The language produced by learners learning a second language is extremely varied. It can range from one learner to another in regard to many factors. These variations can be accounted for by a number of ideas including: first language (L1) interface, age differences, motivation, self-confidence, aptitude, anxiety, gender and social distance. In this essay I will define SLA and then outline five of the main linguistic theories. These outlines will form the basis for my analysis of the differences in language that are produced by learners. Finally, I will consider what level of impact these theories have and how they can account for these differences and, the many difficulties and successes that learners have on their way to learning a second language.
2. What is SLA and what accounts for the language produced by learners? Saville-Troike (2006: 2) defines SLA as not just the learning of a subsequent language to that learnt in childhood but also the study of the processes involved and of those who are learning it. The language produced by learners changes as they learn the language and that language can differ from one student to another, even if they have the same L1. The following theories provide an insight into how and why this language may vary. Some are backed up by empirical data, others are not, but all have their strengths and weaknesses and they all have supporters and critics. 3. The main theories in SLA
3.1. Contrastive Analysis Hypothesis (CAH)
In terms of the principles of CAH, Gass and Selinker (1994: 59) state that it is “a way of comparing languages in order to determine potential errors for the ultimate purpose of isolating what needs to be learned and what does not need to be learned in a second language learning situation”.Saville-Troike (2006: 34-35) explain that it focuses on the differences and similarities between the L1 and the Second Language (L2). This means that the similarities and differences between L1 and L2 play a crucial role in learners’ production. Saville-Troike (2006: 35) also points out that there will be a transfer of elements acquired in the L1 to the target L2. This transfer is considered positive if the same structure exists in both languages and the transfer results in the correct production of language in the L2.
However, it can also be negative if a language structure from the L1 does not exist in the L2 but the structure is transferred leading to the production of incorrect language. Arab students often omit the verb to be. For example, this book mine for this book is mine since both of them have the same meaning in Arabic /هذا الكتابُ لي /həðəlkɪtəbʊlɪ/. This kind of error might be made since the verb to be is rarely used in the present tense in Arabic. Because of this, Arab students may apply the Arabic rule to English. On the other hand, Arabic and English share the same idea regarding the position of object pronouns. The object pronouns are placed after the verb in English and Arabic.
In contrast, with French, they occur before the verb. Mitchell and Myles (1998: 30) say that the predictions of CAH, that all the errors made in learning the L2 are due to interface from L1, were shown to be unfounded. They claim that many studies and research explain convincingly that the majority of errors could not be attributed to the L1. In other words, CAH might not predict learning difficulties, and was only useful in the retrospective explanation of errors. This point considerably weakened its appeal. However, the heightened interest in this area did lead to the origin of Error Analysis. 3.2. Error Analysis (EA) and Interlanguage (IL)
3.2.1 Error Analysis (EA)
Mitchell and Myles (2004: 29-30) consider this approach to be influenced by behaviorism through the use of fundamental distinctions between the learners’ first and second languages to predict errors, adding that EA showed that CA was not able to predict most errors. They claim that the differences between L1 and L2 are not necessarily difficult, citing as an example the difference between English and French in terms of unstressed object pronouns. These cause a problem for English speakers learning French, but not for French speakers learning English. Saville-Troike (2006: 39-40) observes that EA distinguishes between systematic errors, which are due to a lack of L2 knowledge and mistakes, which are made when the knowledge has been processed. She highlights some of EAs shortcomings including: 1. Some people do not make errors because of L1 interface.
2. Focusing only on errors does not provide information regarding what the learner has acquired. 3. Learners may not produce errors because they avoid difficult structures. For example, Arab students avoid using models auxiliaries since they have difficulties in understanding their role in each sentence. They may use I want…, I need …., instead of could I have, I would like ……..? Overall, EA is not good at accounting for variability in SLA data. 3.2.2 Interlanguage (IL)
Saville-Troike (2006: 40-41) states that the term IL was introduced by Selinker in 1972, “to refer to the intermediate states (or interim grammars) of a learner’s language as it moves toward the target L2”. Ellis (1997: 19) hypothesises that the nature of variability changes during the process of L2 development in the stages below: 1. One form for multi-functions e.g., I live in Manchester, last year I live in London, next year I live in Amman. 2. Some forms have been acquired e.g. I live in Manchester, last year I lived in London, next year I lived in Amman. 3. The various forms start to be used systematically. Here the student may write the forms correctly but still use the incorrect forms when speaking. 4. The student uses the forms correctly and consistently.
