ISSA Proceedings 1998 – Democratic Justice, Argumentative Dialogue, And Political Legitimacy
My aim in this paper is to address some links between argumentation theory and political theory. Practitioners in both areas share an important element of common concern, namely, identifying the conditions of rational argumentative dialogue. On the one hand, argumentation theorists have offered models idealizing a preferred structure of discussion aimed at reaching a reasonably well-defended position on some subject, while on the other hand, some political theorists have been concerned, over the past decade or so, to think about social deliberation as part of a defense of democratic legitimacy and social justice. In the present context, the interest of the latter idea, for both sorts of theorists, is that an appealing conception of legitimacy or justice for modern democratic societies might be developed by focusing on the idea of a rational democratic discussion.
My more specific aims are as follows: first, to explain the immediate background in political philosophy to the current concern with the links between dialogue and justice (i.e., John Rawls’s approach and its problems); secondly, to clarify the reasons for thinking that democratic legitimacy is best understood by reference to a model of social discussion; thirdly, to register a general claim about the material preconditions for meaningful participation in democratic discussion aimed at reaching decisions about the terms of political association; and finally, to address several objections to the idea that a model of “deliberative democracy” is at all relevant to our self-understanding as citizens in modern democratic societies.
2. Rawls’s Contractarian Argument and Beyond
In 1971, John Rawls’s book, A Theory of Justice , was published, immediately reinvigorating political philosophy and initiating a series of debates about justice and political justification that have continued to this day. Rawls’s achievement consists in two different variations on some old themes: first, he offers substantive principles of justice, attempting to show that liberty and equality are compatible moral and political values, and secondly, he defends those principles, in part, by means of a social contract argument. For our purposes, it is this argument – rather than Rawls’s specific conclusions – that is the jumping-off point for my discussion.
Rawls’s argument appeals to a hypothetical contractual situation in which individuals choose principles from behind a ‘veil of ignorance’, a device designed to rule out bias and therefore ensure impartiality in the resulting principles. The problem is to choose principles of justice to underpin the main social, economic, and political institutions for a given society, and Rawls’s argument is that we should imagine what principles individuals would choose if they did not know anything about themselves that would enable them to tailor the chosen principles to their own advantage. The principles that would be chosen in this hypothetical so-called ‘original position’ are the principles we should accept because the choosing situation is designed to cohere with our considered judgements about the requirements of justice. One such judgement is that justice is closely linked to impartiality , another is that a person’s life prospects should not be determined by their good or bad luck in respect of natural abilities or social circumstances (Rawls 1971: 18-19).
According to Rawls, as I have said, persons in the original position are situated behind a ‘veil of ignorance’, thereby preventing them from knowing precisely who they are. (They do know some general facts and theories about human psychology and social structures, and they know that their society is characterized by moderate scarcity of resources and limited benevolence of individuals.) Accordingly, in the original position there is nothing to distinguish one person from anyone else: if anyone has a reason to prefer one principle of justice to another, then everyone has that same reason. Hence, at this stage, we are presented with a problem of rational choice, and Rawls (controversially) believes that a rational chooser would adopt a ‘maximin’ decision rule, focusing only on the worst-off position in any resulting social framework and preferring
that framework in which the worst-off are better off than they would be under any other arrangement.
The most important thing for us to notice about Rawls’s account is that its goal is to provide a means of defending principles of justice everyone can accept, yet it does so not by appealing to everyone to participate in a dialogue about justice but by adopting the standpoint of any person selected at random while behind the veil of ignorance (Rawls 1971: 139). So Rawls’s argument is ‘monological’ in form, that is, the thought experiment puts everyone on an equal footing so that each person will reason in precisely the same way. But there are two important objections to this monological approach: first, it fails to account for the different perspectives individuals and groups legitimately bring to the ongoing public dialogue within contemporary pluralist societies, and secondly, it does not provide a satisfactory link between justice and democracy in the justification of the central social institutions of a society, a link whose importance stems from the claim that both justice and democracy are bound up with the morally prior notion that each person’s interests are due equal consideration. These objections point the way to an alternative conceptualization of how to defend principles of justice. On this view, the justice of an arrangement is connected to its legitimacy, and legitimacy in turn is best thought of as arising from a deliberative process in which each person has an effectively equal say in determining their terms of association. Hence we should turn our attention to the topic of ‘democratic justice’ and to the egalitarian dialectical process that is sometimes taken to be a necessary condition of political legitimacy.
