Ser and estar : Corrective input to children’s errors of the Spanish copula verbs

Abstract

Evidence for the role of corrective input as a facilitator of language acquisition is inconclusive. Studies show links between corrective input and grammatical use of some, but not other, language structures. The present study examined relationships between corrective parental input and children’s errors in the acquisition of the Spanish copula verbs ser and estar. Spontaneous speech production data from 10 Chilean children (mean age 2;6, age range 1;10–3;4) interacting with their main caretaker at home were analyzed. The study focused on whether parents provided corrective input when their children produce erroneous ser and estar forms and whether it was beneficial for children’s acquisition of the copulas. Results revealed that parental corrective input was positively linked to grammatical copula use by children, although it was not linked to the number of errors, or to the children’s age. The findings suggest that corrective input is a potential source that may constrain the grammatical acquisition of the Spanish copulas.

Keywords

Corrective input first language input Spanish copulas

Input is the language that children hear from people around them, mostly their parents. In the main, there are two basic approaches towards the input that children receive. One approach minimizes the role of the input and holds that children possess an innate universal grammar (independent of the language of exposure) that guides them when acquiring language and requires only the trigger of exposure (Chomsky, 1995; Marcus, 1993; Pinker, 1984, 1988). The opposite approach holds that language acquisition is a generalized learning mechanism that relies heavily on the statistical distribution of the input data (Gathercole, Sebastián, & Soto, 1999; Tomasello, 2000, 2003). Therefore, an important difference between the two approaches lies in the importance they adjudicate to the input.

These and related issues have led to a considerable literature investigating the role of input as a facilitator of language acquisition in children (Baker & Nelson, 1984; Brown, Cazden, & Bellugi-Klima, 1969; Clark, 2010; Conti-Ramsden, 1990; Hart & Risley, 1999; Nelson, 2000; Saxton, Backley, & Gallaway, 2005; Snow, Perlmann, & Nathan, 1987), and in second language acquisition (Lyster, 2004). However, while most researchers agree that input plays a role in language acquisition, they still hotly debate the extent to which it may be responsible for constraining language learning.

The language input that the child receives has been characterized in several ways (Saxton, 2008). In the current study, we focus on corrective input. That is, we concentrate on the explicit corrections that parents provide immediately after the child has made an error. Hence, corrective input is different from, for example, positive input in that the former only refers to input that occurs immediately following an error while the latter includes all correct forms that appear in the input, independent of the child’s utterances. (It should be noted that not all authors view positive input as grammatical data to which children are exposed; see Cowie [1999], Pullum and Scholz [2002], and Saxton, Kulcsar, Marshall, and Rupra [1998], for a variety of positions on how to characterize the relationship between corrective and positive input.)

In this article, we examine explicit corrective input as constrainers of Spanish-speaking children’s grammatical errors with ser and estar. We concentrate on the corrections that parents provide immediately after a child has produced an error since it has been empirically shown that the provision of a direct and immediate contrast between the child’s and the adult’s talk is what makes parental corrective input especially helpful for recovering from mistakes (Bohannon & Stanowicz, 1989; Saxton, 1997; Saxton et al., 1998, 2005). This is because corrective input has the following strengths: it ensures that the child knows that the adult has the same intended meaning; it localizes the error; and it provides the correct form. Therefore, if some type of input provides the necessary evidence for allowing children to repair their error, it is corrective input (Bohannon & Stanowicz, 1989; Farrar, 1992; Moerk, 1991; Saxton, 2000; Saxton et al., 2005). The available research shows an inconsistent and complex picture in which it is not clear how, and to what extent, corrective input may assist in the acquisition of language by children. For example, while some studies find interesting relationships between corrective input and grammatical use of language structures (Farrar, 1992; Saxton, 1997; Saxton et al., 1998), others have found relationships for some but not other structures (Proctor-Williams, Fey, & Loeb, 2001; Saxton et al., 2005) and there is one study that fails to find any relationship between corrective input and subsequent grammaticality in the use of articles and WH auxiliary inversion in English (Morgan, Bonamo, & Travis, 1995). Together, these studies indicate that parental corrective input may have an influence on some, but not other, grammatical structures and also that children may not receive the same level of corrective input for all structures (articles, verbal morphology, etc.).

