for a charitable reconstruction. My second charge is systematically more important By casting his discussion in terms of connecting links between premises and a conclusion, it may seem that the issues are all purely logical. As he knows, some of them are and some of them are not. The decision whether to accept or reject a counter-example, e.g., Thomson's, is extra-logical. More to the point, decisions concerning what comparisons are relevant or what cases are to be considered like cases are also extra-logical. Sidgwick's Principle of Justice and its descendant, Schwartz's Sufficient Generality Test, are impotent unless backed by substantive principles that specify their domain of application. When they seem to generate substantive conclusions, we can be sure that other principles are at work, and we have a right to know what they are. When these substantive principles are simply slipped past us in the guise of a canon of logical criticism, we can say that the principle of Logical Neutrality has been violated. Notes [1] New York, Random House, 1 980. All page citations are in the text. • Logic and Substance A Reply to Fogelin Thomas Schwartz University of Texas, Austin Fogelin errs less in disagreeing with me than in sup- posing that we disagree. What he attacks is not so much my reconstruction of the abortion argument as the as- sumption that such reconstructions are, like Cartesian first principles, not open to revision, even in principle. What he attacks is not so much my criticism of the abortion argument as the assumption that such criticisms are, like those found in mechanical formal-logic texts, purely logical rather than substantive. Far from having affirmed either assumption, I expressly denied both. Who has been uncharitable? When he is not attacking straw Schwartzes, Fogelin nicely illuminates two issues fundamental to informal logic: First, to what extent can one separate logical criticism of arguments from substantive criticism? Second, to what extent can one separate the criticism of argu- ments from their reconstructionl 5 1 Fogelin suggests that the following modification of (P2') might block Judith Thomson's violinist counter- example: It is wrong deliberately to terminate the life of anything that lacks the Thompson trait but has the status S, is innocent, is nonthreatening, and whose future life (if not terminated) is likely to be worthwhile for itself. To preserve validity, we must add this premise: Normally, a human fetus lacks the Thompson trait The Thompson trait, says Fogelin, is the "basic feature of the Thompson counter-example," which" is that a person is forcefully [forcibly?] and against her will placed in a position which, through none of her own doing, makes her responsible for maintaining another person's life." This maneuver accomplishes nothing. Force is not essential to the violinist counter-example, so the italicized qualification in Fogelin's modification of (P2') does not block that counter-example. To see why, let us alter the example by supposing that you allowed yourself to be connected to the violinist, although without agreeing to remain connected, and that your disconnecting yourself would not make the violinist worse off in any way than he would have been had you not been connected to him if the first place: he would incur no additional pain, op- portunity costs, or any such thing. Surely, then, you are not obligated to remain connected. You began an act of pure charity whose discontinuance would merely and cost- lessly reinstate the status quo ante, for which (we may suppose) you had no responsibility. 2 Fogelin says I did not "extract the underlying principle of the counter-example and add it as a further qualifica- tion to ... (P2')." But in the first place, I did extract the "underlying principle" (Art of Logical Reasoning, pp. 239-40). In the counter-example, I said, two persons, Wand F, are related as follows: 0) F is temporarily using W's body to support F's life. (ii) W had no obligation to let F begin using W's body to suport F's life .. (iii) If W stops letting Fuse W's body to support F's life (thereby bringing about F's death), that will not make F any worse off than F would have been had F never begun using W's body to support F's life. If the violinist example is morally analogous to normal abortion situations, then (i)-(iii) exhaust the essential features of that example, and the" underlying principle" is this: (UPl) Normally, when (i)-(iii) hold, W has the right to terminate W's support of F's life. If the violinist example is not morally analogous to normal abortion situations, then there is a fourth essential feature: (iv) It is not the case that Wand F are a pregnant woman and her fetus. and the "underlying principle" is rather the following: (UP2) Normally, if (i)-(iv) hold, then W has the right to terminate W's support of F's life. To accept (UP2) but deny (UP1) is to hold that between a pregnant woman and her fetus there is some special relationship that obliges the woman to let the fetus use her body to support its life, a relationship that does not obtain in other situations of