Date: Thu, 11 Dec 1997 18:29:25 +0900

Author: Yukiko Ueda (Kanda Univ. of International Studies, COE)
e-mail: ueda@kanda.kuis.ac.jp
Title: NEW: Questions to Elly van Gelderen: 
              Interpretable and Non-Interpretable features Cross-linguistically

1.      This is a conceptual question.  As you mentioned in your paper,
Interpretable features are not deleted or erased after they are checked
because THEY ARE RELEVANT TO THE INTERPRETATIVE COMPONENT.  Chomsky expects
that the interpretative component should be language universal.  Does your
analysis entail Non-universal interpretative component by assuming that
language variation depends on setting up interpretability of each features?

2.      Pro-drop phenomena and Interpretability of features:
        < 1 > 'None of the cases of reduced inflection have a null-subject
since pro needs to be licensed by STRONG FEATURES.' ( the paragraph below
(18)) What do STRONG FEATURES mean here?  Rich morphological agreement features
of I in the traditional sense?
        < 2 > Could you give me further explanation of the relation between
morphological richness (morphological realization), Interpretability of
nominal features, and the possibility of Null Subjects?
        < 3 > Is it possible to say that modern languages which permit a
Null Subject, such as Italian and Japanese, have Interpretable nominal
phi-features (details about subclasses of phi-features are omitted here)? 
Although Japanese permits Null Subject, it does not show any inflection
with respect to phi-features, especially gender and number as illustrated
(1) below.

(1)     a.      (Mary-ga)    ringo-o    tabe-ta.
                     -Nom  apple-Acc eat-Past
                'Mary ate an apple.'
        b.      (otokonokotati-ga)   ringo-o    tabe-ta.
                         boys-Nom apple-Acc eat-Past
                'Boys ate an apple.'

In some cases, 1st and 3rd person distinction can be shown by modal
attachment to the verb as given in (2).

(2)     a.      (watasi-ga)     keeki-o   tabe-ta-i.
                         I-Nom   cake-Acc eat-want-Pres
                'I want to eat a cake.
        b.      (John-ga)       keeki-o     tabe-ta-GAR-u.
               John-Nom    cake-Acc  eat-want-3P(or 2P)-Pres.
                'John seems to want to eat a cake.

Some questions arise concerning your analysis.  Does the Japanese fact in
(2) make you predict that Japanese nominals have an Interpretable
phi-feature with respect to person rather than gender and number and that a
Null Subject is permitted in Japanese because of the Interpretable person
phi-features?  However, we cannot, in fact, observe multiple phi-feature
checking in Japanese unlike French given in (3).
(3)     Les animaux sont tous parties   (= your (6))
The Japanese fact is rather parallel to English exsample in (4) with
respect to reusability of the feature.
(4)     John is intelligent
It seems to me that we cannot claim that the features are not Interpretable
only because they are not morphologically marked more than once.

Classical Japanese subjects are more frequently omitted than modern
Japanese.  I am not sure whether or not classical Japanese verbs are much
sensitive to verbal agreement morphologically, but a subject-verb
honorificational system seems to be more sophisticated than madern
Japanese.

Yukiko Ueda
Kanda Univ. of International Studies, COE
ueda@kanda.kuis.ac.jp