[Postings from 2010, copied from my research website. 8/22/2014, HH]

Because of research of/for one of the USC students, who might be working on "rightward scrambling" in Bangla, I did some research about works on "A-ga V-ta, B-cm." At one point of the chain of reference-finding, I came across the following paper, cited in Takita, Kensuke. 2009. Argument Ellipsis in Japanese Right Dislocation. Ms. to appear in Japanese/Korean
Linguistics 18, ed. Marcel Den Dikken and William McClure, 380–391. Stanford, Calif.: CSLI.

Saito, M. 2007a. Notes on East Asian Argument Ellipsis. Language Research 43:
203-227.

http://www.nanzan-u.ac.jp/LINGUISTICS/staff/saito_mamoru/pdf/saito-2007-LangResearch.pdf

The paper contain remarks such as:

"Hoji 1998 contains illuminating discussion, but as far as I can see, his arguments against
the ellipsis analysis are not at all conclusive."

"Aside from this problem, Hoji’s (1998) approach raises an issue on how far we can
stretch the possible interpretation of pro. It is already non-standard to assume that pro can be
construed as indefinite. If pro is simply a pronoun without phonetic content, we would expect
it to be definite in interpretation."

"The discussion above, I believe, has shown that the ellipsis analysis covers a wide range
of data and has a firm empirical basis."

"Having confirmed the plausibility of the argument ellipsis hypothesis, I would like to turn
to the analysis"

I had to check and see if Hoji 1998 indeed proposed the "pro analysis" of the null object construction (NOC) in Japanese. Much as in the case of my Otagai paper in regard to its main claim, the main point of Hoji 1998 is to show that the NOC in Japanese cannot be analyzed on a par with VP Ellipsis in English (like otagai cannot be analyzed unambiguously as a local anaphor). Whatever the paper suggests as an alternative account of the NOC is not much of interest, as far as I am concerned. But Hoji 1998 now seems to be understood as claiming that the null object is pro; but it is already made clear in Hoji 1995 that I do not believe that empty nominals in Japanese are [+pronominal, -anaphor].

At any rate, with "the plausibility of the argument ellipsis hypothesis" re. the NOC in Japanese having been "confirmed" -- as far as I can tell, not on any *Schema-based predictions that get supported by a confirmed schematic asymmetry, but on some observations that are only (loosely) compatible with the hypothesis -- and with a number of works getting out that accept the main thesis and the research orientation in Saito 2007, it is no wonder that we see works that adopt "the argument ellipsis hypothesis" re. the NOC in Japanese. If someone points out, as a reviewer of a paper or as an author of a paper, that Hoji 1998 has already shown that it cannot be maintained, the authors of such works (or the reviewer) might as well say something like "Gee, the reviewer (or the author) does not seem to be keeping up with recent literature and seems to be unaware of all these recent works."