the (i)-(iii) variety. I n my book (§11.6) I suggested that such a relationship might be promissory, proprietary, or official. I explained these relationships and argued that the real abortion issue- what ought to be debated-is not whether fetuses share a general "protective" status (S) with adult humans, but whether fetuses bear a special "protective" relationship to their mothers. I found it easier and more instructive to discuss the "underlying principle" of the violinist example directly than to use it to qualify (P2'), as Fogelin desires. But let's try it Fogelin's way. Let us say of an entity F that it satisfies (i)-(iv) if there exists a person W such that F and W together satisfy (i)- (iv), and that F does not satisfy (i)-(iv) if there exists no such person W. Similarly for other combinations drawn from conditions (i)-(iv) and their negations. Pursuing Fogelin's recommendation, we might qualify (P2') by restricting it to things that do not satisfy (i)-(iii). To preserve validity, we must then add the premise that a human fetus (normally) does not satisfy (i)-(jji). But that is flatly false. Alternatively, we might add the true premise that a human fetus (normally) does not satisfy (i)-(iv) (rather than (i)-(iii)). We must then qualify (P2') by restricting it to things that do not satisfy (i)-(iv). The qualified premise is equivalent to saying that (P2') holds for things which do not satisfy (i)-(iii) and that (P2') holds for things which satisfy (i)-(iii)-but-not-(iv). The assertion that a thing satisfies (i)-(iii)-but-not-(iv) is, however, just a compli- cated though explicit way of saying that the thing is a fetus. So the modified premise explicitly asserts that (P2') holds for fetuses. That is question-begging. Neither modification, then, is the least bit reasonable. The point would have been complicated and hard to follow for an introductory textbook. For these reasons, and in order more clearly to bring out the "special protective relationship" idea, I chose to examine the "underlying principle" of the violinist counter-example separately instead of trying to add it "as a further quali- fication to ... (P2')." 3 Fogelin's main point, however, is not that the anti- abortion argument can indeed be successfully revised to meet the violinist counter-example, only that I gave an "uncharitable reconstruction" of the argument by not thus revising it. As you just saw, the recommended revision can hardly be called charitable-which again is one reason why, though I took up the substance of the revision and drew a lesson from it, I chose not to present it as a revision. Fogelin does raise an important issue here: In recon- structing an argument, how far must one go toward blocking criticism? 6 Reconstructing an argument consists mainly (though by no means entirely) in supplying validating tacit pre- mises. I offered three criteria-Fidelity, Generosity, and Generality-for choosing among alternative sets of vali- dating tacit premises. Applying these criteria involves evaluating candidate premises, hence blocking certain criticisms. I also offered three diagrammatic heuristics for finding validating premises: the circle-closing strategy, the tree-diagram technique, and the Venn-diagram tech- nique (not for ascertaining validity, but for identifying validating premises and applying the Generalitycriterion). And I suggested that one try to meet criticisms of argu- ments by rereconstructing them-by revising them to block the criticisms. Wherever one stops the process of successive criticism and revision, there always remains the logical possibility (though it may be no more than that) of further revision. Like legal and scientific findings, but unlike Cartesian first principles, argument-reconstructions are in principle open to revision. Still one must eventually stop; no one can say in general exactly where. There are temporal and other cost-constraints to be met; these vary from context to context. Often further revision seems hopeless, or at least unpromising. Sometimes (as in the case at hand) the only revisions that come to mind are plainly less chari- table than the reconstruction one has criticized. And after a while, revision tends to produce something better- regarded as a new argument than as a reconstruction of the original. It is worth emphasizing that argument-reconstruction, though constrained by validity, based on principles of selection, and facilitated by semi-mechanical techniques, is no algorithmic task. Reconstruction depends on evalu- . ation and is always in principle open to revision. I thought I made this point adequately by expressly and repeatedly basing reconstruction on evaluation and by expressly and repeatedly recommending revision in the face of criticism (though not without end). I thought I prevented the point from being lost by devoting the last two paragraphs of my book to it. 4 Fogelin takes issue with my use of the test of Sufficient Generality to criticize (P2"). This test explicates the intuitive idea that one cannot immunize general premises from counter-examples simply by attaching ad hoc restrictions. Look