3. Democratic Legitimacy and Argumentative Discussion
Democratically organized discussion and deliberation are valuable because they help individuals better understand their own interests as well as the interests of others (Christiano 1996: 84). Moreover, where a society’s institutions enable all citizens to discuss matters of public concern, people will be better able to exercise the equal power that is suggested by a prior moral commitment to the equal consideration of persons: in short, someone has power when they actually know which policies will promote their interests (Christiano 1996: 85), and such knowledge is most likely to be gained by everyone when discussion and deliberation are open to all. In a democratic society, each citizen should have an equal say in determining the society’s overall aims. Ideally, this means that deliberation about public policy should be modelled on a discussion procedure that is both rational and egalitarian. I want to address the question of whether such a model is relevant to large, technologically advanced societies, but first we need to know a bit more about this model of democratic deliberative discussion (I direct the reader to two of the more helpful accounts in the recent literature, on which I base much of what follows: Christiano 1996: 116-28 and Cohen 1989). Citizens gain information through social discussion and deliberation in which individuals and groups communicate with each other with the aim of reaching a consensus. But if this process is to be legitimate, the consensus reached (or the process whose ideal end-point is the reaching of consensus) must be in line with certain criteria for procedural legitimacy; otherwise, the so-called “consensus” will lack normative force.
What are these criteria? At the very least, participants must appeal to reasons acceptable to anyone, regardless of their social position, class background, natural talents, and so on. This criterion rules out what we might call ‘persuasion by coercion’, that is, giving someone a reason to adopt a position by threatening good or bad consequences if they act in one way rather than another. The idea here is related to the rationale behind Rawls’s veil of ignorance: when self-interested people know certain specific facts about themselves, they can be expected to reason so as to use that knowledge, perhaps to the detriment of others. When a reason is a reason only for someone in a given social position, the danger is that discussants will engage in bargaining aimed at maximally benefiting themselves, taking the interests of others into account only to the extent necessary to further one’s own interests. It should be fairly clear that bargaining in full knowledge of one’s class position and natural talents, combined with minimal rationality directed at achieving one’s ends, will lead to agreement on terms of association that benefits members of the materially advantaged classes and those better endowed with socially useful talents. Since the socially and naturally disadvantaged have less bargaining power, they have to settle for less. But if we accept this bargaining picture, we commit ourselves to the legitimacy of a social arrangement in which people’s life prospects are largely determined by features of themselves – say, their class position or natural talents – for which they are not responsible. It is difficult to believe that the bargaining model is legitimate, for it more or less ensures that the society’s institutional structure results from a series of threats and offers. On the other hand, if we want to model social legitimacy we should not conceive of society as a bargaining procedure in which the parties aim merely to get as much as they can for themselves, regardless of the basic needs and interests of the other parties. The agreement reached by that procedure might be “rational,” in one sense of that much-contested word, but it will not be reasonable.
Participants, then, need to be able to recognize the force of the reasons given in the discussion. But they also need to adopt a certain attitude of mind. Specifically, they must be willing to listen to the reasons given by those with opinions different from themselves. Each discussant wants to persuade the others of the acceptability of his own view, but he must also be persuadable by the reasons offered by those others. Additionally, the aim of participants should be that people change their views on the basis of reasons offered, and not for any other reason. So rational social discussion is in this sense distinguishable from indoctrination in which the two-fold goal is to bring about some belief in others (regardless of the reasons there might be for that belief) and to close off those others to any future change in view (Christiano 1996: 117).
