Lewis and Counterfactuals

Introduction
Commonly uttered sentences such as ‘an Agent has a reason to Φ’ have been characterised as having two interpretations with differing truth conditions of what constitutes a reason. Broadly speaking, the internalist interpretation holds that the agent has a reason to Φ only if by their Φ-ing they will satisfy, to some degree, a motivation or desire they have (Williams, 1981, p101).  This interpretation can be motivated by the example of music lessons. If an agent has a reason to learn the piano, she must have some motivation or desire to want to play the piano. But if she truly lacks any motivation which could make playing the piano desirable, such as for pleasure, then it seems uncontroversial to believe she has no reason to play the piano. The reason for learning the piano depends on some motivation to make playing it desirable.

​

On the other hand, the externalist interpretation does not require a motivation or desire for there to be a reason to Φ (Williams, 1981, p101). Reasons are not necessarily dependent on the agent’s motivations for its truth claim. Ethical scenarios present a good example to motivate the externalist view. Imagine a mother who has no motivation to care for their baby and would perhaps benefit from the child's death. A virtue of the externalist interpretation is they can say the mother has a reason, regardless of motivations, to care for the child. This makes externalism more ethically intuitive and internalism seem unreasonable.

​

However, Barnard Williams presents a famous case against the possibility of external reasons by demonstrating the scope the internalist interpretation covers with reasons and then arguing external reasons either collapse into being internal reasons or are not reasons at all. Williams' argument could be interpreted in several different ways; therefore, this paper will attempt to provide the strongest interpretation against external reasons. I will first present Williams' argument and then assess one of the best objections by McDowell. I believe Willams is insufficient in demonstrating the impossibility of external reasons.

​

Williams Argument

Motivations
It is important to note Williams's understanding of the motivations for internal reasons is broader than desires alone. What Williams calls an agent's ‘‘subjective motivational set’’ can be desires such as for pleasure or hunger but is broad enough to allow ‘patterns of emotional reaction’ such as the motivation to weep on hearing sad news, ‘dispositions of evaluation’ such as dispositions to drink etc (Williams, 1981, p105). Consequently, an agent has an internal reason to Φ if it satisfies an agent's motivational set in this broader description. This avoids an objection that agents often do not have a desire to Φ but, due to a disposition, have an external reason to Φ which they don’t desire.

​

Normative and Explanatory reasons
A second more crucial element of Williams's argument is his understanding of explanatory and normative reasons.
The first views reasons as merely explanatory, taking a ‘reason for action form’ (Williams, 1981, p102). A reason is anything which explains why an agent Φ-ed. Therefore, reasons cannot be judged as good or bad but rather as explaining why one acted. Williams notes that in explanatory reasons the ‘false and true beliefs on the agent's part cannot alter the form of the explanation’. Meaning explanatory reasons depend upon the agent's belief rather than the real world. Hence Williams's classic example of the Gin drinker having an explanatory reason to drink a glass of petrol he believes is Gin. The explanatory reason here would be the agent's desire for the contents of the glass which he believes is gin. The explanatory reason would not change if it were true or false. 

​

This is not the sort of reason Williams is primarily concerned about but rather his focus is on normative reasons which are concerned with the agent’s ‘Rationality’ and where their Φ-ing may or may not be justified (Williams, 1981, p102). He writes concern for only explanatory reasons ‘looks in the wrong direction’ if we are to treat people as rational agents. Therefore, he argues reasons have a ‘third-personal’ normative element which provides a standard to judge an agent's rationality. (Williams, 1981, p103). A person may believe and act on a reason they do not actually have reason to act on or believe. Consequently, an agent has a reason to act only if their reason conforms to rational norms.

​

I take this as a virtue of Williams' account as it allows the possibility an agent's reasons for actions could be irrational despite it explaining their actions. Following the gin and petrol drinker example, one normative standard Williams argues is that an agent does not have a reason to Φ if the reason was grounded on ‘false belief’(Williams, 1981, p103). More specifically, false belief about Φ-ing being a means to satisfying a motivation, like in the Gin example. Or false belief about having a motivation. If we are treating agents as rational beings it is clear false belief in this way does not constitute a reason since if the agent knew the facts, like the bottle being filled with petrol, they would not be motivated to drink it. Consequently, normative reasons go further than explanatory reasons by allowing reasons for actions to be judged against rational norms to prevent irrational reason from constituting a reason.

