Ling537(Fall06): Prof. Hoji
Outline Draft ver.1
Maki Irie
Is intermediate reconstruction possible? Review of Nishigauchi 2002
There are several issues that you intend to address.
Which one(s) of those do you want to focus on?
I think it would be more effective if you focused on one or two issues.
You can address the other issues too, but I am not sure if you can address all the issues in depth in your term paper.
Hoji-sensei, I understand I should consider the manageability, but I am not sure which issues I should omit (or not focus on). But what I think about this is the following.
I think I should understand the X0 scope bearing elements (Heycock 1995 and Fox 1999,2000), to design experiments, in conjunction with FD-BVA. In addition, I think I will have to refer to referential/non-referential reading of quantifiers (#-cl-no-N, vs. N-cm ... #-cl) since referential/nonreferential readings of quantifying expressions may be of importance in scope interaction between X0 SB and QP. So, I think the discussion on X0 SB and floating quantifier would be necessary for me.
Then, it would be better to omit close discussion on Condition C and A and significance of the validity of Hsub. In addition, maybe it would be better to concentrate on the experiments that are based on SB and BVA readings, not interaction between QPs along with SB?
It should not be very difficult to demonstrate that Hsub's on Conditions A and C are not well supported empirically. Although you may need to do the relevant experiments before you finish your thesis, assuming that you will base it upon your critical review of Nishiaguchi 2004, that can perhaps wait. So, why don't you concentrate for your term paper on the two issues you have noted above, which you would have to discuss together at one point anyway.
You should try to state (i) the basic generalization proposed/adopted in Heycock 1995 (and if possible Fox 1999 (the relevant chapter in Fox 2000 is perhaps the same as Fox 1999; so you can forget about Fox 2000, for now)), and (ii) your current understanding of the specific/non-specific readings in Japanese that are being discussed in Nishigauchi 2004.
1Outline
The main purpose of this project aims at close examination on the hypothesis proposed by Nishigauchi 2002, summarized as follows:
(1)The 'scrambled' NPs are reconstructed not only to their original position but also to the periphery of various projections (edge of CP, vP, and so on) that dominate their original position.
What is the theory behind "to be reconstructed to …" in Nishigauchi 2002?
I.e., how is "being reconstructed to …" expressed in Nishigauchi in theoretical terms?
Theory behind "to be reconstructed to..."
I think that Nishigauchi uses the terms 'reconstruction' 'to be reconstructed to' as primitives. He uses rightward arrows to demonstrate the schematic structures involving reconstruction here and there (for example, (N6) in 2.2, (N38) in 5, and so on), which seem to depict the downward movement of the moved NP at LF. My understanding is that 'reconstruction' by Nishigauchi means LF lowering of an NP moved at surface structure. The possible sites for this lowering are claimed to be the NP's original position or the periphery position(s) between the surface position and the original position. Okay. Try to identify the relevant exposition/remarks in Nishigauchi 2002 that leads you to the above conclusion, and give that in the paper, where you state clearly what you think Nishigauchi 2002 assumes/claims.
Specifically, Nishigauchi 2002 claims that there are [more thanat least three <==at least three? <-- yes, thank you for the correction] positions where a dislocated NP in the scrambling construction can be interpreted. Namely, the dislocated NP in the scrambling construction (NPDL) can be interpreted in its surface position as in (3a), in its original position as in (3c), or in the intermediate position between the surface position and the original position as in (3b):
(2)Surface:NPDL ....[ .... ec .... ] (ec= the position to which NPDL is thematically related.)
(3)LF:
a.NPDL ....[ .... ec .... ] -- no reconstruction
b.... [ NPDL [ .... ec .... ] -- the edge of the intermediate phrase
c.... [ .... NPDL .... ] -- the original position (reconstruction)
The main point of Nishigauchi 2002 is the possibility of the intermediate reconstruction site as in (3b): reconstruction to the intermediate phrase is possible. In fact, he argues that the (3b) type reconstruction is forced if the NPDL is necessarily interpreted in the intermediate position in order to observe binding theory such as condition A and C.
However, his claim and arguments have problems. The first problem lies in the referential and non-referential reading of QPs. The second problem is that his auxiliary hypotheses (Hsub's) may not be correct. Addressing these problems, I would like to argue that his arguments do not show that his main hypothesis (Hmain) is valid, with an emphasis on the importance of the validity of Hsub.
Considering the discussion outlined above, I would like to claim the following hypothesis, against Nishigauchi:
(4)The intermediate reconstruction in scrambling constructions is not possible.
<==If you can establish that something is possible only if 'intermediate reconstruction' is possible, then you will have a negative prediction.
