The role of -e and -agi 329
In (168), ang marks the topic (subject) of the clause. In (168a), the actor, babae
‘woman’ is subject and the verb takes future actor topic morphology (in bold).
In (168b), the theme (goal), bigas ‘rice’ is subject and the verb take future goal
topic morphology. (168c,d) show the same for direction (locative) and benefac-
tive arguments. As illustrated amply in Schachter and Otanes 1972, in Tagalog,
the types of arguments that can be selected as subject and associated with dis-
tinct verbal (voice) morphology include actor, theme (goal), location/goal (di-
rection), beneficiary, instrument, subject matter, time, manner, and others. The
first six of these are the same roles that can be core syntactic arguments in Ma-
durese, and like Tagalog, each is associated with specific verbal morphology
(except when the theme is object and there is no overt affix).
There is, however, some question as to whether or not -e and -agi suffixes
are properly treated as voice morphemes. Most analyses of Indonesian-type
languages do not do so, and the Indonesian voice system is generally characte-
rized as a two-voice system (Wolff 1996, Ross 2002, 2005). There have been
alternative views though. Naylor (1978) explicitly draws the parallel between
Philippine languages and Indonesian. Using the term ‘focus’ (as is standard in
some of the literature), Naylor identifies Indonesian -i as ‘locative focus’ and
- kan as ‘instrument focus’, and posits a null affix for ‘goal focus’ (marking
themes as objects). Verhaar (1984) treats -i and -kan as part of the same system
as the active (meng-) and passive (di-), referring to -i as the ‘locative role mark-
er’ and describing -kan as marking one of five ‘semantic species’–’benefactive’,
‘instrument’, ‘dative-accusative’, ‘accusative-accusative’, and ‘noncausative’.
Additionally, he identifies men...i and men...kan as circumfixes ‘deriving men-
verbs’ (1984:6), underscoring an analysis in which actor voice and -i and -kan
are part of the same system. More recently, Musgrave, while stopping short of
analyzing Indonesian -i and -kan as part of the voice system, does provide four
arguments that ‘suggest that the Indonesian system is related’ to the Philippine
systems (2001:72).
The apparent derivational nature of -e and -agi suffixes (especially with
causatives) contributes to the applicative analysis. But the distinction between
derivation and inflection in Madurese is blurred, and it is far from crystal clear
that actor voice morphology is purely inflectional. As described in Chapter 3
and Chapter 4, many lexical items denoting nominal entities can be used as
verbs when affixed with the actor voice morpheme. Examples of this include:
(169) root actor voice
are' ‘scythe’ ngare' ‘cut with a scythe’
pacol ‘hoe’ macol ‘hoe’
tokol ‘hammer’ nokol ‘hammer’
kaca ‘glass, mirror’ ngaca ‘look in (use) a mirror’