\documentclass[10pt,a4paper]{article} \pagestyle{empty} \usepackage{graphicx} \begin{document} \begin{verbatim} Meaning, Reference & Modality. Assignment 4 Andreas van Cranenburgh, 0440949 Friday, December 18, 2009 Exercise 1 1) That is the man who swam the channel twice on one day This sentence is what Strawson calls an identification statement, which presupposes that the hearer already knows that a person like the one identified exists. This presupposition is only part of Strawson's theory, whereas the other theories trivialize such sentences to an "A is A" tautology. For Russell's theory "that" is difficult to deal with, whereas for Strawson's theory it is natural to interpret "that" as it is being used in an assertion with a supplied context. Although the first "the" does specify a clear uniqueness condition (swimming the channel twice on one day is quite a feat), the second occurrence of "the" points to a specific channel by convention (ie., the British channel), not by description, which is a feature fitting well with proposals which acknowledge the role of context. Donnellan's proposal adds an extra dimension to this, distinguishing between attributive and referential descriptions. While the other proposals would have difficulty accounting for a reaction such as: - No he only swam the channel once, a storm came. Donnellan's proposal deals with this by stating that the description is referential, instead of attributive, which is the standard Russelian way for a description to function. Thus his proposal makes a distinction between semantic and speaker reference, which the other proposals neglect and dismiss as some extra-linguistic phenomenon. 2) Put your garbage here The fact that this is an imperative sentence is already a problem for all of the proposals. There is a lot of information in the message that is not part of the sentence. For example, the message is not intended to convey that the Napolitan mafia is welcome to use this location for their purposes; the message is meant for people from the neighborhood with a reasonable amount of appropriately packaged garbage. Furthermore the sentence is very indexical. It has no subject, so it is unclear whose garbage is talked about. The word `here', curiously, refers to itself, just like the ``you are here'' signs on maps. These aspects make the sentence a good example arguing for the context sensitivity of Strawson's proposal. Relying on the context is preferrable to assuming that there exists an exact and complete eternal logical form hidden somewhere, as assumed by Russell's proposal, which would need to specify such details as precisely how much garbage is reasonable to put there, where exactly it is permissible to put the garbage (demarcating the borderline between `here' and `not here'), etc. Such attention to detail is possible (and needed) in mathematics, but not feasible for everyday language. Humans make sense of this sentence without considering these details, so an account of semantics should not insist on filling them in. Exercise 2 3) George sorted red apples, and Susan did too. Susan must have done whichever it was that George did, because `too' implies a similarity is at play, not just that a similar description holds. This seems to imply that what Susan did depends on a context where George did something, which favors a contextualist reading. On the other hand, I would consider the sentence infelicitous if Susan did a different kind of red apple sorting than George, so it contributes to the intuitions of the truth-conditions, which makes it a semantic issue. Even so, one can always come up with a situation or with a longer utterance to make the sentence correct after all, e.g., George's apple sorting work might be contrasted with Susan's apple shopping as a joke: George sorted red apples, and Susan did too, in a way. What it comes down to is that there is a strong default reading where Susan and George did exactly the same thing, but this is accompanied by an endless stream of possible exceptions and special situations when the sentence is used in a suitable context, due to the elastic nature of language. Exercise 3 4) The last five years every MSc student interviewed an ILLC researcher because the Core Logic professor told him to do so, and the next five years this will unfortunately be the same. Dekker's problem is that the sentence may receive a mixed reading, with 'an ILLC researcher' being read both specifically and non-specifically, e.g., in the last five years the students each interviewed some ILLC researcher of their choice, and in the next five years the Core Logic professor might specify a specific ILLC researcher everyone has to interview. The problem here is that the choice of reading upsets the neatly defined division of labor of structure versus pragmatics, because ruling out such readings requires that pragmatics interferes with logical form. Recanati-Stanley discussion: The minimalists will surely claim that this phenomena can be accounted for with structural explanations (for example, a quantifier ambiguity might introduce an internal referent so that its reading can be re-used later). Contextualists might prefer an account in terms of the context supplying a plausible reading (e.g., there being too many students for the specific reading). I think neither of these explain the issue. I think the mixed reading is blocked by being so strange and unlikely as to require a very explicit and elaborate description, whereas the sentence as it is now is simple and thus goes with a simple interpretation. So the issue is resolved neither by structure nor context, but by shortcuts in interpretation. I also think the focus on scope ambiguities in general to be an artefact and bias of logical languages, so the issue is somewhat artificial; I suppose this makes me a contextualist. Exercise 4 Humpty Dumpty: - Russell-Strawson & Donnellan-Kripke debate: Humpty-Dumpty would prefer Donnellan's and Strawson's side, because they are more ambiguous, giving him more freedom `to be master.' - Recanati-Stanley debate: Humpty-Dumpty is squarely in the contextualist camp, because he is basically arguing for completely free pragmatic enrichment and arbitrary sense-transfer. Alice: Alice would be, mutatis mutandis, in the opposite camps, opting for the more literal and eternal interpretations, if only to win the discussion, perhaps. Humpty Dumpty is sort of right that every word can mean what you want it to mean, consider pathological cases such as poetry where it is licensed to do all kinds of strange things. One variable that constrains this is how difficult it is to recover the intended meaning. What should be kept in mind is that normally one actually wants to communicate effectively, and because of this we rely on conventions which form the path of least resistence. Another thing to keep in mind is that the arbitrariness of language is limited. While single words have a primary, arbitrary meaning, such as `bottle' meaning BOTTLE, most of language consists of derived meanings such as compound terms and metaphors that are motivated somehow by appealing to human imagination. The `amount' of meaning also has to remain tractable, it is not possible to claim some new word stands for the collected works of Shakespeare and hope to impress. Humpty-dumpty and Alice's discussion is ultimately not at all about language, but about him not being co-operative. I think in the end all proposals and formalisms break down somewhere, and that makes it perhaps counterproductive to look for problematic example sentences, especially when they are not particularly realistic and only meant as a `nice knockdown argument.' It is better to measure success using naturalistic data, starting with more common language and leaving obscure ambiguities for later. Much of the philosophy of language literature seems to take too narrow a view of language and ignore how much ambiguity is a cognitive phenomenon, dealt with throughout subjective experience, not purely through linguistic mechanism. \end{verbatim} \end{document}