Public Values in the Right Context

ABSTRACT I am sympathetic to Avner de Shalit’s position that a political philosophy should incorporate public values, but I see their role differently. Philosophers of science standardly distinguish between values being introduced in the context of discovery (inputs into the investigation or arguments) and in the context of justification (acceptance or rejection of substantive claims in light of the arguments or investigation). I argue that de Shalit is wrong to put the public values in the context of discovery; with respect to normative theories (such as political theories), the values should be introduced in the context of justification.

Locking ourselves in an ivory tower is a problem. As Avner de Shalit [2021] draws attention to, despite the commitment many political philosophers have to democratic institutions, political philosophers seldom appeal to the demos themselves. This seems problematic; if these philosophers are normatively committed to institutions which privilege public values, why are the public not incorporated into the process of political philosophy itself? However, while I agree that there is a problem such that these public appeals are merited, I disagree with de Shalit about their place. In doing so, I sketch an important contrast between scientific and normative theories.
To set the stage, consider a distinction influential in the philosophy of science: the difference between the context of discovery and the context of justification. This 'context distinction' was introduced by Hans Reichenbach of the Berlin Circle to indicate parts of the (idealised) scientific process. Roughly speaking, the context of discovery involves investigating phenomena and gathering evidence while the context of justification involves the logical processes of linking the evidence to conclusions or the construction of arguments for substantive conclusions. 1 While there are both objections to the distinction [Nickles 1985] and controversies about how to understand it [Shickore and Steinle 2006], it has provided a basis for discussion in philosophy of science throughout much of the twentieth century [Hoyningen-Huene 1987]. While de Shalit does not appeal to this distinction, I believe using it to understand the process of theory choice provides both clarity for his own view and a framework to explain the contrasts with my own view.
The method that de Shalit suggests to incorporate public values is to consult the public and use the results as inputs into theoretical consideration, in what he calls 'public reflective equilibrium'. In terms of the context distinction, starting with public values instead of the theorists' own is analogous to including public values in the context of discovery (i.e. during hypothesis creation or evidence gathering). In contrast, I will defend the claim that public values should come in at the point of political theory comparison or selection, a point in the process analogous to the context of evaluation (i.e. during the choice of whether to adopt particular substantive claims or conclusions). Indeed, I believe that this could be generalised and that the place for public values is in the context of discovery for normative theorizing in general.
My first objection to de Shalit's position is his suggested method does not accomplish his own goals: trying to meet people in 'a manner which is as random as possible' is not a rigorous way to come to understand the political values which shape our institutions. His method would be neither representative nor systematic. What we would need instead is rigorous analysis of the values and-importantly-the extent of their influence, neither of which is achievable by philosophers asking random people they meet normative questions. However, I believe that de Shalit would be unreceptive to changing his method to a more systematic alternative like representative sampling. That is because such changes would compromise other goals of his; resultant responses would be very coarse-grained and would not admit of Socratic interaction.
Sometimes, de Shalit lowers the ambition of his methodological goals by suggesting that the intention is merely to 'inspire' normative theorizing or to guarantee that the theories 'relate to' public institutions and their justifications. These desiderata are much less demanding than trying to determine what values shape shared institutions. However, these minimal desiderata could be met in all kinds of ways: most obviously, one could come to understand the theories, such as they are, behind public institutions by interviewing or reading interviews with influential politicians and stakeholders. One could even be inspired by reading science fiction or fantasy literature (both of which contain stellar examples of theories of social change and political theory). These will still be less systematic than the social science approach to discovering the views of polities, of course.
However, even if de Shalit were to systematise his suggested method so that the inputs were properly reflective of the public at issue, I would still have a second, more fundamental objection to his method: it unduly limits the importance of public values by restricting them to inputs. This objection will require some unpacking. In order to do so, I will construct a version of his argument and then explain how his solution is unsatisfactory. While I am framing this in a somewhat different manner to de Shalit, I believe these claims to be wholly consistent with his picture.
First, we could grant to de Shalit that normative theorizing, of which political theorizing is a species, requires appeal to values. This contrasts with science, since science, unlike normative theorizing, is built on an empirical basis. In this sense, we can say that normative theorizing grants space for legitimate disagreement without the possibility of demonstrable falsehood and that this is unlike a standard picture of scientific reasoning.
