2. Language attrition - two languages in one brain
3. Factors influencing language attrition
3.1 Psycholinguistic factors
3.2 Sociolinguistic factors
Language attrition is a comparatively new research approach for socio- and psycholinguists all over the world. Their interests go into different directions, like the process of language attrition itself, the linguistic areas affected of forgetting or rather losing a language, and the “Effects of the Second Language on the First” (Cook 2003). But especially one topic divides the minds of the researchers: Which factors influence the loss of a language; and to what extent?
The elaboration at hand tries to answer this basic question with regard to the loss of the first language (from now on referred to as L1) in a foreign language (from now on referred to as L2) environment. Withal the focus lies on the attrition process in adults, although that of children is shortly introduced, too. The material put together and evaluated in this paper originates from several socio- and psycholinguists which conducted research on the different factors that have an influence on language attrition.
The first step to answer this question is to give a short overview on how two language systems can exist in one individual. As a part of this, it will be clarified what exactly is meant by the term “language attrition”, because it is often misunderstood. Afterwards, it will be easier to understand the connection between these two systems and therefore the reason why they are likely to be attrited. Furthermore, a closer look will be taken at the most common, non-pathological factors that have an impact on language attrition, parted into sociolinguistic and psycholinguistic factors. Their degree of influence on this process will be discussed thereafter with regard to the hypothesis: Sociolinguistic factors have a greater impact on the process of language attrition than psycholinguistic factors have.
2. Language attrition - two languages in one brain
As already mentioned in the introduction, researchers are still looking for a final answer to the question why and how people attrite a language. Nevertheless, it is important to know about the structure in bilingual’s minds to understand which factors can influence it. In her book Effects of the Second Language on the First the author and editor Vivian Cook (2003) introduces the most common assumptions on the relationship between two languages in one speaker’s mind and puts them together in her integration continuum.
Cook’s (2003) integration continuum consists of three models. The first one is the separation model, where L1 and L2 form two independent language systems “with no connection between the different languages in the mind” (Cook 2003: 6). As an example for this model, she mentions the foreign language teaching at school, where the new language is taught “with no links to the first” (Cook 2003: 6-7). Teachers hope that in this way the level of the L1 has no influence on the learning of the L2. The second model that is part of the integration continuum is the model of partial integration. It says that there is partially language-specific knowledge in the speaker’s mind, but also an overlapping area with aspects belonging to both language systems (cf. Cook 2003: 8). Finally, the third and last component of the integration continuum then composes the integration model. Here, both languages form one single system whose elements have to be balanced in order to not mix up the two languages (cf. Cook 2003: 7).
Cook’s integration continuum as such now combines these three stages. It is illustrated in the following chart taken from her article Introduction: The Changing L1 in the L2 User’s Mind (cf. Cook 2003: 9). The term “LA” here describes the first language of a speaker, the term “LB” the second, foreign one.
She makes clear that “[it] does not necessarily imply a direction of movement” (Cook 2003: 9) so that it functions in both directions. This means that the speaker can develop 2 from a separated to an integrated bilingual as well as the other way round. This again would mean that the speaker is more likely to attrite when his or her knowledge of two languages develops from one single system into two completely separated systems with no connection to each other. The factors that can influence this transition between the different levels of the continuum are discussed in chapter three.
But the illustration also clarifies that the term attrition does not automatically mean the complete loss of knowledge of one language, because neither the L1 nor the L2 ever disappears in the model. Attrition therefore means rather a “difficulty in access to knowledge” (Wen Cui Zhou 2010: 12). The knowledge is still there, but cannot be activated that easily. Wen Cui Zhou (2010: 123) furthermore argues that “language attrition is not a final state” and that it as a result can be slowed down or even reversed.
3. Factors influencing language attrition
3.1 Psycholinguistic factors
One area of factors which can influence the process of language attrition are the psycholinguistic ones. Other than the sociolinguistic factors, the psycholinguistic ones are not influenceable by the speaker himself, because they deal with the human faculty of speech in general.