​

Rationality and a dynamic motivational set
Since Williams is concerned with normative reasons, his account of rationality is essential for understanding internal reasons and his argument against external reasons. He writes an example of practical reasoning is being able to conclude that ‘one has reason to Φ because Φ -ing would be the most convenient, economical, pleasant etc. way of satisfying some element in S, and this of course is controlled by other elements in S’ (Williams, 1981, p104). Therefore, practical reasoning starts with an end of satisfying one's motivational set and then must evaluate the most fitting means of accomplishing that end. The most fitting way could be the most efficient or ‘pleasant’ option available, but Williams is careful to note that whatever is most fitting must be within one's motivational set. If an agent does not care for what is most economical then correct deliberation would not find the most economical option as fitting as it does not satisfy an agent's motivations. Consequently, practical reason is not just finding the ‘Causal means to end’ but also finding the most fitting means to that end where fitting correlates to an agent’s motivations.

​

Williams's account of deliberation is not limited to the above example but allows other rational processes to assess the most fitting way of satisfying one's motivational set. For example, deliberation can find the maximum way to satisfy one's motivational set by ‘time-ordering’, and attaching ‘weight’ to elements in one's motivational set in cases of conflict etc (Williams, 1981, p104). Furthermore, Williams enforces the role of ‘Imagination’ as a rational aid in deliberation (Williams, 1981, p105). Imagination can draw the mind to scenarios where an agent's motivational set may be most satisfied allowing one to see more clearly what satisfying certain motivations may look like. For example, an agent may have a motivation for pleasure and thus have a motivation to watch movies all day. The agent could use imagination to realise their motivation to watch movies all day would lead to boredom and not satisfy their desire for pleasure. Thus, removing that motivation and perhaps adding a new desire for reading.

​

As a result, practical reasoning not only provides a normative standard for rationality but allows the motivational set to be dynamic. First, it provides a normative standard as an agent is acting rationally if they correctly deliberate in the way described above. Second, an agent's motivational set is dynamic as practical reasoning can shape an agent's motivations to be more integrated. As we saw in the movie and the gin drinker example, correct deliberation, with its rational tools such as imagination, means-end reasoning etc, would remove motivations to drink the contents of the glass as it is not a means to satisfy the desire for gin nor watching movies all day satisfy the motivation for pleasure. Consequently, elements in one's motivation can change.

​

One interpretation of Williams's dynamic view might suggest external reasons are needed for change in a motivational set. An agent may have no motivation to Φ at T1 but after being told they do have reasons to Φ, such as it is a pleasant activity, they gain a new motivation to Φ at T2. It seems an external reason statement could have been true at T1 since it convinced them at T2.
However, Williams's account is not committed to this as motivations are only added or removed because of existing motivations. If deliberation is always from existing motivations then elements in one's motivational set must be hierarchical. There must be fundamental motivations which all other motivations are dependent on. These fundamental motivations allow deliberate to function and which are not deliberated on. The motivation to watch movies all day was due to the more fundamental motivation for pleasure thus when imagination in deliberation discovers that it would be incredibly boring that motivation was removed and a new one which might satisfy the motivation of pleasure was added. It follows that another person may know internal reasons which satisfy an agent's motivational set better than they do (Williams, 1981, p104). And that if you know an agent's fundamental motivations you can know all the reasons for that agent to act. Therefore, practical reasoning is procedural starting from fundamental motivations and deliberating the most fitting way of satisfying those motivations.

​

Internal reasons
Following this account of deliberation, it is worth noting what exactly Williams believes constitutes an internal reason. Internal reasons are ‘discovered in deliberation’ (Williams, 1981, p104).  Consequently, an agent has a reason to Φ only if they correctly deliberate. To correctly deliberate an agent must not have false beliefs that Φ-ing will satisfy their motivations, or false beliefs about their motivations. Rather, practical reasoning starts from fundamental motivations and then, using rational tools like means-end reasoning and imagination, generates new motivation and internal reasons which satisfy their motivational set in the most fitting way. 

​

Williams's argument against external reasons
With Williams's view of rationality and internal reasons in place, we can now assess his argument against external reasons.

