Processing Based Explanations for Poor Grammaticality Judgment Performance by Late Second Language Learners - Coursework Example

Beyond the Critical Period Processing Based Explanations for Poor Grammaticality Judgment Performance by Late Second Language Learners 0 Introduction The aim of this essay is to summarize, analyze, and synthesize an existing article titled “Beyond the Critical Period” which was published by McDonald (2006). The main reason for doing this is to understand this article and its objectives, reviewing its literature to evaluate its theoretical framework. This essay is organized into eight sections in order to achieve this objective by using critical evaluations. Language learning is considered to be different from other types of learning. Learning language is a complex process, yet every normal human being possesses the capacity to learn a language. The first language, or native language, is learned in childhood in a natural way, by exposure to it. The problem lies in learning a second language. Many linguists use the word “acquisition” in the case of the first language. Thus, the first language is “acquired” while the second language is “learned.” But recent research in the field has shown that the same principles underlie the learning of first and second languages. Much research has been done on second language acquisition, and several theories of SLA have been formulated. One such theory is the Critical Period Hypothesis. This research deals with the assessment of the Critical Period Hypothesis and tries to investigate whether any other explanation can account for certain exceptions to the Critical Period Hypothesis. According to this hypothesis, language acquisition must occur early in life if native-like mastery is to be achieved. After the critical period has passed, the age of acquisition is supposed to have no effect, and L2 learners repeatedly showed poor performance in oral and written grammaticality judgment tasks. Late L2 learners are much more likely to accept ungrammatical sentences as more grammatical than are native speakers or early L2 learners. The study was triggered by the observation that not all L2 data conform to a strict interpretation of Critical Period Hypothesis. For example, the age at which L2 performance begins to differ varies in different individuals. It is also determined by similarities and dissimilarities in L1 and L2. A few cases of high performing late learners were found, and studies have also found continuing effects of the age of acquisition on performance well beyond the critical period (Birdsong 1992, Flege et al. 1999); hence the researcher theorized that factors other than the age of acquisition may affect grammaticality judgment in L2 learners. 2.0 Objectives The chief objective of this study is to examine the possibility that difficulties in basic non-syntactic cognitive processes can contribute to poor performance in grammatical tasks, particularly in case of L2 learners. The study also tries to examine whether processing difficulties make it difficult to access and use relevant grammatical information. The study is confined to three types of processing difficulties: memory capacity, decoding ability, and processing speed. These processing difficulties are not grammar specific. An attempt has been made to find out the effect of these processing difficulties on a wide range of simple and complex structures. The age at which an individual is initiated into L2 can also affect L2 performance. . The research question for this study therefore is, whether a decline in language processing ability can provide an alternative explanation for poor L2 learner performance, rather than failure to acquire the grammatical properties themselves? 3.0 Assumptions This study takes into account certain linguistic assumptions, on which the theory is built. For example, it is assumed that people who learn a second language late in life generally perform more poorly on tests of L2 grammar than native speakers or early acquirers. Within subject variability is observed in late L2 learners’ performance, across different tasks and conditions. Variability also occurs on the same tasks across occasions. Studies have found continuing effects of age of acquisition on performance, well beyond the critical period. Few cases of high performing late learners have been found (Birdsong and Molis, 2001) Problems with basic processes can result in difficulties with grammatical processing. This is a non-modular approach which states that non-linguistic processing demands interact with syntactic processing, resulting in a greater impact of processing difficulties on more complex syntactic structures than on simpler ones. (The modular approach predicts no such interaction) A grammaticality judgment task involves many processes beyond grammatical processing. A connection exists between the three factors (working memory capacity, decoding ability and speed of processing) and affects performance in tests involving grammatical processing, even in case of native speakers. Adding stress causes speakers who have adequate knowledge of a particular grammatical construction to not be able to consistently apply it. Late L2 learners are slower at processing their second language than are native speakers. The grammaticality