22(2)pp5-27TeachingSupplement7.pdf TS 22 Daniel H. Cohen tive, as well as "a philosophical , a theoretical , an empirical , an analytical, and a practical component" (p. 276); explain what these terms mean and compare and contrast this approach with that of logical theory defined and presented in class. (10.5) Discuss the pragma-d ialectical distinction between "resolving a difference of opinion" and "settling a dis pute" (pp. 280-81); explain and evaluate this distinc- tion, and compare and contrast it to the distinction (made in so me quarters) be- tween persuas ion and conviction. (10.6) Compare and contrast the pragma-dia- lectical notion of " analytical overview" (pp. 288- 91 ) with the notion of "analysis" of an argument presented in class (e.g., Galileo, pp. 3 11-14). (10.7) Compare and contrast the pragma-dialectical notion of recon struction (pp. 29 1-93) with the notion of reconstruction presented in class (e .g., Galileo pp . 314, 323). ( 10.8) Di scuss the pragma-dialectical concept of " fallacy" by stating and explaining its definition (p. 299) and illustrating it with one or more particular cases (for exam- ple, "begging the question, " p. 301). (* 10.9) Compare and contrast the pragma- dialectical account of "begging the question" (as violation of its Rule 6, p. 30 1) and Finocchiaro 's account (as a special type of " pres uppositiona I di sconnection" called "persuasive disconnection ," in Galileo, p. 318). Informal Logic & the Surprise Exam DANIEL H. COHEN Deparunent of Philosophy Co lby Co llege Waterville, Maine, USA dhcohcn@colby.cdu The Paradox of the Surpri se Exam can be an in valuable pedagogical vehicle fo r deve loping and deployi ng many of the relevant concepts of informal logic, argu- ment analysis, and rudimentary phil osophy of language . These include : the con- cepts of validity and soundness, and the differences between premises and con- clusions, between inferences and implications, between logical entailment and prag- matic implication, and between assertions and other speec h acts. Of most impor- tance, the paradox also provides a good entry into the basics of the three main approaches to argument analysis: the logical , the rhetorical, and the dial ecti cal. Along the way, the students wi ll be introduced to reductio ad absurdum arguing, mathematical induction, and meta-reasoning. Finally. if the recommended reso lu - tion that is provided here is adopted, the treatment culminates in the neighborhood of Godel's In completeness Theorem. © Daniel H. Cohen saklak Text Box Informal Logic & The Surprise Exam TS 23 1. The Paradox Here is one version of the Paradox of the Surprise Exam!: A teacher announces to her class that there w ill be an exam on one of the five days of the next week. Furthermore, it will be a surprise exam in that no one will be able to determine which day the exam will be prior to the day on which it is actually to be given. Were any students able to demonstrate that they have correctly inferred the day of the exam from anything the teacher says, the exam would have to be cance lled beca"se it would not then be a surprise. A clever student from the back of the room remarks to his classmates that at least they can be sure that the exam will not be on Friday. Either we will already have had the exam by next Thursday evening, he explains, in which case there will be no exam on Friday, or else we will not have had the exam by then, in which case we will all know it is coming on the following day - but that would be to no one's surprise. That would contradict the announcement, he concludes, so it would have to be called off. The student next to him, who is a bit cleverer, agrees, and then points out that the exam cannot in fact be given at all! She begins her argument by noting that the surpri se exam cannot be given on Thursday either. On Wednesday evening, she reasons, if we have not had the exam, then only Thursday and Friday would rem ain, and with Friday already eliminated, that means it would have to be Thurs- day - an inference which we are all capable of making. Similar reasoning, she concludes, eliminate s each day of the week in reverse succession . Therefore, there cannot be a surprise exam on any day. As the teacher is pondering their reasoning and nodding her agreement, a third student remarks that the other two students, having convinced themselves that no surprise exam can be forthcoming, are now quite ripe for a surpri se! And any other students who accept their reasoning, could very well be in for a rude awak- ening some day next week. Even Friday is a possibility once we' ve convinced ourselves it is not. Maybe it's just that I'm overly cautious, he adds, but I can't bring myself to accept their conclusions. I don't know exactly what is wrong with their arguments, but [ think there must be something amiss . In the end, that stu- dent was right : the exam was given on the following Thursday and everyone was caught quite by su rprise. W. V. O. Quine defined a paradox as, roughly, an apparently absurd conclusion supported by an apparently cogent argument, further dividing the field into "veridi- cal " and "falsidical " paradoxes according to whether the conclusion is in fact true or false.' The Surprise Exam fits s nugly into the veridical mold: the story in which the exam is given counts is itself an argument to the curious conclusion that a surprise