Another criterion for rational deliberation about political goals is that a range of views should be on offer. The need for a plurality of positions stems from the plural character of free societies themselves: it is highly unlikely that a social arrangement will be legitimate if it fails to address the concerns and perspectives of the diverse viewpoints that develop in the context of free deliberation. One of the problems with the Rawlsian thought experiment with which I began is that it does not make room for this plurality at any basic level. Bringing in a range of views has one important consequence, for our purposes: namely, conflict between perspectives will be inevitable, and there is the persistent worry that consensus can never be reached. I address this problem later on.
Another criterion for reasonable social discussion is what we might call universal comprehensibility. This is the idea that every citizen must be capable of following the arguments given in the process of deciding upon social goals, and that positions are adopted (ideally) on the basis of reasons everyone understands. Yet another criterion is efficiency (Christiano 1996: 118). By this I mean simply that the deliberative process should not take up so much time and effort that the citizens lack the time and energy to pursue other socially useful tasks or purely private activities. Moreover, discussion on any given topic should not consume so much time that other, equally valuable subjects are not discussed.
Perhaps the most interesting criterion is that the process should be guided by the reasons offered (Christiano 1996: 119 and Cohen 1989: 22). Positions should be adopted when they are supported by the best reasons. Despite the obvious importance of this criterion, I will not say much about it here, since the difficult work of deciding which reason is best in a particular instance will likely appeal to considerations at least partly tied to whatever subject-matter is in dispute. (I say, “at least partly,” rather than “wholly,” because any discussion must meet certain general, context-insensitive criteria such as consistency, openness to different viewpoints, and so on.)
Still, what counts as a reason must be a reason that anyone can reasonably accept and that, where expert knowledge in some subject-area is relevant, the experts’ consensus figures centrally. So much for the outline of general criteria for rational democratic discussion aimed at determining a society’s central goals. I want to turn now to consider a necessary condition for implementing such a model of discussion in contemporary societies, namely, the achievement of a roughly equal level of material well-being. We will see, however, that meeting this condition may be necessary, but it is certainly not sufficient for the legitimate reaching of agreement on terms of political association.
It seems fairly clear that the democratic ideal of political equality cannot be realized where there is persistent material inequality. It is more or less impossible for individuals lacking a reasonable share of social and material resources to make their voices heard, especially in societies where relatively few people possess vast resources enabling them to wield great influence upon both the main media of opinion and political representatives. Notice that this state of affairs can persist even where there is no coercive interference with political expression or association; it is simply that some groups speak “so loudly and so much as to deny an effective hearing to contrary voices” (Fishkin 1992: 161). Here the underlying concern for equal consideration and respect should lead democrats to favour (one version of) principles of distributive justice ensuring not only equal civil and political liberties but the material prerequisites for making those liberties practically meaningful. Although this claim is controversial in some quarters, I now will assume its truth in order to focus on a deeper problem for democrats.
4. The Problem of Pluralism
The problem is this: even where everyone had a substantively equal say in democratic deliberation, we would still live in a society “characterized by moral pluralism, and so [we] must contend … with disagreements rooted in differing conceptions of the aims and purposes of human life, and in different allegiances and attachments, differences that can lead to deep and enduring conflicts” (Moon 1993: 86). One way to counter this problem was suggested by Jean-Jacques Rousseau (Rousseau 1762): we could ensure uniformity of opinion by way of a strict regime of censorship along with other mechanisms such as a civil religion that serves as a focus of patriotic devotion. This approach is unacceptable in large, modern societies – Rousseau himself favoured small, simple societies for precisely this reason. In any case, if, like Rousseau, one is concerned foremost with freedom, it seems odd to try to achieve it by setting up an institutional framework in which freedom is directly and intentionally curtailed.
The pluralism problem generates an objection to the model of democratic deliberative discussion. The objection is that the consensus at which the model aims is simply not possible. There is a vast range of opinion on matters of social policy, for instance: think of policies on welfare, taxation, education, and health, not to mention abortion and euthanasia. In the end, I think it is fair to say that if achieving consensus is required by the model, then the model must be rejected. But the obvious reply is that the model need not deny the persistence of deep-seated conflict of opinion about matters of public concern. The utility of the model is not that it promises to settle all conflicts; rather, it is that it might lead to the acceptance by the majority of citizens that such conflicts as are irresolvable are nonetheless defended on all sides by people who can and do appeal to reasons with some persuasive force, even if those reasons are reasonably rejectable by those who do in fact reject them. An additional benefit of the model I have been describing is that it institutionalizes free and open discussion in a way that is aimed at generating respect for co-participants in the process. Consequently, where conflict of opinion is not resolvable, there is greater likelihood that peaceful means will be used to change the views of one’s opponents.
