General purpose
(see also Work Packages)
As the name suggests, we plan to take into account all the ingredients that are necessary to build a complex sentence with an ‘object clause’ after verbs like say, believe or hope and we compare them with relative and adverbial clauses. It is obvious that all these embedding structures have certain properties in common. For example, the embedded clauses are often introduced by complementizers. The type of this complementizer is uniform for relative and complement clauses in some languages (e.g. that in English), whereas complementizers may differ from a clause type to another in other languages, e.g. in German and Greek. On top of that, there is considerable variation in the dialects/spoken variants as well as in former stages. But in general, these complementizers can be traced back to demonstrative pronouns or they consist of invariable particles, e.g. wo or pu for relative clauses in some German and Greek dialects. Adverbial clauses on the other hand employ lexical elements with a rather transparent semantics (while, despite, although etc.) that sets them into a relation to the matrix clause – or better to say to the event described in the matrix. A typical relative clause has a ‘host’ in the matrix as well in that it modifies a constituent which plays its own role in the matrix clause.
With complement clauses, however, it is not evident whether there is such a ‘host’. Sentences such as ‘Mary believes/says/hopes that John is an angel; Maria glaubt/sagt/hofft, dass Hans ein Engel ist; Marie croit/dit/espère que Jean est/soit un ange’ do not contain any overt material in the matrix clause. A recent, prominent idea is that this type of verbs describe a mental state or act with propositional content. This propositional content itself is materialized by the embedded clause, but the mental state (or act) is represented in the matrix clause as an “attitudinal or locutionary object” (Moltmann 2003). This approach will be assessed with respect to the other, traditional idea, namely that complement clauses are arguments.
The project’s major hypothesis is that a uniform analysis is possible in that clausal embedding always involves a “tripartite structure”, including a (potentially silent) “anchor” in the matrix clause, the propositional content in the embedded clause and the complementizer as the mediating element between these two. The project will concentrate on Germanic, Greek and Romance and test the central hypothesis against non-Indo-European languages like Basque and Hungarian.
More specific goals
As said, there is a huge variation among languages when it comes to the lexical source(s) of the complementizer. Besides the demonstrative pronoun (English that, German dass), we find locative particles, equation particles, and question word based ones (French que). One of the goals of this project is thus to systematize this variation by developing a database which will be accessible to the research community.
We will closely investigate the different complementizer types, simplex and complex ones, study their morpho-syntactic structure and their meaning contributions in relation to their etymological source. Languages may also change over time as to which type of complementizer they use in which surroundings. Nevertheless, one would not expect this variation to be ‘random’ (cf. the work on diachrony by Weiß 2019, 2020, 2021, Axel-Tober 2012, 2017 on Germanic, by Brandner & Bräuning 2013 on Germanic dialects, by Manzini & Savoia 2011, 2018 on Romance, by Faure 2021 on Greek). Our work will thus also include diachronic investigations of complementizers, thereby also putting emphasis on the formal relation to the matrix anchors.
We will explore a new and innovative approach to argument structure with special reference to the anchors which we assume to be an integral part of every instance of clausal embedding. The leading hypothesis is that verbal predicates that are compatible with embedded clauses are verbs that are derived from those nouns that are taken in the work by Moltmann as the attitudinal object itself, e.g. belief, claim etc. The type of the derivation induces a syntactic position for a direct object, realized as light noun/expletive (“something/it that is believed”) and that via this anchor (and the respective complementizers), the content of the predicate will be connected with the attitudinal object.
Finally, we will explore in which way our theory can be extended to adverbial and interrogative clauses. Finite and non-finite complementation will be compared in both a semantic and a syntactic perspective. Apart from that, a specific study will address the major challenge to our theory: the well-known extraction asymmetries between complement and adjunct clauses, which have motivated radically different underlying structures in traditional generative syntax.
Methods
Our study will be mainly based on the elaboration of corpora. We aim at constituting a large database that is designed to enable a thorough crosslinguistic comparison. It will be made openly accessible once the project is over. We do not exclude other methods such as introspection, questionnaires and experimentation.