can be announced beforehand. TS 24 Daniel H. Cohen 2. Conceptual Components Four parts have been se parated in this version: (a) the teacher's announcement; (b) the arg ument against the last day; (c.) the second argument eliminating the precedi ng days ; and (d) the response, or meta-arg ument, as it we re, resisting and countering the first two arguments.' When presented thi s way, virtually all of the principle components of argumentati on theory are readily di stingui shab le and avai l- able for ins pection. First, we can consider the matter rrom a narrowly logical perspective . That is, we can consider the studen ts' arguments as proofs , and aSSess their validity and soundness. cogency or fallaciousness. There is an expl icit premise-the teacher 's announcem ent . There are several conclusions-Frida y is not a possibility, no day is a poss ibility, every day is a possibility. And where there are proofs, there are inferences to evaluate: the Reductio ad absurdum by the first student, the math- ematica/ induction by the second, and the meta-reasoning by the third. One lesso n that might be drawn from the first two students' arguments is that if the reasoning is va lid , but the conclusion is false-the exam appare ntly can be given- then the premise must be faulty. T he teacher has so mehow contrad icted herse lf, or poss ibly fail ed to make a genuine assertion. To draw thi s co nclusion is to id entify impli cit, or enlhymematic, assumptions, viz., that the teacher has as- serted someth ing and she can be tru sted .4 There is the air of paradox, or even of se lf-contrad icti on, about announcing a surpri se beforehand . The fact that the exam can be given to the students' surprise, as in the story, shows that everything the teacher said could tum out to be true, so there is no logical contradiction in th e an nouncement. There is, however, some- thin g of a pragmatic contradicti on: when a sentence, p, is asserted, there is the pragmatic implication - albeit not a logi cal entailment-th at it has been asserted in earnest and that it is be li eved. The prior anno uncement of a surprise See ms to violate, by implicature, the Gricean principle of quality, 5 Th is poi nts to a seco nd , broader perspective: the rhetorical. While the teach- er 's announcement is not itse lf an argument that is intended to persuade or con- vince, it becomes part of rational -persuasion dialog ues when th e students appro- pri ate it for use as a prem ise in their argum ents. We can sta rt by asking wheth er the statement announcing the exam was true or fal se. In the event , it turned ou t to be true , but was it true all along, or did it become tru e? Is the an nou ncement best understood as a simple dec larative statement? Or is it better heard as a declaration of intent, a prediction . or so me oth er form of disco urse that is unsuitable for use as a premise in argument'?6 The teacher's statement belon gs in the troublesome category of future cont ingents. but there are versions that manage to avoid th e extraneous issues associated with them . There is, for example, a version in volv in g a "Surprise Egg," an egg that is in one often boxes arranged in a row before us . The boxes are to be opened in sequence, with the information that we will not be able to figure out wh ich box contains the egg before we get to it .' While a surpri se Informal Logic & The Surprise Exam TS 25 exam that has been slotted into a day on the calendar is not yet actualized, the egg in question is already palpably present in one of the boxes. (But which? Not the lastone, surely. And since it cannot be in the tenth, it just as s urely cou ld not be the ninth ... ) Third, we can consider the matter from a more dialectical perspective, focus- ing our attention on the exchanges between the teacher and the students, and between the first two students and the third. In particular, the exchange between the second and third students can be construed as an argument over a thesis (that the exam cannot be given) with a proponent (the second student), an opponent (the third), and an audience (the teacher and the other students). The third stu- dent's response is then seen as an objection. Their dialogue can be imaginatively continued : the second student might answer the third by admitting that his (meta) reasoning might be more practical. If we accept the conclusion that we cannot know the date of the exam beforehand, we will be motivated to study and be prepared each day. But if we accept her reasoning, then they will expect the exam and thus forestall it, by the stated condition that it be a s urpri se. There are two very important points to make here. First, further extrapolation of their dialogue can be easily imagined. As a general rule, argumentation does tend to be open-ended in this way. Objections to arguments are often best co nstrued as invitations to further dialogue' While the products of reasoning may be thought of as static proofs, argumentation itself is a process whose products are events with internal dynamics. Second, there are different audiences that might be tar- geted. Who is supposed to be persuaded: the other students (the exam can be given, so be prepared) or the teacher herself(the exam is expected, so it cannot be given, on pain of contradicting yourself)? One way to measure argument success is by whether the opponent is persuaded; a better way is by whether the target audience (which often includes the opponent) is persuaded. 