Exploring the Relationship between Transfer and
Input in the Acquisition of the Spanish Passives
Joyce Bruhn de Garavito1 and Elena Valenzuela2
1The University of Western Ontario and 2McGill University/McMaster University
1. Introduction
The difference between adjectival and verbal passives has been explored since the 1970s (Wasow
1977; Levin and Rappaport 1986, among others). In Spanish, verbal passives take the copula ser and
adjectival passives take the copula estar.
Schmitt (1992), Lema (1992) and Luján (1981) have suggested that the difference between the
two copular verbs in Spanish, ser and estar, is aspectual. According to Schmitt, ser is unmarked for
aspect and derives its aspectual interpretation from the type of predicate. Estar, in contrast, represents
the result state of an accomplishment verb, which corresponds very closely to what an adjectival
passive is. It is not surprising, therefore, that the two types of passives are realized with the two
different copulas.
Only recently has the acquisition of the two copulas been studied systematically (VanPatten 1985;
1987; Geeslin 2001; 2002a; 2002b). Although not studied so far, the acquisition of the two passives
constitutes an interesting problem for learnability in relation to second languages. In the first place, the
two types of passive exist in most languages, including English. This should give the learners an
advantage when acquiring the passives in Spanish. However, in Spanish the distinction is encoded in
two different verbs which are notoriously difficult to master, in particular because many adjectives and
participles often appear with either one. As has been noted since the 1960s (Stockwell, Bowen and
Martin 1965; Schwartz and Sprouse 1997) a distinction that exists in one language but not in another
may be very difficult to acquire.
This paper will report on the results of an experiment that explores the relationship between
knowledge of the two copular verbs, whose underlying aspectual properties the learner has to derive
from the input, and the possibility of transfer of the underlying properties of the passive. We assume
that if the learner has acquired the aspectual difference between the two copulas s/he will be able to
apply this to the expression of the passive. We therefore tested knowledge of ser and estar by
examining their acceptance with different types of predicates, including the stage level and individual
level predicates. Our main objective was to test knowledge of the two passives, which is what we will
be reporting on here.
2. Passives
Many languages include two types of passives: a verbal passive, that denotes an event and can
therefore express an agent, and an adjectival passive, which denotes a state (Levin and Rappaport
1986). In English, verbal passives are very productive in that it is possible to form a passive from
almost any transitive verb. The formation of an adjectival passive is far more restricted, leading some
researchers to argue that the adjectival passive is linked to changes in the lexicon (see Wasow 1977;
1980). Wasow (1977) developed several tests for adjectival passives, including the possibility of
* This project was made possible thanks to a grant from the Social Sciences and Humanities Council of Canada to
the first author, for which we are truly grateful.
© 2005 Joyce Bruhn de Garavito and Elena Valenzuela. Proceedings of the 7th Generative Approaches to Second
Language Acquisition Conference (GASLA 2004), ed. Laurent Dekydtspotter et al., 13-23. Somerville, MA:
Cascadilla Proceedings Project.
prefixing the participle with un- (e.g. untouched, uninhabited), and inserting the phrase very much
before the participle (e.g. very much respected).
Although not all tests work for Spanish, it seems clear that in Spanish the adjectival passive is
expressed by means of the verb estar ‘to be’ and a past participle, while verbal passives are expressed
with the verb ser, also ‘to be’. Although the tests proposed by Wasow do not always work in Spanish,
there is a relatively strong body of evidence for a similar pattern in both languages. In the first place,
passives formed with estar and a past participle always express a resulting state and the agent, if there
is one, cannot be expressed, as illustrated in (1a). Because estar expresses a state, it generally appears
in the imperfect and not the simple preterit when referring to the past. In contrast, the passive
constructed with ser denotes an event, can take an agent in a by phrase, and its canonical past tense is
the simple preterit, as shown in (1b). Unlike (1a), in which the dinner was already prepared when the
speaker arrived, in (1b) the preparing of the supper is interpreted as an event taking place preceding
the act of arrival.
(1) a. La cena estaba (*estuvo) preparada (*por Carmen) cuando llegué a casa.
‘The dinner was (imperfect) prepared (*by Carmen) when I got home.’
b. La cena fue (#era) preparada por Carmen cuando llegué a casa.
‘The dinner was (preterit) prepared by Carmen when I got home.’
As in English, not all participles can take the prefix in- (Varela 1992; 2000) but when they do they
generally take the verb estar without an agent as in (2a). There are some exceptions, in which the
participle, generally formed from a stative verb, can be used to express an intrinsic characteristic of a
noun as in (2b). In other words, it functions as an individual level predicate and no longer as a real
participle.