In addition, an issue that warrants further investigation is that corrective input may be more helpful at certain ages. For example, Hirsh-Pasek, Treiman, and Schneiderman (1984) found that mothers were more likely to repeat ungrammatical rather than grammatical sentences produced by 2-year-olds, but not by 3- and 4-year-old children. Strapp, Bleakney, Helmick, and Tonkovich (2008) found differential responses to corrective and non-corrective input for 3- and 5-year-olds but not for 4-year-olds in a study of novel nouns and verbs. Therefore, it is important that we investigate whether there are certain ages at which corrective input may be more beneficial to assist in learning language.

Most of the literature that has investigated the effects of corrective input in first language acquisition has been carried out in English, with fewer studies involving other languages (but see Chouinard & Clark, 2003; Laakso & Soininen, 2010; Van Veen, Evers-Vermeul, Sanders, & Van den Bergh, 2009). Although there are some studies, such as Watson-Gegeo and Gegeo (1986) and Ochs (1982), that describe the social context of language acquisition in other languages these investigations focus on the appropriate social use of language rather than on its correct grammatical use and therefore do not address the issue of corrective input directly.

In summary, when studying the role of corrective input in language learning it is desirable that researchers: (1) investigate the range of grammatical structures that will allow us to establish which grammatical structures are (or are not) sensitive to corrective input; (2) examine children of different ages; and (3) study languages other than English in order to understand how particular cultural-environmental parental practices may or may not have an effect on language. The study that we present next addresses these research issues by presenting an analysis of natural production data of parental corrective input to children’s errors in the acquisition of the Spanish copula verbs ser and estar.

The Spanish copulas ser and estar

Ser and estar are the two copulative verbs in Spanish. Investigating whether parental input has an influence on their acquisition is important for two main reasons. First, mastering the appropriate use of ser and estar is a complex task. Children need to consider distributional information, a wide range of syntactic, semantic/aspectual, pragmatic, and discursive factors in order to determine the choice of one or the other form. Therefore, it could be hypothesized that children might find it difficult to choose between the two copulas. The difficulty of the task that the child faces becomes evident when we examine the general distribution of ser and estar as illustrated in (1) below (* stands for ungrammatical and # for infelicitous):

(1) a. María es/#está lista. ‘M. is intelligent.’

b. María #es/está lista. ‘M. is ready.’

c. María es/*está maestra. ‘M. is a teacher.’

d. María es/*está de España. ‘M. is from Spain.’

e. María *es/está comiendo. ‘M. is eating.’

As the examples indicate, ser and estar overlap in some environments but they are not always interchangeable. It is reasonable to expect errors when children begin to learn their varied uses.

The second reason as to why it is important to study parental input is that ser and estar are two of the most frequent words in Spanish (Real Academia Española, CREA, 2009). If the acquisition of language is influenced by environmental factors such as parental input, Spanish-speaking children are exposed to rich information as to how to use ser and estar. That is children have numerous opportunities to compare their own ser and estar productions with those of the adults that surround them.

Despite the fact that ser and estar are two of the most frequent words in Spanish, and children receive copious amounts of information regarding their use, with the exception of Sera (1992) and Silva-Corvalán and Montanari (2008), there are no studies that investigate whether the input may help children learn ser and estar. Although in both of these studies some differences are found between the use of ser and estar by children and adults (e.g., overuse of estar), the researchers found that children and adults used the copulas similarly with regard to frequencies and distribution found in the input.

We examine in this study the errors that young monolingual learners of Chilean Spanish make when using ser and estar in spontaneous speech while interacting with their main caretaker. We investigate how parents react to their errors, and the relationship between parental correction rates and children’s uses of the forms.

