The syntactic representations of predicates and their arguments, in different grammatical voices
Since the mid-1990s, serious questions arose about the syntactic position of external arguments, and their relationship with the lexical predicate. The proliferation of (neo-)Davidsonian analyses of event semantics and the increasing empirical evidence that syntax more transparently feeds semantic interpretation led to the conclusion that the external argument of a predicate is introduced by a separate functional head (e.g., little v, Voice). This research project aims to answer questions about whether Voice and little v are different, how many functional heads there are in the argument structure domain, whether passives are “built on top of” active structures, how middles figure into generative analyses, and whether all arguments are severed from the lexical predicate like external arguments are. In addition, it explores how a passive/middle-like Voice analysis can be extended to tough-constructions and clausal complements of too/enough to solve some of the locality issues raised in their generative analyses.
This paper explores a phenomenon of English in which out- combines with a predicate to form a complex predicate (e.g., outsing, outdo, outrun, outsmart, ...), here called "out-PRED". A thorough investigation uncovers several new generalizations, leading to analyses (i) that out-PRED formation is a productive syntactic process, which build upon the structure for PRED, and (ii) that out- is the core of the out-PRED clause’s extended verbal projection. These findings are derived through a structure in which out- merges with PRED before any argument(s) can merge. This will be further supported by exploring domains in which out-PRED is unavailable; though these domains appear unrelated at the surface, they share deep derivational properties that are incompatible with this sort of out-PRED derivation. The findings of this work have implications for the representation of argument structure more generally, supporting analyses where all arguments of a verb are syntactically severed from it.
Ahn, Byron. 2018, September. How to ‘Suppress’ an Internal Argument. Talk presented at 2018 Annual Meeting of the Linguistics Association of Great Britain. Sheffield, UK.
This talk explores the productive ’out-PRED’ phenomenon in English (e.g. out-sing, out-do, out-run, out-smart), in which a prefix ’out-’ can affix to (a sub-class of) predicates (’PRED’s), interacting with PRED’s argument structure in a surprising way. I draw the following novel generalizations about this phenomenon:
(1) out-PRED inherits all of PRED’s morphophonological irregularities
(2) none of PRED’s internal arguments can surface in out-PRED
(3) if the interpretation of PRED depends on its internal argument, out-PRED is impossible
(4) out-PRED can always be passivized, even when PRED cannot be
Together, these indicate that out- prefixation creates an argument structure that is syntactically distinct from that of PRED. I pursue an analysis in which out- merges with PRED before any of its internal argument(s) can merge, and the newly formed out-PRED projects its own argument structure with its own Voice-related properties such as passivizability. A core component of this analysis is that internal arguments of PRED-type predicates must be severed from the lexical predicate (i.e., introduced in a syntactic position outside of √P/VP). This adds to a growing literature that converges on the result that syntax transparently encodes Neo-Davidsonian semantic argument structure; i.e., all arguments in a predicate are introduced by individual semantic functions, which each in turn map onto individual functional projections in the syntax.
@unpublished{Ahn-LAGB18-out,author={Ahn, Byron},howpublished={Talk presented at 2018 Annual Meeting of the Linguistics Association of Great Britain. Sheffield, UK},title={How to ‘Suppress’ an Internal Argument},year={2018}}
Ahn, Byron. 2015, March. Out-Sourcing Internal Arguments. Talk presented at WCCFL 33, Simon Fraser University.
This talk provides new evidence that internal arguments are severed from the lexical predicate. Data come from the productive English verbal-prefix, out-. Crucially, verbs like out-drink, out-think, out-bounce, out-weigh, etc. (henceforth out-PRED) never appear with the internal arguments of PRED without out- (i.e., drink, think, bounce, weigh):
(1) a. My book club can drink (a lot of beer).
b. My book club can out-drink (*a lot of beer) your book club (*a lot of beer).
c. My book club can out-drink *(your book club).
(2) a. James weighs *(180lbs).
b. James out-weighs (*180lbs) Josh (*180lbs).
c. James out-weighs *(Josh).
Exploring the grammatical properties of out-PRED reveals at least four generalizations:
PRED is syntactically active
out-PRED has an entirely distinct extended verbal projection,
compared to PRED
out-PRED selects a PRED complement, which contains none of PRED’s
internal arguments
If a PRED and its internal argument must be interpreted together,
out-PRED is impossible
These generalizations lead to the conclusion that syntax more closely mirrors a neo-Davidsonian semantics: PRED and its internal arguments are each introduced by unique semantic functions, which correspond with unique syntactic positions, outside of the lexical predicate.
@unpublished{Ahn-WCCFL33,author={Ahn, Byron},howpublished={Talk presented at WCCFL 33, Simon Fraser University},title={Out-Sourcing Internal Arguments},year={2015}}
Ahn, Byron & Craig Sailor. 2014, January. Obligatory Object Gaps in Infintival Clauses. Talk presented at the 2014 annual meeting of the LSA, Minneapolis, MN.
In English non-finite clauses, every obligatory empty category (e.c.) occurs in subject position (e.g. raising-to-subject/object, subject/object control, etc.), with two apparent exceptions: tough constructions (TCs), (1), and object-gap ’too’ (OGTs) constructions, (2):
(1) They are easy for Mary to intimidate e.c..
(2) John was too ugly for Mary to date e.c..
We argue that these are not exceptional: they too involve a subject e.c., because TC and OGT predicates obligatorily select for Middle voice clausal complements. The thematic object e.c. moves to subject-position of the infinitival:
(3) They are easy/too smart [TP e.c. to [VoiceP middle [vP intimidate e.c. ]]]
@unpublished{Ahn-Sailor-lsa2014-object-gaps,author={Ahn, Byron and Sailor, Craig},howpublished={Talk presented at the 2014 annual meeting of the LSA, Minneapolis, MN},title={Obligatory Object Gaps in Infintival Clauses},year={2014}}