(2) a. Esta casa está (*es) inacabada (*por Carmen).
‘This house is unfinished (*by Carmen).’
b. Esa palabra es inacentuada.
‘That word is unaccented.’ (It does not take an accent, not ‘there is no accent on it now).
Another possible test in Spanish is to precede the participle with the modifying adverb (muy) bien,
as shown in (3). Again, this can only be done if the auxiliary is the verb estar and there is no agent.
(3) La carne está (*es) muy bien preparada (*por Carmen).
‘The meat is very well prepared (*by Carmen).
The difference between the two passives is related to aspect, that is, the difference between events
and states. It is therefore not surprising that in Spanish they are realized by different copulas given the
facts that current analyses of ser and estar have brought to light. In the first place, it has been noted
(Varela 1992) that the participles used in verbal passives and those used in adjectival passives differ in
their aspectual interpretation. For example, in (4a), with the verb ser, the interpretation points to the
beginning of the act of building, while the same sentence with the verb estar (4b) points to the end,
that is, the resulting state.
(4) a. La casa será construida en marzo. (Varela, 1992, p. 225)
‘The house will be built in March.’
b. La casa estará construida en marzo.
‘The house will be built in March.’
According to other researchers the difference is not to be found in the participle, but rather in the
nature of the two copulas. Lema (1992), Luján (1981) and Schmitt (1992) argue that the difference
14
between them is aspectual. According to Schmitt (1992) “ser is underspecified with respect to
aspect…estar corresponds to the result state of an accomplishment verb”. This explains why ser can
take individual level predicates, while estar is restricted to stage level predicates. As a consequence,
the subject of ser can be interpreted as generic (5a), while the subject of estar cannot (5b).
(5) a. Los policías son nerviosos.
‘(The) policemen are nervous’ (can refer to policemen in general or to a group of policemen)
b. Los policías están nerviosos.
‘The policemen are nervous’ (can only refer to a specific group)
To summarize, the fact that the two copulas are used in Spanish to express the difference between
the two passives follows naturally from the intrinsic aspectual difference between them. An adjectival
passive expresses a state so it is to be expected that it is realized with the copula used for this purpose,
estar. Only ser can accommodate events such as verbal passives.
3. Learnability
The acquisition of the passive constitutes an interesting problem for acquisition. On the one hand,
the difference between the adjectival and verbal passives exists in English, the first language of the
learners. Assuming the Full Transfer/Full Access Hypothesis (Schwartz and Sprouse 1996; 1997),
there should, in principle, be no problem for L2 learners to acquire the difference in Spanish. On the
other hand, in order for the English learners of Spanish to acquire the two forms of the verb ‘to be’ (ser
and estar), they must map the knowledge they may implicitly have of the structure of these two forms
onto the two copulas. To do so they must first derive from the input the difference between them,
namely, that their distributional properties can be accounted for in terms of aspect. It has been argued
for a long time that a distinction found in the L2 which is not present in the L1 may be difficult to
acquire (Stockwell, Bowen and Martin 1965; Schwartz and Sprouse 1997). In many situations, the
difference between sentences using one or the other copula may be quite subtle. For example, in (6a),
given the presence of the verb ser, the sentence must be interpreted as an event. Therefore, la comida,
the subject, cannot refer to food, its most common interpretation, but to a meal. This is in contrast to
(6b), which simply describes the place the food is to be found in.
(6) a. La comida es en la casa.
‘The dinner is (takes place) at the house.’
b. La comida está en la casa.
‘The food is in the house’
To further complicate the matter, our poor understanding of these two verbs is reflected in our
teaching. Not only do students receive long lists of rules which are impossible to produce under
pressure, but the learners are misinformed. Among other things, they are often given the impression
that, apart from a short list of adjectives that can only appear with estar ( lleno, ‘full’, contento,
‘happy’, etc) most alternate between the two copulas, although sometimes with a change in meaning.
But this does not come close to explaining the difference between a stage level predicate and an
individual level predicate, which is probably unteachable anyway.
The third problem is related to historical change and dialectal variation. Students faced with
teachers from different regions will hear different uses. It has been shown that the use of these verbs is
in flux, not only in the United States, where there may be pressure from English to simplify, but also in
Mexico (Silva-Corvalán 1986). In other regions, usage may be more conservative. For example, in
most dialects, a location must be introduced by estar. In the dialect of one of the authors, from
Colombia, use of the copula with location depends on whether the subject is movable or permanent (el
libro está en la mesa, ‘the book is on the table’, Madrid es en España ‘Madrid is in Spain’).
