Unbounded Dependencies in Palauan Revisited

Veronica Gerassimova 

Stanford University

Background: Palauan is widely cited as a language with unusual unbounded dependencies (UDs). Since Georgopoulos (1985, 1991) (henceforth G.) and Chung and Georgopoulos (1988), it has been assumed that Palauan is one of the few languages which exhibit special marking along the UD chain, or wh-agreement (among other unusual characteristics). I present new data showing that UDs in Palauan do not possess these striking properties. Rather, Palauan UDs belong to a different, well-attested syntactic type, namely languages marking the presence of UDs only on the highest clause of the chain, as in Hausa, Moore, (Haïk 1990), Irish RPs (McCloskey cf. 1997). I discuss the changes in the typological space of UDs across in Western Austronesian, and across language in general resulting from these findings given existing syntactic analyses.

Data and analysis: The key existing assumptions about Palauan UDs can be summarized as follows: (a) verbs in Palauan show a three-way morphological distinction between non-UD sentences (1), UDs involving the subject (2), and UDs involving the object (4); (b) descriptively, subject UDs exhibit gaps (2), (3), while object UDs employ RPs (4), (5); (c) both types of UDs show (the same) morphological marking along the UD chain, by way of a special verb form (3), (5).

However, new data and call for fundamental revisions of all three assumptions: (a) the morphological distinction in UDs is not three- but two-way, since only object UDs exhibit special verb marking. Contrary to the conclusion in Georgopoulos (1985, 1991), the subject head-marking in default clauses is not an inflectional affix, but a clitic therefore verbs in default sentences and subject UDs are formally identical (1), (2).  (b) subject UDs involve RPs, rather than gaps, just like object UDs, since the subject marking clitic functions as a pronominal in UDs as in the alternative (6); (c) it is not the case that all clauses along the UD chain show special marking. Without exception, native speakers reject sentences like (5) as ungrammatical, and change them into (7), with (8) as an alternative.

The pattern emerging from these data is that in Palauan UDs special marking appears only in the highest clause of the chain (7), (8) not on all intermediate clauses as previously reported.

Discussion: First, cross-linguistic observations indicate that while gaps are subject to local syntactic constraints, e.g. islands, RPs are generally immune to them and correlate with the hypothesis that RPs involve semantic binding which is not constrained by the syntax (cf. Sells 1984). Thus the Palauan data underscore the generalization that RPs do not show island effects.

Second, there is one less example of a language showing special marking along the chain. The present study illustrates the benefits of careful linguistic analysis of phenomena with typological significance. Thus, examining closely the UD constructions in the closely related Chamorro might reveal further insights about the form and function of UD marking in Austronesian, and in natural language in general.

Third, languages with UD marking in the highest clause emerge as a typologically important group which deserves a deeper analysis of the form and function of this syntactic pattern.

(1)    Ng=omes     er a   rengalek   a     sensei. [1]

3sg=see        P D  children   D    teacher

The teacher is looking at the children.

Subject UDs (G. 1985)

(2)    A   sensei   [a omes      er a        rengalek] 

D   teacher   D see         P Det  children

The teacher is looking at the children.

(3)    ng=te`a [a l-ilsa       a Miriam   [el       milnguiu   a   buk   er ngii ]]

3s=who  D B-3-see     D M.         C-nf     read             D book          P her

Who did Miriam see reading her book?

Object UDs (G. 1985)

(4)    A     rengalek_i  [a l-omes   er tir_i   a   sensei]

D     children D  B-3-see   P 3pl   D  teacher

The teacher is looking at the children.

(5)    ng+ngera_i  [a omulemadsu    [el   lulengiil   er ngak [el   bo      kuruul  er ngii_i]]]  

3s+what      D B-2-think      C-nf     B-3-wait   P me            C-nf B-Fut     B-1s-do      P  it

What do you think that they were waiting for me to do?

New data

(6)    ng=te`a_i  [a l-ilsa       a Miriam  [el    kmo ng_i=milnguiu    a buk      er ngii ]]

3sg=who     D B-3-see     D M.         C-fin        3sg=read               D book      P her

Who did Miriam see reading her book?

(7)    ng+ngera_i  [a omulemadsu    [el kmo   lulengiil   er ngak [el   bo      kuruul   er ngii_i]]]  

3sg+what     D B-2-think      C-fin          B-3-wait P me       C-nf     B-Fut     B-1s-do      P  it

What do you think that they were waiting for me to do?

(8)    ng+ngera_i  [a omulemadsu    [el kmo   lulengiil   er ngak [el   mo     moruul er ngii_i]]]  

3s+what      D B-2-think      C-fin          B-3-wait P me       C-nf     Fut        1s-do      P  it

What do you think that they were waiting for me to do?              

Bibliography

Cherny, Lynn. (1993). The role of agreement and modality in Palauan. In, Proceedings from the 11th West Coast Conference on Formal Linguistics (WCCFL) (Jonathan Mead, ed.)

Chung, Sandra. 1982. Unbounded dependencies in Chamorro Grammar. Linguistic Inquiry, 13, pp. 39-78.

Chung, Sandra. 1998. The design of agreement: Evidence from Chamorro:.University of Chicago Press.

Chung, Sandra, and Georgopoulos, Carol. 1988. Agreement with gaps in Chamorro and Palauan. In, Agreement in Natural Language: Approaches, Theories, and Descriptions. M. Barlow and C. Ferguson (eds.). CSLI, Stanford.

Georgopoulos, Carol. 1985. Variables in Palauan syntax. NLLT 3, pp. 59-94.

Georgopoulos, C. (1991). Syntactic Variables: Resumptive Pronouns and A’-Binding in Palauan. Dordrecht: Kluwer Academic Publishers.

Goldsmith, John (1981). The structure of wh-questions in Igbo. Linguistic Analysis 7, pp. 367-393.

Haik, Isabele. 1990. Anaphoric, Pronominal and Referential INFL. NLLT 8, pp. 347-374.

Josephs, Lewis. 1975. Palauan Reference Grammar. The University Press of Hawaii.

Maling, Joan and Annie Zaenen. 1978. The non-universality of a surface filter. Linguistic Inquiry 9, pp. 475- 497.

McCloskey, James. (1997). Movement, resumption and complementizer choice. Ms. University of California, Santa Cruz. Paper presented to the Second Celtic Linguistics Conference, University College Dublin, June 1997.

Sells, Peter. (1984). Syntax and Semantics of resumptive Pronouns. PhD dissertation, UMass.

Watanabe, Akira. (1996) Case Absorption and Wh-agreement. Dordrecht: Kluwer Academic Publishers