But there are alternative discussion models, distinguishable by their characteristic handling of the pluralism problem. I want to focus now on one of the more interesting alternatives for dealing with the pluralism of modern societies. This is the approach favoured by Bruce Ackerman. On Ackerman’s view, a just society institutionalizes a public dialogue characterized by what he calls “conversational restraint” (Ackerman 1989). According to this strategy, conflicting and deeply-held moral ideals should be excluded from public dialogue. Instead, the emphasis should be on those beliefs shared by all participants. In this way, no one will be forced to impose on others views those others may reasonably reject. (Remember that directives backed by state authority are ultimately supported by the force of the organs of the state; the aim is to maximize the extent to which that inevitable threat of force is further sustained by reasons the citizens can accept, reasons unrelated to that threat.) Ackerman’s model is compatible with the idea that the terms of political association should be freely accepted. But it fails to satisfy the criterion (mentioned earlier) that views should be accepted on the basis of reasons offered in the discussion. The model prevents participants from questioning competing beliefs; yet if the aim is to pick out commonly held views, the model backfires, since it is precisely such questioning that can lead to the identification of common beliefs (Moon 1993: 77). That is, we identify the positions we share with others by defending claims they question, since we defend those claims by appealing to reasons we believe others will accept. Perhaps paradoxically, common ground is found by testing views that conflict with each other in order to see what can be publicly defended. Ackerman’s ‘conversational restraint’ model fails because it backs away from dialogue in precisely those contexts where dialogue holds out the only non-violent hope of reaching an accommodation between competing views. My tentative conclusion here is that we would be better advised to encourage civilized discussion about conflicting moral ideals, rather than pushing all such ideals off the public agenda.
5. Further Objections
I began with an outline of Rawls’s contractarian method for defending principles of distributive justice, and we were led to the idea of a democratic dialogue as a means of making up for certain weaknesses in the Rawlsian approach. There is a further respect in which the dialogue strategy improves upon the monological approach. The Rawlsian style of contractarian argument is sometimes accused of begging the question. Its purported basis is what would be agreed by individuals in a hypothetical contractual situation, but the principles of justice it aims to produce are not in fact adequately defended by appeal to hypothetical agreement: the correct characterization of the initial choice situation presupposes a substantive view about justice, hence all contractarian justifications of justice are viciously circular: one gets out only what one puts in, so skeptics of one’s substantive conclusions may reasonably reject such a method of persuasion. Moreover, a further problem with the contractarian strategy is that, being hypothetical, it cannot generate actual obligations to abide by the conclusions agreed to. The idea here is that real contracts generate obligations – think of a promise which, once made, obligates (at least prima facie) the promissor to do whatever it is she promised to do – but a hypothetical contract is patently unreal, so it couldn’t in itself generate anything. The movie mogul Sam Goldwyn – the ‘G’ in ‘MGM’ – is supposed to have said that a verbal contract isn’t worth the paper it’s written on; the present point is that a hypothetical contract isn’t worth the paper it’s written on. (Or, to be strictly accurate, a hypothetical contract isn’t worth the paper it’s not written on. See Hampton 1997: 66).
The solution to these difficulties is to conceive of one’s contract as actual rather than hypothetical, thereby enabling agreements reached to have real justificatory force. The deliberative dialogue is meant to be an actual process, so (at least potentially) it packs the relevant justificatory punch. The problem then is to point to a real form of agreement people reach that has the legitimacy-producing features I’ve pointed to in setting out criteria for procedural legitimacy. There are many reasons why we might think that any actual process of generating agreement cannot generate legitimacy; the pluralism problem is among the most difficult. I will close by mentioning another.
