A Persian Cafe, Edward Lord Weeks

Saturday, 14 November 2015

The Uses and Abuses of Signalling

Epistemic Status: I wrote this between about 10pm-2am on Tuesday, and have decided that I'm unlikely to find the motivation to go over it. Hence I'm publishing it, with a proviso that it is more than usually likely to contain mistakes and is undoubtedly too long.

BD Sixsmith has a post discussing the nature of virtue signalling. There's much to agree with and little if anything that I directly disagree with, but the notion of signalling which he inherits is less precise than signalling in the Hansonian sense. This post, then, is my attempt to reintroduce that precision.

What is signalling?

Signalling, in  Robin Hanson's sense, is spending resources in order to communicate information about oneself. There are two differences between this definition and that which Sixsmith* inherits from James Bartholomew ("the way in which many people say or write things to indicate that they are virtuous.") First, Hanson's account covers a wide range of actions, rather than being restricted to speech. (To be fair, Sixsmith makes the observation in his own post that signalling is hardly a property solely of verbal communication). Second, crucial to Hanson's account is the idea that resources are consumed by signalling.

Why and what do people signal?

Or more precisely, what information are they trying to convey about themselves? So far as I am aware, the traits which we signal fall into two categories: commitments/values, and abilities.

When one is aiming to signal a particular trait, it is crucial to the value of one's signalling that someone who did not possess the trait in question could not perform the behaviour which constitutes signalling. If anyone could do it, then there is no way for the signalling behaviour to mark you out as someone who possesses this valuable trait.

The behaviour one chooses in order to signal a trait, then, will depend heavily upon whether the trait in question is an ability or a commitment. If the trait is an ability, then the signalling behaviour should be one that is easy to perform only if one possesses the trait. This is at the heart of Bryan Caplan's signalling model of education: a degree is relatively easy to obtain if one is intelligent, whereas if one is of average or below-average intelligence then the cost is far higher**.

If the trait is a commitment, on the other hand, then the behaviour should be one that is costly to perform, even if one does possess the trait. The logic here is that if one did not place so much import on this commitment, one would not be willing to make the sacrifice which is entailed by the signalling behaviour. A perhaps out-of-date example of this might be buying an expensive engagement ring: one would only be willing to spend so much money if one really did want to be married to the beloved. Back in the days when becoming pregnant out of wedlock was considered shameful, the engagement ring functioned as a signal that the man did not intend to run away if his fiancée did indeed become pregnant.

What is costly varies hugely from person to person. For example, suppose Jim earns £25,000 per annum but manages to donate more than £10k to charity each and every year. Wow! one might think, What a generous guy! Jim's charitable donations successfully signal generosity. But now, suppose that he in fact earns more than a million pounds per annum. Now he looks positively miserly, because £10,000 is no longer a significant cost to him. If Jim wishes to signal generosity, he really needs to up his game.

Thus the cost of a signal may be either high or low. Whether or not it ought to be depends upon the nature of the trait being signalled.

So is virtue an ability or a commitment?

This is a tough question, one that philosophers have been debating for more than two millenia. (Admittedly, that's actually a fairly low standard). Clearly there is some extent to which it is an ability - i.e. the ability to recognise what is the morally right thing to do in a range of situations - but there's an ongoing debate as to whether one needs motivation to act morally, beyond the simple knowledge of which actions are moral.

Fortunately I solved meta-ethics one day on the bus a couple of weeks back, and so can tell you the answer. For perfectly rational, ideal agents: yes, moral beliefs are inherently motivating (though this is actually a rather misleading way of putting it, since something's being motivating comes prior to its being moral). However, humans are not perfectly rational. Most pertinently we suffer from "Akrasia", or weakness of the will. The fact that we recognise we in some sense "should" do something does not mean that we actually will do it. The ability to resist akrasia and perform moral actions, then, is a type of commitment - in this case a commitment to doing good.

Real-world Virtue Signalling

So let us take what a left-wing person might think of as "the virtue of progressivism". How is one to signal this virtue? Progressivism seems less like an ability than a commitment, so the key to a successful signal will be that it is a signal one will make only if one is truly committed to the cause. Such signals might include: long hours of unpaid labour, spent knocking on doors or stuffing envelopes; taking the time to read long and often incomprehensible Marxist tracts (also a way of signalling intelligence); saying things which induce non-progressives to take you less seriously.

