OPR IV.13 The Reasons One Has

As Keith pointed out in his last post, Section 13 is one of the most important, and likely to be one of the most controversial, sections of The Order of Public Reason.  Although there have been a lot of controversial points made in earlier sections, much of the last several sections may have seemed more descriptive than normative and, therefore, less threatening.  This, however, is the section where the normative rubber hits the descriptive road.  The account that Gaus develops here represents a continuity that goes all the way back to at least Value and Justification and is continued in Justificatory Liberalism, those that are familiar with these earlier works will find much that is familiar here, though often in a different or expanded form.  There is a lot in this section so I won’t necessarily be discussing everything, only the most important points.  If there is something that I leave out that is important, we can definitely expand on those points in the comments.  I will spend most of my post today presenting Gaus’ view and save my critical remarks for Wednesday’s post, but feel free to bring up any critical issues that you may have in the comments.

Gaus begins this section by noting that the idea of social morality that he has been advancing seems to rely on an internalist conception of reasons.  As he claims, the debate between internal and external reasons has become a kind of obsession.   Gaus’ conception of reasons stakes out a position in that debate in opposition to what Joseph Raz has called the Classical View of external reasons as facts about properties of action.  On the Classical View, internal reasons are merely beliefs about facts, not reasons themselves.  The internalist, however, argues that reasons for action must be connected to the motivational set of an agent, that is, to her beliefs and desires about the action in question.   The debate between internal and external reasons has, according to Gaus, become confused.  The debate is really about what reasons there are or what reasons exist.  It is, therefore, really a question about the ontology of reasons.  But, as we have seen throughout the Order of Public Reason, social morality in a world of constrained and embodied reasoners is about what reasons we have, not with what reasons there are.  Gaus’ theory then is only inconsistent with externalism if the externalist also holds an externalist theory of what it means to have a reason, basically that to have a reason is for there to be an external reason that applies to that person.  Gaus rejects this form of externalism, the externalist view of having a reason, as implausible.

Externalism about having reasons is implausible because “it misconstrues the relation between having a reason and being a rational agent.” (233) Consider Aristotle, for instance.  The externalist is committed to saying that Aristotle had a reason to accept the truth of particle physics; but surely, Gaus argues, to think that Aristotle had a reason to embrace particle physics is to make a serious mistake.   Reasons are justificatory, but there is no way that Aristotle could be justified in believing particle physics.  Similarly, to claim that a person or a group have a certain moral reason even though that reason is totally inaccessible to them, in the same way that Aristotle’s reason to accept particle physics is inaccessible, is to not only misuse the language of reasons but, more importantly, to misunderstand and “undermine the point of discourse about reasons and rationality.”(235) We use the idea of having a reason, according to Gaus, to make the actions and intentions of other people intelligible.  The externalist view of having a reason severs the idea of a reason from its role in explaining and justifying action.

Rejecting the externalist view of having a reason does not, however, commit Gaus to, what he calls, the Reason Affirmation Thesis that to have a reason is to affirm that one has that reason.  Affirming a reason is neither necessary nor sufficient to having a reason.   The neurotic may affirm reasons that they do not, in fact, have–crazy beliefs, that the world will end if one blinks for instance, do not provide reasons.  Affirmation is also not necessary to have a reason, all that is required is that there be, in the words of Bernard Williams, “a sound deliberative route” from the subjective motivational set that one has to the reason.  We might think of the Reason Affirmation Thesis as the idea that only the reasons that an agent actually claims to have at any time can justify action.  The rejection of the Reason Affirmation Thesis means that Gaus is committed to some amount of idealization of rational agents.  If we notice that we need to go beyond the actual reasons that agents claim to have, we need another standard of what counts as a reason.  The problem is that once we begin to idealize, we move closer and closer to the externalist view of having a reason.  Gaus cites Steve Wall who argues that once we begin to idealize we realize that “a fully rational person will affirm all, and only, the (external) reasons that apply to her.” (237) If Wall is correct; we will have backed into the externalist view of having a reason merely by idealizing.

