Lecture 1 (Feb. 1, 3:30-5:00): A menagerie of Merges
Usually simplification of a syntactic system leads to an increase in its expressive power, as the removal of restrictions widens the potential richness of representations. The proposal, within Minimalist Syntax, to reduce the structure building part of the grammar to the operation Merge, has lead to an explosion in the kinds of derivations and representations admitted by the theory. We have External Merge, Internal Merge, Self Merge, Under Merge, Parallel Merge (and its cousin Sideways Move), Late Merge (and its big sister, Wholesale Late Merger), Morphological Merge(r), as well as an increase in the kinds of structure admitted, with Roll Up, Remnant Roll Up, and Head Roll Up. And the field hasn't yet really got its mitts on Pair Merge and various imaginable versions of that.
This lecture points out the problem, and suggests two avenues to reduction of this richness. One is to impose a restriction on the design of the structure building system: syntax builds hierarchical structures that can be mapped to semantic interpretation with no loss of information. That is, the system does not have the capacity to change structure without semantic effect. The other avenue is to reduce the complexity of the structure building operation itself.
I then tackle head movement operations, showing that structures where head movement has applied do not feed more information to the semantics than those where it hasn't applied: head movement phenomena are, as has been often noted, semantically vacuous. I also show that the word building capacity of head movement has to be separated from the positioning of a head with respect to other elements in structure. But this means the structure changing operation of head movement is needed neither for semantics, nor for morphology.
Lecture 2 (Feb. 2, 3:30-5:00): Roll up, Roll Up, There's Nothing to See
The second lecture shows how to build a syntactic system that separates the building of structures from their labelling via the elimination of all heads from the syntax, except semantically contentful roots (an extension of the system presented in Adger 2013). No heads means no head movement, and I briefly show how word building and positioning can be captured in such a system via what Adger and Svenonius (2010) called second order interface features (essentially following the insights, if not the implementation, of Brody's 2000 Mirror Theory). But this lecture is mainly about Roll-Up derivations, which also do not feed semantics. The theory I present not only rules out head movement, it also makes Roll-Up derivations impossible, though it allows a kind of base generated alternative. I present some evidence from asymmetries between binding and linear order in prepositional phrases that suggest the such a theory is empirically, as well as theoretically, preferable to one employing Roll-Up.
Lecture 3 (Feb. 3, 3:30-5:00): Immemorious Merge
The final lecture turns to a simplification of the structure building operation Merge itself. In Borges' short story, Funes the Memorious, the protagonist has such a prodigious memory that every moment's existence of every object is distinct to him, and generalization becomes impossible. It is the limits on our memory that make abstraction available. Taking a lesson from this, I propose a theory of Merge with radically limited memory. I propose to bifurcate what is classically termed the Workspace, in Minimalist Syntax, into two memory domains, mimicking the separation of computer CPU memory into a cache-register structure. The cache contains the resources necessary for a derivation, and, once a computation has halted, it is the single element in the cache that is submitted to interpretation by systems of meaning and externalization. However Merge does not apply in the cache; rather elements are moved from the cache to a register, which is binary in structure, and Merge operates on the contents of this register, returning the contents to the cache when completed. This allows the recursive step of Merge to just be a general grouping operation, with binarity being a result of the architecture of the memory structure of the system, rather than stipulated in the operation itself. I also show how such a system rules out Parallel Merge and Late Merge, removing the theoretical motivation for such operations that usually figures in their justification.