There are those who believe that democracy is the preferred system of governance because of the epistemic advantages it bestows. To be more precise, there are those who believe that democratic decision rules are better because they “track the truth” more closely than their non-democratic brethren (e.g. autocratic or elitist decision rules). Such decision rules can include voting and deliberation, with deliberation often being preferred for a variety of reasons. Those who defend this position are known as Epistemic Democrats, and although there is no perfect characterisation of their view, I’ll be taking the following as about as canonical as one can get:
The Epistemic Defence of Democracy: Democratic decision rules are justified, and the political decisions made through such procedures are legitimate, because they have more epistemic value than alternative decision rules. This means that decisions made via such processes are, in general, more likely to be right.
Limiting our focus to the deliberative version of epistemic democracy, there is a question about the precise source of the epistemic value of deliberative decision rules. According to some — usually referred to as instrumentalists — deliberative procedures are better because they can get us closer to some procedure-independent standard of truth or rightness. That is to say, for instrumentalists there is a goal or aim or end against which deliberative procedures are to be measured. According to others — referred to as pure proceduralists — there is no procedure-independent standard of truth or rightness. Rather, the deliberative procedure is itself intrinsically epistemically valuable.
I’ve read several defences of instrumentalism over the years, but only one article that defends pure proceduralism (that's not to say there aren't others). That was by Fabienne Peter and it was entitled “Pure Epistemic Proceduralism”. The article itself is rather dense, but I want to zone in on one portion of it in this post (pgs 35-37, for those with a copy). The portion in question purports to offer a basic defence of pure proceduralism, and my hope is that by explicating here I will gain a better understanding of it.
The basic defence is made up of two, ultimately interlocking arguments. Each provides some support for pure proceduralism, but only jointly do they (allegedly) offer an exclusive defence of it. Individually, they provide a partial defence that leaves some room for instrumentalists to salvage their view. The first argument is the argument from reasonable value pluralism, and the second is the argument from knowledge construction. I’ll outline both below, and make some comments along the way.
1. The Argument from Reasonable Value Pluralism
The notion that values are irreducibly plural is a popular one. Isaiah Berlin is perhaps the classic exponent of this notion. If I may quote from him here (I can’t remember the exact source for this quote right now):
“We are urged to look upon life as affording a plurality of values, equally genuine, equally ultimate, above all equally objective; incapable, therefore of being ordered in a timeless hierarchy or judged in terms of some one absolute standard.”
If Berlin is right, and if values are really plural and equally genuine, then we might expect reasonable people to disagree over about what ought to be done. You might think that the well-lived life is one that is dedicated to charity work, and I might think the well-lived life is one that is dedicated to intellectual discovery. According to pluralists, both our views are legitimate, but the differing emphases and potential clashes between them could filter into the democratic process. In a deliberative forum, I might try to argue for the merits of government-funded research programmes, whereas as you might argue for government-funded welfare programmes. With a limited pot of money to share out, one of us might not get what we like.
Peter’s first argument for pure proceduralism takes this possibility — i.e. the possibility that reasonable people may disagree as to which values should be pursued and with how much vigour — seriously. Indeed, she takes it as a foundational premise for assessing the legitimacy of any democratic decision rule. As she sees it, any democratic procedure that fails to accord proper respect to reasonable value pluralism would be illegitimate, and the problem is that an instrumentalist approach fails to accord the proper respect.
This is the gist of the argument. Now let us spell it out in more detail by introducing something called a legitimacy condition. A legitimacy condition is anything that a procedure must satisfy in order to be legitimate. The debate between instrumentalists and pure proceduralists is essentially a debate about the number and weight of the legitimacy conditions that democratic decision rules have to satisfy. Pure proceduralists count fairness, equality of participation, consideration of different views and arguments, and deliberation over outcomes as the key legitimacy conditions. Instrumentalists can count fairness and equality of participation and likelihood of achieving the truth as the key legitimacy conditions. Indeed, at least as I read her, Peter seems to think that instrumentalists accord decisive weight to achieving the truth in the legitimacy calculus.
And therein lies the rub. For if decisive weight is accorded to achieving the truth, respect for pluralism is merely contingent. If a procedure is more likely to achieve truth by not providing a public forum for deliberation, then so be it. To be clear, this is not to say that, in practice, deliberative forums will not be mandated on instrumentalist grounds. It is merely to say that, in principle, they need not be. That is enough to damn them in Peter’s view because to accord proper respect for value pluralism necessarily mandates deliberative procedures.
