Announcements INTD0112 If you haven’t submitted your research proposal to me yet, please do so ASAP. Please Introduction to remember that I have to approve your topic. Don’t procrastinate. Linguistics Homework #5 will be posted by tomorrow morning. Due date is Friday April 20. Make sure you watch Part Two of the Human Lecture #15 th Language movie before you attempt to answer April 12 , 2007 the questions. More syntax Transition I know I have to curb my passion for things Within the principles and parameters framework, syntactic at one point, so this is the last session cross-linguistic variation is attributed to the existence on syntax. We have to move on with other of a number of parameters in UG, each of which has topics on the syllabus. binary options to be set one way or the other on the basis of the primary linguistic data. Today, I want us to discuss 2 topics: a. VOS/OVS/OSV languages. and b. A couple of universal principles of Under this approach, then, acquiring a human grammar: Constraints on movement and language is nothing but a process of parameter fixation. Here’s the table we ended up with last time: Binding Principles. Verb attraction and subject Parameters and languages placement in head-final languages Parameter English Japanese French German Welsh We have seen how the interaction of different parameters in head-initial languages can give HD Head-initial Head-final Head-initial ? Head-initial parameter rise to different languages, e.g., English, Welsh, French, and German. Subject Specifier of Specifier of Specifier of ? Specifier of placement IP IP IP VP Now, one should wonder if we see this same parameter parametric interaction in head-final languages. Verb I down to V ? V up to I V up to I V up to I movement It turns out that, at least as far as we know, parameter such interaction does not exist. V2 No ? No Yes ? parameter 1 Verb attraction and subject Verb attraction and subject placement in head-final languages placement in head-final languages (1) (2) CP CP Let’s consider the interaction of subject ru ru placement and verb attraction with head- IP C IP C finality in tree-geometric terms. ru ru NP I' VP I Here’s how the two trees would look like for Subject ru ru two head-final languages, one of which places VP I NP V' the subject in specifier of IP, and the other ru Subject ru places it in specifier of VP, with verb Object V Object V attraction going either up or down: Subject position is high Subject position is low Verb attraction and subject How about VOS, OVS, and OSV placement in head-final languages word orders then? VOS: Malagasy As you can see, no matter where the subject is, it will manasa ni lamba ny vihavavy precede the verb, and whether V moves up or I moves wash the clothes the woman down, there is no effect on word order. “The woman is washing the clothes.” The parametric approach thus predicts that no comparable word order variation in head-final OVS: Hixkaryana languages can result from the subject placement and kanawa yano toto verb attraction parameters, which seems to be the canoe took person case. “The man took the canoe.” How about VOS, OVS, and OSV VOS and OVS word orders then? VOS and OVS orders share one property: They both have the subject in final position. OSV: Nadëb Perhaps there is a parameter that accounts for samũũyyiqa-wùh subject position in human language, then. howler-monkey people eat Mark Baker suggests a “subject side “People eat howler-monkeys.” parameter”: “Subjects may occur initially or finally in the sentence.” 2 VOS and OVS VOS and OVS (3) (4) CP CP The interaction of the subject side parameter ru ru with the HD parameter should give us VOS C IP IP C ru ru (Mirror Japanese) and OVS (Mirror English): I' NP I’ NP ru Subject ru Subject I VP VP I ru ru VObject Object V Malagasy VOS order Hixkaryana OVS order Predictions? Deriving OSV (5) (6) Now, here’s a question: CP CP Would the verb attraction parameter and the ru ru subject placement parameter be relevant to either IP C IP C of these two language types, or both, or neither? ru ru I' NP VP I ru Subject ru VP I V' NP Correct, it should be relevant for OVS orders, ru ru Subject giving rise to Mirror Welsh. But does it exist? Object V Object V This is what Nadëb and Warao are claimed to be. Hixkaryana OVS order Nadëb/Warao OSV order Summary table for the 6 language types Something just doesn’t seem right Parameter English Japanese Welsh Malagasy Hixkaryana Nadëb That looks like a nice