Download presentation
Presentation is loading. Please wait.
Published byNoah Wilson Modified over 9 years ago
2
Contrastive analysis: an Overview Raung-fu Chung Based on Thu Nguyen
3
Speech error analysis 1. Implosive stop 2. Unreduced vowel, raised vowel 3. Labialisation 4. Stop deletion, cluster simplification 5. Shortened [i:] vowel. No contrastive vowel length in V. 6. Checked stop substitution 7. Stopping, checked stop substitution
4
Linguistic explanations for L2 phonological systems Why certain L2 sounds are easier to acquire than others? What is the role of L1 transfer in L2 sound acquisition? What kinds of L2 sounds/features are more problematic/ more difficult to acquire? Why? Apart from L1 transfer, what other linguistic factor(s) involves in L2 sound acquisition?
5
Linguistic explanations for L2 phonological systems Major theories Contrastive Analysis (Lado, 1957) Speech learning model (Flege, 1995) Similarity Differential Rate Hypothesis (Major and Kim, 1994) Universals: Markedness Differential Hypothesis (Eckman 1977, 1996) Ontogeny Phylogeny Model (Major 2001)
6
1. Transfer and Contrastive analysis The structural systems of the L1 and L2 should be compared and contrasted to determine the areas where they differThe structural systems of the L1 and L2 should be compared and contrasted to determine the areas where they differ Errors can be predicted to occur mainly with the areas of the L2 that differ the most from the L1Errors can be predicted to occur mainly with the areas of the L2 that differ the most from the L1 Language teaching should be focused on the areas where the L1 and L2 systems differ the mostLanguage teaching should be focused on the areas where the L1 and L2 systems differ the most
7
Predictions L2 sound differs from L1, or not exist in L1: difficult –English / ɵ / /æ/ for Chinese. Chinese / u_ / for English learners L2 sounds similar to L1: easy to transfer –English /m/ and /n/ for Chinese Habit formation, transfer of L1 articulatory habit
8
Transfer and Contrastive analysis (cont.) Strong version: Errors, difficulty, and success can be predicted on the basis of a contrastive analysis of the L1 and L2Strong version: Errors, difficulty, and success can be predicted on the basis of a contrastive analysis of the L1 and L2 Weak version: An analysis of the sources of learners’ errors can be facilitated by contrasting the L1 and L2Weak version: An analysis of the sources of learners’ errors can be facilitated by contrasting the L1 and L2
9
Problems of CAH Predictions based on a contrastive analysis often are not borne out in actual learner dataPredictions based on a contrastive analysis often are not borne out in actual learner data Errors that are not predicted are often more numerous than errors that areErrors that are not predicted are often more numerous than errors that are The L1 is not the sole—usually not even the primary—source of errorsThe L1 is not the sole—usually not even the primary—source of errors Language learning is not just habit formation (demise of behaviorism)Language learning is not just habit formation (demise of behaviorism)
10
Problems of CAH, BUT: It is usually similarities and not differences that give rise to incorrect associations between the L1 and L2, resulting in errorsIt is usually similarities and not differences that give rise to incorrect associations between the L1 and L2, resulting in errors Different types of similarities and differences represent different levels of learning problemsDifferent types of similarities and differences represent different levels of learning problems
11
Moderate version of CA: similar phenomena are harder to learn than dissimilar phenomena. The psycholinguistic reason: similar sounds are harder to acquire because gross differences are more often noticed, due to perceptual saliency, whereas minimal differences are less likely to be noticed, resulting in non-learning.
12
Speech Learning Model - Flege Learner’s ability to perceive the sound determined the difficulty of acquisition The ‘equivalent’ or ‘similar’ sounds are difficult to acquire because a speaker perceives and classifies them as equivalent to those in the L1 and no new phonetic category is established Whereas ‘new’ (dissimilar or different) sounds are easier to learn because the speaker perceives these differences and establishes new phonetic categories.
13
Speech Learning Model 2 beginners perceptually assimilate most L2 categories to native ones; if L2 segment is sufficiently dissimilar - a new L2 perceptual category is established over time the new category formation may be blocked by equivalence classification for less dissimilar sounds, so a single perceptual category subsumes both L1 & L2 segments, leading to persistent accented production in L2 or even to shifts in L1 production
14
Which criterion is adopted for ‘similar’/‘new’ phones? A ‘new’ phone is "an L2 phone which does not have a counterpart in L1, and may therefore not be judged as being a realization of an L1 category". Examples: French /y/ for English learners. English /æ/ /θ/ for Chinese Chinese / ɯ, i/ for English learners A ‘similar’ phone is "an L2 phone which is realized in an acoustically different manner than an easily identifiable counterpart in L1". Examples: French /t/ and /u/ for English learners. English /b/ /d/ for Taiwanese
15
Some evidence supporting the SLM Flege (1993): Americans learning French/u/ and /y/: advanced learners produced /y/ authentically (the new/dissimilar sound), but produced /u/ unauthentically (the similar/equivalent sound) Bohn & Flege (1992): German speakers of English didn’t produce the similar English sound/i, I, e/ authentically but produced the dissimilar sound /æ/ authentically.
