There Will Be No Message Discipline

This is one of those posts I write because I so often find myself making the point that I should like to be able to just have it all written out and available to point to. It concerns a genre of article, tweet, opinion, argument… I often encounter in many different spheres. As ever, I won’t give negative examples, so if this doesn’t resonate with you presumably we have just had different experiences and you may go on your way in the peace of the Lord. 

The genre I am concerned with is: people imploring leftists to speak and act in a fashion that is less off putting to the uninitiated. It’s sort of a genre of respectability politics, but lacks the full connotations of that. It’s more just that, say, abolition of the nuclear family is a very unpopular position, so if you lead with a bunch of harsh condemnations of nuclear families as a mode of child rearing you are certainly going to code yourself as non serious or even scary to many listeners. Likewise the left is much more likely to be sensitive to subtle instances of bigotry most people would not care about, to the same effect. The genre I am concerned with takes examples like this and implores leftists to keep this to themselves, to give people a pass on that sort of thing, in the hopes of building an effective coalition for winning and wielding power. 

(An example of the genre, to make it concrete, was the widespread discussion among American twitter users about whether or not the left should hold back from criticism of Biden’s Build Back Better programme back when it seemed like that might be a thing. It is one thing to direct such “hold your tongue!” advice to prominent national figures, but for this to be an instance of the sort of thing I care about we must focus on such advice being offered to internet randos. I saw multiple cases of people on twitter engaged in back and forth wherein both were what I would describe as ordinary schmucks, one was expressing dissatisfaction with the limits of Build Back Better, and the other was telling them that they need to shut up with this as it will only fracture the coalition necessary to pass the bill which was surely an improvement on status quo.  This is the genre of thing I mean.)

I don’t like this genre and this post will explain why.

Now it’s going to turn out to be significant who exactly my audience for this post is, so a word on that upfront. I think the arguments I give here go for most ordinary schmucks having every day conversations; prominently, in today’s day and age, conversations on social media or semi-public fora. It would not apply to: a) people who are extraordinarily visible and likely to be seen as a thought leader or trend setter. b) certain sorts of intimate one-to-one conversations where strong bonds of trust exist between participants. I will highlight where and why these caveats are important as they come up. It is also presupposing an audience whose goals are for the left to gain political victories, since the implied instrumental argument of the genre presupposes that as well. I think this will in fact rule out many journalists (doubly so prominent journalists, for reason a above still applies), since the appropriate goal for journalists might be properly distinct from seeking partisan political victory, and in any case often just is not in fact seeking left victory

I think that straight off the bat anyone is likely to generate the following two objections to the genre. First, an objector may respond: I simply do not interact with enough people to scare people out of a coalition. The viability of a left coalition does not require people to like me in particular, after all; most will in fact never even encounter me. Second, more general considerations re how we ought behave in a democracy might speak against the genre. Too much Machiavellian cleverness in the expression of our beliefs and preferences can just lead to democratic deliberation or aggregation not really taking into account the actual knowledge dispersed throughout the population. Then we are all poorer for it. While I see something to both of these, I want to just set them aside here. Presume that words spoken on social media really are available to everyone, and that strategic behaviour is possible and would be good if effective.

Now, let’s consider the situation of our ordinary leftist schmuck. I suppose they want the left to win, but they also enjoy speaking their mind and engaging in the rough and tumble of online. Importantly for the argument that follows, this is to say: they do not post only because they think it will influence outcomes and care about what happens when it does. All else equal, after all, people do not like censoring themselves and enjoy airing their thoughts. And since they know of themselves that they are nought more than an ordinary schlemiel, they are aware that even if they could persuade everyone that they personally were a reasonable comrade, that would not in fact do much to change people’s opinion of the left as a whole. In fact at a first approximation they are simply irrelevant; no one takes them to be representative nor has any reason to, and the opponents of the left will always be able to find other people (whose speech our ordinary schmuck cannot control) to portray as unreasonable. All this leads to more or less the same odds for the possibility of left coalition resulting whether or not our Joe Schmo is the one seen to say unreasonable things themselves. The relevant facts here being that: one, their decision to moderate their speech cannot anywise make any large number of other people moderate their speech in turn, and two the discrediting anti-coalitional effects of scary (or “unreasonable”) leftists speech don’t depend at all on it being them in particular who engages in the speech.

So we could form a very simple model of this ordinary schmuck as follows. They are faced with a choice between speaking their mind or engaging in message discipline. And (in this simplified model, though nothing depends on this particular simplification) on some issue we are presently campaigning over the left can either win or lose. Most of all they would like to be able to speak their mind and have the left win, but they accept it would be worth engaging in message discipline if that meant the left do win; whereas, disprefferred to both those, if the left are going to lose anyway they would prefer to speak their mind than not.

With numbers representing ordinal preference rankings if you draw that out as a decision table you get 

And it’s not hard to see that our protagonist now has a straightforward dominance argument for speaking their mind.

