Constraint on Merge: The Roots of the Lexical/Functional Divide

The Florida AI Research Society Conference, 2005

2002

The Minimalist Program in syntactic theory seeks to factor cross-linguistic variation out of the syntax, and entirely into other grammatical components, chiefly the lexicon and the interfaces with semantics and phonetics. At the same time, current Minimalist research accords a central role to the operation of" Merge," a generalized transformation combining two autonomous subtrees as daughters of a single node (Chomsky 1995, 1998).

2002

Stabler (2001:258) suggested that “‘asymmetric’ feature checking” is among the mechanisms that “remain to be carefully explored” within his “minimalist grammars” framework. Our paper will undertake exactly this kind of exploration. Concretely, we are going to extend the “minimalist grammars” framework of Stabler (1997) (see also Harkema 2001; Michaelis 2001) by adding the operations adjoin and scramble. These are “asymmetric” counterparts of merge and move, respectively (Section 2). Subsequently, we will formulate our views on the treatment of scrambling and adjunction in German within this extended formalism (Section 3). Finally, we are going to address various linguistically motivated formal constraints on the resulting grammar (Section 4). Formal definitions are given in the Appendix (Section 5).

Due to the programmatic nature of Minimalism, there are of course many directions that it can take. This paper demonstrates that one trend in Minimalism is steering toward Dependency Grammar (DG): bare phrase structure, label-less trees, and specifier-less syntax are all aspects/developments of/within Minimalism that are steering toward a dependency-based (as opposed to a constituency-based) understanding of syntax and grammar. With this trend in mind, a version of Merge is introduced that completes the development. This version generates dependency-based structures in a strictly left-to-right manner, i.e. in the manner that all natural language is spoken and processed. The particular innovation that makes this concept of Merge possible is the catena. The catena is a novel unit of syntactic analysis that is unique to dependency-based structures. A number of empirical considerations support catena-based left-toright Merge, e.g. garden path sentences, restrictions on center embeddings, aspects of coordinate structures, subject-auxiliary inversion, the distribution of pronouns, etc.

Studia Linguistica, 2005

This article discusses the syntax of coordinate structures, in particular the status of initial coordinators, multiple coordination, and the asymmetries between conjuncts with respect to c-command relations. The idea of coordinators as heads-hence the CoP-is endorsed, but not for initial coordinators; rather they figure in a separate 'distributive phrase', which is transparently correlated to the (often ambiguous) semantics of the construction. Furthermore, it is argued that the lack of c-command between conjuncts is an instance of a broader effect, namely the 'invisibility' of paratactic material in generaland of second conjuncts in particular. Therefore, the grammar must have means to attach a paratactic constituent to the rest of the structure in a way that will eventually block c-command relations from the context. As this is not standardly available, we are led to define an operation called b-Merge, which induces a special type of inclusion relation, 'behindance'. Thus, a modern revival of the 3D approach to coordination is brought about.

Natural language & linguistic theory, 1990

5.1 Introduction We take it for granted that child language morphosyntactic development is determined by an emerging internal computational system (what is often called the 'Language Faculty' (LF)). Given this, then by definition, if stages are borne out during which child speech presents immature structures, it becomes incumbent upon the developmental linguist, somehow, to attribute such intermediate stages to a pegged immature computational system. Therefore, as I see it, the task of any sound child syntactic theory is to restrict the computational work-space available for the developing child, in any one stage of development, in ways which fit the child's speech production. Specific to merge, we cite that it is not just one operation, but rather a family of operations—where the type of merge which gets employed is often dependent upon the nature and maturational complexity of the given operation. Merge therefore may follow a gradient typology in its own right, and when issues of maturation come up, an eye on the type of merge that gets employed (child language) becomes a central concern. We also argue that there is a more general developmental (maturational-based) sequence of 'Merge over Move'. This broad sequence also seems to map onto a +/-gradient productivity cline whereby Derivational morphology sides with Merge and Inflectional morphology sides with Move. So, we have a two-prong hypothesis at work: (i) Merge in its narrow scope (developmental ontogeny—as determined by the type of merge employed given the nature and complexity) and (ii) Merge over Move in its broad scope (developmental phylogeny—as based on broad selective typologies/parameters of a given language). 5.2 Movement Applications Movement has recently been defined within MP as a form of merge. But there is not just one type of merge. Rather, merge makes-up a family of distinct movement operations, with their defining aspects being delimited, for the most part, by two crucial factors: (i) Locality of movement (local intra-phrasal vs. distant inter-phrasal), and (ii) Nature of Scope (semantic vs. syntactic). When merge is said to employ the former kind (local/semantic scope), it is said to be external merge. When merge is said to employ the latter (distant/syntactic), it is said to be internal merge (= move). The following section sketches as an overview the two-prong distinction.

Studia Linguistica, 2008

Within early Minimalism, there are Economy Principles such as ÔLast ResortÕ, ÔLeast EffortÕ, and also ÔMerge as Late as PossibleÕ. The latter principle makes it possible to explain grammaticalization, an important phenomenon in language change. The general direction of change is one where lexical items are reanalyzed as grammatical (or functional) categories, e.g. verbs as auxiliaries and prepositions as complementizers. One explanation in terms of Late Merge is that if a lexical item is not immediately relevant for theta-structure, it can wait and merge later, rather than merge early with additional movement. In later Minimalism, move is reformulated as internal merge and not considered more uneconomical than regular merge, renamed as external merge. This paper will provide ample examples of the workings of Late Merge and examine how we can capture the grammaticalization effects through an Economy Principle, Feature Economy, that is at work when children acquire a language as well as when language changes.

Studies in Modern Grammar, 2016

This paper examines the validity of the two claims advocated in Chomsky (2014) and address some inconsistencies in these claims. One claim concerns the proposed unification of the Extended Projection Principle (EPP) and the Empty Category Principle (ECP) under the labeling theory Chomsky advances in the paper. This paper shows that the EPP and the ECP are not in fact given a unified account under labeling theory as proposed but rather they are still approached by the two independent principles, i.e. the EPP by labeling and the ECP by the Phase Impenetrability Condition. The other claim that this paper investigates is on the proposal that the operation Merge can apply freely in relation to other operations in the syntax. This paper shows that at least in certain cases, the timing of the application of Merge cannot be free but should instead be constrained, especially in relation to the operation of Labeling Algorithm.