Practical Reasoning

I: What is practical reason?

0. Lecture Plan

1. What is practical reason?
2. Internal and external reasons for action
3. Practical reason and morality
4. Reasons, epiphany and aspiration

1. Two ways in which reason can be practical

Human beings have the capacity to think (deliberate) about what they are to do. Call this capacity ‘practical reason’ and the characteristic kind of thinking ‘practical reasoning’.

Thought of this kind is practical in two senses:

(1) Its subject matter. The thought is about action.
(2) Its consequences. Practical reasoning produces, or culminates in, action.

The idea of practical reason is at least as old as Aristotle; it plays a crucial role in the structure of Kant’s and Hume’s ethics. It continues to be at the centre of contemporary philosophical work in ethics and allied philosophical subdisciplines.

It is generally compared/contrasted to theoretical (or more rarely, speculative) reason. This is the capacity to think (deliberate) about how things are, what is the case, how the world is, etc.

2. Practical reason as a source of philosophical questions

What must be true of agents and their reasoning in order that their reasoning can give rise to action? In what sense, if any, is deliberation a form of reasoning? Is there deliberation about ends or only about means? What is it to act for a reason? Do our reasons for action depend on our desires? Or might there be reasons for doing things we have, and could have, no desire to do? (What’s at stake here is a bigger question: is morality rationally required?)

Practical reason as a distinctive point of view. When we (human beings, agents) deliberate about action, there is a distinctive attitude we take towards ourselves and our situation, a distinctive standpoint we adopt. What is distinctive of this attitude/standpoint?

3. Contrast with theoretical reason

We can think, reason, deliberate about things other than action: this is supposedly the province of ‘theoretical reason’, the kind of thing (say) natural scientists engage in. This example suggests an account of theoretical reason as essentially impersonal (i.e. it treats its subject matter in terms that are in principle accessible to any rational agent) and is concerned with things like description, explanation and prediction.

Description – e.g. ‘what is happening?’ – involves an answer to the question of how things stand with the world independently of my will. Explanation – why did this happen? why did s/he/you/I do this? – is typically retrospective, concerned with things that have already taken place. Prediction is essentially forward-looking – what is going to happen in the future? what will s/he/you/I do?

But practical reason starts with a distinctively normative question: what am I to do and why? what ought I do and why? what would be best for me to do and why? The ‘why’ here is not the why of explanation but of reasons and therefore, implicitly, involves values. In practical reasoning, we assess, weigh our reasons for action – considerations that count for or against some course of action. Further, practical reasoning is (or seems) essentially first-personal (what am I, or what are we, to do?)

Consider the question ‘What will/shall I do today?’ ‘Will’ suggests an attitude of theoretical reason and the question asks for a prediction (which sounds a little weird in its first-personal form: ask yourself, why?) ‘Shall’ suggests the distinctively practical stance; the answer will express not a prediction (‘I suppose this is what I’m going to find myself doing’) but rather an intention (‘I shall X’ = ‘I intend to X’).

4. Analogy with theoretical reason

Note that the distinction between descriptive and normative can be overdrawn. There is a sense in which even theoretical reason is concerned basically with a normative question: in asking (e.g.) how things stand with the world, we are implicitly asking the question: ‘what ought I to believe?’ And just like with practical reason, we form beliefs in response to (epistemic) reasons – evidence, inference, argument, interpretation, etc. Note that this too is, in a sense, first-personal. Given this analogy, the distinction between theoretical and practical reason can be understand a distinction between two sets of evaluative norms: one for assessing action, one for assessing belief.

The norms that govern the exercise of practical reason are centrally concerned with truth: one believes some proposition about the world iff, and because, it is true. This brings with it some other rational norms/principles: e.g. ‘Believe that p because the evidence supports p’; ‘Don’t believe both p and not-p’; ‘If you believe that p, and that p entails q, then believe that q’. In each of these cases, to violate these norms is to be (theoretically irrational).