16
Problems: Definitions of similar and dissimilar are not always clear-cut Some exceptional findings: German speakers of English (Bohn and Flege 1992) and Korean learners of English (Kim, 1994): actually did better with the similar sounds
17
The Similarity Differential Rate Hypothesis So what does easier or harder mean? Major and Kim(1996): the notion of “difficulty” is indeed the wrong notion, their claims about similarity and dissimilarity concern rate: The Similarity Differential Rate Hypothesis: dissimilar phenomena are acquired at faster rates than similar phenomena, and markedness slow rate Evidence: English / ʤ /and /z/ for Korean learners –/ ʤ / : similar: produced better /easier at initial stage –/z/: dissimilar: reach native target more quickly
18
Remark: Flege’s SLM (1995) implies that L1 persists for similar sounds but did not address the role of other language universal. Major’s SDRH claims the rate is slower for similar sounds also because transfer persists. However, neither SLM nor SDRH address the role and proportion of universal in relation to L1.
19
The Markedness Differential Hypothesis(Eckman (1977, 1996) The areas of difficulty that a language learner will have can be predicated on the basis of a systematic comparison of the grammars of the NL, the TL, and the markedness relations stated in universal grammar
20
The Markedness Differential Hypothesis (cont.) Marked: less frequent in languages of the world and implicational hierarchy, unmarked: more frequent. Only those parts of the L2 grammar which are more marked than the L1 will cause difficulty and the more marked they are, the more difficult they will be. Unmarked phenomena are acquired before marked phenomena
21
The Markedness Differential Hypothesis(cont.) statistical frequencies: The type of /r/ in American English (IPA [r]) is more marked than the type of /l/ in American English: in the languages of the world, [r] represents only 5.6% of the liquids but [l] 42.6% Implicational hierarchy: [st], [sn], [sl] Sonority hierarchy: [st]>[sn]>[sl]: English [st]: more marked than [sn] and [sl] and thus induces more vowel epenthesis from Spanish learners
22
Ontogeny Phylogeny Model (Major 2001) Interaction of Transfer with Universals Combination of similarity, universal grammar (including markedness) and transfer into one model if x (when compared to a sound in L1) is more similar and more marked than y, then x will be acquired after or at a slower rate than y. Interlanguage = parts of L1 +parts of L2 + parts of U that are not already part of L1 and L2.
23
Ontogeny Phylogeny Model Chronological Corollary of the OPM. IL develops chronologically in the following manner: (a) L2 increases, (b) L1 decreases, and (c) U increases and then decreases. Similarity Corollary of the OPM. In similar phenomena, IL develops chronologically in the following manner: (a) L2 increases slowly, (b) L1 decreases slowly, and (c) U increases slowly and then decrease slowly. Thus, the role of L1 is much greater than U, compared to less-similar phenomena. Markedness Corollary of the OPM. In marked phenomena, IL develops chronologically in the following manner: (a) L2 increases slowly, (b) L1 decreases and then decreases slowly, and (c) U increases rapidly and then decreases slowly. Thus except for the earliest stages, the role of U is much greater than L1, compared to less-marked phenomena.
24
StagesNormal phenomena Similar phenomena Marked phenomena Earlier stages Later stages Earlier stages Later stages L2 acquired L1 dominates L1 decreases U minimal U increases U decreases L2 acquired slowly L1 dominates L1 decreases slowly U minimal U increases slowly U decreases slowly L2 acquired slowly L1 dominates L1 decreases L1 decreases slowly U minimal U increases rapidly U decreases slowly
25
Some examples 1.Korean L1-L2 English: Korean has /p/ but no /f/. The learner substitute [p], then [p] and [Φ] (neither English nor Korean) and then finally [f] 2. L1 Hawaiian-L2 English. Hawaiian has only voiceless obstruents /p k ? h/ and CV syllable structure Big: [bik] ->[biki] ->bigi ->[big] 3. L1 Vietnamese –L2 English. Vietnamese only final unreleased voiceless stops /p t k/ tent: [ten_] -> [tent ʰ ə]->[tent ̚ ]
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.