Courses:

The text below includes old notes, exercises, links and reading assignments posted to the course Web site by the faculty member before and during the class.

There are old lecture notes (PDF) available for this course, from previous semesters and from when I taught a course on this material at the Summerschool in Düsseldorf last summer. Over the course of this semester, the new version will be handed out in installments. The notes are eventually supposed to become a central part of a sequel to Heim & Kratzer's textbook Semantics in Generative Grammar. The second volume is projected to be a collaboration of three authors: Kai von Fintel, Irene Heim, and Angelika Kratzer.

As much as possible, readings for this course will be made available in electronic form.

We will read some classics and some newer work. For the classics, there is now a convenient collection edited by Paul Portner and Barbara Partee: Formal Semantics: The Essential Readings. See also Frank Veltman's list of 40 classics in semantics and pragmatics.

This course is the second of the three parts of our graduate introduction to semantics. The others are 24.970 Introduction to Semantics and 24.954 Pragmatics in Linguistic Theory. Like the other courses, this one is not meant as an overview of the field and its current developments. Our aim is to help you develop the ability for semantic analysis, and we think that exploring a few topics in detail together with hands-on practical work is more effective than offering a bird's-eye view of everything. Once you have gained some experience in doing semantic analysis, reading around in the many recent handbooks and in current issues of major journals and attending our seminars and colloquia will give you all you need to prosper. Because we want to focus, we need to make difficult choices as to which topics to cover. We tend to rotate topics from year to year to keep the course fresh. Eventually, we hope to have a text book that would allow you to work through some additional topics not covered in a particular instantiation of the course. Until then, our apologies.

This time around, we will work on a number of topics involving intensionality: (1) modality, conditionals, scope in modal contexts, (2) tense, events, time adverbials, (3) questions.

Eric Schwitzgebel has a draft entry on "Belief" for the online Stanford Encyclopedia of Philosophy.

One of his examples starts like this:

When someone learns a particular fact, for example, when Kai learns that many astronomers no longer classify Pluto as a planet, he or she acquires a new belief -- in this case, Kai acquires the belief that many astronomers no longer classify Pluto as a planet.

I assume that people called John or Mary are desensitized to having their names used in examples by linguists and philosophers. But as the proud bearer of a rather rare name, this is a disconcerting experience. When I saw Quine's example "Tai always eats with chopsticks" (1966, Elementary Logic (Revised Edition), §37, pp. 90-92), that already came close to home. Then, Uli Sauerland started using the name of his son, which happens to be Kai, in his examples, as in "Kai had peas or broccoli last night" (Scalar Implicatures in Complex Sentences). Now this. I can't help but feel that these people are talking about me. Very strange.

Two optional readings for the units on modality are:

Angelika Kratzer's 1981 paper "The Notional Category of Modality."

Ferenc Kiefer's 1987 paper "On Defining Modality."

The semantics for believe given in (23) on pp. 1-12 of the lecture notes says that Mary believes Jan is loyal is true iff *all* the worlds compatible with Mary's beliefs are worlds where Jan is loyal.

Now, what does this mean for negated belief-sentences?

(i) Mary doesn't believe that Jan is loyal.

This should be true iff *not all* of Mary's belief worlds are worlds where Jan is loyal. This would leave it open whether she is agnostic about Jan's loyalty or in fact believes that Jan is not loyal.

In fact, though, we all tend to read that sentence as claiming that Mary believes that Jan is not loyal. Why that is so is an interesting issue. One approach treats this as the result of a syntactic transformation (often called "neg-raising"). Another treats it as a case of pragmatic strengthening. Here are two relevant references:

Horn, Laurence R. "Remarks on neg-raising." In Syntax and Semantics 9: Pragmatics. Edited by Peter Cole. NY: Academic Press, 1978, pp. 129-220.

