LINGUIST List 2.333

Thursday, 4 July 1991

Disc: Circumfixing, IO Agreement

Editor for this issue: <>

Directory

1. Allan C. Wechsler, Circumfixing
2. Tim Stowell, IO Agreement

In Israeli Hebrew, the potential tense verb affixes are
 Sg. Pl.
 m. both f.
1 'a- ne-
2 te- te--i te--u
3 ye- te- ye--u
Were it not for one anomaly, it would be easy to assign glosses to the
prefixes and suffixes individually as:
'a- 1sg.
ne- 1pl.
te- 2
ye- 3
-i f.
-u pl.
But this scheme would predict *ye--i "3fsg." where we in fact see te-.
In a very conservative account, the anomaly would forbid us from
"factoring" the affix te--i, forcing us to treat it as a circumfix.
Most would prefer to regard the affix as a prefix and a suffix acting
together.
This example from a "gray area" prompts me to ask Dr. Fromkin for a
careful definition of a circumfix. How can you distinguish a circumfix
from a prefix/suffix pair?
Arabists may disconfirm me, but I believe there are less gray examples
in Classical Arabic.

RE A. Manaster Ramer's question about references for IO clitics in
Romance, the classical GB literature has several important dissertations
that deal with this subject. They include Osvaldo Jaeggli's MIT disserta-
tion (1980) later publixhed by Foris, Hagit Borer's MIT diss. (also by
Foris), and (a pre-GB source) Judith Strozer's UCLA dissertation from
1977. A large number of papers and dissertations by a variety of other
authors in the early 1980's addressed theoretical questions involving
these clitics, many of them concerned with R. Kayne's generalization that
clitic doubling is possible only with an NP argument that is governed by a
preposition (given a Case-theoretic treatment by Jaeggli). One of the main
dialects of interest was that of River Plate Spanish, which also allows
clitic doubling of a human direct object (also preceded by "a" in Spanish).
--Tim Stowell