hce_cmu
109年
英文
第 47 題
📖 題組：
Part 2. During the past few decades, a group of second-language (L2) acquisition researchers have attempted to account for L2 ultimate attainment in terms of input processing problems. Among these scholars, Sorace (2004) posited that advanced L2 learners’ real-time processing of L2 input (such as spoken words) is often less rapid and less automatic than first-language (L1) speakers; consequently, L2 learners’ ability to access and coordinate multiple sources of knowledge from different domains may be computationally costly. In this regard, when the meaning of a word or grammatical form requires computing information from different linguistic (sub)domains (e.g., syntax, morphology or semantics), near-native speakers may not efficiently access all the required information for real-time interpretation. According to Liu (2009), this reduced efficiency in real-time processing may not significantly disrupt advanced L2 learners’ comprehension; but it may exert a negative impact on the retention of the comprehended information. In a similar vein, Juffs and Harrington (1995) and Hernandez et al (1994) both observed that highly advanced L2 learners performed as well as native speakers in rejecting ungrammatical L2 sentences, but with much longer response time. These scholars thus concluded that although near-native speakers are able to develop a fully amalgamated processing system to reduce cross-linguistic disparities, they were still unable to consistently understand L2 sentences using L2-specific processing strategies. Liu (2009) and Juffs and Harrington (1995) therefore both maintained that processing efficiency, rather than grammatical competence, is the source of difficulty that distinguishes a native speaker from a near-native speaker. To further explore if L2 learners, who are situated in an input-rich environment, are able to acquire a native-like processing system in the L2, McDonald (1987) recruited English-Dutch and Dutch-English bilinguals who learned the L2 at different ages, and had various amounts of exposure to the L2. McDonald found that both beginning and intermediate L2 learners, who had not yet had sufficient exposure to the L2, still utilized L1-based processing strategies while comprehending L2 sentences. As a result, these L2 learners still suffered from L1 interference. However, McDonald observed that both early and late L2 learners in the long exposure group resembled native speakers in terms of the linguistic cues they drew on when comprehending L2 sentences. McDonald concluded that L2 learners, irrespective of their onset age of learning the L2, are able to completely master L2 processing strategies given sufficient L2 exposure. McDonald’s view is not without challenge. Bassetti (2004) posited that when learning an L2 with a mature (L1) processing system already in place, adult L2 learners would need to merge the knowledge systems of their two languages. The more comprehensively the L1 and L2 systems are integrated, the more likely the resulting hybrid knowledge system can be consistently drawn upon to reduce the disparities in the decision space in L2 processing. However, Bassetti argued that the above scenario is extremely difficult, if not impossible, for L2 learners; even if it is possible, a full merger of L1 and L2 knowledge systems is not equivalent to two monolinguals housed in a single brain. Consequently, bilinguals would always exhibit subtle, non-native intuition and judgment in both their L1 and L2. Following the above view, complete native-likeness is not possible in L2 acquisition. Hitherto, researchers still differ in their views on the ultimate (grammatical) attainment in L2 acquisition. The inconclusive finding is attributed to the nature of the participants examined in existing studies. Birdsong (2009) contended that to make a strong case for L2 ultimate attainment, one needs to draw on empirical evidence from the most advanced L2 learners whose language development (in the L2) has reached the highest attainable state. Otherwise, one can always argue that L2 learners’ competence or performance may be subject to change given more exposure, time, and practice. Notwithstanding Birdsong’s suggestion, researchers continue to use various lenient or methodologically convenient criteria recruiting the so-called “advanced L2 learners,” with some using achievement tests (such as scores from final exams) and others using placement tests. The results of these tests can only shed light on L2 learners’ achievement in some limited language subdomains, rather than their overall proficiency profile. Without using stringent standardized “proficiency” tests (such as TOEFL, IELTS, or TOEIC), existing studies might have collected data from L2 learners whose L2 development is still in progress, rather than genuinely advanced L2 learners whose L2 has reached a plateau. Longitudinal data collected from the latter L2 learner population are warranted in order to shed more light on the developmental trajectory that L2 learners take and the upper limit that they can reach. Hitherto, much remains unknown about the attainability of a native-like processing system in L2 acquisition.
