Word order type and alignment type*

In her recent stimulating book NICHOLS (1992) suggests that there are no correlations between word order type and morphological alignment, where by alignment is meant the patterns of argument marking in mono-transitive and intransitive clauses, e.g. accusative, ergative, active, tripartite, hierarchical and neutral. Her findings pertain to the relationship between word order type and the dominant alignment type displayed by a language, dominant alignment being an abstraction over the alignments found with nouns, independent pronouns and verbal agreement. Since consistency in the alignment of the three categories is the exception rather than the norm (see p.8), the question arises of whether the lack of a correlation between word order and alignment established on the basis of dominant alignment also holds for the individual alignments of nouns, pronouns and agreement. The relationships between word order type and alignment that have been proposed in the literature include: a) a correlation between non-neutral alignment and SOV order; GREENBERG's (1963:96) universal 41: "if in a language the verb follows both the nominal subject and nominal object as the dominant order, the language almost always has a case system". b) a tendency for neutral alignment of nouns and agreement in SVO order (e.g. LEHMANN 1978; MALLINSON & BLAKE 1981:179) 1 ; c) a preference for non-neutral alignment of nominals and/or agreement in free word order languages (e.g. MALLINSON & BLAKE 1981; STEELE 1978); d) an association between ergative alignment and non-SVO order (e.g. e) an association between ergative alignment and object-before-subject order (e.g. SASSE 1978, PRIMUS 1991). While these putative relationships may be indeed not cross-linguistically valid, as NICHOLS implies, it seems worth subjecting her claim concerning the absence of a correlation between word order type and alignment type to closer scrutiny. The current paper presents the results of my own investigation of the issue conducted on the basis of a sample of 237 languages, the composition of which is shown in the appendix. 2 The discussion is organized as follows. §1 outlines the word order and alignment typologies used in the investigation and presents the distribution of word order and alignment types among the languages in the sample. In §2 the relationship between word order type and the neutral vs non-neutral alignments of nouns, pronouns and agreement is considered. §3 examines the non-neutral alignments manifested by the three categories relative to word order type. And in §4 my results concerning the relation between word …