Following previous research of parental recasts on the use of the copula in English (Proctor-Williams et al., 2001) and studies in Spanish showing that children have some problems with ser and estar in both the production and comprehension domains (Schmitt, Holtheuer, & Miller, 2004; Schmitt & Miller, 2007; Sera, 1992; Sera, Gathje, & Castillo, 1999), it is predicted that Spanish-speaking parents will also correct incorrect or inappropriate uses of ser and estar. Specifically, we examine whether corrective input associates with fewer errors of ser and estar by children and whether parental corrections are influenced by the age of children and the number of errors produced by them.

Method

Participants and procedure

Spontaneous speech production data from 10 Chilean children (mean age 2;6, age range 1;10–3;4); mean MLUw 2.36 (range 1.4–2.92) interacting with their main caretaker at home were collected, transcribed, and analyzed. Data were collected from two one-hour sessions (in two different sessions approximately one week apart) for nine children and from a one-hour session for one child. A trained research assistant checked 10% of the transcriptions for consistency, with agreement of 95%. Disagreements were resolved by discussion.

The mean length of utterance in words (MLUw) has been shown to correlate well with MLU in morphemes in Spanish (Aguado, 1988; Jackson-Maldonado & Conboy, 2007; Jackson-Maldonado et al., 2003), as in English (Parker & Brorson, 2005) and French (Trudeau & Sutton, 2011). Furthermore, MLUw is not only easier to apply but also more reliable than MLU in morphemes. Calculating MLUw allows the researcher to avoid the problem of determining which morphemes are being used productively by children.

Data from one of the original 11 children who participated in the study were omitted because he had a much higher MLUw than the rest of the children. He had a MLUw of 3.66 while the second highest score was 2.92. Since both the level of corrective input by the carer and the kind of response to corrections by the child can vary according to the child’s developmental stage it was necessary to remove this child from the analysis. We set a cutoff point of an MLUw of 3.0 for developmental consistency.

Coding

Coding errors

Every child utterance that either contained ser or estar, or that would require ser or estar for grammatical correctness, was analyzed. Child utterances were coded for ungrammatical or infelicitous uses of ser and estar and any other structure that formed part of the utterances containing these two verbs (e.g., determiners, prepositions, participles). A trained second coder rechecked all ser and estar utterances produced by the children. Agreement on error type was 94%. Disagreements were resolved by discussion. The types of errors concerning the use of ser and estar were identified as:

(2) a. Omission: When the children omitted the required ser or estar form (e.g., este no pato ‘this no duck’).

b. Commission: When the children used an incorrect form (either ser, estar or any other verb) although either ser or estar was the required form (e.g., *ahí e(s) mamá ‘there be (ser) mommy.’ Note that, in this string, estar, but not ser, is the correct form.)

c. Morphosyntactic: When the children made errors of ser/estar morphology, agreement, word order, and closed class elements such as prepositions, determiners, etc. that are relevant to the correct use of ser and estar (e.g., ¿ceniza cuál son? ‘ash, which are?’ Note that ceniza ‘ash’ is singular and, in this string, es is the correct form.)

d. Semantic/pragmatic: When the children made errors of form-meaning mappings of ser and estar (e.g., Mother: ¿y eso qué es? ‘and what is that?’, Child: azul ‘blue’, Mother: no, ¿qué es eso? ‘no, what is that?’ ¿qué animal es ese? ‘what animal is that?’). Note that the mother asks for an NP but the child answers instead with a color adjective (AdjP). The child answer would be correct if the question were ¿de qué color es eso? ‘what color is that?’ Hence it seems that the child still has some problems associating an NP referent with the typical syntax of ser plus NPs.

