Course Outline—Logic 3003, Spring 2005 Instructor: Bryson Brown Office: TH 222 Phone: 329-2506 Email: brown@uleth.ca Office Hours: Tuesday, Thursday from 11:00-12:00 and by Appt. Logic is often described as a normative study of reasoning, by which we mean a study of how to evaluate reasoning as good, or bad. And reasoning is done (or at least expressed) in language. But ordinary language isn’t very well suited to producing systematic theories of how to reason in it. Though we can often arrive at a pretty wide consensus that a particular argument is persuasive, or that it’s completely unconvincing, it’s very hard to articulate clear and straightforward criteria for these judgments. The result is that there turn out to be a lot of cases where people just plain disagree about whether an argument is good or not. Aside from bias and prejudice and sheer stubborn cussedness, there are perfectly honest reasons for this kind of disagreement. Ordinary language is fuzzy, irregular in its usage, and sometimes outright ambiguous—so the sentences used in an argument may well mean something different to different audiences. Further, each of us brings background beliefs to our evaluation of arguments—since these aren’t often made explicit (sometimes we may not even be aware of them), they can lead us to differ on whether an argument succeeds without being able to see why we differ. This was a particularly acute problem in mathematics, where apparently convincing proofs were occasionally overturned by clever people who found ways to construct odd, unanticipated counter-examples. Math isn’t supposed to be like that, so mathematicians responded by trying to systematize the process of reasoning. Today’s techniques in propositional logic and quantificational logic grew out of this effort, though they have deep roots in philosophy, reaching back to Aristotle, early mathematicians like Pythagoras, and figures like the Stoics, who had interesting ideas about some of the connectives we treat using truth functions today. The result of this work, well begun in the 19th Century by Frege and continued by Russell and Whitehead, Wittgenstein, Post, Carnap, Quine, Lewis, Gödel, Kleene and others, is modern formal logic. Our focus will be on classical logic, which remains the heart of the discipline. In formal logic we use artificial languages with simple, fully-specified structures. This allows us to pass over the ambiguities and fuzziness of ordinary language, though they re-emerge when we try to connect the elegant results of formal proofs back to arguments in our natural languages (which are, of course, the ones we were interested in in the first place). This is the first run of Logic 3003, following a re-working of our logic program. 3003 is intended as a second course in classical formal logic. We will be reviewing some familiar material as we go along, including proof theory (natural deduction) for propositional and quantificational logic. But we will go beyond this practical approach & prove some important facts (called metatheorems) about these systems and their relation to classical semantics for these logics. Some of the course work will involve using a software package called Simon and Simon Says. The software is under development at Simon Fraser University, but this is not a first run, so I’m optimistic that it will work well. The details of the assignments that I will be using Simon for are still being worked out. This is partly because I’m working with this software for the first time, and partly because I’m still trying to work out the balance between review/improvement of your basic logical skills, and studying the systems formally, including proving a few things about them. Grades will be based on two tests (25% each), two written assignments(15% each), and five smallish Simon-based assignments (4% each). My usual grading scale is: A+: 95 and higher; A: 85-94; A-: 80-84; B+: 78-79; B 73-77; B-: 70-72, and so on. I’ll let you know if there are any adjustments to this. Initial Course Schedule (tentative): Week 1: January 6, 2005. Introductory remarks. Get Simon up & running! Proof and Consequence, Chapter 1. Week 2: January 11, 13. Chapter 2. Propositional proof theory: logic as a search for general kinds or structures of good arguments (not topic or context specific); telling good arguments from bad; English connectives; sequents and ; documenting proofs (some discussion here of alternative ways to approach proof construction & documentation); L and the rules of L; working with Simon’s proof editor. Week 3: January 18, 20. Chapters 2,3. Brief review of translation (working with Simon’s formulation editor); structural properties & the general idea of a consequence relation (remarks re. multiple-conclusion systems & duality); compactness. Formal definition of L (our propositional language) R (our rules of deduction), and proofs. Theorems and uniform substitution. First metatheorems (3.1, 3.2). Theorem and Sequent introduction rules. Week 4: January 25, 27. Chapter 3. Effectiveness. Arithmetizing our language & proof theory; searching the proof space. The standard interpretation of L. Tautologies, inconsistencies and substitution instances. N-ary truth functions and expressive completeness. Conjunctive normal forms, disjunctive normal forms. Sheffer strokes, dagger strokes. Major Assignment 1: Distributed January 27, due February 3. Week 5: February 1, 3. Chapter 3. Strong Soundness of L: A proof by strong mathematical induction. (Remarks on mathematical induction.) Week 6: February 8, 10. Chapter 3. Weak & Strong completeness of L: Kalmar’s method. Sequents & truth-table lines: finding a proof for every such sequent. Completeness demonstrated constructively by defining canonical proofs of the tautologies. Strong completeness & conditionals corresponding to sequents. The significance of soundness and completeness; maximality. Week 7: February 15, 17. Why stay with just one proof when you can have two? A Henkin-style weak completeness proof for propositional logic: The Strategy—show that every consistent wff is satisfiable (i.e. has a ‘1’ somewhere in its truth table). Test 1: February 17. Week 8: March 1, 3. Introduction to quantifiers & quantifier rules. Simple translations & reading Lq sentences back into English; UI and EE rules, the importance of arbitrary names, and methods to ensure that arbitrary names are correctly used. Week 9: March 8, 10. Practice with some more quantifier proofs; some discussion of proof technique. Introduction to the formal account of Lq. Week 10: March 15, 17. Interpretation of Lq. Satisfaction of wffs and quantificational models: Domains, interpretations of proper names, and predicate letters, and the treatment of variables and arbitrary names by assignments. Second assignment distributed March 17. Week 11: March 22, 24. Soundness of Lq. I will be away March 24. But I will be in for extended office hours later that afternoon & watching email closely to provide feedback and advice as you work on the second assignment. Week 12: March 29, 31. Second assignment due March 29. Completeness of Lq. The Henkin strategy reviewed, construction of canonical models for sets of sentences, Week 13: April 5, 7. Remarks on the strength and limits of Lq, second-order logic and induction, and incompleteness issues. Final exam (takehome) distributed April 7. Week 14: April 12, 14: No Classes: I will be at a conference in Dubrovnik, Croatia. I will be watching my email and responding to questions while there as best I can. Exam Period: I will be in during the first week of exams—I’d like to schedule a couple of sessions for discussion/ advice on completing the exam, review of any issues that remain puzzling for you all.