An obvious difficulty is that the level of participation required by this account of democratic legitimacy is too high to be realistic in the context of large modern states. This problem is akin to the difficulty with socialism once noted by Oscar Wilde: it would take up too many evenings. In fact, the objection might not be simply that the deliberative democratic ideal is too time-consuming: the problem might be more serious. For one might argue that, even if time could be set aside for discussions about society’s goals in which everyone participates, it might be the case that some people – perhaps the vast majority – are incapable of contributing to such a discussion. In short, the ideal of rational social deliberation cannot be the ideal for a democratic society in which every citizen is meant to contribute to the determination of social aims (Christiano 1996: 123).
In reply to this objection, one must temper the enthusiasm for large-scale participation that might have seemed to characterize the deliberative model. The aim is definitely for more participation than we now see, but it does not require that every detail of state policy be up for discussion. There are experts in many areas necessary to the formation of a rational social policy, and no one is capable of becoming an expert in all of those areas. However, the deliberation model should call for all citizens to be participants in the setting of overall aims for the society, and for this task they need only a general capacity to understand policy directives and institutional mechanisms (Christiano 1996: 169). Most importantly, every citizen is in fact capable of judging for themselves the effects of a policy or the workings of an institution: as Aristotle said in the Politics, Book 3, the best judge of a house’s livability is the person who lives in the house, not the person who built it. The political analogue of this argument is that, even if we rightly make use of experts in devising social and political institutions, citizens themselves should be allowed to decide whether some directive is satisfactory. And if their decision is to be rational and informed, everyone must play at least some role in following the reasons offered in public dialogue. Moreover, democratic participation has the potential for moral education of the entire body of citizens (Christiano 1996: 82-3). In short, the model of deliberative democratic discussion is far from perfect, but no other scheme holds out any hope of generating legitimacy in societies committed to equal consideration for all of their citizens.
One might question the empirical evidence, pro and con, relevant to the deliberative model we have been discussing. On the one hand, it seems implausible to say that this model has ever been implemented to any significant degree in a modern democratic state. But there is one indication that the model is worth further investigation. Consider, in closing, an article in The Economist magazine for May 16th, 1998, p61. The background is as follows: the American political philosopher James Fishkin has recently been defending something called a “deliberative opinion poll” in setting out his ideal of a “self-reflective society” (Fishkin 1992). Roughly, the idea is to get a representative sample of the population together for a long weekend, all expenses paid, and expose them to the details of a specific policy question. At the end of the discussion, in which experts are questioned and the participants debate amongst themselves, a “deliberative poll” is taken. Fishkin’s general approach has much in common with the abstract model I have been defending. It is an attempt to create a forum for discussion among citizens in societies where it is literally impossible for everyone to participate equally on every occasion.
The model would be of potential value if it produced results in conflict with a status quo in which prevailing views are generally in line with influential interests and often directly opposed to the public good. This is where the recent article becomes relevant. The Economist describes an implementation of this vision in the state of Texas. Fishkin himself organized the event. In the case in question, the aim was to discuss the spending plans of public power companies. How did this experiment work? Well, I will finish with a question: Would we have predicted that a representative sample of the population of Houston would opt for an increase in their energy bill in order to pay for wind turbines?
Ackerman, Bruce. 1989. Why Dialogue? Journal of Philosophy 86, 5-22.
Barry, Brian. 1995. Justice as Impartiality , Oxford: Clarendon Press.
Christiano, Thomas. 1996. The Rule of the Many , Oxford: Westview Press.
Cohen, Joshua. 1989. Deliberation and Democratic Legitimacy. In: Alan Hamlin and Philip Pettit (eds.), The Good Polity , Oxford: Blackwell.
Fishkin, James S. 1992. The Dialogue of Justice , New Haven and London: Yale University Press.
Hampton, Jean. 1997. Political Philosophy , Oxford: Westview Press.
Moon, J. Donald. 1993. Constructing Community , Princeton: Princeton University Press.
Rawls, John. 1971. A Theory of Justice , Oxford: Oxford University Press.
Rousseau, Jean-Jacques 1762. The Social Contract.