General research program
In its research program for the years 2008-2013, the ZAS consolidates the projects of the first twelve years, grouping them into larger units to investigate the following two main questions:
Which linguistic operations are responsible for the complexity of human language − at all linguistic levels from sound patterning to discourse structure − in comparison to other communication systems? And how are these operations expressed at different structural levels?
We take as a starting point the assumption that linguistic complexity derives from two complementary mechanisms, the symmetric combination of constituents A+B and the asymmetric embedding (subordination) of one constituent into another A[B]. Both mechanisms can be illustrated with syntactic examples. In the sentence [The dog barked] and [Eva heard the dog] two clausal units are symmetrically coordinated, whereas in [Eva heard that the dog barked] one clause is embedded into the other. Both phenomena can be contrasted with the expression [Eva heard the dog bark] with no inner clause boundary. The projects investigate how the constituent boundaries resulting from coodination and embedding are marked.
Both mechanisms systematically emerge – though with different relative importance – on all linguistic levels, with a tendency for subordinating operations to be found at higher structural levels. For example, phonological structure exhibits not only linear combination in the sequencing of phonemes (e.g., dog: d+a+g), but also hierarchical structure in the formation of syllables (e.g., onset-rime, d [a g]). At the prosodic level, coordination conditions the organization of prosodic phrases, for instance in parsing the clause into two constituents in [[the DOG] [is LOUD]]. However, subordination also plays a role in certain cases, e.g. in parsing the expression [to hear [the DOG] bark]]. At the morphological level, coordination is rare, one example being the case of copulative (or dvandva) compounds such as bittersweet. The majority of cases at the morphological level involve subordinating structures, for example, endocentric compounds such as small talk, which denotes a kind of talk that is modified by small.
While subordinating structures dominate at the syntactic level, coordination can also be found: [[the dog and the cat] slept]. Subordinating syntactic structures reflect semantic relationships. Thus, in Eva heard [that the dog barked], the embedded sentence that the dog barked fills the argument place of the embedding sentence. Also important at the syntactic level is the phenomenon of movement, when the constituents of a sentence can occur in different positions, e.g. Eva loves the cat vs. The cat, Eva loves (but not the dog). Different kinds of movement have been identified, depending on whether the moved constituent is embedded or adjoined. Furthermore, it is well-known that individual languages or registers, like spoken vs. written, differ in their tendencies to allow coordinating paratactic and subordinating hypotactic structures. At the discourse level, we find that the majority of rhetorical relations between discourse parts are subordinating in nature (e.g. whether one element constitutes the reason for the following element), but coordinating structures (e.g. in narrative listings) can also be identified.
During the period 2008-2013, the program aims at investigating which phenomena of linguistic complexity can be attributed to these two structural operations: coordination and embedding. This thematic core is well anchored in the work at ZAS, thereby allowing close cooperation among projects. Structurally distinct levels of language (phonology, morphology, prosody, syntax, semantics) give rise to interfaces which must fulfill certain criteria. For instance, a semantic embedding is often in addition expressed via a syntactic embedding; and a morphological or syntactic coordination is usually paralleled at the prosodic level. Work on the interfaces is a further source of interdisciplinary research and fruitful exchanges among the projects.