The following paper compares and contrasts three different studies in order to investigate, whether grammatical gender agreement in heritage speakers involves incomplete acquisition, and if so, how gender agreement is affected by this and other factors like age of onset and context of acquisition. The papers by Montrul, Foote, & Perpiñán (2008) and Alarcón (2011) analyse Spanish heritage speakers, whereas the study by Polinsky (2008) focuses on American Russian heritage speakers. In this paper, all three studies are analysed in order to compare the participants, the procedure and the specific results concerning grammatical gender under incomplete acquisition across different languages. The analysis of the papers is followed by a general discussion of their main outcome. In the end, I will conclude the overall results and provide an outlook.
Inhaltsverzeichnis
1. Abstract
2. Introduction
3. Study 1
3.1 Participants
3.2 Experiment 1
3.2.1 Task and Materials
3.2.2 Results
3.3 Experiment II
3.3.1 Task and Materials
3.3.2 Results
3.4 Experiment III
3.4.1 Task and Material
3.4.2Results
3.5 Discussion
4. Study II
4.1 Participants
4.2 Experiment 1
4.2.1 Task and Materials
4.2.2 Results
4.3 Experiment II
4.3.1 Task and Materials
4.3.2 Results
4.4 Discussion
5. Study III
5.1 Participants
5.2 Experiment 1
5.2.1 Task and Materials
5.2.2 Results
5.3 Experiment II
5.3.1 Task and Materials
5.3.2Results
5.4 Discussion
6. General Discussion
7. Conclusion
References
1. Abstract
The following paper compares and contrasts three different studies in order to investigate, whether grammatical gender agreement in heritage speakers involves incomplete acquisition, and if so, how gender agreement is affected by this and other factors like age of onset and context of acquisition. The papers by Montrul, Foote, & Perpinan (2008) and Alarcon (2011) analyse Spanish heritage speakers, whereas the study by Polinsky (2008) focuses on American Russian heritage speakers. In this paper, all three studies are analysed in order to compare the participants, the procedure and the specific results concerning grammatical gender under incomplete acquisition across different languages. The analysis of the papers is followed by a general discussion of their main outcome. In the end, I will conclude the overall results and provide an outlook.
2. Introduction
According to the American Heritage College Dictionary the word heritage means that something is acquired from birth or something which is transmitted to a generation by a preceding generation (Montrul, 2016). Learning and mastering a heritage language can take multiple years of exposure to it. During the preschool years the basis of the linguistic competence is established. It expands as a child is further exposed to written language in school and as he or she develops cognitively and socially. Based on Montrul's ideas there are two factors necessary in order to acquire a language. The first is the so called innate linguistic capacity of language and the second is input of the language. Children and adults have to learn both the comprehension and the oral production of a language in order to become a proficient speaker of it. (Montrul, 2016). Another important factor for learning or unlearning a heritage language is the age of onset. The age of onset of a language can be seen as relevant for the acquisition or the incomplete acquisition of a language (Montrul, 2016).
If a heritage speaker was insufficiently exposed to the L1 in his or her early childhood and especially the school years and thus did not receive the minimum input in order to acquire different aspects of the language, like the morphology and the syntax, we are speaking of incomplete acquisition. According to Montrul (2016; as cited in Pudnam, 2013) incomplete acquisition can lead to divergent grammar, which is different from the baseline due to insufficient input. Pudnam and Sanchez (2013; as cited in Pudnam, 2013) state that it is rather unclear at which point of time a child has completely acquired a L1. According to Montrul (2002, 2008, 2009; as cited in Pudnam, 2013) and Polisky (2006; as cited in Pudnam, 2013) the main difference between L1 attrition and incomplete acquisiton of the L1 is that attrition takes place when production errors in a fully acquired mature L1 grammar occur due to an exposure to the L2, whereas incomplete acquisition concernes bilingual children who are exposed to L1 input of low quality during the age of linguistic development, age 0 to 4, or the age of later linguistic development, which takes place between the age of 4 to 13, which results in an incomplete development of the L1 (Montrul, 2009; as cited in Pudnam, 2013).
While studying this field of research the following questions arose: does grammatical gender agreement in heritage speakers involve incomplete acquisition? If so how does this factor affect grammatical gender agreement in heritage speakers and how important are the already mentioned factors like age of onset and context of acquisition for completely or incompletely acquiring grammatical gender agreement?
My hypothesis is that all three papers I am going to analyse will come to the result that incomplete acquisition is a matter of the age of onset and the context of acquisition ofthe bilingual individuals.
This paper is structured as follows. I am going to analyse the reseach papers by Silvina Montrul, Irma V. Alarcon and Maria Polinsky. I will discuss the aim of each study, the procedure and the specific results of them. Then I am going to clarify which approach they support and how the authors reflect their own results. Then I am going to have a general discussion in which I will compare and contrast the main outcomes and the main arguments of the studies. After that I am going to provide some criticism concerning the studies. Finally I will finish this paper by concluding the overall outcome and by answering my research questions and the question whether my hypothesis was confirmed based on my findings.