Tovena, L. M. "Neg-Raising: negation as finite failure?" In Perspectives on Negation and Polarity Items. Edited by J. Hoeksema, H. Rulmann, V. Sanchez Valencia, and T. van der Wouden. Philadelphia: John Benjamins, 2000, pp. 331-356.

There are attitude verbs that resist neg-raising. In fact, in our pragmatic course, where we need to talk about reasoning about people's beliefs, we prefer to use to predicate be convinced that, precisely because it doesn't invite neg-raising.

(ii) Mary is not convinced that Jan is loyal.

Sarah pointed out that (ii) has a nuance to its meaning that is not quite predicted by our semantics. (ii) should be true iff *not all* of Mary's belief worlds are worlds where Jan is loyal. In other words: iff *some* of Mary's belief worlds are worlds where Jan is not loyal. Now note that this is still compatible with Mary in fact being convinced that Jan is not loyal (= *all* of Mary's belief worlds are worlds where Jan is not loyal). But we tend to hear (ii) as expressing doubt on the part of Mary, as if it meant that *some but not all* of Mary's belief worlds are worlds where Jan is loyal.

My immediate response is that this is an instance of the kind of Gricean scalar/quantity implicature that generates the understanding that *some* sometimes means *some but not all*. The idea is that if the speaker knew that Mary is convinced that Jan is not loyal, he or she should have said so. Since they didn't, we deduce that they must think that Mary is not convinced that Jan is not loyal. A similar phenomenon arises with the modal may, which is often read as meaning the same as "may and may not", which is what Larry Horn calls the "two-sided" reading of may. We will explore this in detail in the pragmatics course in the fall.

There is a possible connection here to recent work on the scope of modals, but even if there isn't, this is a good occasion to refer to the following:

von Fintel, Kai, and Sabine Iatridou. "Epistemic Containment." Linguistic Inquiry 34, no. 2 (2003): 173-198.

Butler, Jonny. "What mustn't and can't mustn't and can't mean." Master's thesis, University of York, 2001.

———. "A minimalist treatment of modality." Lingua 113 (2003): 967-996 ("a much polished version of the modal/negation stuff from the MA").

Stowell, Tim. "Tense and Modals." To appear in J. Guéron and J. Lecarme, eds.

I will add this to the lecture notes at some point.

Due Class 6: Exercise 2.1-2.5 from the lecture notes.

Due Class 7: a careful step-by-step derivation of the truth-conditions of sentence (31) on pp. 2-10 of the lecture notes.

I will provide some of the references I mentioned in class for further exploration of topics in modality. Here are the references having to do with Keith DeRose's investigation of the vagueness of epistemic possibility.

DeRose, Keith. "Epistemic Possibilities."

———. "Simple 'might's, indicative possibilities and the open future".

Teller, Paul. "Epistemic Possibility."

Hacking, Ian. "Possibility."

———. "All Kinds of Possibility."

Gibbs, Benjamin. "Real Possibility."

Edgington, Dorothy, "Conditionals." In The Stanford Encyclopedia of Philosophy. Fall 2001 Edition. Edited by Edward N. Zalta.

Peter Suber's page on the 'paradoxes of material implication.'

Here are a couple of optional exercises for those of you who want to try their technical prowess.

In class, we mentioned a number of ways one could get if-clauses to restrict a modal operator like might.

1. The simplest way is to assume that the proposition expressed by the clause under if is intersected (via a generalized version of Predicate Modification) with the set of worlds accessible from the evaluation world. Your first task is to formulate a generalized composition principle of Predicate Modification that could apply in this case.
   
   
2. Follow-up: Discuss what we could do to ensure that the if-clause indeed supplies a set of worlds for the rule you just formulated to work on. After all, until now only Intensional Functional Application was allowed to introduce the intension of an expression into the semantic calculation.
   
   
3. The other two ways amount to high-type wizardry substituting for the simple intersection rule you just formulated. Your next task is to formulate two meanings for if that would do the trick: (i) an if operating on the accessibility function argument of the modal, (ii) an if operating on the modal itself. In each case, your proposed meaning should give the same result for an entire sentence (like If we are on Route 183, we might be in Lockhart) as the simple intersection semantics.