story, except for one HD H-initial H-final H-initial H-initial H-final H-final thing: It just can’t be right. parameter For one thing, our table makes it seem like all Subject side S-initial S-initial S-initial S-final S-final S-final these types of basic word order languages parameter should have the same statistical distribution, Subject Specifier Specifier Specifier Specifier which is obviously not the case. placement Irrelevant Irrelevant of IP of VP of IP of VP parameter Consider their frequencies in Tomlin’s sample Verb attraction I to V Irrelevant V up to I Irrelevant Irrelevant V to I again: parameter 3 Distribution of basic word order Frequency of VSO: Bonus points! types in the world’s languages Nothing that we have said so far explains why VSO Word order # of Languages % languages like Welsh should account for around 9% of human languages. They are not rare; but they are SOV 180 45 not predominant either. SVO 168 42 There is a mathematical explanation for the statistical VSO 37 9 distribution of VSO languages, however, given the assumptions we made in today’s class. If you can VOS 12 3 come up with that, you get 10 points of extra credit. OVS 5 1 Make sure you illustrate your answer with tree diagrams. Let’s set next Tuesday as a deadline for OSV 0 0 this extra credit assignment. Cheating never pays! More bonus points Back to UG principles Another problem with the table is that it cheats a little bit: For one thing, it says that the verb movement parameter is irrelevant to head-final languages like Japanese, but at the In the rest of today’s lecture, I would like us to same time makes use of that specific parameter to explain the difference between Hixkaryana and Nädeb. discuss examples of universal principles of Since the argument for making verb movement irrelevant to grammar, and how they help us account for head-final languages is compelling, we should probably keep certain facts about the syntax and semantics of to that assumption. But then we have to explain how languages like Hixkaryana and Nädeb exist, and why they are so rare. human languages. If you can think of an analysis whereby we explain the word order in these languages, while accounting for the rarity of these languages at the same time, you get 10 more extra credit points. Deadline is next Tuesday as well. Constraints on Wh-movement Constraints on Wh-movement We have already seen examples of wh-movement in But now consider these cases of wh-movement: English (remember “t” is the trace of the moved wh- *Who did you meet Mary and t ? phrase): *Who do you believe the claim that Mary met t? Who did John meet t ? *Which book did Mary talk to the author who wrote t? *Who do you wonder whether Mary met t ? Notice that the distance between the wh-phrase and its original position in the D-structure could extend across *Who did Mary talk to John without meeting t ? several CPs: Who did you say that John met t ? Obviously, wh-movement is not unconstrained. There are cases where the movement is for some reason Who does Mary believe that you said that John met t ? blocked. 4 Islands Islands The substructures out of which wh-movement is Adverbial clauses are islands: blocked are technically called islands. *Who did Mary talk to John [without meeting t] ? Complex NPs are also islands: Coordinate NPs are also islands; *Who do you believe [NP the claim that Mary met t ]? *Who did you meet [NP Mary and t ]? Relative clauses are also islands: Island constraints cannot possibly be learned *Which book did Mary talk to [the author who wrote t]? on the basis of primary linguistic data that the Also, embedded CPs introduced by a wh-word child hears around her. Therefore, they must act as islands blocking wh-movement: be built-in. *Who do you wonder [CP whether Mary met t ]? Link on islands Binding! Let’s revisit some English examples from early in the semester on the difference between reflexives and A discussion of islands is available through other kinds of pronouns. one of the links on the textbook website. Go Before we do that, just a quick note on “convention”: there, click on the Syntax chapter, then click To indicate coreference between two elements in a on the link to Constraints. Remember that you sentence, linguists use the convention of subscripting need to have an account to be able to access both elements with the same index, e.g., the materials. Johni said that hei already had lunch. (John = he) Johni said that hej already had lunch.