From owner-cse191-sp08-list@LISTSERV.BUFFALO.EDU Thu Feb 14 10:43:46 2008 Date: Thu, 14 Feb 2008 10:43:26 -0500 From: "William J. Rapaport" Subject: 191: MODUS TOLLENS QUESTION To: CSE191-SP08-LIST@LISTSERV.BUFFALO.EDU A student writes: | I had a question about Modus Tollens. | if, for example, you had the premises: | | 1: (p^r)-->(s^t) | 2: NOT s | | could you use Modus Tollens to arrive at NOT (p^r) since the NOT s would | make the conclusion false meaning the antecedent would also have to be | false? If so, is it enough to just state MT as the reason behind this, | or is there some other steps that need to be done first? The answer depends on whether you consider yourself a logician or a mathematician. (If you consider yourself a computer scientist or a computer engineer, you get to choose :-) A mathematician would say that you are correct. A logician would say that, technically speaking, you could not cite MT immediately. Instead, you would have to do derive -(s^t) before you could apply MT. To derive -(s^t), you could use addition (our old friend :-) and DeMorgan: 1. ((p ^ r) -> (s ^ t)) 2. -s 3. -s v -t : from 2, by Addition 4. -(s ^ t) : from 3, by DeMorgan 5. -(p ^ r) : from 1,4, by MT Now, if you were a computer that only knew how to use MT, Add, and DeM, would you be able to jump immediately from line 2 to line 5, the way a (human) mathematician could? Possibly, depending on how you were programmed. But most likely, such a computer would have to behave like a logician, and follow the rules without taking any shortcuts. (Possibly, if it was capable of learning, it could learn such a shortcut, but that would require some very clever AI programming; sounds like a nice project for anyone looking for one :-) By the way, for more information on "teaching" logic to a computer by programming it to do logic, take a look at (a now somewhat outdated, but still useful text), co-authored by one or two people whom some of you may have heard of: Schagrin, Morton L.; Rapaport, William J.; & Dipert, Randall R. (1985), Logic: A Computer Approach (New York: McGraw-Hill) * SCI/ENGR Book Collection BC138 .S32 198