The norms that govern the exercise of practical reason, by contrast, are concerned not (or not directly) with truth but with the goodness, desirability or value of actions. Nevertheless, one might think there are also norms of practical rationality.

A further (quasi-)distinction: theoretical reasoning produces (or changes) our beliefs; practical reasoning produces our actions. But note that practical reasoning itself produces action via mental states (e.g. pre-eminently, our intentions…) Also, these modifications are fallible and often difficult to predict: there are cases of both practical and epistemic ‘akrasia’.

5. Direction of fit

To have an intention is, roughly, to have settled on plan which one then seek sto realise through one’s actions. But this makes intention quite unlike belief. Beliefs are representational propositional attitudes. They aim, as it were, to ‘fit the world’. If one discovers that the world is not how one previously believed it to be, one will feel (and usually acknowledge) a certain rational pressure to modify one’s beliefs accordingly. Crudely: when our beliefs and the world don’t match, our beliefs need to change.

Desires and intentions are generally thought to have the opposite ‘direction of fit’. With intentions, we aim to ‘make the world fit our desires’. If the world is not the way one intends for it to be, we (as it were) change the world. (Think of Elizabeth Anscombe’s example of the shopping list…)

Preparatory reading for next week:

Bernard Williams, “Internal and External Reasons,” in Moral Luck (Cambridge: Cambridge University Press, 1981), 101–13.

Motivation and Practical Reasoning II: Internal and External Reasons

0. Recap

Theoretical vs. practical (uses of) reason; direction of fit.

  1. Two kinds of statement about reasons for action

If there is such a thing as practical reason, then that must mean that there is something it is for an action (rather than a belief) to be rational or irrational. One form of words in which we sometimes express this thought is to say that someone has a reason to do something, or, to put it more formally, A has a reason to φ. What does this mean? And what are its truth conditions (i.e. what has to be the case for such a statement to be true)?

Bernard Williams’s classic paper, ‘Internal and External Reasons’, proposes that there are two ways sentences of the form A has a reason to φ’ or ‘There is a reason for A to φ’ can be interpreted.

On the first, the truth of the sentence implies, very roughly, that A has some motive which will be served or furthered by his φ-ing, and if this turns out not to be so the sentence is false: there is a condition relating to the agent’s aims, and if this is not satisfied it is not true to say, on this interpretation, that he has a reason to φ. On the second interpretation, there is no such condition, and the reason-sentence will not be falsified by the absence of an appropriate motive.

Williams terms the first the ‘internal’, the second the ‘external’, interpretation. He goes on to propose the following, deliberately rudimentary, ‘sub-Humean’ interpretation of the internal reasons statements.

A has reason to φ iff A has some desire the satisfaction of which will be served by his φ-ing.

Of course this is too crude. Consider this obvious objection: A wants a gin and tonic; A believes this bottle contains gin (it actually contains petrol); therefore, A wants to mix what’s in this bottle with tonic and drink. According to the sub-Humean model, A has reason to mix petrol with tonic and drink it. But intuitively, she doesn’t. So the sub-Humean model must be wrong.

Further, it’s not only our desires that we deliberate from – there are also our life goals, commitments, loyalties, patterns of emotional valuation, projects… These are all part of our subjective motivational set (S). In light of these complexities, Williams proposes this (sensible) revision:

A has reason to φ iff A would be motivated to φ, if he deliberated rationally from his S and was fully informed.

Williams ultimately ends up with this formulation:

A has reason to φ only if there is a sound deliberative route from A’s S to a motivation to φ.

The notion of a sound deliberative route is, to be sure, vague. But Williams thinks that it is a vagueness in the right place; after all, the phenomena themselves are vague and it is a virtue of a theory if it preserves that vagueness in the right way. Williams does not take a mechanical view of deliberation as a series of logical deductions in the style of the Aristotelian practical syllogism; it could well involve (as it generally does, in life, involve) the exercise of imagination. Deliberation can both uncover pre-existing but previously unknown motivations and sometimes provoke new ones. But it does not do so in any predictable or obviously law-governed way.