The 10 children included in this study produced a total of 169 errors. Table 1 shows the distribution of errors in the current corpus. Children made commission, omission, morphosyntactic, and semantic/pragmatic errors. It is important to note that children omitted ser and estar at similar rates. This suggests that they do not seem to treat ser and estar as having different status with respect to their required presence in the sentence (but see Liceras, Fernández-Fuertes, & Alba de la Fuente, 2012, for claims that children omit ser more than estar in Spanish and Catalan). Interestingly, as the examples below show, the non-adult-like uses of ser as the auxiliary in the progressive and to locate object/person subjects indicate that when children are not sure about whether ser or estar has to be used, they default to ser. That is, in the current study children overused ser when the required form was estar. This is shown in (3):

(3) a. E(s)te e(s) durm(i)en(d)o. (1;10)

‘This is (ser) sleeping’

b. E(s) com(i)endo plátano (1;10)

‘(It) is (ser) eating banana’

c. Este es caminando también. (2;11)

‘This one is (ser) walking too.’

d. Allí e(s) mamá. (2;2)

‘There is (ser) mommy.’

e. Ahí e(s) yo. (2;2)

‘There is (ser) I.’

f. E(s) ahí mi chanchito. (2;10)

‘(It) is (ser) there my piggy.’

g. Aquí es comida. (3;6)

‘Here is (ser) food.’

Table 1.

Distribution of errors per category type produced by children.

Omission Ser	Omission Estar	Morphosyntax Ser	Morphosyntax Estar	Commission	Sem/Prag	Total
31	33	61	22	14	8	169

It has been noted in the previous literature that non-adult uses of ser and estar are not frequent in studies of natural production (Liceras et al., 2012; López-Ornat, 1994; Sera, 1992). However, children do make some errors that offer valuable information regarding how they progress in the acquisition of ser and estar. Most studies have reported that Spanish-speaking children overuse estar (see Liceras et al., 2012; López-Ornat, 1994; Schmitt et al., 2004; Sera, 1992; Sera et al., 1999). Interestingly, contrary to what previous research has concluded, the current study suggests that ser is the default copula in the early productions of young monolingual Spanish-speaking children. The contradictory findings are not unexpected since the studies use different methodologies and include children of different ages (from 1;10 to 11 years of age). Hence it is plausible that children undergo different stages in which ser and estar are produced and understood differently.

Coding corrections

Consistent with Saxton’s (1997) Contrast Theory of Negative Input, it was determined that a correction was made if the parental reply contained an immediate repair of the specific faulty structure and also provided the correct target structure. Agreement on parental reply type was 96%. Disagreements were resolved by discussion. For example, in (4), the child omits auxiliary estar and the mother repairs the child’s utterance by including the omitted structure:

(4) Mother: ¿Y dónde está tu guagua? (1;10)

‘Where is your baby?’

Child: No está.

‘(She) is not.’

Mother: ¿No está?

‘(She) is not?’

Child: A tuto.

‘To sleeping.’

Mother: Está haciendo tuto.

‘(She) is sleeping.’

In contrast, in (5), the mother does not repair the morphological error because she omits the copula altogether. Hence, parental replies such as (5) were not included in the coding:

(5) Child: E iguale. (2;4)

‘I(s) same (pl).’

Mother: Iguales las panties.

‘Same (pl) the panties.’

Results

Children’s errors

The 10 children included in this study produced a total of 585 forms of ser and estar. From these, they produced a total of 169 errors regarding the use of ser and estar (Table 2).

Table 2.

Ser/estar productions and errors by children and rate of parental correction.

Child	Age	MLUw	Total ser/estar prod. (TCP)	Number of child errors	% child errors	Errors corrected by parents (%)
1	1;10	2.26	75	20	26.7	10 (50.0)
2	1;11	2.02	33	19	57.6	11 (57.9)
3	2;2	1.4	41	17	41.5	0 (0.0)
4	2;4	1.76	34	13	38.2	4 (30.8)
5	2;4	2.52	54	18	33.2	10 (55.6)
6	2;10	2.76	60	21	35.0	8 (38.1)
7	2;11	2.76	121	5	4.1	4 (80.0)
8	2;11	2.56	42	7	16.7	6 (85.7)
9	2;11	2.64	56	14	25.0	4 (28.6)
10	3;5	2.92	69	35	50.7	9 (25.7)
Total			585	169

Note: TCP = total copula production.

