selfless virtue is a disingenuous cover for self-interested social power play. The premise I want to work from, in contrast, is that the public representation of we-intentions are mostly genuine expressions of we-intentions. The question, then, is how the sincere expressions of we-intentions can be systematically distorted to give the appearance of being insincere expressions of I-intentions, and what implications this has for the ability of we-intentions to be realised in we-actions.

This question turns on the difference between the way our goals are represented to ourselves (i.e. as intentions) and the way they are represented to others, and the possible divergence of these representations. Conventionally speaking we have a kind of first-person authority with regard to our intentions—that is, the default is to take people’s intentions to be whatever they sincerely believe they are. But the public (or third-person) understanding of a person’s goals must be inferred from their actions, which will include but not be limited to their statements of intent. In addition to a person’s intentions we must also consider their interests, and the extent to which their actions serve their interests is a different question from whether their actions line up with their stated intentions.

In a system that rewards virtue every selfless act also serves self-interest, whether that is its intention or not. Whenever there is a virtuous act in such a system, it is ambiguous from the third-person perspective whether it is motivated by a we-intention or an I-intention. Nevertheless, the usual thought is that there is some fact of the matter about whether a given selfless act is indeed selflessly motivated. The contemporary situation in which virtue-signalling becomes significant is exacerbated by the fact that what is rewarded is not virtue itself but its performance, the reward typically being something like cultural capital or social media clout. The left-critical view then sees the absence of follow-through in action as evidence that the original performance was I-motivated.

As I see it, the trouble with this view is that it does not do justice to the non-linear aspects of practical reasoning. When we reason about how to act, a critical input is our beliefs about how everyone else is reasoning, including whether we believe them to be reasoning from self-interest or we-interest. There is no point acting on a we-intention to your own immediate detriment if you do not believe others will also be acting on the same we-intention. (That the alignment of immediate self-interest is not enough to secure cooperative action in many situations has been well illustrated by various game theoretic conundrums like the prisoner’s dilemma and the tragedy of the commons.) Conversely, if you perceive everyone around you to be already reasoning from self-interest, then there is no point in not doing so yourself. Our experience of our own agency is mediated by our understanding of the agency of others, based on the inferences we make from our understanding of their actions, interests, and intentions.

And of course, this is why signalling virtuous intentions to one another is an essential feature of the ethical language-game. It is not a problem in itself. However, when this language-game is transported into a communication structure that systematically rewards individuals for the performance of virtue, it undermines the ability of we-intentions to unambiguously signify as we-intentions, even when they are completely sincere. It seems likely to me that the kind of thing that gets roasted as virtue-signalling is often just the sincere signalling of a we-intention. If many of them don’t get followed through in decisive action, it may just be for the simple reason that it’s not clear enough to the signaller that similar sentiments do not mask tacit I-intentions. Which, ironically, then makes the original we-intention appear to mask an I-intention: the script writes itself. According to this logic, what we have been calling virtue-signalling is a distributed and self-producing process arising from the overlaying of a self-interested reward structure on the communications channels we attempt to use for self-disinterested reasoning1.

Anyone who’s ever engaged with some kind of social justice or ethical cause in the social media age will be familiar with the odd sense of being rewarded for it without ever wanting to be, often in ways so subtle that the reward is not even noticed until the moment it accidentally gets cashed in. The feeling is like having one’s sincerity drained away from outside without having had any say in the matter. It is processes like these that seem to me to fuel the rapid oscillation between quasi-religious earnest and depressive apathy that are characteristic of contemporary political and moral discourse. In the bouncing between these two poles what is lost is the possibility of slow politics, the kind of gradual accumulation of collective will that is now nostalgically yearned for, perhaps epitomised in the progressive imagination by the civil rights movement of the 1950’s and 60’s.

The deep reasons for this situation seem to me to stem not from some kind of moral failings on the part of individuals, but from the systematic distortion of the communications structures we use. I think that the reason social media appears to be full of virtue-signalling is just that it doesn’t really support anything else. All content is always-already storyfied, its own communicative content subsumed by the series it appears in, a new variation on an aesthetic. These are clout trading spaces, by design. Any communications space which forces participants to trade reasons on the model of a marketplace—i.e. as distinct individuals with ontologically prior needs, desires, beliefs and interests—will necessarily undermine forms of reasoning in which intentions are the outputs of group decision processes, not their inputs.


  1. Since writing this I’ve discovered the concept of a “grift”, explained nicely in this twitter thread:

The argument I’ve made in this post is that the economic logic of contemporary communication structures has the effect of making sincere attempts to build collective motivation appear to be potential grifts (by rewarding them), which in turn has the effect of turning them into real grifts (by undermining their ability to build the collective motivation required to address the real problem).