There are also a variety of things which one can do to suggest that one is progressive, but which involve lower costs and will therefore be taken less seriously. Mocking the Daily Mail is a British national pastime, and serves poorly as an indication of one's progressivism. Ditto expressing concern for the poor, which can be done equally well by Tories. These are simply too commonplace.

Alternatively, suppose one wishes to signal that one cares for the global poor. The most obvious way to help them is to give money to a charity which works in the third world, but this is not the most visible activity. Since signalling is an act of communication, it must of necessity be visible to the desired audience. Hence instead of doing overtime at work and donating one's extra pay, one might perform an impressive, embarrassing, or otherwise unusual activity and solicit donations.

An even more extreme way to demonstrate one's commitment to helping the people of the third world is to actually travel there to help. Some students do this, and we talk about it "builds moral character". (Yeah, Bryan Caplan might say, it builds moral character in exactly the same way that going to university builds intelligence.) In recent years there has been increasing suspicion that these trips are actually more leisure than work, and the status of the people who go on these trips has taken a corresponding hit.

Is everything signalling?

No. Sixsmith is of course correct to note that "When teenagers get involved in musical subcultures and buy new clothes, and cut their hair, and paint their nails or pierce their lips it is informed by their need to appeal to their peers, yet it would be foolish to think that they don’t like the music." Perhaps one might try to salvage more of the signalling view by claiming that appreciation for a particular aesthetic is in fact an ability of sorts, but this is surely pushing the model too far.

Socially Useful Signalling

Sixsmith claims that "The problem of virtue signalling in the modern age is in large part a problem of scale. When the world was not so criss-crossed with roads, flights and wireless connections one demonstrated one’s virtues in one’s own community. On such a local scale, one’s beliefs regarding, say, the proper treatment of poor people in one’s neighbourhood had to be actualised in one’s behaviour. Modest and productive virtues could take precedence." 

Perhaps this was true among the general populace. I imagine it varied from community to community, since while the principles behind signalling are the same across communities, the specific acts which are socially accepted as signals vary greatly. My concern here is with signalling as it was (and still is) practised by aristocrats and nations. The principle "virtues" these actors have traditionally sought to convey are power, wealth***, sophistication, and piety. Hence they have built large palaces, commissioned great cathedrals and works of art, and many other things besides. Some of these (such as many of the Oxbridge colleges, or the Szechenyi Chain Bridge) have been of great use to society. (Though how much of that is post hoc rationalisation? If we could send money back into the Middle Ages, would we want it spent upon founding Elizabeth College Oxford, or would we want it spent upon bringing peasants out of dire poverty?) Much of it, however, was spent upon buildings that are nice to look at, but don't really serve any purpose other than attracting tourists.

Perhaps the world's most beautiful COMPLETE WASTE OF RESOURCES.
I really don't know what gets us out of this trap of socially-wasteful signalling, perhaps other than unilateral action by people already recognised as possessing certain virtues. Effective Altruism appears to be achieving a beneficial form of wealth-signalling, but there are many other things which we signal using behaviours which are wasteful if not outright damaging. Sixsmith suggests "spending more time in our communities and less on Twitter," but this is of course reliant upon his belief that signalling within communities is usually socially beneficial. This is not to say that he is wrong, merely that I reserve judgement either in favour or against.

A second problem with signalling is that it is inevitably a positional good: my signal makes yours less useful, either by showing that it's easy enough that I can also perform it or sufficiently low-cost that I am also willing to perform it. Again I don't know how to tackle this: the ideal solution is a Pigovian tax, and indeed this provides the most solid justification for heavy taxation of luxuries. But such taxes are difficult to implement (Marathon-running tax, anyone?) and easy to mess up. Nor does this give us any guidance upon what to do as individuals (apart from thinking up new ways of signalling which have yet to be exploited by all and sundry).

* I would use his first name, but am uncertain as to his preference between Ben, Benjamin, or BD.

** For purposes of simplicity, I'm leaving out the extent to which (in Caplan's view) a degree also signals conscientiousness and/or conformity.

*** Wealth signalling: also known as Conspicuous Consumption.

No comments:

Post a Comment