No CrossRef data available.
Article contents
Kent Bach on Good Arguments
Published online by Cambridge University Press: 01 January 2020
Extract
I take two passages in a recent paper by Kent Bach—‘Newcomb's Problem: The $1,000,000 Solution,’ Canadian Journal of Philosophy 17 (1987) 409-25—as occasions for several observations about practical arguments and senses in which they may ‘work’ and be ‘good.’
First Passage
…one can only be amused by those advocates of BOTH who…realize that takers of BOTH almost always get but $1K whereas takers of ONE almost always get $1M, and proceed to bemoan the fact that rational people do so much worse than irrational ones. Despite their logical scruples, they seem to have a curiously low standard of what constitutes a good argument. Evidently they would rather be right than rich. One would think that a solution requires not merely a seemingly irrefutable argument but an argument that works, one whose use is likely to pay off to the tune of at least $1M. (412)
- Type
- Research Article
- Information
- Copyright
- Copyright © The Authors 1989
References
1 Though they are somewhat beside the object of this paper, I interject here comments on Bach’s purported ‘$1,000,000 solution,’ which ‘solution,’ we may note, is said to be specifically for Newcomb Problems in which the predictor ‘makes his predictions by plugging [a] psychological profile [for the agent]...into a powerful psychological theory’ (413).
Bach’s ‘meta-argument for ONE’ (412) is for the conclusion that in any such problem there is not for BOTH, though there is for ONE, a ‘solution’ in sense of a ‘not merely irrefutable argument but an argument that works, one whose use is likely to pay off to the tune of at least $1M’ (412). By a ‘solution’ we can gather that Bach means an argument that is ‘good’ in a sense additional to the three I have distinguished, that is furthermore not a simple compound of these three. Being ‘seemingly irrefutable’ is not the same as being ‘good1.’ And being an argument ‘whose use is likely to pay off to the tune of at least $1M,’ may be being an argument whose use on the occasion would yield evidence for such a pay-off, which is different from being either ‘good2’ or good3.’
In any case, however, and regardless of the exact details of Bach’s ‘solution-concept,’ there is a problem with his meta-argument. For though he presents reasons for the conclusion ‘that there is no good argument for BOTH’ (414), he does not present reasons for thinking that there is (in his sense) a ‘good’ argument for ONE. Even if all arguments for ONE ‘work’ in his sense (as he implies-‘on the assumption that PR anticipated you would use it, you can use any argument for ONE’ [424]), Bach gives no reason for the conclusion that there is even one argument (first-order, meta-, meta-meta-, or what have you) for ONE that is ‘seemingly irrefutable.’ He has in fact not addressed this issue, or done anything to counter even the extreme opposite view, held by most two boxers, even if rarely expressed, that every argument for ONE, once clarified and reduced to its fundamentals, will be found to contain some ‘obvious error’ or other (my emphasis: Corliss G. Swain, ‘Cutting a Gordian Knot: The Solution to Newcomb’s Problem,’ Philosophical Studies 53 [1988], 391). Bach has not taken seriously the possibility that there is no ‘solution’ in his sense to Newcomb’s Problem. Subject to uncertainty concerning his sense of ‘solution,’ that there is no ‘solution’ to Newcomb’s Problem is what all two-boxers believe.
2 Gauthier, David ‘Maximization Constrained: The Rationality of Cooperation,’ in Campbell, Richmond and Sowden, Lanning Paradoxes of Rationality and Cooperation: Prisoner’s Dilemma and Newcomb’s Problem (Vancouver: The University of British Columbia Press 1985), 89Google Scholar; and Gauthier, David Morals by Agreement (Oxford: Clarendon Press 1986), 183Google Scholar
3 ‘Budweiser’ is lifted from the following:
To Joseph T. Wearn Gainesville, Fla.
Dear Joe: 7 December 1965
I make a practice of swiping one sheet of stationery from every first rate hotel where I stop, like Castle Hill, and this gives tone to my correspondence.
I am writing simply to report a development of the story you told me about the boy who told his schoolmaster that alligators ate herons, pigs, small dogs, and beer bottles. While drifting south this morning on route 17, trending towards Brunswick, I regaled my wife with this yarn, hoping to relieve the tedium of mid-morning on a national highway. She listened attentively and made no comment. About five minutes later she said, “I wonder how an alligator eliminates a beer bottle.” “That’s simple,” I replied. “He schlitz.”
I did not get a very strong response to this witticism, and we knocked off another couple of miles in silence. Then I asked Katharine, “Do you know how an alligator feels after he has passed a beer bottle?” She said, no, she didn’t know. “He feels sadder budweiser,” I said.
The response was still rather weak, and silence fell upon us again.
A few minutes later, my wife broke the awful stillness. “Pabst he does, and pabst he doesn’t.”
It seemed necessary to tell you about this, without trying to tell you of our enjoyment of our stay with you, which will come later, if no thing happens to interrupt our southing.
Yrs Andy
(Letters of E. B. White, Dorothy Lobrano Guth, ed. [New York: Harper & Row 1976], 537)
4 If there is implicit in Gauthier’s writings a ‘meta-argument’ for ONE in certain Newcomb Problems, it is a simpler argument than Bach’s (see note 1 above). If there is such an argument implicit in Gauthier’s writings, it is one that claims, quite without regard for conditions like that of ‘seeming irrefutability,’ that there is a one-box ‘rational solution,’ but not a two-box one-an argument that makes this claim solely on the ground that there is a one-box (first-order) argument that ‘works3.’ Rationality, at the level of particular choices, is-or can at least seem to be-in Gauthier’s writings, ‘whatever works3.’
I thank Paul Abela for questions that led to this paper, and Richmond Campbell for comments on its penultimate version. My paper ‘Infallible Predictors’ (Philosophical Review 97 [1988], 3-24) especially remarks in it on ideas of Terence Horgan’s, may be found relevant to parts of Bach’s paper on which I have not commented.