15
Finally, the use of the verbal passive in Spanish is quite rare, given that there is another form of
the passive, the impersonal, which is generally the preferred choice. In one of the tasks presented to the
participants of this subject, monolingual Spanish speakers preferred to translate an English verbal
passive by a Spanish impersonal.
Our understanding of the acquisition of the two copulas is just beginning. According to VanPatten
(VanPatten 1985; VanPatten 1987) learners go through stages in the acquisition of the copulas, with
exclusive use of ser at the beginning, and use of estar progressively spreading through the next three
stages. Geeslin (2000; 2002), following Silva Corvalán (Silva-Corvalán 1986; 1994) uses two
constraints, the Frame of Reference Constraint and the Susceptibility to Change Constraint, to show
that second language acquisition mirrors language change. However, a great deal needs to be done.
This paper aims at contributing to our understanding of one of the uses of the two copulas, to
distinguish the two types of passives.
3. Experiment
3.1 Participants
The experiment included two groups: A group of 9 English L1 learners of Spanish as a second
language and a group of 10 native speakers of Spanish. The L2 group was made up of learners who
began to study Spanish at or after puberty. Some of them were living in a Spanish speaking country at
the time of testing, others were in North America. All were interviewed orally, and the taped
interviews were subsequently judged by two native speakers on four criteria (morphology, syntax,
phonology, and vocabulary), who gave them an overall rating between 7 and 10 (where 10= native
speaker).
The control group consisted of 10 Spanish L1 speakers. None of these participants was an early
bilingual although all spoke English to a certain degree. They came from different Hispanic countries,
with a majority from Colombia.
3.2 Description of tests
The participants completed three tasks. The first of these was a grammaticality judgement task
consisting of 70 sentences ranging over 14 types, 5 tokens for each type. However, 20 of these
sentences (4 types) looked at the copulas with stage level and individual level predicates, which we
will not be reporting on in this paper. This leaves a total of 50 sentences, with 10 types. They
contrasted the use of the two copulas in passive constructions regarding the presence or absence of an
agent, as shown in (7), aspect (simple past vs. imperfect) (8), participles with the prefix in- (9). Each
sentence appeared twice, once with ser and once with estar.
(7) a. Aquí la comida es/*está preparada por un cocinero professional.
here the food is prepared by a cook professional
‘The food here is prepared by a professional cook.’
b. Aquí la comida es/está preparada antes de la llegada de los clientes.
here the food is prepared before the arrival of the clients
‘Here the food is prepared before the arrival of the clients.’
(8) a. El libro fue/*estuvo escrito en Inglaterra.
the book was (pret) written in England
‘The book was written in England.’
b. El artículo *era/estaba escrito en ingles.
the article was (imp) written in English
‘The article was written in English.’
16
(9) El libro estaba/*era inacabado.
the book was unfinished
‘The book was unfinished.’
The second task completed was a selection task. The participants read a short scenario, and were
then given two sentences, one with ser and one with estar. They were asked to judge whether one or
the other was acceptable, whether neither was acceptable or whether both were acceptable. In this paper
we will be reporting the results of 10 scenarios which examined the interpretation of the subject of
each of the copulas. Recall that if the copula ser is used the subject may be interpreted as generic. In
other words, if the story forced a generic interpretation, the participants could not choose the verb
estar. There were 5 scenarios of this type (10). On the other hand, if the subject referred to a specific
element, then from a syntactic point of view both ser and estar are possible, though 5 stories were
constructed in such a way that there was a strong preference for estar (11).
(10) A Carlos no le gusta la comida del Perú. Siempre se queja. Carlos dice:
a. En el Perú el pescado es preparado crudo.
b. En el Perú el pescado está preparado crudo.
c. Ni a ni b
d. Ambas, a y b
‘Carlos does not like the food from Peru. He always complains. Carlos says:
a. In Peru fish is (ser) prepared raw. (Correct answer)
b. In Peru fish (estar) is prepared raw. (Incorrect answer)
c. Neither a nor b. (Posible answer)
d. Both a and b. (Incorrect answer)
(11) El Sr. Ramírez no quiere tomarse el café. No le gusta el sabor. El Sr. Ramírez dice:
a. Este café es preparado a la americana.
b. Este café está preparado a la america.
c. Ni a ni b
d. Ambas, a y b
‘Mr. Ramírez does not want to drink his coffee. He does not like the taste. Mr. Ramírez says:
a. This coffee is (ser) prepared American style. (Incorrect answer)
b. This coffee is (estar) prepared American style. (Correct answer)
c. Neither a nor b (Incorrect answer)
d. Both, a and b. (Incorrect answer)
In this paper we will not be reporting on the results of the third test, which was a translation task.
