Acquisition of Spanish Mood Selection in Complement Clauses
Requena, P. E.1; Dracos, M.2, & Miller, K. L.3
1University of Montana, 2Baylor University, 3The Pennsylvania State University
Complement clauses are clauses which constitute arguments of a complement-taking verb.
(1) I believe this → ‘This’ is an argument of ‘believe’ (DO)
(2) I believe [that John was there too] → ‘that John was there too’ is an argument of
‘believe’ (DO)
In Spanish, the finite complement clauses require the selection of either the Indicative Mood
(IND) or the Subjunctive Mood (SUBJ).
SENTENCE COMPLETION TASK
PARTICIPANTS
A total of 66 monolingual Spanish-speaking children (4;2 to 10;3) participated in a Sentence Completion
Task. 13 adults were also tested.
Here we ask: How do children acquire adult-like mood selection in complement clauses
with particular complement-taking predicates?
Child Age Group
SEMANTIC CLASSES
Based on semantic class (Lastra & Butragueño, 2012), complement clauses with Certainty
predicates (e.g. saber ‘know’) select IND, whereas the SUBJ is selected to varying degrees in
complement clauses expressing:
(1) Volition (e.g. querer que ‘want that’),
(2) Comment (e.g. es importante que ‘it is important that’, estar contento de que ‘be happy
that’), and
(3) Uncertainty (e.g. dudar que ‘doubts that’, no creer que ‘not believe that’)
GRADIENCE IN MOOD SELECTION
Far from a direct one-to-one correspondence between mood (IND vs. SUBJ) and semantic
class (Terrell & Hooper, 1974), Kempchinsky (2009) proposes that “there is a continuum of
complementation types, in both syntactic and semantic terms.” Such gradience could be
captured by means of analyses of frequencies of use in spoken corpora or in elicitation
studies (e.g. Gudmestad 2010).
100%
95%
84%
80%
60%
53%
40%
20%
11%
Volition
Comment
Uncertainty
Certainty
0%
Rate of Subjunctive (Gudmestad 2010)
Mood selection varies depending on semantic/pragmatic and lexical factors even within a
certain class, as well as across individual speakers (Blake 1981, 1983, 1985; Montrul 2004).
DEVELOPMENT OF COMPLEMENT CLAUSES
Comprehension and production of complementation happens very early (De Villiers, Roeper,
& Vainikka, 1990; Thornton & Crain, 1994). Children’s earliest uses of what look like
complement clauses in spoken corpora seem to result from the use of fixed hedges (‘I think,’
‘You know,’ ‘That means’) thus making acquisition evolve in a piecemeal fashion (Thomson &
Mulac 1991, Kidd, Lieven & Tomasello 2006).
When do children reach adult-like performance in mood selection?
The answer to this question depends on the particular construction (lexical effect, Blake
1983) and early uses may also consist of fossilized expressions.
Volition: querer ‘want’ constitutes one of the earliest uses together with negative
imperatives and some adverbial uses (e.g. queré(s) [:quieres] que te haga[SUBJ] otro globo?
‘Do you want that I make another balloon for you?’, VIC 2;11) (González 1970). Blake (1983)
shows that adult like performance is reached between 7;0 and 9;0.
Comment & Uncertainty: Given that these complement-taking predicates refer to beliefs and
cognitive states, they emerge later than Volition (Desire). Mood selection with these
predicates has been suggested to reach adult levels around age 10;0 (Blake 1983), even
though in Blake’s study children were approximating adult-like performance, but did not
reach it at that age either.
n
Age Range
Mean Age
4-5 yrs
21
4;2 – 5;6
4;9
6-7 yrs
20
6;4 – 7;8
6;10
9-10 yrs
25
9;4 – 10;3
9;9
100
100
80
79
77
66
54
50
40
36
30
20
1
0
0
4-5 yrs
6-7 yrs
VOLITION
Comment Condition – Expected Response: SUBJ
El papá y el hijo están en una fiesta. Al hijo le encanta bailar, y eso al papá lo pone contento. ¡Mirá
su cara! Completá lo que digo: Siempre, el papá está contento de que el hijo baila-IND /baile-SUBJ.
Father and son are at a party. The son loves to dance, and this makes his father happy. Look at his
face! Complete what I say: Always, the father is happy that his son dances.
Uncertainty Condition – Expected Response: SUBJ
Antes de comenzar el partido, este chico se lastimó el pie y le duele mucho. ¿Va a jugar el partido?