Williams argues that a condition for a reason to be true is that it must be able to ‘figure in an explanation of that action’ (Williams, 1981, p106). If it is true, that there is a reason to Φ it must have the possibility of it being a reason worth acting upon in a particular case. If it is a reason for action in particular cases then it must have explained that action. Therefore, all reasons for action must be able to figure into an action explanation. If reasons require the ability to explain someone’s acting in a particular case, then it must be able to ‘motivate him to act’ since the action would never take place without the agent being motivated to perform the action (Williams, 1981, 107). Consequently, external reasons must have the potential to motivate an agent to act.

​

One interpretation of Williams's argument suggests external reasons fail to meet the explanatory condition of reasons. External reason statements are distinguished from internal reasons by the fact they are not dependent on an agent's motivational set. The external theorist believes reasons are true regardless of motivations or an agent's ability to see an external reason as something worth doing. Consequently, External reasons fail to meet this criterion.

​

However, I do not believe we should interpret William's argument this way. Williams' concern for normativity in actions suggests reasons only need the possibility of figuring into an action explanation. There are internal reasons that an agent may never know because of poor practical reasoning skills or false beliefs. Nevertheless, Williams argues these are true internal reasons for action despite not figuring into an explanation in a particular case. Consequently, reasons do not have to motivate an agent in an explanatory reason for action but only have the potential in optimum conditions. Therefore, the explanatory reasons theorist only has to provide an account where a person in the optimum rational conditions would be motivated.

Williams argues that an external reason could motivate an agent to Φ only if an agent “comes to believe the reason statement … because, in some way, he is considering the matter aright.” (Williams, 1981, p108). Meaning, that an agent who has no possible existing motivation to Φ could be motivated to Φ if they come to believe the external reason was true. On believing the external reason a new motivation, and new internal reason, would be produced which could figure into an action explanation. This argument seems very plausible since believing the truth of a reason could provide enough motivation to act on that reason.

​

An objection one could raise is that the external reason statement does not fulfil the action explanation condition rather the new internal reason does. Without the new motivation, the external reason is false.
However, Williams emphasises that the agent comes to believe the reason statement because they consider the ‘matter aright’ (Williams, 1981, p109). This means that the agent comes to believe the external reason was true and an action worth doing all along. This preserves the external reason as the agent recognises it was true even before they had any possible motivation. On coming to believe the external reason, a new motivation and internal reason is produced. Consequently, it is not that the external reason becomes an internal reason but rather the truth of the external reason is believed as true all along which produces the new motivation and internal reason. As a result, the external reason fulfils the action explanation condition for reasons as the truth of the external reason all along produced the motivation for action. 

​

This leaves the question of how one believes an external reason as aright? Williams' answer to this question is the most important part of his argument. He believes an agent comes to believe the external reason aright only by rationality (Williams, 1981, p109). Williams' account of rationality, as we have seen, requires existing motivation. Deliberation is procedural starting from an existing motivational set and finding the best way to satisfy it. Therefore, the external reason theorist cannot appeal to practical reason as driving the agent to find the external reason true all along as it requires an existing motivation which would make it an internal reason. Consequently, there cannot be any external reasons.

This presents a strong argument as it seems the external theorist is forced to either say that an agent comes to see the external reason as ‘aright’ by rationality or a non-rational method. Williams argues that the externalist cannot believe an external reason thought some ‘irrational’ way since this would defeat the purpose of a reason (Williams, 1981, p110). Consequently, he argues externalism must be equivalated to “if the agent rationally deliberated, then, whatever motivations he originally had, he would come to be motivated to Φ” (109). If this is true, then external reasons collapse into internal reasons as it requires the agent to have existing motivations for them to believe the external reason statement.

​

Objections to Williams's Argument

McDowells Objection
One of the strongest objections to Williams's argument is against the claim practical reasoning is the only option for an agent to accept an external reason statement as true all along. McDowell presents an alternative option of how an agent could come to believe an external reason aright, without appealing to existing motivations or irrationality, namely ‘conversion’ (McDowell, 1995, p100). For example:

​

First, imagine an agent, David, who was brought up with the correct ‘ethical upbringing’, perhaps an ‘Aristotelian’ one, so that their habits, desires, and dispositions conform to a particular ethical behaviour and way of viewing the world (McDowell, 1995, p101). If his upbringing is successful, then David will have a ‘Correct conception’ of the world. He will consider certain reasons for actions as true. Such as, you should drink in moderation because wit is a desirable quality over drunkenness, or you should care for your children because there are meaningful goods like a filial relationship. David sees these reasons as aright due to his upbringing. 

