The (Ir)relevance of Truth to RationalityThe main purpose of my dissertation is to argue that rationality is insensitive to the truth in significant and surprising ways. The three main claims I aim to establish are that: [1] that in order for an agent to act for the reason that r, it need not be the case that r; [2] in order for an agent to act for the reason that r, the agent must believe that r; and [3] in order to explain an action done for the reason that r, we need only suppose that the agent believes that r. I think that the most plausible picture of both practical and epistemic rationality will be one that packages together and traces out the consequences of these three claims.
While much of the project concerns acting for reasons, the last two chapters of the dissertation start toward transplanting this package of views into epistemology, and examining potential implications for normative reasons. The major implication, I claim, is a view on which normative reasons, like motivating reasons, need not be facts; and a theory of epistemic reasons on which neither evidence nor the basing relation are factive. Read below for more specific details and chapter summaries, and please email me for manuscripts! |
Chapter I |
Acting in the Light of a Falsehood. I start out in this chapter by getting back on the table a view widely thought to be a nonstarter in action theory. This is the view that in order for an agent to act for the reason that r, it need not be the case that r. I argue that conceptual analysis and casuistry suggest that this Nonfactive View is actually the intuitive view. However, theorists have been persuaded away from it (and toward some kind of Factive View) by the combination of two thoughts. The first is that the reasons for which agents act can play an explanatory role with respect to actions; and the second is that something that is not the case cannot explain anything. It is a common thought that these two axioms taken together are inconsistent with the Nonfactive View. But I demonstrate that this reasoning is faulty, and that the Nonfactive View is consistent with these axioms. To meet the heart of the objection (that the Nonfactive View cannot give good reasons explanations in certain cases), I begin to illustrate some of the explanatory resources the Nonfactive View has at its disposal – but I put off giving a fuller illustration until the end of the third chapter. In the first chapter I content myself with concluding that, since the Nonfactive View is the intuitive view, and the best argument against it fails, we should endorse the Nonfactive View: it is possible to, as I say, act in the light of a falsehood. While I think that this Nonfactive View holds for both acting and believing for reasons, I restrict myself in this chapter to action.
|
Chapter II |
Acting in Unbelief. A complete theory of acting for reasons will not just declare that reasons need not be facts, and show how reasons can play a role in explaining actions. It will also answer other significant questions about the nature of action and reasons. One significant question concerns what cognitive constraint an agent must stand in with respect to some reason in order to act for that reason. This is the question that I focus my attention on in the second chapter of my dissertation. Though I think this is an important question in actin theory, it has yet to receive any direct and independent treatment; so one of my two main aims in this chapter is to initiate and set the parameters for a thorough discussion about the cognitive constraint.
After distinguishing the issue of the cognitive constraint from surrounding issues, I set about arguing for the view that: in order for an agent to act for the reason that r, the agent need only believe that r. I begin, in similar style to the first chapter, by observing that conceptual analysis and casuistry suggest that this Belief View is the plausible view. I then consider arguments that the Belief View is too weak – we need a stronger cognitive constraint on acting for a reason, such as knowledge. I show that, of the two main arguments given in favor of the Knowledge View, one is invalid, and the other is unsound. I then consider the possibility that the Belief View itself is too strong, by tending to an apparent counterexample to the view. I suggest that the example can be explained away; and since we have strong theoretical reason not to weaken the cognitive constraint beyond belief, we should do our best to explain it away. Since, as I have argued, independent discussion of the cognitive constraint is new, I conclude by imagining potential implications of the Belief View for the rest of the literature on reasons. In particular, I argue that if the Belief View is true, then the view that reasons for acting are psychological states of agents is highly implausible. So, somewhat surprisingly, in settling the issue of the cognitive constraint on acting for reasons, we also find good reason to deny a significant and widely held view about the ontology of reasons. |
Chapter III |
Arguments from Error. In the third chapter I proceed to draw a lesson from the thoughts in the first two chapters. It is a common thought that theorizing about a certain class of cases, which have been called "error cases", is crucial to settling several central debates in action theory. Error cases are those, roughly, in which an agent errs in some way in arriving at a view about the way the world is, and then acts in the light of that view of the world. Theorists divide on how they respond to error cases. But despite their differences, many theorists claim to glean motivation for their views from these cases; and they all do so in the same way. They all claim that while their theory has no trouble giving good reasons-explanations of actions in such cases, other theories are explanatorily deficient.
