The Argument from Rational Action

Rationality, Time, and Self

In this chapter I will take up and defend an argument mentioned in Chap. 3. There we saw that many actions will only be rational if their agents have tensed and/or first-personal beliefs. Mellor tried to account for this need in terms of the causal role of beliefs. However, I raised some doubts about the adequacy of this answer because it presupposed an asymmetry that did not in fact exist and it seemed to neglect the manner in which these actions are responsive to reasons, or rational. I took up this theme in the subsequent chapters. I argued that reasons, both in their role of justifying and motivating, are facts. I then argued that acting for a reason was a matter of being aware of that reason, and then forming a goal, choosing to act, and acting in response to that reason. I used this notion of acting for a reason to clarify the notion of rationality. A rational action, in the paradigm or non-error case, is an action which an agent does for reasons which provide all things considered reason for that action. In an error case, when an agent mistakenly takes a reason to obtain which does not obtain, then that agent may qualify as rational on the basis of the similarity between that case and a paradigm case.

Download Free Magazines

I will bring these elements together in this chapter to argue that if rationality demands tensed and first-personal beliefs in place of tenseless and non-first-personal beliefs, then this must be because the former involve an awareness of facts that the latter does not. Both tenseless and non-first-personal  beliefs can be involved in acting for reasons. What dis- tinguishes them from tensed and first-personal beliefs in the eyes of ratio- nality is simply the reasons they are an awareness of. It follows from this that the tenseless theory of time is wrong because it denies that tensed beliefs capture facts not captured by tenseless beliefs. In addition, we are moved towards an emergentist theory of the self according to which it is something over and above the body which can be readily captured by third-personal language. This does require a realist and pluralist ontology which I will also show to be independently plausible.

7.1     The Argument from Action

My first argument against the tenseless theory of time and towards the emergence of the self concerns rational actions and runs as follows:

P1. In some cases an agent must have a tensed belief in order to act ratio- nally and no tenseless belief can satisfy this requirement.
P2. In some cases an agent must have a first-personal belief in order to act rationally and no non-first-personal belief can fulfil this requirement.
P3. The reasons that motivate actions (m-reasons) and the reasons that justify actions (j-reasons) are facts.
P4. In non-error cases of rational action if an agent is required to have a belief and another belief cannot satisfy that requirement, then the for- mer must be an awareness of a j-reason that the latter is not.
C1/P5. If there are non-error cases of rational action in which an agent is required to have a tensed belief and a tenseless belief cannot satisfy that requirement, then the former must be an awareness of a j-reason that the latter is not. (From P4.)
C2/P6. If there are non-error cases of rational action in which an agent is required to have a first-personal belief and a non-first-personal belief cannot satisfy that requirement, then the former must be an awareness of a j-reason that the latter is not. (From P4.)
C3/P7. If there are non-error cases of rational action in which an agent is required to have a tensed belief and a tenseless belief cannot satisfy that
The Argument from Rational Action requirement, then the former must be an awareness of a fact that the latter is not. (From P3 & P5.)
C4/P8. If there are non-error cases of rational action in which an agent is required to have a first-personal belief and a non-first-personal belief cannot satisfy that requirement, then the former must be an awareness of a fact that the latter is not. (From P3 & P6.)
P9. There are non-error cases of rational action in which an agent is required to have a tensed belief and no tenseless belief can fulfil that requirement.
P10. There are non-error cases of rational action in which an agent is required to have a first-personal belief and no non-first-personal belief can fulfil that requirement.
C5/P11. Agents are aware of facts in tensed beliefs that they are not aware of in any tenseless beliefs. (From P7 & P9.)
C6. Agents are aware of facts in first-personal beliefs that they are not aware of in any non-first-personal beliefs. (From P8 & P10.)
P12. According to the tenseless theory of time there are no facts captured by tensed beliefs that cannot be captured by tenseless ones.
C7. The tenseless theory of time is wrong. (From P11 & P12.)

I shall refer to this as the argument from action. In this chapter I will focus on the temporal case. It is slightly less clear what conclusions we should draw about the self and I will say more about these in Chap. 10.1

I will now set about justifying the premises that I take to need external support, namely P1-4, P9, P10 and P12. The other premises gain their support from one or more of these.