El entrenador no cree. Completá lo que digo: El entrenador no cree que el chico… juega-IND
/juegue-SUBJ.
Before the start of the game, this boy injured his foot and it hurts a lot. Is he going to play in the
game? The coach doesn’t believe so. Complete what I say: The coach does not believe that the boy
is going to play the game.
Control Condition – Expected Response: IND
Cuando los nenes no hacen la tarea, los papás se preocupan, ¿no es cierto? Este papá no está
preocupado porque sabe que a su hijo le encanta hacer la tarea. Completá lo que digo: El papá sabe
que el nene siempre hace-IND la tarea.
When children don’t do their homework, parents get all worried. Isn’t it true? This dad is not worried
because he knows that his son loves doing his homework. Complete what I say: The dad knows that
the boy always does the homework.
96
85
70
60
100 100
10
PROCEDURE
Volition Condition – Expected Response: SUBJ
La mamá está preparando la comida, y la nena está mirando la televisión. La mamá necesita ayuda
para poner la mesa. Completá lo que digo: La mamá quiere que la hija pone-IND/ponga-SUBJ la
mesa.*
The mother is preparing food, and the girl is watching television. The mom needs help to set the
table. Complete what I say: The mom wants that her daughter set the table.
100
93
90
Percentage of Use of Subjunctive
SPANISH MOOD SELECTION IN COMPLEMENT CLAUSES
The acquisition literature is inconclusive and the norm of mood selection used to interpret it ‘confusing’
(Blake 1983). However, one important aspect of mood selection has been highlighted, namely the fact that
THERE IS GREAT VARIATION IN MOOD SELECTION BASED ON INDIVIDUAL LEXICAL ITEMS. Thus, a study of
this phenomenon should isolate individual lexical items, test them more than once in the experiment (cp.
Blake 1983), and make sure that a clear adult norm for mood selection is used to interpret child results.
Group Means for SUBJ use in each Condition
0
9-10 yrs
Age Group
COMMENT
0
UNCERTAINTY
Adults
CERTAINTY
Percentage of Response Types in the Uncertainty Condition
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Figure 1. Sample trial per conditions (*the underlined sections are the expected responses to be provided by the participant)
RESULTS
Overall use of the SUBJ by adults was different from that of the two younger child groups, but not from the
oldest child group.
Volition: Adults: categorical SUBJ / Children: adult-like (except 5/21 four-year-olds). ➔ This shows a clearer
adult norm (vs. Blake 1983) and much earlier acquisition with querer ‘want’ than the age at which Blake
(1983) reports that IND errors start to decrease (7;0).
Comment: Adults: categorical SUBJ / Children: adult-like by 9;0-10;0 ➔ This shows a clearer adult norm (vs.
Blake 1983) and earlier acquisition with estar X que ‘is happy that’ than the age at which Blake (1983)
reports that IND errors start to decrease (10;0-11;0) but at which children in that study still do not pattern
like adults.
Certainty: Adults: categorical IND/ Children: adult-like (except for one trial of one four-year-old). ➔ This
shows early acquisition and a clearer pattern even for adults than Blake (1983).
Uncertainty: Adults: VARIABLE (77% SUBJ) / Children: VARIABLE (seem to display a ‘U-shaped’ pattern in
subjunctive choice, but qualitative analysis shows that they are not adult like until age 9;0-10;0). ➔ Similar
to Blake (1983).
4-5 yrs (n=21)
SUBJ
6-7 yrs (n=20)
VA A + Inf
9-10 yrs (n=25)
Age Group
INDIC (pres)
INDIC (pret)
Adults (n=13)
Negation
DISCUSSION
How do children acquire adult-like mood selection in complement clauses with particular
complement-taking predicates?
- Children acquire Spanish mood selection in a piecemeal fashion. Our results indicate that children
may display adult-like performance with particular constructions from the earliest stages. Thus,
the key is to examine particular complement clause matrices.
When do children reach adult-like performance in mood selection?
- This depends on the lexical verb and on the semantic complexity.
- Methodological Implications: The study of Mood selection in complement clauses needs to
concentrate on a limited number of lexical items at a time and test them more than once. This
provides clearer adult norms in the light of which child data can be interpreted. The study of adultlike performance needs to take into account alternative (non-target) responses.
THE 41st ANNUAL BOSTON UNIVERSITY CONFERENCE ON LANGUAGE DEVELOPMENT