​

Second, imagine another agent, Jones, who was not brought up properly but ‘Slipped through the net’ (McDowell, 1995, p101). Jones' motivations are skewed because of their upbringing, and they do not value things the properly brought-up person sees as having value. Jones’ current motivational set, even after correct deliberation, does not value caring for one's child or drinking in moderation. Consequently, Jones has motivations of the wrong kind and does not desire what has real value in the world, whatever it may be.

​

Now imagine Jones has a Conversion which is an ‘intelligible shift in motivational orientation’ not caused by any existing motivation or deliberative processes (McDowell, 1995, p103). They will gain the motivations that the properly brought up person has and see value in the world which they previously did not see. Jones will come to believe external reasons were true all along and believe he failed to see the real value in Φ-ing which was always there. The conversion is dependent on the truth of the external reason which leads to Jones gaining new motivations. Consequently, conversion offers an alternative option of seeing an external reason aright without appealing to deliberative reasoning. The converted Jones now believes an external reason was true all along and so has a motivation that satisfies the action explanation conditions Williams originally proposed. Therefore, there can be external reasons.

​

McDowell's conversion example demonstrates two ways of viewing normative reasons.
Williams' view of normative reasons holds an agent has a reason to Φ only if the reason was produced by correct deliberation which assesses the most fitting way of satisfying their motivational set. As we have already shown, Williams' account of rationality is not concerned with an agent's fundamental motivation but only with correct deliberations and true beliefs which will satisfy those motivations.

However, McDowell’s normative standard for rationality argues an agent has a reason if they give ‘all relevant considerations the force they are credited with in a correct picture of one's practical predicament’ (McDowell, 1995, p106). The ‘Correct Picture’ is what has real value in the world which the correct ethical upbringing allows one to see. Therefore, Normative reasons are not just concerned with correct deliberation from existing motivations but that those motivations are of the right kind.

​

Irrationality Objection
One objection is that the conversion is irrational as it requires an agent to believe an external reason without reason in the deliberative sense. (McDowell, 1995, p103).
However, I believe conversion is no more irrational than if the agent always had the same fundamental motivations the properly brought-up person had. Parents do not provide deliberative arguments for action to give children motivations, rather a child gains dispositions and motivations through upbringing. Conversion allows the possibility that a person brought up with one set of motivations could convert to gaining a new set of motivations. As we have argued, Williams' account of rationality is procedural, starting from fundamental motivations which are not subjected to or changeable by practical rationality but allow it to function. Therefore, an agent gaining new motivations in conversion is no less rational than if they had those fundamental motivations to begin with. Consequently, conversion offers an alternative option of seeing an external reason aright without appealing to deliberative reasoning or an irrational means.

Pettit and Smith Objection
A second related but more difficult objection is that McDowell is placing a rational bias on an ethical upbringing. Pettit and Smith write ‘Reasons must have to do with  ... having desires that meet norms of reason’ and suggest those desires being a certain kind are not norms of rationality but rather of norms of Aristotelian ethics, ‘etiquette’, ‘Biology’ etc (Pettit & Smith, 2006, p31). They argue that norms of rationality are only concerned with starting from existing motivations and using ‘rational principles’ like means-end reason or imagination to deduce reasons for action. Normative rationality is limited to these rational principles and McDowells example of the Aristotelian upbringing is not a rational norm but rather what it would be good for an agent to do in respect of an Aristotelian upbringing.

​

Rebuttal to Pettit and Smith
Although it is arguable an Aristotelian upbringing and its motivational set is the correct kind for a rational agent, I believe we can show clear cases where an upbringing, with certain values and motivations, is not one a rational agent would have. Consequently, this shows the contents of fundamental motivations matter and can be judged against rational norms. The motivations of a rational properly brought-up agent constitute an external reason that an improperly brought-up agent could come to believe true all along if they had the correct ethical and rational upbringing. The improperly brought up agent lacks a ‘correct picture’ of the world the rational properly brought up person has (McDowell, 1995, p106).

