Abstract
>This chapter is devoted to a problem that involves both the question of determining the morally relevant outcome of an action, discussed in Chapter 4, and the problem of determining the relevant alternatives for an agent in a situation, discussed in Chapter 6. Roughly put, the problem is whether an agent’s own simultaneous or future actions should be allowed to determine what he ought to do in a certain situation.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Jackson & Pargetter, p. 234.
Actualism is also defended by Holly Smith in an earlier paper (Smith(1)), while possibilism is advocated by I. Lloyd Humberstone (Humberstone), and Patricia S. Greenspan (Greenspan). (None of these writers use the terms ‘actualism’ and ‘possibilism’.)
Smith(2), p. 197.
Smith(2), p. 202.
Smith(2),p. 201.
In Carlson, pp. 122–123 I interpreted Smith as in fact holding that maximal sequences of actions always last throughout the remainder of the agent’s life. Jan österberg convinced me that this interpretation is probably wrong.
Jackson & Pargetter, p. 251. They do not define a ‘most relevantly specific action’, but they give the following example: “Writing and mailing a cheque is more specific than writing a cheque; writing a cheque in blue ink instead of black and mailing it is more specific than either, but in a way which is (usually) irrelevant to value”. (Jackson & Pargetter, p. 250, note 16.)
Jackson & Pargetter, p. 251.
Jackson & Pargetter, pp. 237–240.
It might be thought that KO*-KW* are immune to Jackson’s and Pargetter’s first argument, anyway, since these principles do not imply that Procrastinate ought to say yes, or even that it is right to say yes. An action’s belonging to a best sequence is therefore not sufficient for its being obligatory or right. But KO*-KW* nevertheless treat membership in a best sequence somewhat differently from membership in a worst sequence. Membership in a best sequence is necessary for Tightness, while membership in a worst sequence is neither necessary nor sufficient for wrongness.
If avoiding the worst sequence is more important than carrying out the best sequence, saying no might have higher moral status than saying yes, since the former action, unlike the latter, is sufficient for avoiding the worst sequence. But this is not Jackson’s and Pargetter’s position.
Jackson & Pargetter, p. 239, note 6.
I have here employed Bergström’s notion of a ‘strongly sufficient condition’. (Bergström(1), p. 95.)
Clause (ii) is needed since there are cases where P acts contrary to Q’s advice. If a is right and b is wrong, and P will do a only if Q advises him to do b, it may be right for Q to tell P to do b although b is wrong. Clause (iii) is needed since Q’s act of advising P to do a may have external effects which make it right, although a is wrong. Suppose, for example, that some madman sincerely threatens to kill ten people unless Q advises P to do something wrong.
In Jackson(l), Jackson advocates a more moderate form of actualism. I briefly discuss his proposal in Carlson, pp. 129–130, note 17.
As Jackson and Pargetter acknowledge, the conjunction of their two principles often yields incompatible prescriptions. “Our claim is [...] that if [Jones] asks Tell me what to do now’, the answer is ‘Do not change lanes and do not accelerate’ — despite the answer to ‘Tell me whether or not to accelerate’ being ‘Accelerate’.” (Jackson & Pargetter, p. 250.) Once again, I find this conclusion unacceptable. I think that a correct (set of) principle(s) should be such that if an agent would know what it (they) prescribe(s) for her, she would know what to do. (Cf. 6.3.) The prescriptions ~c&~a and a would together only cause Jones confusion. They offer no real guidance for action, since they ‘pull in opposite directions’, to use Bergström’s phrase. (Bergstrom(7), pp. 89–90.)
Jackson & Pargetter, p. 241, note 7.
For the most part, Jackson and Pargetter talk about what an agent ought to do ‘at’ a time, but there are some formulations that suggest this alternative interpretation of their principle: “Clearly [an action] must be an option at (or over) whatever time it is [...]” (Jackson & Pargetter, p. 233), “what an agent ought to do for the rest of his or her life [is] the best [...] out of the set of options that occupy that period oftime” (p. 252), “if you want the answer as to what an agent ought to do at or during some time, look at all the [...] actions possible at or during that time” (p. 255). (Italics mine.)
Jackson and Pargetter apparently regard it as in general meaningful to make prescriptions for whole lives. I would deny this, since extremely long sequences of actions are seldom performable, according to CA7. (See 5.5.)
Sobel(3), p. 196, Sobel(4), p. 113. However, Sobel does not accept S as a moral principle, since he rejects consequentialism in general. He regards the principles under discussion rather as “principles of rational choice”. (Sobel(4), p. 114.)
Sobel gives the following definition of (immediate) performability, or ‘openness’: “An action x whose first moment would be t is open at t if and only if there is a sequence of actions, with first member a beginning of x, intermediate members continuations of x, and last member a conclusion of x, such that at t the agent of x can ‘at will’ make sure that the first member takes place and, for each opening sub-sequence, were its members already insured, then the agent of x would be able ‘at will’ to make sure that the next member took place.” (Sobel(3), p. 198.)
Sobel(3), pp. 198–199.