One of these factors is age. It involves the age until when the L1 and L2 were acquired respectively learned as well as the age at the onset of attrition. Concerning this aspect, the first thing to mention is the difference between adult und child L1 attrition. It is hard to find evidence for L1 attrition in adult bilinguals because its effects are small (cf. Schoenmakers-Klein Gunnewiek 1998; cited in Opitz 2011: 209) and often subjectively experienced rather than objectively measurable (cf. Dostert 2009: 13). In child L1 attrition, in comparison, in fact a complete loss of a language is possible (cf. Dostert 2009: 14). Köpke and Schmid (2004: 10) even state that “the younger the children are when their language environments change, the faster and deeper they will attrite” (cited in Wen Cui Zhou 2010: 12). One reason for this enhanced child L1 attrition is the critical period which ranges from birth to puberty. Researchers argue that only after this period a first language is completely acquired. During the critical period the language system is not completely stable yet. Nevertheless, it is difficult to do research on child L1 attrition, because it is “the most important step [...] to find out what children have acquired” (Wen Cui Zhou 2010: 18) of their L1 during the critical period, and this is a question that cannot be answered that easily.
This is closely related to the fact that the “level of achievement at the onset of attrition” (Schmid 2002: 20) is very important. Wen Cui Zhou (2004) summarizes in her thesis Birth Language Attrition/Retention and Relearning: A Longitudinal Study on Perception of Mandarin Chinese Tones among School Age Chinese Adoptees in the Netherlands the results of taking a child away from its familiar linguistic environment during the critical period. She writes that “with birth language input disappearing suddenly, they have to reset their mind to learn a completely new language; as a result, their limited and unstable birth language is attriting drastically” (Wen Cui Zhou 2004: 18). This means that with their L1 not completely acquired, children do not have a real chance to maintain their birth language. Adults in contrast have a stable and completely acquired L1, which they will never lose entirely. But it is not only the break with a linguistic environment during the critical period that has a significant influence on L1 attrition. Also older people are vulnerable to attrite. One cause for this is “the language spoken to children and to elderly people” (Seliger & Vago 1991: 41) as Caporael and Culbertson (1986) suggest on the basis of their research. In both cases the language is different from that spoken to middle-aged people. The linguistic demands put on the elderly are much lower, so that they are also likely to attrite due to their reduced input. Researchers agree that the age question is a major one when it comes to the analysis of attrition.
The other psycholinguistic aspect is the language aptitude. There are people that have a language aptitude lying above or under the average and therefore, these people score accordingly better or worse when it comes to linguistic tests. The influence of the language aptitude factor on the process of language attrition is nevertheless controversial. While in her thesis Between Attrition and Acquisition: the Dynamics between Two Languages in Adult Migrants Cherciov (2011: 40) at first, based on preceding studies, argues “that greater language aptitude should not only lead to higher L2 proficiency, but also prevent L1 attrition”, she also mentions the trade-off hypothesis which says that the amount of attrition raises with the proficiency of the L2 (cf. Opitz 2005, cited in Cherciov 2011: 40). After her survey she then says that language aptitude can, in comparison to other factors, only play a minor role in speeding up or slowing down the language attrition process (cf. Cherciov 2011: 123).
3.2 Sociolinguistic factors
Besides the psycholinguistic factors there are also sociolinguistic ones that can influence the process of attrition. They involve all cultural and personal attitudes towards a language. One of these sociolinguistic aspects is immigration and the associated language input. In her thesis Wen Cui Zhou (2010: 10) argues that “language development is a dynamic process” which is caused by “input and real life use”. If this input and real life use is missing, attrition is the consequence, because it is also a sort of language development - in this case a development to the negative (cf. Wen Cui Zhou 2010: 10). This is also what happens to immigrants when coming to the new country. As already said in chapter 3.1, for children it is nearly impossible to maintain their L1 during the critical period without any input from their usual environment. This is quite similar to what happens to adults: If there is no one to talk to in the L1, the lack of input and lack of use of the language leads to a degradation of the L1; but not to a complete loss like it is sometimes the case in child L1 attrition.