In this chapter I demonstrate that this kind of argument is flawed. I examine several of these arguments and demonstrate that they fail. I proceed to argue that we can abstract away from the particulars of each argument to see that the kind of argument in question is flawed – so each instance of it is bound to fail. The fatal flaw is that there is quite a large gap between what can be said about error cases themselves and the quite general claims about acting for reasons that theorists claim these cases establish. Furthermore, this gap is one that can only be bridged by placing unmotivated and implausible constraints on action explanation. This being so, I suggest that theorists will have to look outside of error cases to settle the central debates in the theory of reasons. |
Chapter IV |
Motivating Reason to Slow the Factive Turn. By the end of the third chapter, I take myself to have established the three main claims I mentioned at the very beginning: the Nonfactive View of acting for reasons is true; the Belief View is true; and the combination of these two views is suggestive of an Anti-Psychologistic view of reasons that can give perfectly good reasons-explanations of actions. In the fourth chapter, things get a bit more ambitious, as the ultimate suggestion of the chapter is that the materials of these first three chapters give us the necessary resources for constructing a novel argument against the view that evidence is factive.
As factive views of reasons have been gaining favor in epistemology, theories of practical reason have been increasingly considering the possibility of nonfactive views. I aim to slow the factive turn in epistemology, as it were, by appealing to points about the nature of practical reasons. I begin by arguing that the best theory of reasons will hold just one position on the factivity of reasons, which holds across normative realms – practical, epistemic, moral, and others. I then proceed to argue that, if motivating reasons need not be facts (as I established in Chapter One), then normative reasons need not be facts. What follows is a connection that I suspect many will find surprising: these two claims together show that in order to establish the view that normative epistemic reasons (or "evidence," if you like) must be facts, one must discredit the view that practical motivating reasons need not be facts. I treat two serious objections which try to tease apart rationality and reasons-responsiveness. If there are no better objections than the two that I see off, I conclude, we have found a novel argument against the view that normative epistemic reasons must be facts (and, by extension, the view that normative reasons in general must be facts). |
Chapter V |
A Practical Theory of Epistemic Reasons. I conclude in the fifth chapter by completing the epistemic transplant that the fourth chapter began. So, I aim to defend the views that: [1] in order for A to believe for the reason that r, it need not be the case that r; [2] in order for A to believe for the reason that r, A need only believe that r; and [3] we can explain any belief-for-the-reason-that-r by appealing to r. As is the case with practical reasons, I think that these three views are highly suggestive of a view on which motivating reasons for belief are not psychological states of agents, and on which normative reasons for belief need not be facts.
Now, if the claim of uniformity that I argue for in Chapter Four stands, I get all of these claims for free (just by the work done in the first three chapters). But I take some time in this chapter to lend some independent plausibility to these claims. I begin by suggesting that the debate about the nature of acting for reasons mirrors debates about the epistemic basing relation. So, I argue that the epistemic basing relation is not factive: in order for A's belief that p to be based on some reason r, it need not be the case that r. I suggest that this has some interesting consequences in epistemology – for example, it follows that the basing relation cannot be causal. I proceed to argue for the Belief View about epistemic reasons: in order for A's belief that p to be based on some reason r, A need only believe that r. I then show that, as is the case about reasons for acting, this all lends itself to a very natural picture on which beliefs-for-reasons can be explained by appeal to the reasons in question. I then turn my attention to the normative dimension. If what I said in Chapter Four is correct, there is a novel and compelling argument that evidence is not factive. That is: in order for some proposition e to be evidence for some proposition h, it need not be the case that e. I suggest that this is tantamount to claiming that propositional justification can have its source in falsehoods; and since one can form a belief for the reason that r when it is not the case that r, it follows that doxastic justification can also have its source in falsehoods. This straightforwardly puts me at odds with what has been called the "Knowledge-First" program in epistemology – but not (as I argue) for all the usual reasons. In particular, I conclude by demonstrating that my theory here is consistent with (and indeed, I endorse) externalism about epistemic reasons. |