This is because this whole way of representing the problem presupposed that the leftist schmo isn’t really able to affect whether or not the left win or lose by their particular speech actions on social media or cetera. Now in fact I think that is usually true for the vast majority of people the vast majority of the time. As per my caveats, of course, if you are an unusually influential person or engaged in trusting intimate conversation with someone who is likely to be decisive themselves then actually this might not go, so our argument will have to be more complicated. So we will consider the more complicated case in a moment. But I actually think that most of us actually do have this dominance argument for speaking our mind available to us.

But let us give the proponents of arguments in the genre their due. Though I think they lack good reason for this it’s clear they do usually presuppose that there is some ability to influence the course of events here. In decision theory we would say there is “act-state dependence”, that how I behave affects the probability of being in one state of the world rather than another. And in fact as I have represented our ordinary leftist schmuck it captures the idea that if they were someone whose speech decides whether we are in the leftist win versus leftist lose scenario they should indeed engage in message discipline. In essence this would be like choosing between the two circled outcomes in the diagram below, in which case one should clearly engage in message discipline. So I do not think this representation entirely begs the question against the genre.

Now very rarely will any of us, even quite influential persons, be this decisive. So our question must be: under what sort of circumstances is one’s behaviour influential enough to make this a plausible consideration? Here our simple ordinal preference table will not do as a representation, since to sensibly reason about this we must have a more sophisticated idea of exactly how good or bad the various outcomes are, and what their respective probabilities are.

Any precision in setting the values of these would be arbitrary and artificial so I will not try. Instead I think it is enough to note some very general considerations here. In general there are two sorts of factors that will make this a sensible move. First, if the conditional probability of the left winning given that one engages in message discipline is very much greater than the conditional probability of the left winning given that one speaks one’s mind. Second, if the utility of the left winning is so much greater than that of the left losing that anything with any tiny probability of shifting the balance towards the left winning is ipso facto justified. Of course it could be with some combination of the two they balance out in the middle, but in fact I do not think this is relevant. For anything other than a highly negative response to the first condition is exceedingly unlikely.

I note again my assumption that we are talking about an internet rando engaged in ordinary social media posting. I think it is clear that for such a person it is not just that they are not decisive, but that even if we grant their behaviour is not strictly irrelevant it is still exceedingly unlikely to make a difference. This is because if the left is on track to form a winning coalition the probability that it would not occur if a random schmuck tweets something insensitive is just vanishingly small. Social forces are not so easily derailed, and in a tolerably large democracy (and a non-democracy of more than 2 people) most of us simply make no difference. This is familiar to many in another guise as the paradox of voting; so let us dub it the paradox of posting. There is no escape from the paradox of posting by noting the bare fact that one is not entirely irrelevant to how things go; still one is far too insignificant to really justify the arguments of the genre.

This leaves the hopes for the genre of argument I dislike pinned on the utilities point. Now I think it is possible we do live in something like the Political Pascal’s Wager scenario, of the difference between victory and defeat so huge that it justifies leftists in engaging in self-sacrificial behaviour. I have blogged a bit about this here. So it is not that I simply reject this as implausible. But it is that I think that if this is what is supposed to justify making arguments in the genre it is quite mistaken, and in fact as a consideration it probably goes in the other direction. 

For if the issue is that leftist victory heralds the possibility of utopia, or leftist defeat catastrophe, or both, then this should also factor into one’s choice as to whether to engage in the activity of offering the argument itself. In particular one would have to weigh whether one’s chances of persuading enough people who are themselves sufficiently influential that their behaviour matters outweighs the risk of by one’s words contributing to the narrative that the left are unreasonable and actually dissuading some onlookers. “See, even leftist agree that leftists sound crazy!” sort of thing. At a minimum I do not think people have seriously grappled with this, so in the extreme utilities case they are at least being irresponsible. But in fact I think that for most ordinary non-taste-maker schmucks talking to broad audiences in public fora wherein their words are observable but they lack established relationships of trust with those they seek to persuade - the probabilities go the wrong way. Basically you are very unlikely to make a difference either way, but my guess is you are more likely to be picked up and highlighted by enemies of the left than you are to persuade internet strangers with good political speech.


So there we have it. I actually think in most cases the very simple dominance argument is enough: the genre of arguments to the effect that one ought moderate their speech for instrumental reasons is refuted by the simple fact that speaking my mind dominates the alternative. But even if one tried to take act-state dependence into account in some sophisticated way, I still do not think the conditions hold which are necessary for offering arguments from the genre to be a good idea.

If anything will get us out of this paradox of posting it is coordination. Something like a party with a mechanism for adopting a mass line — in particular something which assured me that my engaging in message discipline is either evidence others will also be, or will causally contribute to making it more likely they will. But absent organised solidarity we shall all keep furiously tweeting into an indifferent political void, as the lone and level timeline stretches far away