3.3. The Monitor Model Theory
Mitchell and Myles (1998: 35) point out Krashen’s theory was based on five hypotheses which are: 1. Acquisition – Learning hypothesis Gass and Selinker (1994:144) refer to Krashen’s assertion that ‘acquisition’ and ‘learning’ are separate knowledge, and that language acquisition is a subconscious process. The acquirers of language are not consciously aware of the grammatical rules of the language, but they rather develop a kind of correctness. This is certainly the case for young children learning their L1. On the other hand, language learning refers to the conscious knowledge of L2. The learners know the rules, they are aware of them, and are able to talk about them. Gass and Selinker (1994: 148) criticise this hypothesis. They claim that it does not show evidence of the distinction between acquisition and learning as two separate systems. However, Krashen said that many can produce language fluently without having been taught any rules and there are many that know the rules but are unable to apply them whilst speaking (Lightbown and Spader 1999: 38).
Krashen’s hypothesis states that what learners learn is available as a monitor (Saville-Troike (2006: 45). Learners will make changes and edit what they are going to produce. The language that learners have consciously learnt works as an editor in situations where they have sufficient time to edit, are focused on form and know the rule (Gass and Selinker 1994: 145-146). This conscious editor is called the Monitor. There are variations in use of the monitor that affect the language that learners produce. Acquired language skills can lead to improved fluency but overuse of the monitor can lead to a reduction in fluency (Krashen 1988: 30-31). Moreover, Krashen (1988: 30-31) believes that there is individual variation among language learners with regard to ‘monitor’ use.
He claims that the learners who use the ‘monitor’ all the time are ‘over-users’, often producing stilted language whereas, ‘under-users’ will often speak quickly but with a lot of errors. Learners who use the monitor appropriately are considered ‘optimal-users’. These find a good balance between speed and accuracy, continuing to refer to want they have learnt but acknowledging the importance of communication. He emphasise that lack of self-confidence is the major cause for the over-use of the ‘monitor’. Gass and Selinker (1994: 149) criticise this hypothesis as they believe that the monitor is only useful in production but it is useless in comprehension since it consists of learned knowledge that is used to edit utterances.
1.Natural Order Hypothesis
According to the natural order hypothesis the acquisition of grammatical structures (rules) proceeds in a predictable order (Gass and Selinker 1994: 145). They add that in a given language, some grammatical structures generally tend to be acquired early while others are acquired late regardless of the L1. They say “the natural order was determined by a synthesis of the results of the morphemes order studies and are a result of the acquired system, without interference from the learned system”. Krashen cited the example that many advanced students in English will still not be able to apply the rule for the third person singular verb, where an –s has to be added to the verb, when speaking quickly.
According to the input hypothesis, SLA cannot take place without sufficient and necessary comprehensible input (Mitchell and Myles 2004: 165). Acquirers develop competency over time by receiving comprehensible input to move their present level to the next. Gass and Selinker (1994: 146) emphasise that this hypothesis is central to Krashen’s description of acquisition and is a complement to the Natural Order Hypothesis.
1.Affective Filter Hypothesis
Krashen’s hypothesis suggests that not everyone has the same ability in learning a second language and that self-confidence, motivation and anxiety all affect language acquisition (Gass and Selinker 1994: 148). He proposed that an Affective filter acts as a barrier to language input. Krashen (1988: 38) explains that a number of affective variables play a crucial role in SLA. These variables include motivation, self-confidence and anxiety. He claims that learners who are highly motivated, self-confident and less anxious are better equipped for success in SLA. Low motivation, low self-esteem, and anxiety contribute to raise the affective filter which prevents comprehensible input from being used for acquisition. In other words, if the filter is high, the input will not pass through and subsequently there will be no acquisition.
On the other hand, if the filter is low and the input is understood, the input will take place and acquisition will have taken place. Gass and Selinker (1994: 148) say that the filter and filter hypotheses explain the failure of SLA according to two parameters: insufficient input and high affective filter, or both. Gass and Selinker (1994: 150) criticize the Filter Hypothesis because it does not explain how it works? Or how the input filter works? However, others see that it as something that can be seen and applied in the classroom and that it can explain why some students learn and produce better language than others (Lightbown and Spader 1999: 40).
3.4. Universal Grammar (UG)
The definition of UG by Chomsky (1976, as cited by Cook, 2001: 181-182) is “the system of principles, conditions, and rules that are elements or properties of all human languages … the essence of human language”. According to Chomsky, there are principles, which allow or prevent a specific structure from occurring in all human languages, and parameters, which govern ways in which human languages differ, usually expressed as a limited choice between two options. These principles and parameters are built in the human mind. In other words, children have an innate faculty that instructs them while learning of language (Mitchell and Myles, 2004: 33).