judgment test involved tokens of ten core properties of English and whether they are present or absent in the constructions: Articles Regular Past tense Third person singular Regular plural Present progressive tense Irregular past tense Irregular plural Wh- questions Yes/No questions Word order 4.0 Review of Previous Studies The researcher has thoroughly studied previous works on L2 acquisition and this research is based on a sound theoretical platform. Three kinds of previous studies have been reviewed: (i) those showing evidence for poor performance on grammaticality judgment tasks by late L2 learners. ( ii ) those arguing for and against a critical period for language acquisition. (iii) Those arguing that during language processing, both linguistic and non-linguistic information needs to be integrated, and this puts heavy demands on the processor. This last hypothesis is relevant to the study because it assumes that L2 learners have a generally reduced ability to integrate linguistic and non-linguistic information during processing. Previous studies have shown variability in L2 learners’ performance across different tasks and conditions and also on the same tasks across occasions. The researcher has reviewed studies on L1 acquisition and takes into consideration findings that even native speakers cannot access and apply grammatical knowledge under stressful conditions. 5.0 Method The researcher used the experimental method for this study. Two experiments were conducted. The researcher used three measures of language processing ability that are assumed to be independent of grammatical knowledge. The first measure is that of working memory span. In this the participants hear sets of 3, 4, 5 and 6 words which are concrete nouns referring to objects of different sizes. After hearing them they have to repeat them according to size, going from smallest to largest. Hence, if they hear goldfish, pig and needle in one set, they have to repeat needle, goldfish and pig. For testing the decoding ability, the participants hear words that have been broken down into sound segments called gates. The gates are presented one by one until the participant is able to identify the word. E.g. t,te, tel, tele, telev, television. For testing the speed at which people can process a language, participants hear a word followed by a sentence which contains it. They have to press the button as soon as they identify the word in the sentence. In the first experiment, the researcher compares performance on the language processing measures with performance on an auditory grammaticality judgment task by native speakers of English and L2 learners. In the second experiment, the researcher has tried to examine the effect of stressors on working memory, decoding ability and speed of processing ability of the participants. The conclusion is drawn on the basis of a comparison between the performance of L2 learners and stressed native speakers. 5.1 Experiment 1 In the first experiment a heterogeneous group of 50 late L2 learners having different L2 backgrounds performed tasks that measured their L2 working memory capacity, their L2 lexical decoding ability and their speed of L2 processing as well as L2 grammaticality performance. They had all moved to the US at age 12 or older, and had all been resident for at least one year. Most of them were university undergraduate students. The working memory capacity was measured by assigning a task in which the learners were given a list of words to remember. They had to re-order them in terms of size and repeat them from size smallest to largest. The decoding ability was tested by assigning a gating task. It involved breaking the words into small segments starting from the beginning of the word. Participants were presented each word gate by gate until they identified it. One practice item was followed by 10 multi-syllabic words. The words were always presented in the same order. The dependent measure calculated from the gating task was the average number of gates needed to identify the word correctly. Speed of processing was measured through a word detection task. Participants were given a target word which they had to identify in a sentence and push a button when they heard it. A total of 104 sentences were presented in which 50 were grammatical and 54 were ungrammatical. The word to be detected was a content word, other than the first or last word in a sentence. The sentences were presented in a random order to each participant, but the restriction was that each presentation half had an equal number of grammatical and ungrammatical sentences. The sentences were presented aurally and the participants responded by pressing a button on the computer. This enabled the recording of both accuracy and speed of response. The participants had to answer questions about their background, age when exposed to English, their mother tongue, previous exposure to English, duration of their stay in US etc. They also provided self ratings on their first and second language ability. 