Given that the recording time from one of the children was only one hour (as opposed to two hourly sessions recorded for the other children), the data were normalized by doubling the number of errors and the total ser/estar production for that child (child 5 in Table 2).

Since the children differ in their frequency of use of ser and estar, it is important to consider their erroneous forms in relation to their total production of ser and estar. The most striking result is the great individual variability in the percentages of error. While one child produces erroneous ser/estar forms only 4% of the time, another does so more than half of the time (58%). One might also expect that the children that produce more instances of ser and estar are more prone to make more errors. However, the current data show that this is not the case. The percentage of errors is negatively correlated with total ser/estar production (r = –.632, p < .05). Children who produce more ser/estar forms make fewer errors.

Adult corrective input

An important result is that all parents, except one, corrected their children for ser and estar errors (see Table 2). When we look at the percentages of errors corrected, we see that the amount of correction provided varies immensely from child to child. At one extreme we find a mother who never corrects her child and at the other extreme we find two mothers that correct their children approximately 80% of the time.

Examining the total number of errors we see that the children received corrective input 45% of the time (see Table 3). The rate of correction of ser and estar errors was similar, with 44% of ser errors corrected compared to 46% of estar errors corrected. Therefore, in general, parents do not seem to attend differentially to ser versus estar errors.

Table 3.

Rate of parental correction for children’s ser and estar errors.

Errors corrected	M	SD	Range
Number errors	6.6	3.56	0–11
% Ser errors	43.78	31.8	0–100
% Estar errors	46.2	32.24	0–90
% Ser/estar errors	45.24	26.03	0–85.7

Once we have determined that some children receive corrective input we are faced with the question of whether children could use this information in order to improve their grammatical use of ser and estar. Do children who receive less corrective input make more mistakes than those who receive more? In other words, if corrective input has some effect on the grammaticality of ser and estar, we would expect to find a negative correlation between the amount of corrected input and the number of errors produced by the children.

In order to measure the effects of parental correction, the variable of corrective input was correlated with the variable of children’s error. Since not all children produce the same number of ser/estar utterances, the number of errors was calculated in relation to their total ser/estar (copula) production (TCP). Therefore the two variables correlated are percentage of errors (relative to their total ser/estar production [TCP]) by children and percentage of those errors corrected by parents. A significant negative correlation between percentage of errors by children and the percentage of corrections provided by parents is observed (N = 10, r = –.561, p = .046, one-tailed). This represents a large size effect (Field, 2009). Therefore the result is consistent with the idea that children may benefit from parental corrections when acquiring ser and estar.

Another issue that has been discussed in the acquisition literature is whether parents show a difference in their correction rates relative to the children’s age. Hirsh-Pasek et al. (1984) found that mothers were more sensitive to the ungrammatical utterances of younger children. They only found a significant association between repetitions (with most repetitions including a correction of the child’s error) and grammaticality for 2-year-olds and not for 3-, 4- and 5-year-olds. They explain that mothers may repeat more of their younger children’s ill-formed sentences since ungrammatical sentences are more numerous at early ages.

To investigate this issue we performed two correlational analyses: one for measuring the relationship between the variables of age and percentage corrected and another one considering the number of errors produced by children and their ages. Our results show that (1) younger children were not corrected more than older children (N = 10, r = .077, p = .416), and (2) younger children did not produce more errors than older children (N = 10, r = –.125, p = .366). Therefore we did not find evidence for supporting either the idea that younger children make more ser and estar errors than older children or for the idea that parents correct their younger children’s errors with ser and estar more than older children.

