1

In "Reasons and Persons", Derek Parfit introduces a "self-interest" theory of rationality, which says you should ensure that your life goes as well as possible. In §3 (p. 9 in my edition), he claims that:

  1. The self-interest theory says that the only rational thing to do is to always be selfish (he refers to this as being "never self-denying"). (Quote: "S claims the following. If such a person was never self-denying, he would have the disposition that is supremely rational.")
  2. The self-interest theory says that it is irrational to always be selfish, because sometimes being selfish leads to worse outcomes overall. (Quote continued: "But it would be irrational for this person to cause himself to have, or to keep, this disposition. It would be rational for him to cause himself to have, or to keep, the other disposition, since this would be better for him.")

He explains this contradiction by saying that being rational is a formal aim of the self-interest theory (because it's a theory about rationality) but not a substantive aim, but I don't understand why this resolves the problem. I also disagree with the first claim: I feel that the things that are rational according to a theory of rationality should be precisely the things that the theory tells you to do, so by the second claim, the self-interest theory doesn't tell you that being always selfish is rational.

What is Parfit's justification for the first claim? Do we have different implicit definitions of what rationality means? And how does his formal/substantive distinction help resolve the contradiction?

Josh Hunt
  • 121
  • 3
  • 1
    "Substantive aim" is what a disposition to act directly aims at (first order level), "formal aim" is what guides selecting suitable dispositions (meta level). Selfish interest (or maximizing utility generally) does not work as a substantive aim because there is often no way to tell what will end up selfishly beneficial, and "first order selfish" actions often produce inferior outcomes. So the self-interest theory of rationality has nothing cogent to tell us at first order level, and has to be applied "formally", to selecting dispositions at the meta level. – Conifold Apr 01 '21 at 20:25
  • Thanks, that certainly makes intuitive sense / agrees with how I understand his point (the issue with being selfish is the short-termism). With that interpretation, it still feels like he's using the word "rational" in two different ways, though: in the first claim he's using it to refer to the first-order level but in the second claim he's using it to refer to the second-order level. – Josh Hunt Apr 02 '21 at 07:44
  • 1
    It is similar to the Liar paradox. We are none the wiser about distinguishing first order and meta level until applying "everything is a lie" to itself shows that the idea is self-defeating. In this case, the selfish disposition is self-defeating when applied to itself, only then we wise up about the levels. – Conifold Apr 02 '21 at 11:28

3 Answers3

1

In philosophy the justification for your above first claim ultimately lies in a school of thought in epistemology about the self, namely Rational egoism. Please note it's a kind of normative (prescribes what one ought to do), not descriptive egoism (describe what one naturally does). Also it should not be confused with Ethical egoism, which explores from a purely moral perspective, not epistemic rationality perspective discussed in Parfit's "Reasons and Persons". Further more Rational egoism implies and supports Rational choice theory which is a popular framework for understanding and often formally modeling social and economic behavior.

Interestingly, in my linked source, it specifically mentioned Parfit's criticism of this view by using a pension scheme argument.

Two objections to rational egoism are given by the English philosopher Derek Parfit, who discusses the theory at length in Reasons and Persons (1984).[15] First, from the rational egoist point of view, it is rational to contribute to a pension scheme now, even though this is detrimental to one's present interests (which are to spend the money now). But it seems equally reasonable to maximize one's interests now, given that one's reasons are not only relative to him, but to him as he is now (and not his future self, who is argued to be a "different" person). Parfit also argues that since the connections between the present mental state and the mental state of one's future self may decrease, it is not plausible to claim that one should be indifferent between one's present and future self.

Finally regarding your Quote: "S claims the following. If such a person was never self-denying, he would have the disposition that is supremely rational." I would say the key word here is disposition, so the author seems based rationality talked about here on some subjective disposition or sensual feeling. Personally I don't fully agree with the quote, since sometimes one needs to deny oneself so that one can be corrected to being rational.

Double Knot
  • 3,893
  • 2
  • 6
  • 15
1

Having read a bit further in the book, I think I have come to an answer that I'm happier with.

Later on, Parfit discusses consequentialism, which says that people should aim to make outcomes as good as possible (overall). He points out that it is possible that the most moral disposition might cause you to act immorally. He gives the example of a mother who chooses to spend money giving treats to her child rather than donating the money to a stranger who needs it more. Consequentialism says that spending the money in a suboptimal way is immoral, but the mother only does this because she loves her child; and if she stopped loving the child, then this might lead to worse outcomes overall. Therefore, loving your children could be an example of a moral disposition that leads to immoral acts.

(This presupposes a model of behaviour where you choose your disposition and this mostly determines your actions.)

I think it is the same type of thinking that applies to the self-interest theory: the theory says that it is always irrational to self-deny, but that it is rational to have a disposition that sometimes causes you to self-deny.

Josh Hunt
  • 121
  • 3
  • Glad now you're happier with Parfit's self-interest theory. Once down to the normative ethics level such as you mentioned Parfit's consequentialism, I agree rationality may loose connection with normative moral decision. This is essentially Hume's is-ought problem (https://en.wikipedia.org/wiki/Is%E2%80%93ought_problem), also disposition and emotion are frequently used in many consequentialist's utility calculation. But in my above pension scheme argument (purely epistemic without any ethics/morality involved), he seems also rejects folk rationality, any thought on this? – Double Knot Apr 11 '21 at 22:38
  • In the pension scheme argument, we ask whether you should deny yourself a small benefit now in exchange for a larger benefit later. In the self-interest theory example of being abandoned in a desert and unable to promise to pay a stranger £20 in exchange for being driven home (because by the time you get home, it would be in your interest not to pay) the issue is that you are unable to exchange a small short-term loss in exchange for a big long-term gain, because you can't control your future self. In this sense, his objection to both is the same: you don't stay the same person over time. – Josh Hunt Apr 12 '21 at 17:26
  • (But I haven't read the part of the book that deals with the pension scheme argument yet, so this is just guesswork :) ) – Josh Hunt Apr 12 '21 at 17:27
0

Paradox of self-interest

Simplified postulate of self-interest says : "Rational agent would always act in self-interest" . We could immediately spot two vague terms in this definition : what is rational, and what is self-interest.

Definition of rationality is indeed tricky: it involves processing available information (which by the way could be false) according to the laws of logic, without twisting them (for example by emotions) and objectively. In other words, rationality is impartial, reasonable being would and should judge about everything (including itself) with certain distance and level-headedness . For example rational man would admit (at least privately to himself) that there are better candidates for the job (if such candidates exist) , that competition has better product etc ..

What is a self-interest is even trickier to determine. Let's take two examples : Germans considered it is in their (selfish) self-interest to invade USSR in WW2. Of course, as a result Germany was ruined. Rebellious kid thinks it is in his interest to take puberty blocking drugs and to underway "sex-reassignment" surgery. Years latter he figures he was simply castrated, drugs ruined his bones and internal organs and is close to suicide. Obviously, in both cases, narrow interpretation of self-interest as simply trying to do what currently pleases us, without much thought of consequences, could end in something completely opposite to self-interest.

What we can conclude from this is that rational agent would not give in to vulgar selfishness, because that is not in his interest. This is somewhat paradoxical, but it stems from objective self-realization, including understanding of own limitations: self-denying could be actually best course of action. In case of both Germany and transgender kid, self-denying now would save lot of pain latter. Thus, we arrive to the classical paradox of every "selfish" philosophy of living, one described by Parafit: it is irrational to always be selfish.

rs.29
  • 1,196
  • 5
  • 9