1 Adjective Classes in Typological Perspective 35
(2) Japanese is a dependent-marking language; as discussed in §8, there are two
adjective classes, one verb-like and one noun-like. This suggests a rather specula-
tive historical scenario:
- STAGE ONE. Japanese lacked dependent marking. It probably also lacked head
marking, showing syntactic function by the ordering of phrasal constituents
within a clause. There was a single class of adjectives (the present inflected class),
similar to verbs in their grammatical behaviour. Japanese thus conformed with
the correlation in (33). - STAGE TWO. The language developed dependent marking. In association with
this, it developed a second class of adjectives (the present uninflected class),
which are 'non-verb-like'.
The following points can be adduced to support this scenario:
- The marking of the function of NPs in a clause is by syntactic particles, rather
than by case suffixes. This is a little surprising, since Japanese is a fairly synthet-
ic language with verbs taking a variety of suffixes. It is consistent with depend-
ent marking having been introduced rather recently. Indeed, Shibatani (1990:
333-57) states that although the topic-marking particle wa is present in the ear-
liest records (from about the eighth century CE), the particles ga, marking sub-
ject, and o, marking object, developed fairly recently from other grammatical el-
ements (the object marker evolving before the subject marker). - The verb-like inflected adjective class appears to be archaic, being restricted to
native lexemes; although large, it does not accept loans. It includes all AGE, COL-
OUR, and SPEED items, and most from the DIMENSION and PHYSICAL PROPERTY
types (the types that are typically associated with an adjective class). - The noun-like uninflected class is now bigger than the inflected class and is
growing; it accepts all kinds of loans. This class includes some DIMENSION and
PHYSICAL PROPERTY terms, and most of the HUMAN PROPENSITY adjectives.
This scenario is speculative but not implausible. It suggests that, as with the
Australian languages Emmi and Nunggubuyu, once a language shifts its profile
with respect to head and dependent marking, then there will be a tendency to
reorientate the grammatical properties of adjectives in accordance with the cor-
relation in (33). In Nunggubuyu, the adjective class has had its grammatical pos-
sibilities extended so that it may now accept subject pronominal prefixes, like an
intransitive verb (it does not yet directly accept tense and other verbal suffixes; this
would be the next step). In Japanese, a new adjective class has been established,
which is steadily increasing in size.
Korean is a clear exception to the correlation in (33), being dependent-marking
and having just one adjective class, which is 'verb-like'. The speculative scenario just
suggested for Japanese could be extended to Korean—supposing that the language
originally lacked both head and dependent marking, and had a class of'verb-like'