One way of thinking about this is that normative theorizing often takes place within, or makes fundamental appeals to, first-order normative theories such as deontology, virtue ethics, or consequentialism. Roughly speaking, such first-order normative theories adopt or draw attention to certain aspects of the moral world (we can be inclusive by saying that particularist theories draw attention to all of the aspects). This is a schematic simplification, but it helps illustrate the actual heterogeneous landscape of normative starting points.
Insofar as this is the case, and assuming that there is space for reasonable disagreement when choosing among first-order normative theories, normative theorizing takes place against some background values. There is, in this manner, an unavoidable appeal to some normative values [Mintz-Woo 2021].
Let us grant his claim that the values of political philosophers should not be privileged in this process, either for practical reasons (arguments are more readily applicable or more convincing when addressing shared assumptions) or for theoretical reasons (some are sceptical that the substantive values of philosophers are more likely to be true simply due to their profession). On whichever basis we grant or accept his claim, we see an important difference between normative theorizing (as in the case of political philosophy) and non-normative reasoning (as in the case of scientific reasoning): while the non-normative evidence is (perhaps) properly the domain of scientists, normative values seem to be not particularly limited to normative theorists. This distinction could explain de Shalit's intuition that political philosophy differs from other philosophical areas: when a metaphysical or philosophy of mind question is at stake, most people have no idea what to think about it, whereas when a political philosophy question is at stake, most people have an opinion and do reason about it. [de Shalit 2021: 13] Whether normative discussions should include the public is a live debate, and it is discussed by many, including in attached commentaries. However, I am going to grant to de Shalit arguendo the point that the values of political philosophers may not have a privileged role and consider solutions.
His solution is to bring those values in at the ground level by randomly talking to people in society. However, this does not play to the strengths of the philosopher (or at least to many philosophers)-few of us have the wit and wisdom of Socrates, say; the strength of the philosopher lies in forming valid arguments and in drawing surprising conclusions. An additional challenge is that is it not clear who counts as part of society.
However, the most important objection to this solution is that these public values are only brought in as considerations that enter into the theorizing, that is, as inputs. That artificially limits their importance in the public reflective equilibrium process. Instead, the more systematic-and less contingent-method is one where the philosophers produce arguments and make them available, but where the arguments are not evaluated by philosophers (alone) but the public. Philosophers generate and publish these arguments, but it is up to both decision-makers and the public to evaluate the arguments.
On this model, political philosophers work out the implications of various theories (and principles)-whether they accept those theories or not. Other political philosophers determine whether the suggested implications do indeed follow from the theories under consideration. Some of these implications may be found to be consistent with pretheoretical intuitions and others may be found to be, to a greater or lesser extent, counterexamples. For instance, some may draw implications about political inclusion of non-human animals by expanding the scope of fundamental rights [Donaldson and Kymlicka 2012] and some may draw implications for climate policies from normative principles about global commons problems [Mintz-Woo 2018]. It is up to other normative theorists to determine whether these putative implications actually follow. If they do follow, it will be up to others-including the public-to see whether these packages of theories plus implications support or undermine the underlying theories, including in comparison to other packages. 2 This is therefore also a version of public reflective equilibrium, but less limited than de Shalit's model (we could call it public final reflective equilibrium); the public is able not only to enter into the equilibrium by evaluating the premises, but by considering the weight of the arguments as a whole. We should expect there to be value-laden things to say about conclusions as well as premises in the context of justification and-more importantly-we should expect there to be interesting things to say about arguments when evaluated in juxtaposition to each other. What looks like a strong argument from some first-order normative premises might look relatively unmotivated or implausible compared to another. Or we might find that different arguments converge on similar conclusions, something which should strengthen our credence in those conclusions; recognising this is not so easily achieved by only considering what follows from a given set of pre-existing (public) values.