  1. Reasons and explanation

If there are such things as reasons, Williams thinks, it must be because agents (sometimes) act for reasons. In this sense, reasons are, in some sense, explanatory. But not all explanatory reasons are justifying.

If A, mistaking petrol for gin, mixed it with gin and drank it, we could certainly say what her reason for doing this was. In other words, we could answer the question ‘Why did A drink petrol?’ or ‘What was the reason for A drinking petrol?’ – roughly, she believed it was gin and she wanted a gin and tonic. But this is hardly a justifying reason for A to drink a petrol-and-tonic. Contrast the questions ‘Why did A φ?’ and ‘Why φ?’

The idealisation involved in Williams’s revised formulation above (‘deliberated rationally and was fully informed’, ‘sound deliberative route’) is an attempt to relate the explanatory and justificatory dimension of internal reason statements. What it is for some fact to be an internal reason for action just is for it to be able to motivate under certain idealised conditions.

Williams is then trying to specify ‘the essential meaning of our thoughts and claims about practical reasons’.[1] This is what we are, basically, saying when we make claims about (justifying) reasons for actions.

  1. Could there be external reasons?

Consider the case of Owen Wingrave, a character in a short story by Henry James:

Owen’s family urge on him the necessity and importance of his joining the army, since all his male ancestors were soldiers, and family pride requires him to do the same. Owen Wingrave has no motivation to join the army at all, and all his desires lead in another direction: he hates everything about military life and what it means. His family might have expressed themselves by saying that there was a reason for Owen to join the army. Knowing that there was nothing in Owen’s S which would lead, through deliberative reasoning, to his doing this would not make them withdraw the claim or admit that they made it under a misapprehension. They mean it in an external sense. What is that sense? [‘Internal and External Reasons’, 106]

Williams thinks a reason for action, if it is a real reason, should be capable of being someone’s reason for acting on some particular occasion, and therefore, could be part of an explanation of that action. But, of course, no external reason statement could by itself offer an explanation of anyone’s action since it is the distinguishing mark of external reason statements that they are true independently of the agent’s motivations. But how can we explain an agent’s actions (i.e. intentional actions) by appeal to something that, ex hypothesi, does not motivate him to act?

Williams thinks most external reasons statements are mere ‘bluff’, rhetorical devices that intend (maybe unconsciously) to change the agent’s ‘S’ in some way so as to introduce some motivational element that might give their claim about her reasons some traction. Or, similarly, they might be ‘proleptic’ or ‘optimistic internal reasons statements’: an external reasons statement that claims to be true independently of the agent’s motivations is, at the time it is made, false, but can sometimes subsequently ‘make itself true’.

  1. Williams’s argument against external reasons
  1. For A to have a reason to ø, that reason must (of itself) provide a possible explanation of her behaviour (at least under idealised conditions).
  2. For a reason to explain A’s behaviour, it must motivate her.
  3. All motivation involves desire.
  4. External reasons would not be grounded in A’s desires.
  5. Hence, there are no external reasons.
  6. Internal reasons are grounded in A’s desires, for all A’s (i.e. agents).
  7. Hence, all reasons for action must be internal reasons.

Williams has an internal analysis of (true) reasons statements that makes them intelligible; he is challenging the external reasons theorist to give an account of what external reason statements mean that maintains the link between reasons and explanation.

But maybe the external reasons theorist should just refuse to pick up the gauntlet and deny what Williams assumes, that (1) reasons statements are potentially explanatory and (2) reasons statements can be analysed in terms of something yet more basic. Rather, the external reasons theorist could say, the notion of a reason is primitive and not susceptible to analysis in terms of anything yet more primitive.

  1. Next week: What’s at stake in ethics?

Reasons, blame, punishment, conversion and the objectivity of morality.