Taking (4) as Hmain, I would like to design the two sets of experiments in the following:
(5)a.Interaction between a scope-bearing X0 element (SB) and BVA reading.
b.Interaction between a SB and QP.
Can you provide the schematic structure of each of these experiments on the basis of your current understanding?
<-- (5a)
I have come up with the following proposition that can be led to a negative prediction so far, in conjunction with SB and BVA:
Hmain: (4)
Hsub: FD-based BVA and SB and referential/nonreferential readings
(i)In the following configuration, if BVA(QP1, soko) is available, then SB cannot take scope over QP1. If SB takes scope over QP1, then BVA(QP1, soko) is not available.
If the clause-boundedness of the movement or whatever is involved in Deep OS is valid, then the launching site of the QP1 cannot be the position of the ec. If BVA(QP, soko) is available below, where could the QP originate? Before you try to answer the question, you may first want to construct actual examples and see what judgments you might get.
QP1-ni X-ga [[...soko...]-ga ... ec ... SB to] V
(ec is the position to which QP-ni is related thematically) QP1 is an FD-QP.
Fine. As long as you have an independent test for determining when an SB can or cannot take scope over a given QP, you are in business. (I am assuming that we have a fairly solid test for determining when FD-BVA(A, B) is availability or not available.) So, you should work on that.
You should also consider the Surface OS cases; that would be the case of reconstruction of BVA, and you can see how the SB-Q interpretation interacts with the availability of the FD-BVA.
[Forget about this for now ==>
(ii)reconstruction possibilities
1. no reconstruction
QP1-ni X-ga [[...soko...]-ga ... ec ... SB to] V
BVA(QP1, soko) -- OK; SB>QP1 -- not OK, as QP1 stays in the matrix clause. <==Is not the same as the point in (i)?
2.reconstruction to the original position
X-ga [[...soko...]-ga ... QP1-ni... SB to] V
BVA(QP1, soko) -- not OK; SB >QP1 -- OK
3.intermediate reconstruction (which is not possible)
X-ga [QP1-ni (edge of embedded CP) [...soko...]-ga ... ec ... SB to] V
BVA(QP1, soko) -- OK; SB>QP1 -- OK as QP1 is in the embedded clause.]
I think Nishigauchi's claim that intermediate reconstruction is possible is not written in a falsifiable way. However, I think the hypothesis in (4) is falsifiable.
I need to figure out what kind of readings are expected in the case of SB > QP and QP > SB to make a testable prediction. Yes.
<-- (5b)
I am not so sure as (5a) about this case, and I need to study more. But, I tried to provide the structure on the basis of my current understanding.
Hmain: (4)
Hsub: Ringe-san's theory of QPs and scope interaction and SB and referential/nonreferential readings
(iii)In the following configuration, if the scope QP1 > QP2 is available, then SB cannot take scope over QP1. If SB takes scope over QP1, then QP1>QP2 is not available.
QP1-ni X-ga [QP2-ga ... ec ... SB to] V
(ec is the position to which QP-ni is related thematically) QP1 is an FD-QP.
Again, what is crucially needed is an independent test for determining when an SB can or cannot take scope over a given QP, like "three>want" vs. "want>three." Can you think of other "QPs" for a test like that? I guess people have talked about "want>only" vs. "only>want." Also "can>only" vs. "only>can," discussed in some works by Koizumi and Tada (and possibly others?)
[Again, forget about this for now. ==>
(iv)reconstruction possibilities
1. no reconstruction
QP1-ni X-ga [QP2-ga ... ec ... SB to] V
QP1 > QP2 -- OK; SB>QP1 -- not OK as QP1 stays in the matrix clause.
2.reconstruction to the original position
X-ga [QP2-ga ... QP1-ni... SB to] V
QP2>QP1 -- not OK (I have to learn whether or not it is the case); SB >QP1 -- OK
3.intermediate reconstruction (which is not possible)
X-ga [QP1-ni (edge of embedded CP) QP2-ga ... ec ... V to] SB
QP1>QP2 -- OK; SB>QP1 -- OK as QP1 is in the embedded clause.]
I need to study Ringe-san's work closely, in addition to Heycock. But I'm thinking about omitting this part as I said at the beginning of this document.
I will show the result of the experiments, and discuss the implication of the results.
2Nishigauchi 2002
In this section, I will summarize Nishigauchi 2002.
2.1Claim
2.2Interaction with SB
2.2.1Condition C
2.2.2Condition A
2.3Wh-marker
2.3.1Condition C
2.3.2Condition A
3Issues
In this section, I would like to discuss three issues.