Feel free to turn in your answers. You will get brownie points.

Draft Chapter 5 (on De Dicto -- De Re) available. (PDF)

For what it's worth, here is a short outline of what to expect this semester:

Part 1: Modality

• Chapter 1: Intensionality
• Chapter 2: Modality
• Chapter 3: Conditionals
• Chapter 4: Ordering
• Chapter 5: De Re - De Dicto
• Chapter 6: The Third Reading
• Chapter 7: The Argument Structure of Modals [not this year]
• Chapter 8: Intensional Transitives [not this year]

Part 2: Tense and Aspect

• Chapter 9: Temporality
• Chapter 10: The Referential Analysis of Tense
• Chapter 11: Intervals
• Chapter 12: Temporal Adverbials
• Chapter 13: The Perfect
• Chapter 14: Imperfects

Part 3: Questions

Part 4: Plurals [not this year]

I have uploaded an updated version of the bibliographic guide to topics relevant to modals and conditionals, which you may use for your explorations.

Chapters 1 and 4 have been taken offline. What you should use instead are Chapter 12 from Irene Heim and Angelika Kratzer, Semantics in Generative Grammar (Blackwell, 1997) and Angelika Kratzer, "The Notional Category of Modality," in Paul Portner and Barbara Partee, Formal Semantics (Blackwell, 2002), pp. 289–323.

We are currently working on Chapter 6, which deals with the "Third Reading." Unfortunately, the new version of the lecture notes is not at all ready. So for now, the official reading is Section 6 of the old lecture notes from last summer, pp. 70-83.

A supporting reading is Orin Percus, "Constraints on Some Other Variables in Syntax." Natural Language Semantics 8, no. 3 (2001): 173-229.

Robert Binnick's Bibliography of Tense and Aspect

Foundations of Temporal Logic - The WWW-site for Prior-Studies

Stanford Encyclopedia of Philosophy :

Galton, Antony. "Temporal Logic." In The Stanford Encyclopedia of Philosophy. Winter 1999 Edition. Edited by Edward N. Zalta.

Copeland, B. Jack. "Arthur Prior." In The Stanford Encyclopedia of Philosophy. Winter 1999 Edition. Edited by Edward N. Zalta.

Two papers mentioned:

Partee, Barbara. "Some structural analogies between tenses and pronouns in English." Journal of Philosophy 70 (1973): 601-609.

Musan, Renate. "Tense, Predicates, and Lifetime Effects." Natural Language Semantics 5, no. 3 (1997): 271-301.

Please read the Partee paper for the next class.

The von Stechow paper I mentioned in Class 19 is:

von Stechow, Arnim. "Temporal Prepositional Phrases with Quantifiers: Some Additions to Pratt and Francez (2001)."

The paper this takes off from is:

Pratt, Ian, and Nissim Francez. "Temporal Prepositions and Temporal Generalized Quantifiers." Linguistics and Philosophy 24, no. 2 (2001): 187-222.

The Iatridou et. al. paper should be read before Class 20.

Seth asked in class what else Hamblin had worked on. In semantics, Hamblin is mostly, and justly, famous for his early montagovian paper on questions: C. L. Hamblin. "Questions in Montague English." Foundations of Language 10 (1973): 41-53. [By the way, Foundations of Language was the precursor journal of Linguistics and Philosophy.]

Here are two pages about Hamblin (who died in 1985 and was both a philosopher and apparently a pioneer in computer science):

1. Charles L. Hamblin
   
   
2. Charles L. Hamblin

Two tidbits: "According to his obituary in the Sydney Morning Herald, Hamblin spoke 25 languages, mostly from the Asian-Pacific region, but also including ancient Greek. ...At the time of his death, he was apparently attempting to set words of Wittgenstein to music."