Discussion

In the current study most parents provided corrective input in response to their children’s erroneous uses of ser and estar. However, their rate of correction varied greatly from child to child and did not seem to relate to either the child’s age or the amount of errors produced. These findings are not in line with previous work of parental input in which parents are reported to provide corrections more frequently to younger children who appear to be less proficient users of language (Chouinard & Clark, 2003; Hirsh-Pasek et al., 1984; Penner, 1987). A potential explanation for the different results is that the age range of children included in this study was smaller (1;10–3;5) than that of children who participated in former studies (from 2- to 5-year-olds). However, this range still represents a period of substantial progress in early language use.

Another important finding was the significant association between corrective input provided by parents and the grammatical use of ser and estar by children, suggesting that children may use corrective input in order to limit their error production or to recover from their errors. In the light of this result, even if we accept the no-negative evidence assumption (that is, that parents do not provide explicit negative evidence), we cannot exclude the possibility that corrective input may help children to address ser and estar errors.

It is important to note that we do not interpret the current results as showing that corrective input is the only type of input that plays a role in the grammatical acquisition of ser and estar. These are frequent forms in Spanish, and given that children are exposed to massive amounts of data regarding their appropriate use, it is plausible that positive input may be informative for children’s eventual mastery of the morphological and syntactic properties of ser and estar. In fact, our results show that one child did not receive correction in response to copulative errors. Hence, if this pattern of paternal response is constant, that child will need to recover from his errors by using non-corrective input. This line of thinking is also consistent with learnability proposals that view frequencies of exposure to language as a key to explaining how the language learner revises and changes his or her grammar (see Kirby, 1999; Kirby, Smith, & Brighton, 2004; Yang, 2002, 2010).

Similarly, the current study does not exclude the possibility that children may use corrective and non-corrective input differentially. Experimental research conducted by Saxton (1997), Saxton et al. (2005), and Strapp et al. (2008) indicates that corrective evidence is much more efficient than non-corrective evidence for acquiring correct novel verb forms.

In summary, the current study showed that most parents provided corrective input when their children made ser and estar errors and that the children who were exposed to more corrective input made fewer errors of ser and estar. This suggests that children may benefit from corrections. Further research that contains a larger data set and examines different grammatical structures will provide insight into the issue of whether some structures are more sensitive to corrections than others. Similarly, in order to make more general and accurate claims about parental corrective input and the input in general, future studies also need to compare parental responses to ungrammatical uses with responses to children’s grammatical uses of ser and estar.

Footnotes

Funding and acknowledgements

This article was possible thanks to the support of an Australian Postgraduate Award and the Associative Research Program of Conicyt by the Chilean Government (CONICYT- PIA CIE-05). We thank the assistance of Dr Emlyn Williams (from the ANU Statistical Unit), who kindly guided us through the statistical analysis of the data. We would also like to thank the anonymous reviewers for helpful comments on an earlier draft.

References

Aguado

(1988). Appraisal of the morpho-syntactic competence in a 2.5 year old child. Infancia y Aprendizaje, 43, 75–95.

Baker

Nelson

(1984). Recasting and related conversational techniques for triggering syntactic advances by young children. First Language, 5, 3–22.

Bohannon

Stanowicz

(1989). Bidirectional effects of imitation and repetition in conversation: A synthesis within a cognitive model. In Speidel

G. E.

Nelson

K. E.

(Eds.), The many faces of imitation in language learning (pp. 121–150). New York, NY: Springer-Verlag.

Brown

Cazden

Bellugi-Klima

(1969). The child’s grammar from I to III. In Hill

J. P.

(Ed.), Minnesota symposia on child psychology (Vol. 2, pp. 28–73). Minneapolis: University of Minnesota Press.

Chomsky

(1995). The minimalist program. Cambridge, MA: MIT Press.

Chouinard

Clark

(2003). Adult reformulations of child errors as negative evidence. Journal of Child Language, 30, 637–669.