To be more precise, Sobel believes that a principle should at least not prescribe fully specific actions, which would not be completed if begun. (Sobel(4), p, 124. Cf. note 25, below.) Sobel gives no general definition of a ‘fully specific action’, but I believe he would say that an action a is fully specific iff it does not have two mutually incompatible variants, which are both time-identical with a.
Bergström(7), pp. 95–96.
Bergström(7), pp. 97-98. In the same paper, Bergström gives another argument against Principle S. If ‘realism is the key’, as Sobel maintains, even S seems to be a halfway house. Why not take a step further and demand security of obligatory actions? An action is secure iff it is performable and such that, if begun, it would not be stopped short of completion. Assume that there are two incompatible variants of a 1, in Case 10, let us call them a 11 and a 12. They are both time-identical with a 1 Suppose further that if P were to do a 1 he would in fact do a 11 and that a 11 would be followed by d 2. On the other hand, if P were to do a 12, he would later do c 2. Under these assumptions S prescribes a 1&c 2 (this action is now securable but not secure), although beginning this action would lead to a 1&d 2, which is worse than b 1&d 2. This seems to go against the realism advocated by Sobel. (Bergström(7), p. 96.) Sobel’s answer to this objection is that fully specific actions (see note 23) cannot be securable without being secure, and the important question is what fully specific actions a principle prescribes. (a x&c 2 is not fully specific, since a x has two incompatible variants, which are time-identical with a v) Concerning actions that are not fully specific, prescriptions like this one can according to Sobel be tolerated. (Sobel(4), p. 124)
Sobel points out that this case involves a ‘present mistake’, rather than a ‘future mistake’, “for the sleepy agent does not it seems prefer what would be best. [... I]n this case aiming high would not work because present errors would persist, though perhaps aiming high would work if only these errors were not present in the first place.” (Sobel(4), p. 118.) Nevertheless, he believes that Principle S is right in regarding also present mistakes as relevant. (Sobel(4), pp. 119–122.)
The life (a 1 d 2) is not optimal among those securable at t 0, since the better life (b 1 d 2) is securable at t0. Even if we make no presuppositions about which of the lives (b 1 c 2) and (b 1 d 2) that is securable at t 0, however, a 1 is not obligatory, according to S, if P does a 1&d 2. This is because the worst life b 1 might secure, viz. (b 1 c 2), is equally good as (a 1 d 2). Hence a 1 does not satisfy clause (ii) of Principle S, even if (b 1 d 2) is not securable at t 0.
Sobel discusses, but rejects, also two more possibilistic principles, named B and B*. (Sobel(3), p. 196, Sobel(4), p. 113, Sobel(4), p. 116.) I criticize these principles in Carlson, pp. 132–133.
Feldman, p. 38. Feldman, like myself, regards actions as states of affairs, and I shall speak as if MO prescribes actions, rather than ‘seeing to’ the obtaining of states of affairs, since the former mode of expression is less awkward. Presumably, you can at t see to the obtaining of one of your own actions a iff you can perform a at a time not earlier than t. When I say that P ought, according to MO, to do a at t, I mean that MO prescribes that P as of t sees to the obtaining of a, and that a is performable by P at t.
On a plausible account of performability, not even Prawitz’s Principle T1 satisfies OL, despite its possibilistic character. (Cf. 7.7.)
Sobel(3), p. 201. Bergström agrees with Sobel on this point. (Bergström(7), p. 99.)
Like Sobel, Thomason is mainly concerned with principles as methods of deliberation, rather than as criteria of Tightness.
Thomason, p. 186.
See Parfit, Part Three.
If this view of (what is important about) personal identity is to be exactly reflected in our attitudes towards our own future actions, we should have a whole spectrum of such attitudes. Our actions in the very remote future should perhaps be treated very much like the actions of other people, while our actions in the slightly nearer future should be treated slightly more like our present actions, and so on. (Rabinowicz pointed this out to me.) But if such a spectrum of attitudes are to be reflected in our normative principles, these principles would, I believe, become hopelessly complicated.
Mckinsey, p. 392. My notation differs slightly from McKinsey’s.
McKinsey, p. 391.
Note that also L3 contains a lower percentage of fulfilled L-obligations than L5.
Here, as well as in the principles IO-IW (see below), POG-PWG, and IOG-IWG (see 9.2), I use ‘contained in’, according to Prawitz’s definition, stated in 6.4.
Note that possible lives are actions, in my sense of ‘action’, and may thus have variants.
PO-PW, but not IO-IW, require that an action is contained in an immediately performable action in order to have moral status.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 1995 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Carlson, E. (1995). Actualism and Possibilism. In: Consequentialism Reconsidered. Theory and Decision Library, vol 20. Springer, Dordrecht. https://doi.org/10.1007/978-94-015-8553-8_7
Download citation
DOI: https://doi.org/10.1007/978-94-015-8553-8_7
Publisher Name: Springer, Dordrecht
Print ISBN: 978-90-481-4571-3
Online ISBN: 978-94-015-8553-8
eBook Packages: Springer Book Archive