Recommended reading: Bernard Williams, “Internal Reasons and the Obscurity of Blame” in Making Sense of Humanity (Cambridge: Cambridge University Press, 1995), 35–45; Kate Manne, “Internalism About Reasons: Sad but True?,” Philosophical Studies 167, no. 1 (2014): 89–117; John McDowell, “Might There Be External Reasons?” in Mind, Value, and Reality (Cambridge, MA: Harvard University Press, 1998), 95–111.

[1] This is not how Williams is usually understood, but I think this is the correct reading of his argument. See S. Finlay, 2009, “The Obscurity of Internal Reasons,” Philosophers’ Imprint, 9 (7): 1–22 for a defence.


Motivation and Practical Reasoning III: Morality and Practical Reason

0. Recap

Reason can be practical as well as theoretical (‘what shall I do?’ vs ‘what shall I think?’) Actions, like beliefs, can be or fail to be rational. For an action to be rational, there must be reason to do it; more precisely, an agent must have reason to do it. The notion of a reason here is not (just) explanatory but justificatory or normative. What are the truth conditions of statements about an agent’s (normative) reasons for action? Internalists about reasons believe that there must be some connection between what an agent has reason to do and what that agent could be motivated to do under suitably idealised conditions. Externalists deny this.

  1. Why be an externalist?

Here is one route to externalism. Among the many kinds of reasons we have are moral reasons (e.g. our reason not to kill or steal or lie). Moral reasons are categorical: they apply to us whether we like it or not, i.e. our having reason not to kill or steal or lie doesn’t depend on our wanting not to do these things. That’s what marks them as moral, rather than (say) rules of a club; if we don’t like the rules of a club, we can just resign; there is no resigning from morality.

Nevertheless, morality is not irrational. That must mean that it is rational to be moral. And since morality applies to everyone, that must mean that everyone has reason to be moral. However, not everyone has motivations friendly to morality. This means that our moral reasons don’t depend on the presence of the relevant motivations, what internalists take to be a necessary condition of our having reasons. So, if moral reasons exist, and a quite plausible psychological thesis is true, then internalism must be false and externalism true.

In short, internalism undergenerates reasons, i.e. it fails to acknowledge that we have several of the reasons we (intuitively, pre-theoretically believe that we) have. Moral reasons are a particularly significant instance of this.

  1. Some internalist strategies

(1) Deny that moral reasons are categorical? But this would be to give up on what many moral philosophers think distinguishes the moral from the non-moral.

(2) Deny that anyone lacks the relevant psychological states? But there seem to be obvious counterexamples, e.g. psychopaths.

(3) Deny that everyone has reason to be moral, but everyone has reason to be moral insofar as they are rational? But why think the formal principles of rationality by themselves bring with them any substantive content?

  1. Some better internalist strategies: Mitigation

Accept that internalists can’t account for the universal, categorical force of moral reasons, but say that this was the point in the first place and that it isn’t so bad a consequence after all (i.e. what the externalist regards as a reductio, internalists treat as a modus ponens). Why assume a priori that our theory of practical reason must vindicate everything about the intuitive (naïve?) picture of morality? But (4) leaves internalists needing to explain away the intuitive absurdity or counterintuitiveness of their position. Can they do that?

  1. External reasons claims as false but useful?

Internalists can acknowledge that we have plenty of intuitions that provide grist for the externalist’s mill, but that those intuitions can be ‘explained away’ as part of not the semantics of reasons claims, but their pragmatics. In other words, internalism gives the correct account of the meaning and truth conditions of our reasons discourse. But our reasons discourse also involves a ‘pragmatics’, i.e. achieves effects in the world beyond that of stating truths.