Moreover, Cherciov (2011: 30) states that “language skills may deteriorate in the first decade, but then remain fairly stable after this period”. She proves this thesis with a study by de Bot and Clyne (1994), where immigrants were tested twice, before and after an interval of 17 years. The results could not prove any further attrition in comparison to the results taken 17 years before, what implies that after a certain degree of attrition, the process slows down or even stops (cf. Cherciov 2011: 30). But she also mentions that the importance of this factor of time in the L2 country, which is one subcategory of immigration, is quite ambivalent. Its importance depends on the “degree of contact with the L1”, the age factor and the “efforts [made] to acquire the L2” (Cherciov 2011: 31). De Bot, Gommans and Rossing (1991) make this even clearer by saying that “’time’ only becomes relevant when there is not much contact with the [L1]” (in: Seliger & Vago 1991: 94).
In addition to the input from and the time spend with another language, the output, which means the contact with the L1, is a variable that can slow down as well as speed up the process of language attrition. It depends on two different attitudes during the attrition process: Is the speaker ashamed of talking in the attriting L1? Or does he / she not see any problem with attrition? (cf. Schmid 2002: 23) The degree of self-confidence here can have a huge impact on the contact with the first language. Another influence is the attitude towards the L1. If the first language - in the eyes of the speaker and his / her environment - has prestige, the speaker is more likely to use it wherever possible. If not, the speaker will even downplay the use of that language (cf. Schmid 2002: 23-24).
Besides all that, the motivation in general plays a very important role when learning a new language and therefore maybe attriting another one. Here the keyword immigration comes in again. Schmid (2002) mentions that when moving to another country, the status ‘immigrant’ often has a negative connotation. This can have the effect that immigrants become outsiders and reject their L1. As a result the wish for assimilation comes up and makes these people focus on their L2, while they attrite their L1 at the same time (cf. Schmid 2002: 26). In this case the motivation to learn a new language and leave the other one behind is very high. The reason for this is that language behaviour is a marker of identity. If immigrants want to belong to a certain group of people, the language is an aspect they can change. Other aspects like age, sex or ethnicity are not changeable (cf. Schmid 2002: 27). This is also the reason why Wen Cui Zhou names “a negative attitude towards the L1 and a desire to assimilate rapidly into the dominant society [as] the most important initiators of the attrition process” (Wen Cui Zhou 2010: 123).
Furthermore, the education plays a role when looking at adult L1 attrition, even if it is only a minor one. Cherciov (2011: 28) argues that a higher education is connected to a larger vocabulary, which again, in combination with the generally good education, can lead to better means to visit people in the L1 country. This means that again the L1 input increases, whose importance was already mentioned above. Additionally, she states that the level of education has an indirect influence on the test performance of the experimentees. She mentions this in respect of the C-test1. As explanation Cherciov writes that less educated experimentees are possibly not that familiar with those test procedures (cf. Cherciov 2011: 28) - a statement that should be raised to question. The comprehension of test procedures cannot automatically be related to a lack of education, because it can also depend on other factors, like the formulation of the question or simply a lack of knowlegde to answer the question.
The last and probably least important sociolinguistic factor is gender. Schmid (2002) is one of few to include this variable. It is rather a sociolinguistic than a psycholinguistic one, because it gets its validity only in a cultural context. This means that gender influences the process of language attrition only when it comes to gender roles. The cultural context focusses primarily on women: Are they allowed or expected to work? Do they communicate with other people in the L2 community? (cf. Schmid 2002: 22)
1 The C-Test is a variety of a fill-in-the-blank text. It examines in the first place the grammar and vocabulary knowledge of a person. (cf. Raatz et al. on http://www.c-test.de)