One might wonder about the normative premise that is implicit in all of this. This premise must state something like “we ought to respect reasonable value pluralism”, but why should we do this? Peter offers an argument, which she take from Amartya Sen, that claims that people ought not to be treated as patients, as mere passive carriers of well-being, but rather as agents that have an interest in the autonomous formulation and pursuit of their own ends. In other words, respect for reasonable value pluralism stems from respect for individual autonomy.
This gives us the following argument:
- (1) It is fact that there is an irreducible, but reasonable, pluralism about values.
- (2) We ought to respect reasonable value pluralism.
- (3) If we wish to respect reasonable value pluralism, then (necessarily) people must be allowed to participate in the evaluation and determination of alternative social policies (i.e. there needs to be a deliberative forum for decision-making).
- (4) Therefore, we necessarily ought to have deliberative forums for decision-making.
- (5) Instrumentalists do not think we necessarily ought to have deliberative forums for decision-making.
- (6) Therefore, instrumentalism is wrong.
Written out in this manner, I see several problems with this argument. The first is the assumption of reasonable value pluralism. I’m perhaps overly influenced by something I read by Gerald Gaus years ago on this, but I think pluralism is at least questionable. For starters, it’s not always clear what is meant by pluralism. The Berlin quote I gave above runs together several concepts that should arguably be kept distinct. Perhaps the best conception of pluralism holds that values are intransitive. But even if we accept this definition, it’s not all that clear that pluralism is true. Most arguments for pluralism (that I have seen) tend to revolve around dilemmas in which one has to choose between two (allegedly) equally desirable or equally awful outcomes, which are allegedly irresolvable. I won’t get into it here, but I think those kinds of arguments are insufficient to prove pluralism. For one thing, equally desirable does not mean intransitive.
In addition to this, I’m not entirely convinced by premise (5). I think it would be possible for an instrumentalist to necessarily value deliberation, if it could be shown that deliberation was essential, not merely contingent to achieving the truth. In other words, if they can show that deliberative democratic processes are always going to be more likely to attain the truth than alternatives. And indeed, there are a number of arguments out there suggesting that might be the case. Deliberation often involves the pooling of information, the gathering of evidence, and the weighing of arguments. Each of these might be essential for achieving the truth. Interestingly, Peter’s second argument, which I’ll get to momentarily, seems to endorse this point of view.
Finally, I’m not sure that pluralism is even really necessary for this argument. Given that the key normative premise is supported by reference to individual autonomy and agency, I wonder if the argument might get off the ground with that alone. In other words, I wonder if respect for autonomy, moral equality and agency would mandate some kind of deliberative process. Given that pluralism is, at least for me, a controversial thesis, I would be happier with this option.
2. The Knowledge Construction Argument
Peter’s second argument for pure proceduralism appeals to the essential role that deliberative processes play in the construction of knowledge. Again, Peter appeals to some of Sen’s work to support this point. To quote from the man himself:
the practice of democracy gives the citizens an opportunity to learn from each other . . . . Even the idea of “needs” (including the understanding of “economic needs”) requires public discussion and exchange of information, views and analyses. In this sense, democracy has constructive importance, in addition to the intrinsic value it has in the lives of the citizens and its instrumental role in political decisions.
So the idea here, as alluded to above, is that deliberative procedures allow for the pooling and exchange of information, both of which are essential to producing knowledge. Of course, that just opens the door to the instrumentalist argument I outlined: If deliberative procedures really are essential for the production of knowledge (where knowledge is understood, minimally, as “true belief”), then it looks like one could be instrumentalist about deliberative procedures and still respect reasonable value pluralism.
But, as Peter sees it, this all depends on how one conceives of the constructive process. Later in the article, she partly endorses the Deweyan conception of democracy. According to this conception, deliberative democratic processes are not merely instrumentally essential to the attainment of knowledge, rather they are part and parcel of what constitutes knowledge in the first place. This flows from Dewey’s broader pragmatic theory of truth, which abjures the more traditional correspondence theory of truth, and holds that truth is not merely a correspondence between beliefs and reality, but rather a functional product of the correct procedures. It would take too much time to get into the debate over this conception of truth here, but hopefully what I have said is enough to get the gist of the position.
In the end, Peter finds the Deweyan conception problematic because it too fails to give proper weight to reasonable value pluralism. So she comes up with her own theory which partly incorporates the Deweyan view, but still respects reasonable value pluralism. I won’t outline that view here.
These then are the two arguments Peter uses to supply a basic defence of pure proceduralism. I'm still not convinced since I still think there is room for the instrumentalist view within this framework, but maybe I'm just missing the point.