Birds fly Tweety is a bird Tweety flies DEFEASIBLE NON-MONOTONIC PRESUMPTIVE?
What does “Birds fly” actually mean? DEFAULT CONDITIONAL: (Walton 1992, 37) For a typical x, if x has property F, then we can presume, subject to default in non-typical cases, that x has property G. Plausible generalization: (Walton 1994, 164) “Typically, F are G.” Normally you can expect F to be G in a typical case, but an F may fail to be a G in an exceptional case. Does it mean “All birds fly” ? I don’t fly, so “All birds fly” is false Does it mean “Most birds fly” ? Walton denies that this is an inductive argument, but rarely argues for it. Inductive arguments are also defeasible and non-monotonic. What is presumption? What does Walton mean by “we can presume” and “you can expect?” Taken to mean “there is a reason” these arguments are NOT defeasible. Also, unless we have strong grounds to think that x is an exception, we do have strong grounds to think that x is typical, because it is analytic to say that most birds are typical in the usual sense of typical. Walton’s distinction between presumptive and inductive arguments demands a reading of typical that cannot be glossed in terms of probability. It is NOT simply a question of whether you can give a determinate numerical probability; an argument can still be inductive without this.
What does “Birds fly” actually mean? I think that Walton bases his notion of presumption on a Toulmin warrant. “By the ‘force’ of a modal term I mean the practical implications of its use: the force of the term ‘cannot’ includes, for instance, the implied general injunction that something-or- other has to be ruled out in this-or- that way and for such-a-reason.” Statements such as “Birds fly” are warrants and have the force of licensing inferences from “__ is a bird” to “__ can (probably) fly”, but although they do not state statistical information, they are backed by it. Although “Birds fly” is asserted itself, the statement it licenses is not uttered with the force of assertion but with a force between assertion and supposition. The modal terms “probably” and “plausibly” indicate this force. (D) Petersen is a Swede. (W) A Swede is almost certainly not a Roman Catholic. So, (C) Petersen is almost certainly not a Roman Catholic. is contrasted by Toulmin with (D) Petersen is a Swede. (B) The proportion of Roman Catholic Swedes is minute. So, (C) Petersen is almost certainly not a Roman Catholic. It is not that “Petersen is a Roman Catholic” has a probability but that it is uttered with the implicit proviso that the speaker knows he might be wrong. Most statements made by E are true. A is a statement made by E. Therefore (probably), A is true. Walton is deviating from Toulmin in saying that these are different types of argument. Walton adds to the force that it can shift the burden of proof in a dialogue. In the diagrams, the defendant is shown to have the presumption of innocence and the plaintiff has the burden of proof. Likewise for Walton,“Birds fly”when used as part of an argumentation scheme has a presumption in its favour and the burden of proof is shifted to the opponent, who shifts the burden of proof back by asking one of the schemes critical questions. If all the critical questions are answered, then the proponent wins: the conclusion stands. Salmon’s argument for expert testimony is similar to, and would be taken by Toulmin to be an instance of, the argument form (D), (B) so (C)
What does “Birds fly” actually mean? This makes it too easy to shift the burden of proof. The only difference between this and the fallacy ad ignorantiam is the providing of an argumentation scheme, but not necessarily a good argumentation scheme. You do not know whether the argumentation scheme is any good until the critical questions have been asked and answered. This difference, then, is superficial. Toulmin sensibly requires a backing. It is possible to be justified in believing there to be a backing even without “having” it. Most statements made by E are true. A is a statement made by E. Therefore (probably), A is true. Why does Walton insist that there cannot be an inductive argument? Why must presumption be a completely distinct argumentation? Walton says: “In a case where one expert makes the statement A, and another makes the statement not-A, it follows that A is probably true, and also not-A is probably true. However, such a situation is not possible in the probability calculus.” In other words, he seems to think that we would have two equally good inductive arguments with equally good but contradictory conclusions. But this is a problem only if you take both conclusions to be true. They are both true if you do not detach the conclusion, but if you do detach then one of them must be false. This means simply that the inductive argument did not consider all relevant data and does not imply it to be a “bad” argument in any way.
Can presumption solve this problem anyway? It does not seem to me any more reasonable to suppose that one can simultaneously presume that two contradictory conclusions are correct or say that they are both plausible, than it did to say they were both probable. The fact that the conclusion that Tweety flies is defeated when Tweety is a penguin can be handled perfectly straightforwardly with an inductive argument and the fact the the relative frequency of penguins to birds that fly is zero. “Birds fly” then means just “Most birds fly”. There is no reason at all to assume that there is any distinct form of reasoning or any such speech act as a presumption. There is nothing wrong with the presumptive argumentation schemes as such, but Walton is wrong to insist that they are anything distinct. There are only deductive and inductive arguments. If the argument is read with the conclusion undetached as saying what the epistemic probability of the conclusion is, then the argument is deductive. If the argument is read as with the conclusion detached, then the argument is inductive; inductive arguments are, as Walton admits, also defeasible and non-monotonic.