Consider examples of the type shown in (3c) in which a single character S-J word forms a verb when accompanied by the native suffix -su(ru). At first glance, the
verb愛するaisuru‘love’appears to simply be the one-character S-J word愛ai‘love’ withsuru‘do’attached. The single character愛aifunctions as an independent noun andsurucan generally be added to two-character S-J VNs to form verbs, as in運転 する unten-suru, soaisuru appears to similarly be simply a junction of 愛 ai and suru. However, Kageyama (1980) showed that there are morphological and syntactic differences between愛する aisuru‘love’and恋する koisuru‘love’, which is formed on the native morpheme 恋 koi ‘love’. First, syntactically, aisuru is a transitive verb that takes an accusative object whilekoisuruselects dative case for its object. Morphologically,koiand suruinkoisurucan be separated as in (4a), but separatingaiandsururesults in ungrammaticality, as shown in (4b).
(4) a. Takesi wa Naomi ni koi o si-te ir-u.
Takeshi TOP Naomi DAT love ACC do-GER be-PRS
‘Takeshi loves Naomi.’
b. *Takesi wa Naomi ni ai o si-te iru.
Takeshi TOP Naomi DAT love ACC do-GER be-PRS
Koisurucan thus be analyzed in relation to the phrasekoi o suru, butaisurucannot be made into *ai o suruand must instead be analyzed as a single word.
In fact, comparison of the conjugational forms ofaisuruandkoisurushows this difference clearly; while the negative ofkoisuruis the same as the independent verb suru, that ofai-suruis different.
(5) a. suru‘do’ – si-nai[do-NEG], *sanai b. koisuru – koisi-nai[love-NEG], *koisanai c. aisuru – ais-anai[love-NEG], *aisinai
Besides koi-suru, other verbs that take the pattern shown in (5b) include 損する son-suru ‘suffer a loss’ and 楽する raku-suru ‘live comfortably’. In contrast, the pattern shown in (5c) withaisuruis the same as that shown in the negative forms of native verbs likekakus-u‘conceal’, kakus-anai‘not conceal’(cf. *kakusinai) and watas-u‘hand over’, watas-anai‘not hand over’(cf. *watasinai). Kageyama (1980) concludes from these facts that the stem ofaisuruis notai-butais-.
Miyake (2010) lists a large number of examples of verbs taking the form“one- character S-J word +suru”in their citation forms, some of which are given in (6).
(The English glosses give the meanings of the one-character S-J words when used as verbs.)
Sino-Japanese words 97
(6) a. 愛ai‘love’,介kai‘interfere’,期ki‘expect’,害gai‘harm’,帰ki‘attribute’, 供 kyoo‘serve’,熟zyuku‘ripen’,称 syoo‘call’,属 zoku‘belong’,題 dai
‘entitle’,利 ri‘benefit’,託 taku‘entrust’,要 yoo‘need’,有 yuu‘have’, 適 teki‘fit’,類 rui‘be similar’
b. 屈 kuQ‘give in’,察 saQ‘conjecture’,接 seQ‘touch’,発 haQ‘issue’, 熱 neQ‘heat’,達 taQ‘reach’,反 haN‘go against’,面meN‘face’
Based on their conjugations and accent patterns, Miyake (2010) analyzes all of these uniformly as single character S-J words with the verbalizing suffix-sattached.
However, examining the morphology of the negative forms closely shows that (6) contains both forms with negatives in -sanaiand forms with -sinai. That is, while all the words in (6a) have negative forms in -sanai, all those in (6b) have -sinai. The single character S-J words in (6a) end in vowels whereas those of (6b) end in con- sonants (Q represents the moraic consonant and N the moraic nasal). The general- ization can be made that the words in both (6a) and (6b) have in common that they are formed of a single character S-J morpheme with a suffix -sattached and their negative forms appear as different allomorphs depending on whether the S-J mor- pheme ends in a vowel or a consonant.
In this section we examined the lexical categories of one-character S-J words and identified a verbalizing suffix -s. This -s can turn a single character S-J word into a verb, but it cannot attach to S-J words of two or more characters. As will be shown in the next section, when S-J words of two or more characters are used as verbs, the form -suruis used without exception.
3 Characteristics of two-character Sino-Japanese words
This section outlines the characteristics of two-character S-J words composed of two morphemes. Words like挨拶aisatufor which the meanings of the separate elements 挨 aiand拶 satucannot be specified are excluded. Also not considered are words formed by reduplication like粛々(と)syukusyuku(to) ‘solemnly’,満々(と)manman (to) ‘brimming with’, or 段々(と) dandan(to) ‘gradually’. Authors like Nomura (1988) working in the framework of traditional Japanese grammar have focussed on the productivity and internal structure of two-character S-J words within the tradi- tional kokugogaku approach (Section 3.1). More recent works, such as Kageyama (1980, 1993) and Kobayashi (2004) are based on theoretical morphology and have shifted focus to syntactic aspects such as how two-character S-J words function in a sentence and their ability to take arguments (Sections 3.2 and 3.3).