The Entertainer - An Ethnographic Case Study on Identity Projection by an 11-year-old Moroccan-Dutch Child
- Anna Bruggeman, University of York
- Old College Lorimer Room, University of Edinburgh
The aim of my research is to examine whether identity projection is linked to linguistic
performance and how multi-levelled identities are constructed through use of language. I undertook
a small ethnographic case study focusing on the speech of one 11-year-old Moroccan-Dutch
bilingual child whom I observed in the classroom and during computer tasks and board game
settings. Additionally, I interviewed both the subject and one of his teachers and used results from
a friendship questionnaire filled in by his classmates.
The linguistic variable under investigation was grammatical gender of the definite and
demonstrative determiner, relative determiner and attributive adjective in Dutch (cf. Hulk & Cornips
2006; Cornips 2008; Cornips & Hulk 2008). There were two reasons for choosing this variable in
particular: firstly, non-target use of grammatical gender has been found to be a marker of ‘ethnic’
language use (Nortier & Dorleijn 2008; Cornips 2008). For the subject, growing up in an ethnic
minority environment, variation could thus be expected to carry social significance. Secondly, if this
indeed proved to be the case, it would provide an argument against incomplete acquisition (cf.
Cornips, Van der Hoek and Verwer 2006; Cornips 2008).
The subject’s productions involving grammatical gender were noted down. Subsequently,
results were organised in two different ways: i) according to social context (classroom, etc.) and ii)
according to interactional stance (Bucholtz & Hall 2010). The total number of relevant productions
was 312, of which 31 were non-target. The distribution across different social contexts showed that
variation was most prominent in the classroom (11%) and least common in the interview context
(4%). For interactional stance, classification of non-target items only was carried out in order to
consider the significance of each of these instances individually. The items were judged as
belonging to one the following stances: ‘cool kid’, ‘story teller’, ‘uncooperative’, and ‘serious’
(‘default’ constituting the unclear cases). The lowest number of non-target grammatical gender
productions occurred with ‘serious’ stance and relatively many instances were observed when a
‘cool kid’ identity was adopted. With regard to the type of instance of grammatical gender, the large
majority of non-target use concerned the demonstrative determiner, with adjective, relative
pronoun and definite determiner accounting for the remaining third.
The results showed that the way the subject positioned himself had a large effect on the
degree of non-target use. In addition, the words produced most inconsistently (i.e. with both many
target and non-target functional elements) are very frequent ones. Based on these facts, I argue
that the subject’s variation with regard to grammatical gender is primarily constrained by the
degree of attention he wishes to attract. However, there are many limitations to this study, not least
of all the limited amount of tokens. Additionally, it has to be kept in mind that most judgments are
based on personal observation and may have only partially uncovered the subject’s attitudes.
performance and how multi-levelled identities are constructed through use of language. I undertook
a small ethnographic case study focusing on the speech of one 11-year-old Moroccan-Dutch
bilingual child whom I observed in the classroom and during computer tasks and board game
settings. Additionally, I interviewed both the subject and one of his teachers and used results from
a friendship questionnaire filled in by his classmates.
The linguistic variable under investigation was grammatical gender of the definite and
demonstrative determiner, relative determiner and attributive adjective in Dutch (cf. Hulk & Cornips
2006; Cornips 2008; Cornips & Hulk 2008). There were two reasons for choosing this variable in
particular: firstly, non-target use of grammatical gender has been found to be a marker of ‘ethnic’
language use (Nortier & Dorleijn 2008; Cornips 2008). For the subject, growing up in an ethnic
minority environment, variation could thus be expected to carry social significance. Secondly, if this
indeed proved to be the case, it would provide an argument against incomplete acquisition (cf.
Cornips, Van der Hoek and Verwer 2006; Cornips 2008).
The subject’s productions involving grammatical gender were noted down. Subsequently,
results were organised in two different ways: i) according to social context (classroom, etc.) and ii)
according to interactional stance (Bucholtz & Hall 2010). The total number of relevant productions
was 312, of which 31 were non-target. The distribution across different social contexts showed that
variation was most prominent in the classroom (11%) and least common in the interview context
(4%). For interactional stance, classification of non-target items only was carried out in order to
consider the significance of each of these instances individually. The items were judged as
belonging to one the following stances: ‘cool kid’, ‘story teller’, ‘uncooperative’, and ‘serious’
(‘default’ constituting the unclear cases). The lowest number of non-target grammatical gender
productions occurred with ‘serious’ stance and relatively many instances were observed when a
‘cool kid’ identity was adopted. With regard to the type of instance of grammatical gender, the large
majority of non-target use concerned the demonstrative determiner, with adjective, relative
pronoun and definite determiner accounting for the remaining third.
The results showed that the way the subject positioned himself had a large effect on the
degree of non-target use. In addition, the words produced most inconsistently (i.e. with both many
target and non-target functional elements) are very frequent ones. Based on these facts, I argue
that the subject’s variation with regard to grammatical gender is primarily constrained by the
degree of attention he wishes to attract. However, there are many limitations to this study, not least
of all the limited amount of tokens. Additionally, it has to be kept in mind that most judgments are
based on personal observation and may have only partially uncovered the subject’s attitudes.