Fictionalism: We can have quite vigorous arguments about fictional characters (‘How many children had Lady Macbeth?’) but we don’t really think she existed. Our arguments here are really arguments about something else, e.g. what lies between the lines of the text of some fiction, such as Shakespeare’s Macbeth. Similarly, we sometime speak as if there are external reasons, but they are a sort of useful fiction. Two possibilities: We don’t really believe our own external reasons claims (and this is good, because they’re false) – call it ‘descriptive fictionalism’. Or, we do believe our own external reasons claims (but either: this is bad because they’re false and we should stop literally believing them (revisionary fictionalism); or, this is fine, because they wouldn’t be so effective if we weren’t so sincere (self-deceiving fictionalism)).

Bluff: Sometimes, we have to overstate our case to be rhetorically effective, and it can be rhetorically effective to say (falsely) that someone has reason to do something. This doesn’t mean the claims are true.

Prolepsis: Sometimes, our reasons claims can be self-fulfilling. When I tell someone that he has reason to do something even though he lacks the relevant motivations, he may well acquire a reason to do that thing because he does have a desire to avoid my disapproval or blame or sanction. My reason-claim was false when I made it, but it became true; indeed, it made itself true.

One can be more or less hopeful about whether this is realistic, and about how often this sort of ‘conversion’ can be effected. But is this enough?

  1. Arguments from reactive attitudes

The idea of a reason is implicit in a number of our everyday practices, e.g. our practices of holding other people responsible. These practices involve a familiar set of reactive attitudes: resentment, indignation, gratitude. These are the attitudes we take to agents, rather than to objects. When I resent or blame someone, I am generally making a judgement about them, e.g. ‘he wronged me, and he had reason not to’.

Consider the following argument:

  1. If someone does something morally wrong to me, then I can legitimately blame him. (Moral wrong-doing is sufficient for blameworthiness.)
  2. I can legitimately blame someone for doing something only if he had reason to do otherwise. (Having reason to do otherwise is a necessary condition of blameworthiness.)
  3. If someone does something morally wrong to me, he has reason to do otherwise. (There is always reason not to do what is morally wrong – from (1) and (2).)
  4. Internalism implies that (3) is false.
  5. Therefore, either internalism is false or (3) is false (and therefore, (1) and/or (2) must be false).

In other words, internalists have to say that not all wrong-doers are blameworthy (i.e. they affirm (2) but reject (1) – leading them to reject (3)). Is this a reductio of their position or simply a plausible implication?

Recall Bernard Williams’s example of the man who is nasty to his wife and fails to respond to any attempt to persuade him to behave any better. He – by stipulation – lacks the requisite motivational states. No matter how much we argue with him, he cannot acknowledge having reason to treat her better.

Williams suggests that we simply accept this. (Kate Manne, e.g., thinks this correctly accounts for the sense of exhaustion and futility we sometimes feel when arguing with certain sorts of people.) This man genuinely has no reason to behave any better. Does this mean we can’t blame him? And isn’t that an implausible, even alarming conclusion?

Williams and Manne think:

(1) We can blame him; but only in the proleptic sense. We hope that by blaming him, we will give him a new reason he lacked before.

(2) There is no shortage of other things we can say and feel about him: that he’s malicious, brutal, callous, a sexist… But what we cannot (rightly) say is that he has reason to be otherwise. The whole trouble with him, and people like him, is precisely that they lack reason to be otherwise.

(3)  If his poor treatment turns into abuse, for instance, we needn’t have a moment’s hesitation to enforce the law or worse. The fact that he doesn’t have reason to behave differently doesn’t mean we – and his wife – don’t have reason to stop him. On the internalist view, everyone’s reasons needn’t nicely align. It would be nice if our reasons ‘stuck’ to all agents. But, as Williams puts it, the only kind of ‘glue’ we have that could effect this is social, political and legal.

The internalist’s vision of the world is agonistic: there is conflict all the way down, and nothing in the nature of reason(s) means that there can’t be. If the internalist is right, this may be an important respect in which practical reason is unlike theoretical reason; there is not the same pressure to interpersonal alignment and consistency.

Next week: Moral Motivation

Recommended reading: tbc

Motivation and Practical Reasoning IV: to be uploaded