One has to do with referential and non-referential readings of the NP containing QP in it. I would like to compare the differences between (6) and (7), the former of which Nishigauchi uses as QP in his observation:
(6)a.[NPnannin-no (...) gakusei]-o
how:many:cl-gen student-acc
b.[NP (...) Q-gen (...) N]-CM
(7)a.[NP(...) gakusei]-o (...) nannin
student-acc how:many:cl
b.[NP (...) N]-CM (...) Q
He seems to assume that (6) itself is ambiguous, in the sense that it can be interpreted as nonreferential, namely it asks the number of the students without presupposing a specific set of the students, or as referential, that is, the speaker presupposes a specific set of students and asks how many of the students of that set. However, it is suggested by Hoji and Ishii (2004: 110 (28)) that (6)(6) and (7) (7) patterns differently in that the (6) type QP is, or tend to be, used to refer to a specific entity while the (7) type, involving a floating quantifier, is not.xxx <==You need to provide the minimal info. here. .
The second issue addresses Condition A and Condition C in Japanese, which is are crucially used in Nishigauchi 2002. I would like to point out the dubious status of these conditions in Japanese, and argue, along with Ueyama 1998: xxx <-- I think I need to understand Ueyama 1998 chapter 2 Appendix A.3.2 regarding Condition C and reconstruction effects, and chapter 4 Appendix C (condition C/D), that they should be avoided to be used when designing experiments. (Well, at least in the case of Condition C.)
The third issue is regarding X0 scope-bearing elements (SBs). I would like to review Heycock 1995 and Fox 1999, 2000 so that I can incorporate this into my experimental design.
3.1Referential/nonreferential readings
3.2Condition A and Condition C
3.3Scope-bearing elements
4Hypotheses and predictions
The Hmain of this section is the following:
(4)The intermediate reconstruction in scrambling constructions is not possible.
I would like to make predictions after I understand Heycock and Fox better. I would like to make the following two sets of the experiments:
(5)a.Interaction between a scope-bearing X0 element (SB) and BVA reading.
b.Interaction between a SB and QP.
5Experiments and results
This section will demonstrate the experiments and their results.
5.1Scope-bearing element and BVA
5.2Scope-bearing X0 and XP
5.3Results
6Conclusion
This section concludes the paper.
References
Reading List
Base
Nishigauchi, T. 2002. Scrambling and reconstruction at LF. Gengo Kenkyu 121: 49-106. (SOFT and HARD: Soft copy of the version 2000 is available, and I photocopied the published version)
Super-important
referential/nonreferential reading with X0 scope-bearing elements
Heycock, C. 1995. Asymmetries in reconstruction. Linguistic Inquiry 26: 547-570. (HARD: I photocopied this.)
Fox, D. 1999. Reconstruction, binding theory, and interpretation of chains. Linguistic Inquiry 30: 157-196. (SOFT)
Fox, D. 2000. Economy and semantic interpretation. MIT Press. (NOT YET-- I haven't got this, it's checked out from the library.)
Important
specific/non-specific reading of quantifiers
Nishigauchi, T. and A Uchibori. 1992. Japanese bare NPs and syntax-semantics correspondencesin quantification. Ms. UConn and OsakaUniversity. (NOT YET: I am not sure if I can get this paper...)
Hoji, H and Y. Ishii. 2004. What gets mapped to the tripartite structure of quantification in Japanese. B. Schmeiser, V. Chand, A. Kelleher and A. Rodriguez eds. WCCFL 23: 101-114. and references therein. (SOFT)
Scrambling
Ueyama, A. 1998. Two types of dependency. Doctoral dissertation. USC.(SOFT)
Ueyama, A. 2003. Two types of scrambling constructions in Japanese." In A. Barss & T. Langendoen eds., Anaphora: A Reference Guide. Cambridge: Blackwell. (SOFT and I have the book)
Saito, M. 1992. Long distance scrambling in Japanese, Journal of East Asian Linguistics 1:69–118.(SOFT)
Maybe?
If scope interaction between QPs becomes necessary
Hayashishita. J.-R. 2004. Syntactic and non-syntactic scope. Doctoral dissertation. USC. (SOFT)
These may say something about his view of reconstruction?
Nishigauchi, T. 1990. Quantification in the Theory of Grammar, Dordrecht: Kluwer. (NOT YET)
Nishigauchi, T. 1999. Quantification and wh-constructions, in Natsuko Tsujimura ed., The Handbookof Japanese Linguistics, Oxford: Blackwell.(HARD: I just got this book.)
If pronoun-binding by wh-phrase becomes important:
Ruys, E. G. 2000. Weak crossover as a scope phenomenon. Linguistic Inquiry 31: 513-539. (NOT YET, but it should be available on line)
1/7Outline Draft (Ling537)
Irie