3. Study I
The first research paper I am going to analyse is the study 'Gender Agreement in Adult Second Language Learners and Heritage Speakers: The Effects of Age and Context of Acquisition' by Silvina Montrul. The aim ofthe study is to gain knowledge about gender agreement in Spanish noun phrases. The main question is whether there is an advantage for heritage speakers due to age of onset of bilingualism. Furthermore the author intends to clarify whether there are differences in the overall written comprehension and oral production of gender agreement in noun phrases between adult L2 learners and Spanish heritage speakers. If the L2 learners and the heritage speakers produce gender agreement errors, another aim would be to find out what kind of error patterns in terms of gender, domain of agreement and noun ending they are producing and where they differ from each other (Montrul et al., 2008).
3.1 Participants
The participants of the study consisted of 72 L2 learners of Spanish, 69 Spanish heritage speakers and 22 monolingual native speakers of Spanish. Each L2 learners was a native speakers of English. English was their dominant language at home during childhood and they started to acquire Spanish as an L2 at the age of 12 to 20 years, usually in high school. Most of them were enrolled in Spanish language classes at the University of Illinois and the advanced learners were graduate students or even Spanish language instructors. The participants age ranged from 18to30 years at the time of the data collection (Montrul et al., 2008).
Because they are the largest Spanish-speaking group in the area, all heritage speakers had to be of Mexican-American descent. Another criteria was that each of the heritage speakers had to be born and schooled in the United States and they had to have been exposed to English before they reached preschool age, otherwise they were excluded from the experiment. Just as the L2 learners, each heritage speaker was a graduade or undergraduade student at the University of Illinois and was enrolled in a Spanish language class (Montrul et al., 2008).
The monolingual comparison or baseline group consisted of native Spanish speaking individuals from Spain, Argentina and Mexico. One half was tested abroad, the other half were recent arrivals in the United States. The speakers of these regional varieties were chosen because their dialects fit those of the Spanish language instructors of the University of Illinois, who teach many of the Spanish language classes. Moreover, these dialects were chosen because they have only few lexical differences concerning gender agreement. The members of this group performed homogenously when it came to a written proficiency measure and several experimental tasks, showing that there were no differences between the regional dialects within the group. The age of the group members ranged from 21 to 57 years (Montrul et al., 2008).
3.2 Experiment I
In order to answer her research questions, Montrul conducted three experiments. The first experiment aimed to find out whether heritage speakers are more accurate than L2 speakers when it comes to comprehending gender features (Montrul et al., 2008).
3.2.1 Task and Materials
The participants were asked to solve a task consisting 32 target sentences and 13 distractors, including sentence structures with noun drop, gender and number manipulation, and exemplified the four noun conditions, masculine singular, masculine plural, feminine singular and feminine plural (Montrul et al., 2008).
The task was conducted on a computer using a Web interface. The 45 sentences were randomly presented on the computer screen, followed by three pictures. The participants were asked to read a piece of conversation and had to choose the correct picture afterwards. The participants had to do so on the basis of gender and number of the pictures. They had to avoid distractor items, which were used to make sure that the participants had understood the task (Montrul et al., 2008).
3.2.2 Results
The results showed that the monolingual Spanish speakers did produce very few incorrect responses, confirming that native Spanish speakers do not have problems with gender agreement. Therefore the native-speaker group was excluded from the statistical comparison (Montrul et al., 2008). In order to show the points of distribution and correlation between the scores on the Written Picture Identification Task and the proficiency test for the L2 learners and the heritage speakers, the authors used scatter plots and a repeated-measures factorial ANOVA (Analysis of variance). They showed that more L2 learners scored below 80% on the proficiency test than heritage speakers scored above 80% accuracy in the Written Picture Identification Task. The result suggested that the correlation was more robust for the heritage speakers than for the L2 learners. Another table, which focusses on the performance of noun drop sentences by gender, was evaluated and showed that the L2 learners were more accurate than the heritage speakers. In general it can be said that both L2 learners and heritage speakers produced more gender errors when it comes to identifying feminine nouns and matching them with noun drop sentences. More low or intermediate learners of Spanish as L2 than Spanish heritage speakers performed above 80% accuracy in the test (Montrul et al., 2008).
3.3 Experiment II
The second experiment was supposed to test the direct opposite of the first experiment. The aim was to test the “written recognition of the correct masculine or feminine form of determiners and adjectives based on the ending of the noun“ (Montrul et al., 2008). The authors wanted to test whether the heritage speakers would perform more accurate than the L2 speakers when it comes to identifying the correct gender agreement even if these were manipulated. The same participants were tested as in the first experiment (Montrul et al., 2008).
[...]