5.1.1. Results: Experiment 1 Table 1: Mean scores on language processing measures and the gjt compared Mean scores NS L2ers Sig diff? Working memory (no of sets of words recalled) 4.18 3.83 Yes Working memory (no of actual words recalled) 45.72 42.40 Yes Decoding ability (no of ‘gates’ required before identification) 3.86 4.26 Yes Processing speed (response latency in ms) (identification of a word in a sentence) 2296 2359 Yes Gjt accuracy (%) 98% 91% Only on one measure Gjt response latency (in ms) 3383 3591 Yes The L2ers are significantly worse on all measures than the NS (but only weakly so on judging the grammaticality of sentences) 5.2. Experiment 2: In the second experiment, 300 native speakers of English were divided into 6 groups according to different levels of stress put on their processing capacity while undertaking the judgment task. Group 1 – No stress was put on them Group 2 – They heard 4 digits followed by the sentence to be judged. Then they judged the sentence and recalled the digits. This was the low memory stress group. Group 3 – They followed the same procedure as followed by Group 2, but heard 7 digits because it was a high memory stress group. Group 4 – This group heard sentences through white noise. This was the noise stress group. Group 5 – They had to respond quickly, before the deadline. This was the response deadline group. Group 6 – They heard the sentences which were produced at twice the normal speed. This was the compressed speech group. In the second experiment, the aim was to study the effect of stressors on the performance of native speakers. Several previous studies have dealt with the effect of various types of stress on syntactic performance. Participants were required to remember 4 to 7 digits (for low and high memory load respectively) while judging sentences. Participants in the noise stress condition heard all sentences with an overlay of white noise. Participants in the response deadline stress condition had to respond within a given time limit. Participants in the compressed speech condition heard the sentences at twice the rate of speech. 5.2.1 Results The participants in the second experiment were native speakers and had excellent knowledge of English vocabulary and syntax. But adding a stressor affected their performance significantly. This poor performance was not due to lack of relevant grammatical knowledge. It indicates that grammatical knowledge cannot always be automatically and consistently applied under stress, that is, under situations involving processing difficulties. Experiment 2 Table 2: Comparison of the performance of the different NS groups on accuracy in judging grammaticality and response speed in doing so. NS groups by type of stress (50 in each group) No stress Low memory Stress High memory stress Noise Fast response Compressed Speech Gjt accuracy (%) 98% 98% 97%* 90%* 90%* 95%* Response latencies 3382 3719* 4030* 3460 2576* 2256* * means significantly different from the ‘no stress’ group All stressed groups, apart from the ‘low memory stress’ group, are significantly different on accuracy. McDonald then compares the accuracy performance on each property in the gjt of the NS group with the greatest stress effect – the noise group – with the accuracy performance of the L2ers: Table 3: Comparison of L2er and noise stressed NS accuracy in judging specific grammatical properties in % Property L2ers NS under noise stress Significant? Articles 82 74 No Reg past 87 90 No 3p –s 89 87 No Reg plur 90 90 No Pres prog 93 92 No Irreg past 91 94 No Irreg plur 87 99 Yes Wh-question 93 86 Yes Y/N question 93 93 No Word order 96 97 No The addition of noise to the task for NS appears largely to reproduce the effects of being an L2 speaker 6.0. Observations It can be concluded that stressful processing conditions make it difficult for native speakers to access and apply grammatical knowledge to heard sentences. The poor performance of late L2 learners on grammaticality judgment tasks can be attributed to processing, rather than lack of grammatical knowledge, if it is assumed that their L2 processing capacities are reduced in the same way as native speakers listening to the language under white noise. Thus we can conclude that processing difficulties, rather than age of acquisition is the driving force behind poor grammaticality judgment and performance in late L2 learners. Differing results in grammaticality judgment are caused due to difference in syntactic constructions. Poor grammatical performance by late L2 learners is actually at least partially a result of problems with accessibility of grammatical knowledge due to poor L2 memory capacity, lexical decoding and/or processing speed. Stressful processing conditions make it difficult to access and apply relevant grammatical knowledge. The paper adds to the literature that shows that general processing problems, rather than a deficit in some syntax specific process or module, offers a viable explanation for populations that exhibit poor grammatical performance. 