P12 comes straight forwardly from the definition of the tenseless the- ory that was given in Chap. 2 and the manner in which beliefs are char- acterized. A tensed belief will be a belief an agent has when that agent is disposed to honestly express or affirm uses of tensed language. It will be a belief that an agent communicates by using tensed language. It is there- fore natural to take the beliefs and the language to share contents and to capture the same facts. Therefore, if tensed beliefs capture facts that tense- less ones do not,  then  tensed language captures facts that  tenseless language does not. The definition of the tenseless theory is that it denies that tensed language captures facts that tenseless  language does not. Therefore, we arrive at P12.

P1 and P2 come from Chap. 3. There we saw that Perry’s argument for the essential indexical has showed that some actions require their agent to have tensed beliefs. A consideration of this and Mellor’s arguments led to the conclusion that the tensed beliefs were essential specifically for some rational actions. If an agent is going to act rationally and in a timely fash- ion, then that agent must have a tensed belief. For instance, I must believe that ‘it is now 1pm’ or something similar and tensed if I am going to act rationally to fulfil my goal of listening to the 1 o’clock news. I might have tenseless beliefs such as ‘listening to the news requires the radio to be on’ and ‘in order for the radio to be on at 1 o’clock it must be turned on before 1 o’clock’ all morning, but in order to act rationally in turning on the radio I must also gain a tensed belief. No tenseless belief can fulfil this role. Therefore, P1 is correct.

Chapter 3 also showed that analogous remarks apply for a first-personal indexicals. An agent must have a first-personal belief if that agent is going to act rationally and for considerations that apply to him or her in par- ticular. For instance, if I am going to collect my order I must believe ‘I have ticket 114’ or something similar and first-personal. No non-first- personal belief can fulfil this role. For example, the belief ‘Olley has ticket 114’ will not suffice because I might not know that I am Olley. Therefore, P2 is correct.

The discussion of Chap. 4 showed that reasons are facts. More specifi- cally, it was shown that the reasons that make an action appropriate or justified are facts, and that the reasons that an agent acts for are facts. It is the fact that cycling gives me pleasure that justifies and motivates me to go cycling on a particular occasion. This gives us P3.

7.2     Justification of P9 & P10

P9 and P10 say that there are non-error cases of rational action which require tensed and first-personal beliefs. My reason for taking these prem- ises to be correct is that denying them leaves one in an unacceptable position.

A great majority of our actions are timely actions and/or actions done for reasons specific to ourselves. That is, a great many of our actions fall
The Argument from Rational Action within the scope of Perry’s essential indexical arguments. (I have argued in Chap. 3 that all of Perry’s examples—so those concerning ‘here’, ‘this’, or ‘you’ beliefs too—can be interpreted as concerning tensed and/or first- personal beliefs, and few of our actions escape all of his arguments.) To deny P9 and P10 would thus entail that a great majority of our actions are error cases.

Because reasons are facts saying that the great majority of our actions are error cases implies that they are cases in which we are mistaken about what facts obtain. It would follow that people were generally mistaken in their world view. But, this is something that we should deny and that the tenseless theory has been careful to deny. In its old form this denial occurred in the idea that tensed beliefs are reducible to tenseless ones. In its new form it occurs in the idea that tensed beliefs have tenseless truth- conditions. Either way we are encouraged to think that our tensed char- acterizations get the world right.

If the tenseless theorist tried to avoid my argument by insisting that most of our actions are error cases, then this means the contents of our tensed and first-personal beliefs are mistaken. (Essential indexical cases can be constructed using all sorts of tensed or first-personal beliefs.) If, as it seems, the content of our perceptual beliefs are all present tensed,2 it follows that most or all of our perceptual beliefs are mistaken. Our per- ceptual beliefs get the world wrong. This mistake would thus infect all of our empirical data and hence even the sciences—commonly  revered by tenseless theorists—would  be shown to be built on error. Whilst I am ready to admit that we do not get everything right, error on this scale is surely unacceptable.

Moreover, the tensed and/or first-personal beliefs are required for the actions at issue to be rational actions. This means that the reasons that they are an awareness of, or would be if these were not error cases, are reasons that are essential for the actions. They are reasons required to justify the actions. It follows that if all of these actions are error cases, then they are without reasons that would be essential for justifying them. In other words, essential justificatory reasons would fail to obtain for any of these actions. The actions would hence all be unreasonable actions. However, this sounds wrong. Furthermore, it also implies that we ought to be acting quite differently to the ways we do act. But, it is quite incomprehensible to fathom what such a radical change in people’s  actions ought to be. If I shouldn’t turn the radio on at 1pm, or go to the meeting at 2pm, or buy milk when I run out, or any of these mundane everyday actions, I am lost to imagine what I ought to do. (One could not avoid these consequences by denying that the actions at issue were rational actions, because, as I argued in Chap. 4, it is not possible for the majority of our actions to be irrational.3)

In short, I take it that P9 and P10 must be correct because to deny them is to insist that the great majority of our actions are unreasonable. Further, it would be to suggest that the great majority of our world view is mistaken. I do not think that these consequences are acceptable and hence P9 and P10 must be.