​

To illustrate this point, Anscombe presents a case where an agent desires a ‘saucer of mud’ (Anscombe, 1957, §37). If the content of motivations is not subjected to rational norms then we can suppose that Jones, the agent who was not brought up properly, could have a fundamental motivation for a ‘saucer of mud’. The internalist would then give an account of correct deliberating finding the most fitting way of satisfying this motivation. Anscombe notes that if someone said they wanted a ‘saucer of mud’ we would want to know ‘What aspect of the object desired is desirable?’. The internalist would say Jones just has a fundamental motivation for a ‘saucer of mud’ but Anscombe argues this is ‘Nonsense’. I agree that such a motivation would seem irrational and uttered by a madman. This is because some motivations, like the desire for a ‘saucer of mud’, require there to be something actually desirable about it. The object of motivation cannot be anything if we are to treat agents as rational.

​

This presents a good argument against Pettit and Smith's view that the content of motivations is not subjected to rational norms. Although it is arguable the motivations of an Aristotelian upbringing constitute the motivations a rational agent would have, it is more rational than an agent with a saucer-of-mud-desiring upbringing. Therefore, the motivations the properly brought up agent has are not just norms of Aristotelian ethics but are norms of rationality. This is not to commit us to the view all motivations require an explanation. A motivation for happiness doesn’t require an explanation for what aspect of happiness is desirable. But if anything can be a motivation then a saucer of mud or a thimble of milk etc could be an agent's fundamental motivation. I believe we have just as much a case that rational norms should prevent these motivations then rational norms preventing explanatory reasons in the Gin and petrol cases constituting a reason for an agent. Consequently, if a person can be brought up without the correct ethical upbringing and has irrational motivations then a conversion to the right kind of motivations would allow an external reason to have been true. Therefore, Williams's argument fails to show external reasons cannot exist.  

​

Conclusion 

In conclusion, I have presented what I believe is the correct interpretation of Williams' argument against external reason. I then assess its strongest objection by McDowell which shows an agent can come to believe an external reason was true all along through conversion. Consequently, we have shown that the content of motivations is a normative and thus external reason for action even if deliberation does not find it a reason. Therefore, there are external reasons.

​

Bibliography:

• Anscombe, E. (1957). Intention. Cambridge, Mass.: Harvard University Press. URL: https://www.scribd.com/document/201921189/G-E-M-Anscombe-Intention (12/05/2024)

• Davidson, Donald. “Actions, Reasons, and Causes.” The Journal of Philosophy, 60(23), 685–700. https://doi.org/10.2307/2023177. Accessed 12 May 2024.

• Everson, S. (2024). Action and Agency seminars weeks 7, University of York.

• Everson, S. (2024). Action and Agency seminars weeks 8, University of York.

• Everson, S. (2024). Action and Agency seminars weeks 9, University of York.

• Finlay, Stephen and Mark Schroeder. "Reasons for Action: Internal vs. External." The Stanford Encyclopedia of Philosophy (Fall 2017 Edition), Edward N. Zalta (ed.). URL: https://plato.stanford.edu/archives/fall2017/entries/reasons-internal-external/.

• McDowell, J. (1995). Might there be external reasons? In J.E.J. Altham and R. Harrison (Eds.), World, Mind, and Ethics: Essays on the Ethical Philosophy of Bernard Williams (pp. 68–85). Cambridge: Cambridge University Press.

• Millgram, Elijah. “Williams’ Argument Against External Reasons.” Noûs, 30(2), 197–220. https://doi.org/10.2307/2216293. Accessed 6 May 2024.

• Pettit, Philip, and Michael Smith. (2006). External Reasons. In McDowell and His Critics (pp. 142–179). URL: https://www.academia.edu/22630919/External_Reasons

• Williams, B. (1981). Internal and external reasons. In Moral Luck: Philosophical Papers 1973–1980 (pp. 101–113). Cambridge: Cambridge University Press.

• Wiggins, D. (1976). Deliberation and Practical Reason. Proceedings of the Aristotelian Society, New Series, 76(1975 - 1976), 29-51+viii (24 pages).