The idea of “virtue signalling” has been mainstreamed as a way of criticizing performative displays of moral conduct. It’s often used by people when what they really mean is “hypocrisy”.
However, the signalling of virtue, far from being a vice, is an essential component of the very idea of morality. Morality is not, and never has been, a purely individual matter, but is negotiated through shame, honor, and acceptability in a community.
To put it as plainly as possible: it’s better for the world if you encourage others to do good, even if you make no effort yourself. Buddhism has always accepted that it is good to do good oneself and also good to encourage others to do good. Obviously it’s best to do both, so we can make a morality matrix, from best to worst:
- do good and encourage others to do good
- do good but don’t encourage others to do good
- don’t do good but encourage others to do good
- don’t do good and don’t encourage others to do good
Virtue signaling is incomplete, and can be hypocritical, but in and of itself it is good.
It is striking that there is no equivalent phrase for “vice signalling”, yet it is very much a thing. A recent example is the South Dakota governor Kristi Noem’s story of killing her dog.
The key point here is: why did she talk about it? She simply had to say nothing. It’s a deliberate choice, the advertising and promoting of one’s own vices as way of provoking and challenging those who stand for virtue. She got lots of publicity out of it, and lots of support from the kind of people who like killing things.
The idea of “virtue signalling” is a linguistic hack pushed by people who want to undermine virtue itself. It’s the same strategy as used with terms like “SJWs”, “political correctness”, “wokeness”, and so on. The purpose is to undermine morality, or at least the form of morality based on compassion, in favor of the morality of power. Here’s a good article about it.
I believe that the idea of “virtue signalling” stems ultimately from the philosophy of Rene Girard. Girard taught that all desire is “mimetic”, which essentially means that we want something because someone else wants it. You can see how this applies in morality: we want to do good because we see that others want to do good. By making it socially unacceptable to “signal virtue” we undermine the basis of morality itself.
That’s not, to be sure, what Girard was trying to do; rather, he was trying provide logical support for Catholicism. He was a French Catholic philosopher based at Stanford, largely unknown in wider circles, but to whom many of the alt-right thought leaders are devoted, notably Peter Thiel.
I think Girard’s ideas—whether used correctly or not—permeate alt-right circles in ways that those outside don’t notice. I happened to study Girard in my research on mythology, and now I commonly recognize his ideas among the alt-right.
The alt-right have picked up on this idea, among others, and are using it as a mimetic hack to undermine moral discourse. The very fact that you say you did something good shows that you are bad actually. Thus no-one can speak of goodness, and the social reinforcement for virtue is dissolved.
In Buddhism, our first focus is to do good. When we do, it’s good to speak of that to encourage others. If people get annoyed by that, that’s their problem. 
) There’s a lot to unpack. A couple of reactions:
– this brings me full circle to the AI discussion. You alluded to this in the Stochastic Parrotts essay about Morality: AI can fake a platform for discourse about public morality; i.e, it can fake virtue signalling (oxymoron) while rendering useless the socially intelligent and “grooming” – if not depressing – role played by gossip (and public shaming, ostracizing, etc., which you mention in the OP). That is, people don’t learn how to create and sustain discourse about public morality because AI will do it for them (ostensibly). But AI can only shape the discourse based on how it’s been programed and with what data, fundamentally.