7.3     Justification of P4

Beliefs entered our discussion of rational action on the following basis. To act rationally an agent must act for a reason that justifies that action. (This might need to be altered to accommodate error cases but there is no need to add this complexity here.) To act for a reason an agent must be aware of it. Beliefs provide awareness. Therefore, beliefs are required for rational action.

This reveals the role beliefs have in our consideration of the rationality of actions. By considering this role that beliefs play we can uncover what their rational import is. More specifically, we can see that there are three elements to their rational import. First, beliefs must provide awareness. Second, the awareness  provided must be of justifying reasons. Third, beliefs must enable an agent to act for those reasons.

It follows that if two beliefs are of differing rational import they must differ in one or more of the following three ways. First, one must be a state of awareness whilst the other is not, RIB1. Second, one must be a state of awareness of a justifying reason whilst the other is not, RIB2. Third, one must enable the agent to act for a reason whilst the other does not, RIB3.
P4 simply states that RIB2 must occur when two beliefs are of differ- ing rational import. Therefore, one way to defend P4 is to rule out RIB1

The Argument from Rational Action

and RIB3. RIB1 is not actually an option at all. According to the umbrella notion of belief that has been assumed throughout this book beliefs sim- ply are states of awareness. In the rest of this section I will thus argue against RIB3. In short, I will argue that tensed and tenseless beliefs alike enable an agent to act for a reason, and first-personal and non-first- personal beliefs alike enable an agent to act for a reason.

We saw in Chaps. 5 and 6 that a paradigm case of rational action will involve an agent being aware of a reason for an action, and then choosing to perform that action for that reason, adopting the goal of performing that action, and acting. We thus have four components of acting for a reason: belief, choice, desire, and action. The question of whether a belief can enable an agent to act for a reason is hence the question of whether it can combine with choice, desire, and action in the right way. Accepting RIB3 would involve arguing that some beliefs cannot combine with choice, desire, and action in the right way. However, this is highly implau- sible. It would mean that an agent could be aware of a reason and yet for it to be impossible for that agent to act for that reason.

An agent can fail to recognize that he or she is aware of the same thing on two different occasions. One might hence wonder if this is enough to establish RIB3: in one state of awareness an agent is aware of a reason and is enabled to act for it; however, in the other state of awareness the agent fails to recognize that he or she is aware of the same thing or hence to act. But, this argument is no good. Crucially, an agent is not required to rec- ognize that he or she has been aware of a reason on another occasion in order to be able to act for that reason. Furthermore, even if we supposed such recognition was necessary it does not establish RIB3 or deny P4. Firstly, while it is possible to fail to recognize that one is aware of the same thing on two different occasions it is not necessary to do so. But, it would have to be necessary if it were to ground the impossibility of acting for reasons through specific beliefs. Secondly, such cases of misrecognition appear to be possible because different aspects of that thing are presented on each occasion. For example, I might not recognize that I am looking at the same house on two occasions because on one I was aware of the front aspect and on the other I was aware of the rear aspect. But, being presented with different aspects in this way just is being presented with different facts. For example, being close to the rear of a house is a different fact from being close to the front of a house. Therefore, we are left with the idea that the different beliefs provide awareness of different facts and hence we are led back to P4 and away from the idea that possible failures of recognition can justify RIB3.4

It  appears that  an agent’s  beliefs can vary independently of their choices, desires, and actions, contra RIB3. For example, I can believe that walking gives me pleasure, and can choose not to walk, desire not to walk, and not go walking for this reason (perhaps I feel in need of some self-discipline). Alternatively I might choose to go walking, desire to go walking, and go walking for this reason. Moreover, I might believe that walking  gives me pleasure and choose to cycle, desire to cycle, and go cycling. That is, I can choose to go cycling because walking gives me plea- sure. We can make this case appear plausible by adding in some other beliefs to the effect that people who gain pleasure from walking gain pleasure from cycling. However, there is no need to do this. If we do not, then I will be acting irrationally. The fact that one gains pleasure from walking is not a j-reason to go cycling. This makes it hard to understand such a case because we generally try to understand people by assuming them to be rational. But, this does not show that the case is impossible. The requirement that we take other agents to be rational in order to understand them is a proportional requirement. We must take another to be rational in a good proportion of cases, but we do not need to take that person to be rational in every case. It is clear we can understand others well enough to pick out their occasional irrational actions as irrational on the basis of a background of their rational actions. In fact, there are exam- ples in psychology of apparent widespread irrationality of this sort. It is common for people to perform actions for reasons that are not reasons to do those things or are not reasons that justify those actions.5