Clark

E. V.

(2010). Adult offer, word-class, and child uptake in early lexical acquisition. First Language, 30, 250–269.

Conti-Ramsden

(1990). Young children’s conversation with their mothers and fathers: Differences in breakdown and repair. Journal of Child Language, 17, 115–130.

Cowie

(1999). What’s within? Nativism reconsidered. Oxford, UK: Oxford University Press.

10.

Farrar

M. J.

(1992). Negative evidence and grammatical morpheme acquisition. Developmental Psychology, 28, 90–98.

11.

Field

(2009). Discovering statistics using SPSS. London, UK: Sage.

12.

Gathercole

Sebastián

Soto

(1999). The early acquisition of Spanish verbal morphology: Across-the-board or piecemeal knowledge? International Journal of Bilingualism, 2&3, 133–182.

13.

Hart

Risley

T. R.

(1999). The social world of children learning to talk. Baltimore, MD: Paul H. Brookes.

14.

Hirsh-Pasek

Treiman

Schneiderman

(1984). Brown & Hanlon revisited: Mothers’ sensitivity to ungrammatical forms. Journal of Child Language, 2, 81–88.

15.

Jackson-Maldonado

Conboy

B. T.

(2007). Utterance length measures for Spanish-speaking toddlers: The morpheme vs. word issue revisited. In Centeno

J. G.

Obler

L. K.

Anderson

(Eds.), Studying communication disorders in Spanish speakers: Theoretical, research, & clinical aspects (pp. 142–155). Clevedon, UK: Multilingual Matters Ltd.

16.

Jackson-Maldonado

Thal

Fenson

Marchman

V. A.

Newton

Conboy

(2003). MacArthur inventarios del desarrollo de habilidades comunicativas. User’s guide and technical manual. Baltimore, MD: Paul H. Brookes.

17.

Kirby

(1999). Function, selection, and innateness: The emergence of language universals. New York, NY: Oxford University Press.

18.

Kirby

Smith

Brighton

(2004). From UG to universals: Linguistic adaptation through iterated learning. Studies in Language, 28, 587–607.

19.

Laakso

Soininen

(2010). Mother-initiated repair sequences in interactions of 3-year-old children. First Language, 30, 329–353.

20.

Liceras

J. M.

Fernández-Fuertes

Alba de la Fuente

(2012). Overt subjects and copula omission in the Spanish and the English grammar of English–Spanish bilinguals: On the locus and directionality of interlinguistic influence. First Language, 32, 88–115.

21.

López Ornat

(1994). La adquisición de la lengua española. Madrid, Spain: Siglo 21.

22.

Lyster

(2004). Differential effects of prompts and recasts in form-focused instruction. Studies in Second Language Acquisition, 26, 399–432.

23.

Marcus

G. F.

(1993). Negative evidence in language acquisition. Cognition, 46, 53–85.

24.

Moerk

(1991). Positive evidence for negative evidence. First Language, 11, 219–251.

25.

Morgan

Bonamo

Travis

(1995). Negative evidence on negative evidence. Developmental Psychology, 13, 101–107.

26.

Nelson

K. E.

(2000). Methods for stimulating and measuring lexical and syntactic advances: Why Fiffins and Lobsters can tag along with other recast friends. In Menn

Bernstein Ratner

(Eds.), Methods for studying language production (pp. 113–146). Hillsdale, NJ: Lawrence Erlbaum Associates.

27.

Ochs

(1982). Talking to children in Western Samoa. Language in Society, 2, 77–104.

28.

Parker

M. D.

Brorson

(2005). A comparative study between mean length of utterance in morphemes (MLUm) and mean length of utterance in words (MLUw). First Language, 25, 365–376.

29.

Penner

(1987). Parental responses to grammatical and ungrammatical child utterances. Child Development, 58, 376–384.

30.

Pinker

(1984). Language learnability and language development. Cambridge, MA: Harvard University Press.

31.

