1
2
Alessia Serafini , Francesca Franzon ,
2
1
Davide Bertocci , Giorgio Arcara and
1,2
Chiara Zanini
1
Department of Neurosciences NPSRR, University of Padova
2
Department of Linguistic and Literary Studies DISLL, University of Padova
[email protected] — [email protected] —
[email protected]
Theoretical Background:


Syntax-centred theories (Pollock, 1989; Kayne, 1989; Chomsky, 2000)
Non-syntax-centered theories postulate an interaction with semantics for what concerns some instances of agreement (e.g. Corbett, 2006 and modern psycholinguistic models, i. a. Haskell & MacDonald, 2003; Badecker & Kuminiak, 2007).
Two kinds of Agreement?


Quantification expressions such as Italian un sacco/mucchio di + noun (‘a lot of ...’) show two possibilities concerning Number agreement: the verb can agree either with the quantifier (syntactic agreement) or with the noun that it modifies (semantic agreement).
Crucially, while the semantic agreement is always possible, the option of performing the syntactic agreement seems to be constrained by some semantic properties of the modified nouns, such as concreteness.

Aim: verifying the acceptability of both agreement configurations with respect to the concreteness of the modified noun.
Verb agrees with:
Concreteness of noun:
1a
quantifier
[+ concrete]
Un sacco di limoni è stato venduto al mercato.
‘A lot of lemons was sold at the marketplace’
1b
noun
[+ concrete]
Un sacco di limoni sono stati venduti al mercato.
‘A lot of lemons were sold to the marketplace’
2a
quantifier
[- concrete]
?Un sacco di concetti è stato chiarito a lezione.
‘A lot of concepts was explained at the class’
2b
noun
[-concrete]
Un sacco di concetti sono stati chiariti a lezione.
‘A lot of concepts were explained at the class’
Materials and Methods:
Participants: 40 native speakers of Italian, aged 19 –
25, with 13 to 16 years of education.
80 passive sentences, 20 per each condition (10 with
the quantifier un sacco di and 10 with un mucchio di)
140 fillers displaying un sacco/mucchio di as well as
other quantifiers were added.
All quantification expressions and all inflected verbs
were balanced with respect to length and frequency by
means of it-WaC corpus (Baroni et al., 2009 2009).
Examples
Task: the sentences were presented divided in four chunks: quantifier, plural noun, verb, adjunct. The chunks appeared on the screen one at a time. Once a participant had finished reading a chunk, he/she had to press a key: the chunk disappeared and the following one was displayed. After having read the last chunk, they
had to give a grammaticality judgment about the whole sentence on a Likert scale ranging from 1 to 4 (1= absolutely not acceptable; 4= completely acceptable).
Results:
A mixed models analysis revealed:

No significant difference in RTs in correspondence of the 3rd chunk in any of the four conditions.

With respect to the RTs of the whole sentence, the significant coefficients were: RTs of the whole sentence
[beta = -0.000072, t = -4.138]; condition 1b vs. condition 1a [beta = 1.20, t = 6.975]; condition 2a vs. condition 1a [beta = -0.39, t = -3.80]; condition 2b vs. condition 1a [beta = 1.78, t = 7.31].

With respect to time spent to score the grammaticality of sentences (STs), the significant coefficients were:
RTs of the whole sentence [beta = 44 0.00008, t = 4.651]; rating score 2 vs rating score 1 [beta = 0.49, t =
8.45]; rating score 3 vs rating score 1 [beta = 0.49, t = 8.45]; STs in condition 2b vs STs in condition 1a [beta
= 0.46, t = 7.51].


The higher the STs the higher the RTs of the whole sentence.
The shorter STs occurred in correspondence of polar judgments (1 or 4), i.e., when the sentences is either
judged as absolutely implausible or as absolutely correct.
Discussion
Agreement between the verb and the noun of the quantification expression is generally preferred (receiving higher rates and requiring shorter times to be rated
as well).
Agreement between the verb and sacco or mucchio is only accepted when these words receive a fully referential interpretation. Referential interpretation of
these words is driven by the semantic content of the following noun: when this is [–concrete] il sacco, ‘the bag’, can be only a figurative one – which accounts
for the low rate received by (2a).
The lack of significant differences between the reading times at the third chunk in the different conditions may reflect the fact that in both configurations there is
only one and the same mechanism of Agreement at work.
Agreement has only syntactic basis: syntax always drives the interpretability of the semantic features (such as Concreteness) of the probe rather than integrating some semantic information in some cases and not in others. However, further studies are needed in order to shed light on such phenomena.
References
Badecker, W. & Kuminiak, F. (2007) Morphology, agreement and working memory retrieval in sentence production: Evidence from gender and case in Slovak. Journal of Memory and Language 56: 65-85.
Baroni M., Bernardini S., Ferraresi, A. & Zanchetta E. (2009) The WaCky Wide Web:A Collection of Very Large Linguistically Processed Web-Crawled Corpora, Language Resources and Evaluation 43:3.
Corbett (2006) Agreement. Cambridge University Press, Cambridge.
Haskell, T. R. & MacDonald, M. C. (2003) Conflicting cues and competition in subject–verb agreement. Journal of Memory and Language 48: 760-778.
Pollock, J. Y. (1989) Verb movement, universal grammar, and the structure of IP. Linguistic Inquiry 20: 365-424.