Specificity in Madurese applicatives
William D. Davies
University of Iowa
This paper examines the Madurese applicative suffix -aghi, as illustrated in (1b), and argues that
a notion of specificity plays an important role in applied structures heretofore largely
unrecognized in the literature.
(1) a. Bhibbhi' ngerra' rotè kaangghuy embu'.
aunt
AV.slice bread for
mother
‘Auntie sliced bread for Mother.’
b. Bhibbhi' ngerra'-aghi embu' rotè.
aunt
AV.slice-AGHI mother bread
‘Auntie sliced bread for Mother.’
Other than the voice system, applicative suffixes are the most frequent topic in the literature
on the morphology/syntax of Indonesian-type languages. Of these, Indonesian -kan far outstrips
the others (Dardjowidjojo 1967; Verhaar 1984; Arka 1992; Kaswanti Purwo 1995, 1997;
Postman 2002; Cole & Son 2004; Son & Cole 2008; and others). Just as in Indonesian, the
analogues in the other languages appear to mark a variety of functions in sentences, including
causatives, benefactives, and others, such as the Madurese benefactive structure in (1b). Because
they appear to behave so much like -kan, the analogues in related languages have generally been
assumed to function in the same way and thus have escaped very close scrutiny in the literature,
although that has started to change recently beginning with Davies 2013.
While similarities among the applicative constructions are confirmed to a large degree,
careful examination reveals some differences between Madurese -aghi and what has been
reported for -kan. Differences arise in the interpretation of some structures. For example, both
(1a) and (1b) are ambiguous between an interpretation in which mother receives (and possesses)
the bread and an interpretation in which auntie was acting on mother’s behalf, the so-called
‘deputative’ benefactive (Van Valin & LaPolla 1997). Son & Cole (2008) report that the
Indonesian analogue of (1a) shows the same ambiguity, while (1b) is unambiguous, having only
the interpretation in which mother possesses the bread after it is sliced.
More important are the differences in interpretation between basic and applied structures.
(2) a. Wati a-carèta ka Marlèna bâb èbu'-na.
Wati AV-talk to Marlena about mother-DEF
‘Wati talked to Marlena about her mother.’
b. Wati nyarèta'-aghi èbu'-na
ka Marlèna.
Wati AV.talk-AGHI mother-DEF to Marlena
‘Wati talked about her mother to Marlena.’
In the basic structure (2a), Wati’s mother is likely the general topic of discussion or the speaker
may not know precisely what about Wati’s mother is being discussed. Conversely, (2b) carries
the implication that there is specific information that the speaker knows of. The same notion of
specificity in benefactive and other applied structures, as will be shown.
This specificity provides an account for the distribution of wh-in situ questions. While the
prepositional benefactive object can be questioned in situ (3a), the applied object cannot (3b).
(3) a. Ita ngèbâ
kotha’ kaangghuy sapa?
Ita AV.carry box for
who
‘Who did Ita carry the box for?’
b.*Ita ngèbâ’-âghi sapa kotha’?
As wh-words are quintessential indefinite, non-specific pronouns, the ungrammaticality of (3b)
follows naturally. As will be shown, other applied constructions show a similar fact pattern. The
applied object can be questioned in a cleft construction.
(4) Sapa sè è-ghibâ’-âghi Ita kotha’.
who REL OV-carry-AGHI Ita box
‘Who is it that Ita carried the box for?’
The cleft includes the presupposition that some set of elements that satisfies the state described
in the clause. So, while the element may be indefinite, it is not non-specific.
Turning to discourse evidence, we show that the distribution of -aghi converges with
Donohue’s (2001) observation that applicatives in Tukang Besi largely occur in environments
where the applied object has some level of topic or thematic prominence. Inasmuch as topics are
definite, they are specific. Thus the implication of specificity inherent in -aghi’s distribution
reinforces or perhaps helps account for Donohue’s observation.
The paper concludes by looking at the voice system and applicatives. Like other Indonesiantype languages, subjects in Madurese must be specific. We consider the fact that voice and
applicative morphology include specificity requirements and how it might bear on Davies’
(2005) conjecture that the applicatives might be a vestige of a richer voice system in Madurese.
Selected References
Arka, I Wayan. 1992. Morpholexical aspects of the -kan causative in Indonesian. Syndney:
University of Syndney MA thesis.
Dardjowidjojo, Soennjono. 1967. Indonesian syntax. Washington, DC: Georgetown University
dissertation.
Davies, William D. 2005. The richness of Madurese voice, The many faces of Austronesian voice
systems: Some new empirical studies, ed. by I W. Arka & M. Ross, 193-216. Pacific
Linguistics.
Davies, William D. 2013. Ruminations on a Madurese applicative affix. ISLOJ 4. Bung
Hatta University, Padang, Indonesia.
Donohue, Mark. 2001. Coding choices in argument structure: Austronesian applicatives in texts.
Studies in Language 25:217-254.
Kaswanti Purwo, Bambang. 1995. The two prototypes of ditransitive verbs: The Indonesian
evidence. Discourse, grammar and typology, ed. by W. Abraham, T. Givón & S.A.
Thompson. Amsterdam: John Benjamins.
Kaswanti Purwo, Bambang. 1997. The direct object in bi-transitive clauses in Indonesian.
Grammatical relations: A functionalist perspective, ed. by M. Shibatani. Amsterdam: John
Benjamins.
Postman, Whitney. 2002. Thematic role assignment in Indonesian: A case study of agrammatic
aphasia. Ithaca, NY: Cornell University dissertation.
Son, Minjeong & Peter Cole. 2008. An event-based account of -kan constructions in Standard
Indonesian. Language 84:120-160.
Van Valin, Robert & Randy LaPolla. 1997. Syntax. Cambridge University Press.
Verhaar, John M.W. 1984. Affixation in contemporary Indonesian. Towards a description of
contemporary Indonesian:Prelimiary studies, Part 1, ed. by Bambang Kaswanti Purwo.
Jakarta: Badan Penyelenggara Seri Nusa.