Saville-Troike (2006: 48-49) gives an example of a principle that Chomsky posited which is that every phrase in every language has the same elements including a head. For example, a noun phrase has to have a noun, a verb phrase has to have a verb and prepositional phrase has to have a preposition. On the other hand, an example of parameter is the direction of the head. For example, Arabic is a head last language and English is a head first language. According to Mitchell and Myles (1998: 61-68), UG can account for variations in learner language as follows:
1. No access hypothesis
This hypothesis suggests that UG becomes less accessible with age and therefore its involvement will not be available to adult learners. Chomsky believes there is a critical period for language acquisition and UGs application. Adult L2 learners have to be prepared to apply more general problem-solving skills. Evidence by Johnson and Newport (1989, as cited by Mitchell and Myles, 1998: 65) showed that immigrant children mostly become native-like speakers of L2, but their parents very rarely do. I believe this supports Chomsky’s hypothesis.
2. Full access hypothesis
Mitchell and Myles (1998: 61) state that the processes of L1 and L2 acquisition are very similar. The differences noticed between them are due to the difference in cognitive maturity and in the learner’s needs. It is clear that L2 learners acquire principles and parameter settings of L2 which are not similar to L1 settings. Evidence given by Flynn (1996 as cited by Mitchell and Myles 1998: 66) explained that Japanese L1 learners of English as L2 successfully acquire L2 head parameter settings. They use principles in English which do not operate in Japanese.
3. Indirect access hypothesis
Mitchell and Myles, (1998: 61-62) point out that access to UG is only available to learners indirectly via the L1. They say “there will be just one instantiation (i.e. one working example) of UG which will be available to the L2 learner, with the parameters already fixed to the settings which apply in the L1”. Evidence given by Schachter (1996 as cited in Mitchell and Myles, 1998: 67) showed L2 learners’ failure to acquire principles absent in their L1 and/or failure to reset parameters.
4. Partial access hypothesis
Mitchell and Myles (1998: 62) say that some aspects of UG are still available and others are not. They give an example stating that principles may still be available but parameter settings may not. In addition, White (2003:1-2) represents the application of the idea of UG to the area of SLA. She argues that SLA is constrained by principles and parameters of UG which is well explained in his book “Second Language Acquisition and Universal Grammar”. In terms of criticism, Mitchell and Myles (1998: 70) say that UG as a whole has been exclusively concerned with syntax and the developmental linguistic route followed by learners when learning a L2. Thus, the social and psychological variables that affect the rate of the learning process are beyond its remit and therefore ignored.
3.5. Socio-linguistic theories
Mitchell and Myles (1998: 163) define sociolinguistics as the study of the effect of all aspects of soceity on the language in use. I will focus on the sociocultural theory discussed in Lantolf (1994). Lantolf (1994: 418) emphasises that the origin of sociocultural theory refers to Vygotsky’s ideas. In terms of variations in learner language, Vygotsky (1978 as cited in Mitchell and Myles, 1998: 146 ) defines the Zone of Proximal Development(ZPD) as ” the difference between the child’s developmental level as determined by independent problem solving and the higher level of potential development as determined through problem solving under adult guidance or in collaboration with more capable peers’.
Saville-Troike (2006: 112) says that one way is to help learners within the ZPD is through scaffolding which is defined as verbal guidance which an expert helps a learner to solve a specific task or collaboration of peers to solve a task that is difficult for any one of them individually. This means that little collaboration or guidance are the main reason for variation in learner language. For example, talk between peers could be helpful as in the following example: Student 1: could I say I am loving you, daddy?
Student 2: I am loving ………..
Student 1: yes, I do not stop loving my daddy.
Student 2: love is a state verb
Student 1: yes, so I am love you, daddy.
Student 2: I think simple present form with state verbs?
Student 1: Ah, I love you, daddy.
To sum up, it is clear that not one individual theory on its own can account for all the variations in learners’ language. Each one has valid points and I have shown some of the variations in language these hypotheses may produce. However, in a lot of cases, there is a lack of empirical evidence and further investigation into these theories may identify new learning and teaching methods. Teaching methods have to take into account that L2 learners are varied. Learners do not have the same characteristics so they do not all acquire a L2 in the same way and at the same rate.
Motivation, aptitude, age, social background and self-confidence affect the learners’ abilities. At the current time, and with the knowledge that is available to us, I think it is important for teachers to consider the most important aspects of each theory when preparing their lessons. Clearly not all theories will be addressed in every lesson, but with careful thought and consideration, the ideas may be applied and the results will show whether or not they are effective for that particular group of students.