Acting for a reason or a fact one takes to be a reason has a certain phenomenology. Agents can normally tell you when they are acting for rea- sons and what those reasons are. In the terminology of Chap. 4, one might say having an m-reason has a certain phenomenology whether or not that m-reason is also a j-reason. This can enable one to know when another is acting for a reason and hence when he or she is acting irratio- nally. If I cannot make sense of another’s actions I might ask him what his reasons for performing those actions were. It might turn out that the facts he refers to are not j-reasons for those actions at all and I might hence judge him to be irrational. But, crucially, it is not necessary that I con- clude that he did not in fact decide or desire to act. It is clear people are sometimes wrong about what a particular fact justifies or what it is a j-reason for. If we can talk about being more or less off target in this respect, then each of these are as possible as the other even if they are not equally common. Irrational actions show just how freely beliefs can vary independently of choices, desires, and actions in a way that conflicts with RIB3. All manner of beliefs are capable of both combining with or not combining with desires, choices, and actions. Such combinations might not be expected, but they are possible.

It can be appropriate to take an agent’s movements to be simply non- rational so that it is inappropriate  to judge that agent in terms of rational- ity at all. Breathing is an example of a non-rational movement. Being manipulated by someone else is a different example. However, these cases are clearly different to the ones I just described as irrational. In the cases of irrational action the agents have their action as a goal and they choose to perform that action on the basis of a fact they are aware of (which as it happens is no j-reason for the action that they perform). These condi- tions are lacking in the examples of non-rational actions I outlined and it appears to be this very lack which renders them non-rational.

When considering the argument from rational action our focus is on tensed and first-personal beliefs. The relevant specifications of RIB3 are thus as follows. First, a tensed belief enables acting for a reason whilst a tenseless belief does not. Second, a first-personal belief enables acting for a reason whilst a non-first-personal one does not. But, both of these claims seem simply to be false. Perry’s arguments for the essential indexi- cals utilize specific sorts of cases. For example, we need a tensed belief when we must act at a specific time. It thus appears that there will be cases of rational actions in which an agent does not need a tensed belief. These might be cases in which an agent does not need to act at a specific time or in which the agent has a reason to do something at any time at all. If this is right, it shows that tensed beliefs are not required simply in order for a person to act rationally. It shows that beliefs other than tensed ones can combine with choices, desires, and actions in the relevant way in rational  cases of acting for a reason. Similarly, the need for first-personal beliefs appears to arise in cases where it is oneself in particular who has a reason to act. So if there are cases in which this is not so, cases in which anyone or everyone ought to act, then these will be cases where first- personal beliefs are not required for rational actions or acting for a reason. These will be cases in which non-first-personal beliefs combine with desires, beliefs and actions in the right way. Moreover, it is common for tenseless and non-first-personal beliefs to play roles in many actions. For example, when I am discussing mathematics with a friend I might well make use of a tenseless non-first-personal  belief such as ‘four plus four is equal to eight’. Tenseless and non-first-personal  beliefs readily combine with choices, desires and actions in an agent’s acting for a reason.

It is worth remembering that in thinking that an action is done for a reason we are not thinking that the action is caused by the reason or caused in a specific way.6 It follows that it is inappropriate  to say that only beliefs with a particular causal shape can combine with choices, desires, and actions in the right way to enable an agent to act for a reason. Rather, if we are able to say that an agent desires, chooses and acts on the basis of what they believe, then we can also say that if there are any causal require- ments, then these must also have been met.

RIB3 is false. Tenseless and non-first-personal  beliefs can enable agents to act for reasons just as tensed and first-personal beliefs can. This becomes particularly clear when we consider irrational actions or actions outside the scope of Perry’s arguments. It follows that if tensed beliefs are required for a rational action in place of any tenseless beliefs, then the former must involve an awareness of reasons that the latter lacks. Similarly, if first- personal beliefs are required for rational actions in place of any non-first- personal beliefs then the former must involve an awareness of reasons that the latter lacks. P4 is correct.

Source: Rationality, Time, and Self Kindle Edition by Olley (F.O.C.H.) Pearson (Author)