3. The Principle of Meta-Rationality The third student's argument is noteworthy because it is a meta-argument. His reasoning exemplifies the Principle of Meta-Rationality (PMR), viz., that part of reasoning rationally is reasoning about rationality . In particular, he uses a meta- argument to reject the apparently good argument of his classmate: I don't know what is wrong with her argument, but the conclusion is unacceptable, so I will reject it on the assumptiol1-{)r faith, perhaps-that there really is something wrong with it. In general, it may well be a sign of strength to bow before a good argu- ment,' but in some circumstances, it can also be a sign of strength to resist an argument--even a good argument! To see exactly what goes wrong, consider the first student's argument, that Friday is out as a possibility . The reasoning is logically unimpeachable, but it can be challenged both rhetorically and dialectically. Suppose, to change the story a little, the teacher said, on Thursday, "Tomorrow you will have a test in one of your TS 26 Daniel H. Cohen classes. It will be a surp ri se in that you wi ll not know beforehand whether it will be in history, mathematics, or science-but it will be in sc ience." What can the students conclude? If they believe what the teacher says about the exam being a surprise, they cannot believe what she says about the subject being science. Conversely, if they believe what she says about a science exam, they cannot believe that it will be a surprise. If, as a result, they do not believe what she says at all, she can indeed give a surprise science exam, making her announce- ment true. But it is only true in retrospect. Its truth is contingent on the students not believing her! Her sentence, as a simple assertion, is true if and on ly if the students do not believe it. Thus, any argument they could construct that uses her assertion as a premise-which is to say, as implic it ly accepted-would, ipso/acto, be an argument with a false premise! However valid the argument may be, it cannot be sound. Thus, the pragmatic contradiction noted earlier manifests in a G6delian incom- pleteness result: the teacher's announcement is either an unbelieved truth or a believed falsehood , simi larto the Godel sentence that must be either an unprovable truth or a false theorem. In eithe r case, the announcement is unavailable for use as a premise. That is to say, there is a rhetorical transgression committed by the teacher rather than a logical fallacy comm itted by the students. She has pretended to give them infonnation they can use, but the communication was fraudulent. This was compounded by a dialectical offense. when, as in the original version, the teacher appeared to accept the first two students' arguments but then gave the exam-to their surprise-anyway. 4. Conclusion In addition to all of the issues that can be raised in connection with the Paradox of the Surprise Exam, there is a lso a pedagogical benefit to be had from actually announcing (rather than merely considering or exploring) Surprise Exams: the heightened st udent anxiety that resu lts is accompanied by a heightened level of daily preparation. There is also a significant price to be paid: the heightened level of daily preparation comes with that heightened anxiety, and the long term costs of that anxiety generally outweigh the short term benefits of the preparat ion. Differ- ent students respond differently, though, so this is commended as an area for further empirical investigation. Informal Logic & The Surprise Exam TS 27 Notes I Al so known as the Paradox of tile Unexpected Hanging, this paradox appears in many versions . Even a partial bibliography would extend to several pages, and include the names of such distinguished philosopher.; and logicians as L. J. Cohen, W.V.O. Quine, M. Scriven, A.J . Ayer, D, Kaplan and R. Montague, K. Popper, F. Fitch, andJ . Bennett. 1 W.V.O. Quine introduces this terminology in the title essay of Ways of Paradox and Other Essays (Cambridge, MA: Harvard University Press, 1976). 3 Generally, the paradox is presented as a story in which the exam is indeed given in the end. That stage has been preceded here by the fairl y explicit meta-argument, the peda- gogical purposes of which become clear. 4 This is the gist of Quine's own solution in, lI On a Supposed Antinomy," reprinted in Ways of Parada, and Other Essays . S H.P. Grice, "Logic and Conversation," in Syntax and Semantics, vol. 3, P. Cole and J. Morgan, eds. (New York : Academic Press, 1975). 6 G. Ryle 's discussion ofan argwnent for fatalism . "It Was to Be," the second lecture in Dilemmas (Cambridge: Cambridge University Press, 1954), might be a helpful supple- ment here. 7 The Surprise Egg version appears, with several others, in Martin Gardner's The Unex- pected Hanging and Other Mathematical Diversions (New York: Simon and Schuster, 1969) . • This is the point behind the claim that, ideally, "the charge of fallacy serves to extend the argument, not to cutoff debate." R.H. Johnson and J. A. Blair, Logical Self Defense (Toronto: McGraw-Hili Ryerson, 1977 (I" ed .), p. 200. ' A.F. Snoeek-Henkemans, Poster for the 4th International Conference on Argumenta- tion , International Society for the Study of Argumentation, Amsterdam , June 16-19, 1998.