Gaus argues, however, that Wall’s approach simply identifies rationality with ability to recognize all of the reasons that apply to one’s situation–this is merely to revert back to the original externalist view of having a reason, with all the problems which that view entails.  If this idealized conception of rationality were, instead, something akin to the idea of a perfect epistemic agent with full information, a rational Hercules, the externalist view would be viable again, though with a different justification.  Even this model of idealized rationality, the rational Hercules, however, could still not justify the externalist view of having a reason.  If we were transformed into rational Herculi, Gaus argues, we would only converge on the same set of beliefs and reasons under two conditions:

  1. If we “initially believe precisely the same things and had precisely the same values, ends, goals, and so on” (239)  or
  2. Our initial beliefs have no effect on what we come to believe once we are transformed into fully rational agents.

The first condition, that we initially believed exactly the same thing and had the same goals, values, etc. is so implausible that we can easily reject it; the second condition, that our set of beliefs are not causally linked to a discrete set of earlier beliefs, is also implausible.  Starting from different initial belief sets, merely idealizing our rationality and transforming ourselves into epistemic heroes will not guarantee or even tend to make our final beliefs converge.

Whatever the process of idealization that we choose, the fact remains that, epistemically, we begin from where we are and any process of epistemic improvement can only operate on the material of our initial set of beliefs.  Small differences in initial beliefs will, no doubt, be multiplied as we continue to refine and make coherent that initial set of beliefs.  Furthermore, coherence, as Gaus argues, “is more than deductive consistency and closure.”(241) The fact that we make our belief sets coherent by working on the material of our initial, differing beliefs all but guarantees that the convergence of our belief sets and, hence, reasons will be extraordinarily rare, even given radical rational idealization.  To add to this difficulty, the process of rational belief revision, even for an idealized agent is also path-dependent.  This fact leads to the conclusion that the, “criterion of a fully rational choice is indeterminate: depending on the order in which the choice is made, full rationality may lead to a variety of outcomes.”(243)

Here again we see two key themes that reoccur throughout the Order of Public Reason: path-dependence and indeterminacy.  In this case, the ideal of full rationality is shown to be indeterminate partly because of path-dependence, whereas later in the work we will see that path-dependence can often be a solution to indeterminacy.   Once we see that full rationality as a means to generating a determinate thing-to-do is, in fact, plagued by indeterminacy, we are led to reject what Gaus calls “the myth of full rationality.”  This is the myth that if only we were fully rational, that is, if only we could shrug off our epistemic limitations and have all the relevant information, we would all converge on the same morality, principles of justice, theories of government.  This conclusion leads us to another reoccurring theme, the fact of reasonable and inevitable disagreement.  Even fully rational agents would be led to endorse different theories of morality, justice, and government so an appeal to full rationality will not solve our problems in the normative realm.

The solution, as we will see Wednesday, is to realize that our moral reasons are not are not and cannot be dependent on some idealized conception of full rationality.  Rather than working from an idealized notion of rationality backwards to an account of what reasons a real moral reasoner has, Gaus will build up what it means to have a moral reason from the kind of reasoning that those of us who are not epistemic giants typically use.  This strategy follows a similar strategy that we saw in the last several sections.  Gaus built his conception of moral rules and rule following out of our practices of rule following.  He will attempt a similar project with regards to moral reasons in this section.  On Wednesday we will see his account of “real rationality” and how it connects with the type of moral justification that we owe to one another.

This entry was posted in Books, Posts, Reading Group and tagged , , , , . Bookmark the permalink.

3 Responses to OPR IV.13 The Reasons One Has

  1. Mats Volberg says:

    I am having some doubts about the conditions when rational Herculi are guaranteed to converge.

    Let us imagine two detectives investigating a murder. The senior detective has slightly racist views and thus believes that the Mexican gardener did it, and there is some evidence to suggest it. The junior detective has just went through a divorce and thus believes that the wife of the victim did it, and there is some evidence to suggest it.

