Prefixation in English and Catalan - Departament de Filologia ...

presence of a prefix (see Scalise (1988b) for the same conclusion for Italian). To explain
this fact I can only suggest that –ment has a particular feature [F] which needs to be
satisfied and that the prefix has the relevant feature [F]. However, I am aware that this
suggestion is only descriptively adequate since it explains why –ment seems to be
potentiated by en-, but it does not say anything about the nature of the feature.
After discussing some points left unresolved from the previous sections, and
before ending the present one, now I would like to briefly present another proposal, i.e.
HK’s (1993, 1998, 2002), to see whether it can handle the data satisfactorily. HK adopt
a syntactic approach to the representation of lexical argument structure. Vs are derived
by conflation 30 of a N or A into an empty phonological V base, thus giving it
phonological content. The structural types of lexical argument structure relevant here
are those associated with the morphosyntactic category A and N, given that now I will
focus on how HK’s theory can derive deadjectival and denominal Vs in English and
Catalan. Recall that in V-to-V prefixation no systematic pattern was found, which
explains its omission in the following discussion. Although HK’s theory may seem to
cope with the data adequately at first sight, there are some questions which cannot be
answered within their syntactic approach.
As already said, conflation explains the formation of deadjectival Vs. The
phonological matrix of the A replaces that of the V, which can be empty like in clean
(66a), or partially empty as in enrich which has a prefix or thicken which has a suffix.
For the latter cases, HK assume that the host V is bipartite, consisting of an empty
phonological matrix together with an overt matrix corresponding to that of the prefix or
suffix (66b/c) (HK 1998: 85).
(66) a. V b. V c. V
/ \ / \ / \
V A V A V A
| | / \ / \
[Ø] clean pref [Ø] [Ø] suf
30 Note that the discussion that follows is based on HK (1993, 1998). The same results, though, would be
obtained by using HK’s more recent version. Let me just point out one remarkable difference between
their earlier and later accounts, namely their use of the term conflation. In the more recent version, it does
not refer to a movement operation. In HK’s terms, ‘it is merely the binding relation that holds between the
semantic features of a V (phonologically overt now) and features of the nominal head of its complement
(HK 2002: 103)
51