7.0 Strengths and Weaknesses The study definitely has certain strengths. It has thoroughly reviewed the previous literature, and all assumptions are based on research findings. In order to better understand the effect of memory load on grammatical performance, this study tested more syntactic structures than previous studies have. Pilot tests were conducted to eliminate participants who missed 10% or more responses during trials, in order to ensure homogeneity of the group. To examine the effect of L2 heterogeneity, the same experiment was conducted on a group of 24 learners who all shared Spanish as their L1. For the high memory load stress test, the signal-to-noise ratio was adjusted during pilot testing. The research has tested the effect of a variety of stressors on a large set of different types of syntactic structures. This study tests the relative degree of vulnerability of ten different constructions across the groups. Native speakers were used in the second experiment to ensure that they had excellent knowledge of English vocabulary and syntax. Throughout the study, the emphasis is specifically on the fact that L2 memory capacity, L2 decoding, and/or L2 processing speed, which are thought to underlie L2 grammatical difficulties, are not the factors involved in non-linguistic tasks or the factors measured in L1. There is not the slightest diversification from the chief objective. At the same time, a few weaknesses affect the reliability and validity of this research. In the first place, the effects of only three cognitive processing difficulties—namely, L2 memory span, decoding ability, and L2 processing speed—have been studied. These are not the ONLY three cognitive processes responsible for grammaticality judgment. There are several others. The test for L2 decoding ability, which consists of gating tasks, does not seem to be convincing. Decoding syntactic structures requires other cognitive and linguistic abilities, which cannot be tested through gating tests that involve isolated words. Processing speed in individuals regarding linguistic, mathematical, or any other kind of mental task depends more often on logical sense and development of individuals than on any purely linguistic ability. The effect of the Critical Period Hypothesis in the case of L2 learners cannot be completely ruled out. 8.0 Conclusion The study deals with an important theory of second language acquisition. There are many instances that support the Critical Period Hypothesis, which emphasizes the age at which a second language is acquired. This theory has not only linguistic, but also biological evidence. But the researcher has taken into account previous research on second language acquisition and has pointed out several cases where the Critical Period Hypothesis does not hold true. With this as an assumption, the researcher has theorized that there can be an alternative explanation, which can also account for the exceptions to the Critical Period Hypothesis. He discusses the possibility that language acquisition is closely related to processing ability, which can be affected by stress; and even native speakers, who have competence in the language, are sometimes unable to access and use grammatical knowledge. The same phenomena were observed in L2 learners who could perform grammaticality judgment tasks offline or in supportive conditions. The researcher carried out two experiments on L2 learners and native speakers who were put under stress, and a wider variety of complex semantic structures was selected, unlike earlier experiments. On the basis of the observations the researcher succeeded in proving that the age factor may be important, yet language acquisition is closely related to processing ability. The processing ability may be affected by memory load or decoding capacity, in which case an individual is unable to access and use grammatical knowledge. The conclusions drawn by the researcher are convincing and in accordance with previous studies. Further evidence shows that the more complex structures require better processing ability. Hence performance varies according to different language structures. Performance is poorer in case of difficult structures. Evaluation One more potential confound in this study is the decision to use L2 learners from a variety of L1 backgrounds. It is known that where properties of the target language have equivalents in a speaker’s L1 (e.g. articles, inflectional tense/agreement marking) L2 learners can perform better than those whose L1 lacks these properties. Some L1 learners may have found some of the tasks simple because of similarities in their L1 and the target language. All the participants did not have this advantage. Hence, the study should be replicated with a homogenous group of L2 learners whose L1 lacks the relevant properties. References Birdsong, D., 1992. Ultimate attainment in second language acquisition. Language, 68 (4), 706-755. Birdsong, D. and Molis, M. 2001. On the evidence for maturational constraints in second-language acquisition. Journal of Memory and Language, 44 (2), 235-249. Flege, J., Yeni-Komshian, G. and Liu, S. 1999. Age constraints on second-language acquisition, Journal of Memory and Language, 41 (1), 78-104. Johnson, J. S. 1992. Critical period effects in second language acquisition: The effect of written versus auditory materials on the assessment of grammatical competence. Language Learning, 42, 217–248. Johnson, J. and Newport, E. 1989. Critical period effects in second language learning: The influence of maturational state on the acquisition of English as a second language. Cognitive Psychology, 21 (1), 60-99. McDonald, J. 2006. Beyond the critical period: Processing-based explanations for poor grammaticality judgment performance by late second language learners. Journal of Memory and Language, 55 (3), 381-401.  Read More