    But their current beliefs (values, goals, and so on) make them blind to certain other pieces of evicence which prove that the gardener nor the wife could not have done it, and that the best friend is the culprit. So if both of the detectives become rational Herculi then they will see that evidence and converge, despite starting with different initial sets.

    (I understand that in the moral case this presumes that there is “the right” position and that morality could not be plural, such that there are more than one different but incompatible moral position. And that Gaus argues for the plurality of morality in the appendix, but if his wrong there, could that mean he is wrong here?)

  2. Gaus needs to be able to say that agents can have false beliefs and defective epistemic norms (26). Otherwise, given the number of people with crazy or repellant beliefs and defective epistemic norms and heuristics, no rules could possibly be justified, much less morally attractive ones. Some idealization of actual people is necessary. In this section (13.1), he appears to idealize actual agents to the extent that he requires that 1.they not have crazy beliefs, 2. that they have reasons to which they are committed by sound deliberative routes from their current, non-crazy beliefs (236), and 3. that they are committed to coherence as a set of epistemic norms (240). My worry here is that it seems to me that any idealization conflicts with his argument against externalism. His argument against externalism as a theory about what reasons an agent has appears to be that the concept of reason is tied to our efforts to understand and explain other people’s actions. To the extent that we understand other people’s actions by understanding the reasons which could plausibly be said to motivate them, then the reasons one has must consist of the set of their current beliefs (future beliefs, even those arrived at by deliberation from current beliefs, cannot motivate one’s actions or justify one’s beliefs now). The force of Gaus’s examples of Aristotle not having a reason to believe modern theoretical physics or the pre-modern Eskimaux nothaving reasons to refrain from genocide depend for their force on the way in which both Aristotle and the Eskimaux could have no access to these reasons, since they depend on concepts (e.g. quantum mechanics, race, ethnic group), experiences (modern experimental results, Holocaust), etc. which did not exist when the agents were thinking and acting. Such concepts could not, then, possibly motivate their actions or influence their beliefs. Idealizations of agents are not part of the attempt to understand their beliefs and actions; at most they can be part of an attempt to understand that part of an agent’s beliefs and actions which could make sense, e.g. showing how a person’s fear of horses makes sense if one sees how he had (mistakenly) come to think of all horses as stand-ins for his father. Rather, idealizations are parts of efforts to judge agents or to make claims about what they ought to be, believe, and do. To the extent to which Gaus idealizes agents, he opens the door, as Wall notes, to further idealization, which may result in imputing beliefs and epistemic practices to them which they do not have, which they may not recognize, and which could not therefore explain their beliefs and actions. It seems to me that once idealization enters the picture, we no longer have a debate about what reasons agents have or don’t have, but about how rational we can expect most agents to be – how much deliberation can we expect of them for example; what the norms of rational agency are – is the principle of conservation a good epistemic norm? Is path dependency an insuperable limitation on rational agents or can we expect them to overcome it at times (it seems odd to think that it is justifiable that the temporal sequence in which one encounters evidence for a claim should influence one’s overall judgment about the truth of the claim); and what people ought to believe (are all horses my father?). We are no longer asking what reasons an agent has, but what he ought to have given the limits of what he could have. In the abstract these limits are constraining, e.g. an agent cannot master all of the information available on any particular question or draw all of the deductions from his current set of beliefs. But about any particular piece of information, deduction, or moral or political controversy, there seems little that any particular agent could not end up believing Someone could get a Ph.D. in political theory and believe whatever it is that any particular theorist believes or read all the books his interlocutors have. He could come to see his beliefs in God or the free market or a socialist society as childish fantasy if given sufficient time and exposure to cogent arguments. If this is the terrain of the debate, then the point of the recourse to internalism, to tie moral and political justification to the perspectives of actual people, is largely lost. Rather, theorists should just argue about what people ought to believe by showing what reasons there are.

  3. Does Jerry offer a characterization of a reason (for belief or action) anywhere? I can’t find one.

Leave a Reply