Part 2. During the past few decades, a group of second-language (L2) acquisition researchers have attempted to account for L2 ultimate attainment in terms of input processing problems. Among these scholars, Sorace (2004) posited that advanced L2 learners’ real-time processing of L2 input (such as spoken words) is often less rapid and less automatic than first-language (L1) speakers; consequently, L2 learners’ ability to access and coordinate multiple sources of knowledge from different domains may be computationally costly. In this regard, when the meaning of a word or grammatical form requires computing information from different linguistic (sub)domains (e.g., syntax, morphology or semantics), near-native speakers may not efficiently access all the required information for real-time interpretation. According to Liu (2009), this reduced efficiency in real-time processing may not significantly disrupt advanced L2 learners’ comprehension; but it may exert a negative impact on the retention of the comprehended information. In a similar vein, Juffs and Harrington (1995) and Hernandez et al (1994) both observed that highly advanced L2 learners performed as well as native speakers in rejecting ungrammatical L2 sentences, but with much longer response time. These scholars thus concluded that although near-native speakers are able to develop a fully amalgamated processing system to reduce cross-linguistic disparities, they were still unable to consistently understand L2 sentences using L2-specific processing strategies. Liu (2009) and Juffs and Harrington (1995) therefore both maintained that processing efficiency, rather than grammatical competence, is the source of difficulty that distinguishes a native speaker from a near-native speaker. To further explore if L2 learners, who are situated in an input-rich environment, are able to acquire a native-like processing system in the L2, McDonald (1987) recruited English-Dutch and Dutch-English bilinguals who learned the L2 at different ages, and had various amounts of exposure to the L2. McDonald found that both beginning and intermediate L2 learners, who had not yet had sufficient exposure to the L2, still utilized L1-based processing strategies while comprehending L2 sentences. As a result, these L2 learners still suffered from L1 interference. However, McDonald observed that both early and late L2 learners in the long exposure group resembled native speakers in terms of the linguistic cues they drew on when comprehending L2 sentences. McDonald concluded that L2 learners, irrespective of their onset age of learning the L2, are able to completely master L2 processing strategies given sufficient L2 exposure. McDonald’s view is not without challenge. Bassetti (2004) posited that when learning an L2 with a mature (L1) processing system already in place, adult L2 learners would need to merge the knowledge systems of their two languages. The more comprehensively the L1 and L2 systems are integrated, the more likely the resulting hybrid knowledge system can be consistently drawn upon to reduce the disparities in the decision space in L2 processing. However, Bassetti argued that the above scenario is extremely difficult, if not impossible, for L2 learners; even if it is possible, a full merger of L1 and L2 knowledge systems is not equivalent to two monolinguals housed in a single brain. Consequently, bilinguals would always exhibit subtle, non-native intuition and judgment in both their L1 and L2. Following the above view, complete native-likeness is not possible in L2 acquisition. Hitherto, researchers still differ in their views on the ultimate (grammatical) attainment in L2 acquisition. The inconclusive finding is attributed to the nature of the participants examined in existing studies. Birdsong (2009) contended that to make a strong case for L2 ultimate attainment, one needs to draw on empirical evidence from the most advanced L2 learners whose language development (in the L2) has reached the highest attainable state. Otherwise, one can always argue that L2 learners’ competence or performance may be subject to change given more exposure, time, and practice. Notwithstanding Birdsong’s suggestion, researchers continue to use various lenient or methodologically convenient criteria recruiting the so-called “advanced L2 learners,” with some using achievement tests (such as scores from final exams) and others using placement tests. The results of these tests can only shed light on L2 learners’ achievement in some limited language subdomains, rather than their overall proficiency profile. Without using stringent standardized “proficiency” tests (such as TOEFL, IELTS, or TOEIC), existing studies might have collected data from L2 learners whose L2 development is still in progress, rather than genuinely advanced L2 learners whose L2 has reached a plateau. Longitudinal data collected from the latter L2 learner population are warranted in order to shed more light on the developmental trajectory that L2 learners take and the upper limit that they can reach. Hitherto, much remains unknown about the attainability of a native-like processing system in L2 acquisition.
Which of the following statements can be inferred from the passage?
- A All the views in this passage are strictly based on theoretical deduction without any empirical evidence.
- B Before the methodological issue can be resolved, a clearer picture regarding the best attainable state in the L2 cannot be obtained.
- C Whether future research can produce fruitful insights depends crucially on analytical innovation.
- D Results from achievement tests provide a valid and complete picture of L2 learners’ overall proficiency profile.
- E All existing studies have seriously considered Birdsong’s suggestion.
思路引導 VIP
若要判斷一個研究結論是否能代表整個人群，我們通常會先檢查「受試者的選取標準」是否夠嚴謹。請回想文章最後一段：作者提到許多研究在定義『高階學習者』時，標準是否一致？如果衡量標準（如考試類型）本身存在差異或限制，這會如何影響我們對『學習極限』所下的最終定論？
🤖
AI 詳解
AI 專屬家教
恭喜你精準捕捉到了文章的核心論點！這道題目難度屬於 Medium 到 Hard 之間，主要考驗讀者能否從最後一段的學術檢討中，推導出作者對現狀的邏輯評價。你能選對 (B)，說明你不僅讀懂了表層的文字，還透徹理解了「研究方法」與「結論可信度」之間的因果連結。
研究方法與結論的因果聯繫
文章末段明確指出，現有研究之所以出現「不一致的結果」（inconclusive finding），是因為研究者在招募受試者時，使用了較為寬鬆或方便的方法，例如使用課堂小考（achievement tests）而非嚴謹的標準化能力測試（proficiency tests）。這導致我們無法確定受試者是否真的達到了語言發展的「最高狀態」。因此，只要這種**研究設計上的瑕疵（methodological issue）**未解決，我們就無法撥雲見日，看清第二外語習得究竟能達到何種極限。這正是選項 (B) 所傳達的深層含義。
▼ 還有更多解析內容