Against Number Harmony for Non-Paradigmatic SE
Keywords: syntax, agreement, number, romance, spanish
TL;DR: This abstract discusses the agreement patterns of "se" sentences in Spanish and argues that they cannot be reduced to a post-syntactic process of harmony
Abstract: 1. PROPOSAL: This paper discusses the agreement behavior of “non-paradigmatic se” (NP-SE) sentences. Ormazabal & Romero (2024) (O&R24) address this matter by postulating that agreement in NP-SE may operate in the phonology, through a process of “number harmony”. We provide evidence to support the idea that NP-SE agreement is captured by AGREE (Chomsky 2000 and ff.), its variation following from a “accessibility scale”.
2. AGREEMENT IN NP-SE: Much of O&R24’s discussion revolves around the pair in (1)-(2), from Spanish. As they rightly point out, number agreement varies, in part due to anymacy of the NP in the VP internal position.
(1) Se {*censuró / censuraron} los documentos (2) Se {censuró / *censuraron} a los oponentes
se censor-3sg / 3pl the documents se censor-3s g / 3pl ACC the oponents
‘The documents were censored’ ‘The oponents were censored’
O&R24 point out that there is a more general asymmetry, depending on whether the internal argument is preverbal (and null) or postverbal: the former shows a much more stable behavior (number agreement being obligatory), whereas the latter exhibits an erratic nature, which they argue falls within a “post-PF procedure that we call Number Harmony”. In (3), the NP argument is preverbal, so O&R24 asume it is dislocated, “se” occupies [Spec,TP], and agreement manifests as standard verbal morphology (in (3b)) or as an accusative clitic (in (3a)).
(3) a. (Los documentos) se los censuró b. (Los documentos) se censuraron
the documents se cl-them censored-3sg the documents se censored-3.pl
‘The documents were censored’ ‘The documents were censored’
3. NUMBER HARMONY: O&R24 point out that “it is surprisingly common to find examples where agreement is not triggered by arguments, but in fact by temporal DP-modifiers,” as in (4):
(4) a. Se bailan los lunes b. Se abren los domingos c. Se trabajan los festivos
se dance-3.pl the Mondays se open-3pl the Sundays se work-3pl the holidays
‘We/People dance on Monday’ ‘We open on Sunday’ ‘We work on holiday’
O&R24 conclude that “subject agreement behavior is completely unexpected. Numbers are big enough to dismiss them as performance errors.” They further note that, unlike the data in (4), temporal nominal adjuncts fail to trigger agreement in other contexts, including impersonal verbs and unaccusative sentences (see (5)).
(5) a. Llueve / *Llueven todas las tardes / los domingos b. Cayeron / *Cayó almohadillas
rain-3sg / rain-3pl all the afternoons the Sundays fell-3pl / fell-3sg small pillows
‘It rains every afternoons / on Sunday’ ‘Small pillows fell’
The conclusion of O&R24 is as follows: “agreement facts in NP-SE do not work as predicted by theories based on syntactic agreement […] For postverbal NPs, the verb may show up marked with a default singular number or it may harmonize postsyntactically with a plural NP.” Such harmonization is defined as follows:
(6) CONDITIONS ON NUMBER HARMONY (NH)
a. Syntax plays a minimal role: it simply provides a structure where the set of φ-features in T includes a [person] value supplied by “se” but no [number], and where there is an NP nearby in postverbal position.
b. Agreement follows two different paths in that context: either it takes a default value (7a) or it adopts the value of the closest nominal (7b). When the nominal is singular, the verb never shows up in plural (7c).
c.
(7) a. Se censuró los documentos b. Se censuraron los documentos c. *Se censuraron el documento
se censor-3.sg the document se censor-3pl. the documents se censor.3pl the document
‘Documents were censored’ ‘Documents were censored’ ‘The document were censored’
Crucial here is th condition (6a), which takes HM to be computed in linear proximity terms. Although there are grammatical processes in which adjacency may indeed play a role, the possibility that agreement can also resort to linear order metrics raises a series of conceptual and empirical questions that we address in the next section.
Submission Number: 51
Loading