Pinker

(1988). Learnability theory and the acquisition of a first language. In Kessel

(Ed.), The development of language and language researchers: Essays in honor of Roger Brown (pp. 97–119). Hillsdale, NJ: Lawrence Erlbaum Associates.

32.

Proctor-Williams

Fey

Loeb

(2001). Parental recasts and production of copulas and articles by children with specific language impairment and typical language. American Journal of Speech-Language Pathology, 10, 155–168.

33.

Pullum

Scholz

(2002). Empirical assessment of stimulus poverty arguments. The Linguistic Review, 19, 9–50.

34.

Real Academia Española: Banco de datos (CREA) (2009) Corpus de referencia del español actual. Retrieved from http://www.rae.es

35.

Saxton

(1997). The contrast theory of negative input. Journal of Child Language, 24, 139–161.

36.

Saxton

(2000). Negative evidence and negative feedback: Immediate effects on the grammaticality of child speech. First Language, 3, 221–252.

37.

Saxton

(2008). What’s in a name? Coming to terms with the child’s linguistic environment. Journal of Child Language, 35, 677–686.

38.

Saxton

Backley

Gallaway

(2005). Negative input for grammatical errors: Effects after a lag of 12 weeks. Journal of Child Language, 32, 643–672.

39.

Saxton

Kulcsar

Marshall

Rupra

(1998). Longer-term effects of corrective input: An experimental approach. Journal of Child Language, 25, 701–721.

40.

Schmitt

Holtheuer

Miller

(2004). Acquisition of copulas ser and estar in Spanish: Learning lexico-semantics, syntax and discourse. Boston University Conference on Language Development (BUCLD) 28, proceedings online supplement. Boston, MA, 31 October – 2 November 2003. Retrieved from http://www.bu.edu/bucld/files/2011/05/28-schmitt.pdf

41.

Schmitt

Miller

(2007). Making discourse-dependent decisions: The case of copulas ser and estar in Spanish. Lingua, 117, 1907–1929.

42.

Sera

(1992). To be or to be. Journal of Memory and Language, 31, 408–427.

43.

Sera

M. D.

Gathje

Castillo

(1999). Language and ontological knowledge: The contrast between objects and events made by English and Spanish speakers. Journal of Memory and Language, 41, 303–326.

44.

Silva-Corvalán

Montanari

(2008). The acquisition of ser, estar (and be) by a Spanish-English bilingual child: The early stages. Bilingualism: Language and Cognition, 11, 341–360.

45.

Snow

Perlmann

Nathan

(1987). Why routines are different: Toward multiple-factors model of the relation between input and language acquisition. In Nelson

van Kleeck

(Eds.), Children’s language (Vol. 6, pp. 65–97). Hillsdale, NJ: Lawrence Erlbaum Associates.

46.

Strapp

Bleakney

Helmick

Tonkovich

(2008). Developmental differences in the effects of negative and positive evidence. First Language, 28, 35–53.

47.

Tomasello

(2000). Do young children have adult syntactic competence? Cognition, 74, 209–253.

48.

Tomasello

(2003). Constructing a language: A usage based theory of language acquisition. Cambridge, MA: Harvard University Press.

49.

Trudeau

Sutton

(2011). Expressive vocabulary and early grammar of 16- to 30-month-old children acquiring Quebec French. First Language, 31, 480–507.

50.

Van Veen

Evers-Vermeul

Sanders

Van den Bergh

(2009). Parental input and connective acquisition: A growth curve analysis. First Language, 29, 266–288.

51.

Watson-Gegeo

Gegeo

(1986). Calling out and repeating routines in Kwara’ae children’s language socialization. In Schieffelin

B. D.

Ochs

(Eds.), Language socialization across cultures (pp. 17–50). Cambridge, UK: Cambridge University Press.

52.

Yang

(2002). Knowledge and learning in natural language. Oxford, UK: Oxford University Press.

53.

Yang

(2010). Three factors in language variation. Lingua, 120, 1160–1177.