Bilingual First Language Acquisition with a View to Attrition
- Stephanie Maia, University of Edinburgh
- Old College Lorimer Room, University of Edinburgh
Vulnerability at the interfaces between syntax and discourse pragmatics has been attested in
different forms of linguistic development including bilingual first language acquisition, monolingual
first language acquisition and native language attrition (e.g. Müller and Hulk, 2001, Serratrice 2005,
Sorace and Serratrice 2008). One particularly fruitful area of study has been that of the null and
overt pronoun conditions found in romance languages. In European Portuguese, the null subject
condition is used to refer to the subject of the matrix clause whilst the overt pronoun refers to the
object of a matrix clause. Studies in Italian, which exhibits the same tendencies, have shown that
in the process of acquisition, there is some considerable optionality in the overt pronoun condition,
leading to delayed development in bilingual children. During native language attrition, similar
difficulties can be found in the overt condition with regards optionality (e.g. Tsimpli et al 2004).
If the same effects can be seen in both acquisition and attrition, it seems important that
linguists persue an explanation which takes a unitary approach. Carminati (2002) proposes that it
is the common processing strategies employed in pronoun resolution which gives lead to this
optionality. Language general processing strategies compete with language specific processing
strategies, reinforcement of a language general processing strategy over that of a language
specific processing strategy; or lack of experience of a language specific processing strategy can
be said to create this sort of optionality.
This study takes a 2.6 year old Portuguese-English bilingual, Isabel, whose Portuguese input
has been provided by an attrited native speaker of Portuguese. Isabel’s preferences in anaphora
resolution were assessed using a ‘comprehension test’. The experiment consisted of two puppets.
The puppets would have a ‘conversation’ which one would then report to the Portuguese
‘puppeteer’. The puppeteer would the relay the information to Isabel in either a null or overt
pronominal form:
“Aqui está o Panda e aqui está o Tigre.” [Puppets converse]
“O Tigre disse que tem fome! Quem está tem fome?”
Results from the study show that, although Isabel has little experience of the null subject
condition, she correctly associates the null subject form with the matrix subject. In the case of the
overt pronoun, Isabel struggles and wrongly assigns the overt pronominal form to the matrix
subject as often as she does to the object.
These results are supportive of Carminati (2002) processing strategies. The fact that Isabel is
dominant in English, and has received attrited input does not seem, at this stage, to affect the
strategies she applies to anaphor resolution. Taken with evidence collected from experiments in
other languages, such as those in Italian, we can see a picture emerging where a unitary approach
and processing strategies certainly seem to be an exciting and plausible way forward.
different forms of linguistic development including bilingual first language acquisition, monolingual
first language acquisition and native language attrition (e.g. Müller and Hulk, 2001, Serratrice 2005,
Sorace and Serratrice 2008). One particularly fruitful area of study has been that of the null and
overt pronoun conditions found in romance languages. In European Portuguese, the null subject
condition is used to refer to the subject of the matrix clause whilst the overt pronoun refers to the
object of a matrix clause. Studies in Italian, which exhibits the same tendencies, have shown that
in the process of acquisition, there is some considerable optionality in the overt pronoun condition,
leading to delayed development in bilingual children. During native language attrition, similar
difficulties can be found in the overt condition with regards optionality (e.g. Tsimpli et al 2004).
If the same effects can be seen in both acquisition and attrition, it seems important that
linguists persue an explanation which takes a unitary approach. Carminati (2002) proposes that it
is the common processing strategies employed in pronoun resolution which gives lead to this
optionality. Language general processing strategies compete with language specific processing
strategies, reinforcement of a language general processing strategy over that of a language
specific processing strategy; or lack of experience of a language specific processing strategy can
be said to create this sort of optionality.
This study takes a 2.6 year old Portuguese-English bilingual, Isabel, whose Portuguese input
has been provided by an attrited native speaker of Portuguese. Isabel’s preferences in anaphora
resolution were assessed using a ‘comprehension test’. The experiment consisted of two puppets.
The puppets would have a ‘conversation’ which one would then report to the Portuguese
‘puppeteer’. The puppeteer would the relay the information to Isabel in either a null or overt
pronominal form:
“Aqui está o Panda e aqui está o Tigre.” [Puppets converse]
“O Tigre disse que tem fome! Quem está tem fome?”
Results from the study show that, although Isabel has little experience of the null subject
condition, she correctly associates the null subject form with the matrix subject. In the case of the
overt pronoun, Isabel struggles and wrongly assigns the overt pronominal form to the matrix
subject as often as she does to the object.
These results are supportive of Carminati (2002) processing strategies. The fact that Isabel is
dominant in English, and has received attrited input does not seem, at this stage, to affect the
strategies she applies to anaphor resolution. Taken with evidence collected from experiments in
other languages, such as those in Italian, we can see a picture emerging where a unitary approach
and processing strategies certainly seem to be an exciting and plausible way forward.