closely for Fogelin's criticism. Look beyond his colorful epithets ("logical haymaker," "too quick and easy," "fishy"). He never criticizes the test of Sufficient Generality itself. He never argues that (P2") passes the test. He never defends (P2"). All he says by way of criticism is that the judgments of relevant similarity that I make when I apply the test (or that anyone must make when applying the test) are substantive, or extra-logical, not purely logical. I couldn't agree more. Where am I supposed to have denied this obvious fact? Most criticisms of arguments are (and must be) partly substantive, hence extra-logical. Take the use of counter- examples: Logic (in the broad sense of my book and this journal) dictates that a counter-example refutes a premise. Judith Thomson's violinist story is a counter-example to (P2') only if you have a right to unplug yourself from the violinist in the situation described. I think it is obvious that you do have such a right. But in saying so I make a substantive judgment-a moral one-that logic alone does not compel me to make. Logic is like the law. To decide whether someone is guilty of a crime, legal principles must be combined with extra-legal judgments. The law limits the range of extra- legal judgments that are relevant. It does not produce those judgments. To reconstruct and criticize an argu- ment, logical principles must be combined with extra- logical judgments that are relevant. I t does not produce those judgments. Maybe Fogelin merely wishes to charge me with not acknowledging this point, on which we agree. But I did acknowledge it, repeatedly. My stated purpose (see Preface) was to articulate the procedure typically used by philosophers for reconstructing and criticizing arguments. Although logically constrained, that procedure obviously depends heavily on extra-logical judgments. When dis- cussing the reconstruction of arguments, I expressly and repeatedly call for judgments of the reasonableness of premises. I open Chapter 9 (p. 175) by pointing out at length that the evaluation of premises requires sub- stantive knowledge and that the principles I present specify how to use such knowledge. I end my book with these words: "Logic alone is limited .... Without topical knowledge or creativity, logic will never lead us anywhere. Without logic, however, topical knowledge and creativity will often lead us astray." 5 Although mistargeted, Fogelin's discussion brings out with force and clarity two questions that informal logicians ought to ponder: 1) If argument-criticisms are almost always partly logical and partly substantive, when can we say that an arguer has reasoned badly, as opposed to having made a substantive mistake? (Perhaps an error is one of reasoning rather than substance to the extent that the error is nonobvious, and so requires some effort at demonstration, even to those who accept all the critically relevant sub- stantive judgments.) 2) How can teachers of informal logic avoid foisting (possibly controversial) substantive judgments on their students under the guise of teaching them how to reason better? 7 Note This is a reply to the article by Robert Fogelin, "Charitable Reconstruction and logical Neutrality", printed above. Fogelin refers to my book, The Art of logical Reasoning (New York: Random House, 1981), and to a talk I gave at a conference at Carnegie-Mellon University on logic and liberal learning [reported in !IN, ii.l, pp. 9-131. That talk now appears as a paper, "logic asa liberal Art," in Teaching Philosophy 4 (1981): 231-248. Page reference~ in this rpply are to The Art of logical Reasoning. Editors' Note: Professor Schwartz is in the Department of Government at the University of Texas, Austin .• Argumentum ad Hominem: Aut Bonum aut Malum? John Hoaglund Christopher Newport College "To deliberately take the life of a human being is murder. It is surely repulsive from a moral standpoint, and it is everywhere illegal, punishable by death or life imprisonment. Now the fetus is a human being. The science of genetics has proven conclusively thata unique human being is established by the 23 chromosome pairs constituted at conception. Why, as early as 3 to 4 weeks, a heartbeat is discernable, and at 8 weeks brain-wave activity can be monitored. So abortion has got to be unthinkable. Whoever deprives this fetus of life deli- berately kills a defenseless human being, and that is murder." Bob Dolan was talking to a group of friends. The topic arose in response to news reports of a proposed amend- ment to the U.S. Constitution (an amendment which Bob favored) enabling states to make abortion illegal despite the 1973 Supreme Court decision in Roe v. Wade. Sally Clark was not all all convinced. With more than a hint of excitement in her voice, she replied: "Look, Bob, I'm sick of hearing men preach about the evils of abortion. And I do mean preach. Everyone knows you're a Catholic and that Catholics oppose abortion as a matter of religious faith. Fortunately we live in a nation that separates state and church, so that we don't have to have someone else's religious beliefs forced on us."