One objection to this is that if the public values are brought in the context of justification only, then in fact we will get arguments that are grounded in first-order normative theories which are of interest to, or considered plausible by, political philosophers. 3 It is very likely that these theories may diverge greatly from political views held by the public. If philosophers are not working with a representative group of theories, they may generate a biased distribution of arguments. While I have advocated bringing in those values in the context of justification only, I have two responses to his objection. The first is that it is not necessary for the distribution of arguments to reflect their prevalence in the public to perform this justificatory process. As long as all plausible positions are in contention, it does not matter whether some have been worked out in their implications more than others, since we can still evaluate a given set of packages of theories plus implications against another. Indeed, I suspect that including the public in the context of justification would help prune some of the more outré principles that philosophers endorse-as more philosophers point to more and more implications of a given theory, the potential counterexamples often mount in such a way that what was initially merely counterintuitive becomes straightforwardly absurd, at least for a non-philosopher evaluating that theory.
However, someone worried about this might respond that not including the public in the context of discovery could lead to some theories not being in contention. If the view is not even considered, then it cannot be evaluated in the context of justification; this would be extremely costly if the view was plausibly true! Here, I think we enter a theoretical trade-off. The question is how much theoretical space is unexplored if we do not talk to the public before drawing implications versus how much extra cost is involved in involving the public in the context of discovery. My personal view is that most of the views considered by the public will be discussed in the general philosophical discussion, even if less usually endorsed among philosophers, so not too much theoretical information is lost by not incorporating the public in the context of discovery. Taking my stand also comes with the theoretical simplicity of involving the public at only one point-albeit a more important point. However, I am somewhat agnostic on whether to include the public in both contexts as opposed to only the context of justification since the gains and losses in this trade-off may be ultimately turn on empirical information. What I am committed to is that they should be involved in the context of justification.
While de Shalit is concerned that philosophers respond mostly to each other, my suggested model takes intersubjective agreement as a strength and removes intersubjective disagreement as a barrier. There is nothing inconsistent with various theorists working from or with different first-order normative theories. In fact, a theorist can work with premises that she takes to be suspect in order to tease out whether her worries really follow from the theory under consideration. Focusing on values held by others limits the potential scope for fruitful and creative arguments. Obviously, this is not to gainsay the minimal claim that inspiration can come from random citizens. However, inspiration can come from anywhere, and I am disagreeing that there is a special place for inspiration from random citizens. Most importantly, I am endorsing the philosophy of science conclusion that the context of discovery should (or at least legitimately can) be driven by the interests of the theorists. In short, wherever a theorist thinks she can show that a particular theory has a surprising conclusion, she should do so, whether that argument is based on theories she endorses or her interlocuters endorse or random citizens endorse or whatever may be the case. But limiting her basis for theorising to public values, I suggest, is both undermotivated and arbitrarily narrow.
This also generates several interesting distinctions between normative theorizing and scientific reasoning. First, scientific theorizing does not involve this kind of public appeal at the point of evaluation; if the conclusions are validly drawn from the evidence, we do not think the evidence is subject to reasonable disagreement in the way that first-order normative theories may be. Roughly speaking, this could be justified by the intuition that, if the evidence is empirical, it does not make sense for the layperson to dispute it. In contrast, if the 'evidence' is first-order normative theory, it may well be a reasonable subject of disagreement.
Second, many of the empiricists who endorsed the context distinction were concerned about values appearing in the context of justification. They thought, as de Shalit may mean when he writes 'a perfectly scientific manner', that justification should be done without involving the values of the scientist. Some thought that the justification process would, ideally, be purely formal or logical.
In contrast, when considering normative theories, there is less reason to think that justification is a purely formal or logical process. Political and other normative theories should be considered in conjunction with each other, both because choosing theories involves substantive evaluation of the theories, but also values in weighing theoretical commitments contained within one theory against another.
There are two upshots to this proposal. The first is that we can see a way of incorporating public values where they are both (a) necessary and (b) contested, but informed by philosophical argumentation and rigour. Instead of making the theorist the ultimate arbiter of the reflective equilibrium process, this makes the theories open to public evaluation. The second is that it inverts the traditional philosophy of science conclusion, which is that contexts of discovery must be value-laden and need to incorporate values but that contexts of justification ideally are not valueladen. In normative reasoning, we get an interesting inversion from empirical reasoning, where-even in the ideal case-justification should involve public values.

Disclosure Statement
No potential conflict of interest was reported by the author.