4. Results
4.1. Grammaticality vs. ungrammaticality
Figure 1 shows the overall results for grammaticality in the grammaticality judgment task. As
the figure shows, both groups distinguish between grammatical and ungrammatical sentences. A two
factor analysis of variance, repeated measures, between grammaticality and L1 groups yields no
significant difference by L1 group, (F(1, 17)=.001, p=.9759). However, it does show a significant
difference for grammaticality (F(13, 221)=42.567, p=.0001) and for interaction (F(13, 221)=4.006,
p=.0001). We will now turn to the comparisons of each of the contrasts.
17
1
1.5
2
2.5
3
3.5
4
4.5
5
L2 speakers L1 speakers
Groups
M
e
a
n
r
e
sp
o
n
se
s
Grammatical
Ungrammatical
Figure 1: Grammaticality vs. ungrammaticality.
4.2. Agentivity
Recall that only the passive with ser, which denotes an event, can take an agent. Four types of
sentences examined this property: (a) ser without an agent, (b) estar without an agent, (c) ser with an
agent, and (d) estar with an agent. Only (d) is ungrammatical. Figure 2 illustrates the results.
1
1.5
2
2.5
3
3.5
4
4.5
5
ser-agent estar -agent ser +agent *estar+agent
Sentence type
M
e
a
n
r
e
sp
o
n
se
s
L2 speakers
L1 speakers
Figure 2: Agentivity
The first and third set of columns show that the L2 speakers do not accept verbal passives, with or
without an agent. This is in contrast to the monolinguals who accepted them as grammatical. However,
18
both groups accept estar without an agent (second set of columns), although the controls do so to a
lesser extent. Finally, the controls clearly reject the ungrammatical estar + agent sentences, but the L2
speakers seem to prefer them slightly to the ser + agent sentences. In fact, there is no significant
difference between the responses of the L2 speakers to the contrast between sentences with an agent,
that is, those with ser which are grammatical and those with estar, which are ungrammatical
(F(1,8)=.14, p=.718). It seems that this distinction has not been acquired.
4.3. Aspect
Recall that because verbal passives refer to events with an end point, they generally appear in the
simple past (preterit). In contrast, adjectival passives refer to a state, and therefore, under normal
conditions, they appear in the imperfect. This contrast was measured with four sets of sentences: (a)
ser in the preterit, grammatical; (b) ser in the imperfect, ungrammatical; (c) estar in the preterit,
ungrammatical; and (d), estar in the imperfect, grammatical. Figure 3 shows the results of these
contrasts.
1
1.5
2
2.5
3
3.5
4
4.5
5
ser pret. *ser imp. *estar pret. estar imp.
Sentence type
M
e
a
n
r
e
sp
o
n
se
s
L2 speakers
L1 speakers
Figure 3: aspect
As we see in Figure 3, both groups of speakers make the correct distinctions in the case of aspect,
correctly accepting the preterit for the verbal passives and rejecting the imperfect; correctly rejecting
the preterit for the adjectival passives and accepting the imperfect.
4.4. Prefix –in
Similarly to English –un, the presence of the prefix –in on the participle in Spanish generally
signals an adjectival passive, therefore the copula used must be the verb estar. This contrast was tested
with two sets of sentences: (a) ser + an –in participle, ungrammatical; (b) estar + an –in participle,
grammatical. The results are illustrated in Figure 4.
19
1
1.5
2
2.5
3
3.5
4
4.5
5
*ser+pref. estar+ pref.
Sentence types
M
e
a
n
r
e
sp
o
n
se
s
L2 speakers
L1 speakers
Figure 4: Prefix –in
Both groups of speakers seem to h
本文档为【spanish passive】,请使用软件OFFICE或WPS软件打开。作品中的文字与图均可以修改和编辑,
图片更改请在作品中右键图片并更换,文字修改请直接点击文字进行修改,也可以新增和删除文档中的内容。
该文档来自用户分享,如有侵权行为请发邮件ishare@vip.sina.com联系网站客服,我们会及时删除。
[版权声明] 本站所有资料为用户分享产生,若发现您的权利被侵害,请联系客服邮件isharekefu@iask.cn,我们尽快处理。
本作品所展示的图片、画像、字体、音乐的版权可能需版权方额外授权,请谨慎使用。
网站提供的党政主题相关内容(国旗、国徽、党徽..)目的在于配合国家政策宣传,仅限个人学习分享使用,禁止用于任何广告和商用目的。