PHONOLOGICAL PERCEPTION AND PROCESSING IN A SECONDLANGUAGE

MIN WANG and CHUCHU LI

University of Maryland, College Park

Phonological perception and processing is not only important in spoken language recognition and comprehension, but also important in the development of reading ability. This review paper examined phonological perception and processing in a second language (L2) among infants, school-aged children, and adults. Infants are born with the ability to discriminate universal sets of phonetic distinctions, but this ability may decline or be lost later on due to lack of corresponding linguistic experience. For both children and adults, their L2 phonological processing skill is influenced by their L1 knowledge and not only L1 phonology but also the orthography impacts the L2 phonological processing. L1 phonological skills contribute to the development of L2 reading skills over and above the within-L2 phonological skills. Limitations and future research direction are discussed.

INTRODUCTION

Phonological perception refers to the process by which the sounds of a language are heard, interpreted and understood. Phonological processing refers to the use of phonological information in processing written and spoken language (see Wagner and Torgesen 1987 for a review). Previous studies have investigated phonological perception via examining listeners’ ability to discriminate speech sounds in terms of phonetic categories, for example, researchers have examined whether listeners are able to discriminate voiced and voiceless stops, for example, /b/ and /p/ (Werkeretal. 1981; Wilson 1987; Yeung and Werker 2009). Previous research on phonological processing included a large number of studies that investigated phonological awareness; phonological recoding in lexical access and phonetic recoding to maintain information in working memory. Phonological awareness is a well-studied topic among both monolingual and second language (L2) learners. Phonological awareness refers to an individual’s ability to reflect on, think about and manipulate linguistic sounds at different levels, such as syllable level, onset-rime and phoneme level. According to Treiman and her colleagues’ linguistic structure hypothesis (Treimanetal. 1995), the syllable is at the top of the hierarchical structure-it is the largest and most accessible unit. The phoneme is at the bottom of the hierarchical structure-it is the smallest unit and is a later-developing one for children. Between syllables and phonemes lie intermediate onset and rime units.

Phonological perception and processing is not only important in spoken language recognition and comprehension, but also important in the development of reading ability. For example, a large volume of literature has shown that phonological awareness including both large and small units is an important and reliable predictor of children’s later reading ability (Bryantetal. 1990; Ehrietal. 2001; Vellutino and Scanlon 2001; Wagner and Torgesen 1987). Research on phonological perception and processing will enhance our understanding about the underlying mechanism as to how phonological information is acquired, hence helping us develop better instructional approaches to improve second language (L2) learning.

According to the data ofEthnologueorganization(http:∥www.ethnologue.com), there are 6909 distinct languages in the world. Each language has its own phonological characteristics. For example, languages such as Chinese and Thai are tonal language, whereas English and Dutch are not. In Japanese, two phonemes, /r/ and /l/, are not distinguished whereas they are in some other languages such as English. Russian distinguishes soft and hard consonants such as /mj/ in мясо /mjasa/ and /m/ in мыло /mila/ based on whether they are pronounced with palatal secondary articulation, whereas English does not. Knowledge about phonology is important in the acquisition of a language. Phonology consists of segments and suprasegments. A segment refers to any discrete unit that can be identified in the stream of speech, such as consonants and vowels. In contrast, a suprasegment, such as pitch or stress, is defined as a vocal effect which extends over more than one segment (Crystal 2008).

Segmental information conveys differences in meaning between words in a language. However, in some languages, suprasegmental information carries lexical function as well. For example in English, both the phonemic information (a type of segmental information) inrebelsuch as /r/ /b/ /e/, and the stress information (i.e. the suprasegmental information) are important. The stress information, whether it falls on the first or second syllable, determines whether the wordrebelis a noun or a verb. In Chinese, in the syllablemǎ,maprovides the syllable segmental information, but it is the tone marker above the vowel (i.e. the suprasegmetnal information) that determines that the syllable meanshorseinstead ofmother. In sum, both segmental and suprasegmental information plays an important role in phonological perception and processing.

L2 PHONOLOGICAL PERCEPTION AND PROCESSING IN ADULTS

There are three sections in the review of L2 phonological perception and processing in adults. The first section focuses on adult L2 learns’ phonological perception. Studies that investigate adult learners’ performance on discriminating and categorizing L2 phonemes are reviewed (Wilson 1987; Navarraetal. 2005). The second section discusses the influence of L1 on adult learners’ performance on L2 phonemic perception. Studies that showed evidence of language transfer in phonological perception are reviewed and discussed. The third section discusses research about L2 phonological processing in adults, studies that examined adult learners’ L2 phonological awareness are reviewed and discussed.

L2 phonological perception

ExperimentaltasksinL2phonologicalperception

Various tasks have been used to investigate non-native speakers’ ability to discriminate non-native phoneme contrasts, and the most common two tasks are 1) the phoneme categorization task which requires participants to make explicit judgments regarding whether a phoneme belongs to a particular category and 2) an AX discrimination task which requires participants to judge whether two auditory sounds are same or not. Using both tasks, Pallier, Bosch, and Sebastia′n-Galle′s (1997) showed that native Spanish speakers perform worse on discriminating the Catalan /?/-/e/ contrast than native Catalan speakers. In a phoneme categorization task, both native Spanish and Catalan speakers were introduced two Catalan words:pereandpera. The first vowel ofpereis /e/ whereas the first vowel ofperais /?/. Subsequently, they were presented with a series of seven vowel stimuli (s1-s7) along a continuum from /e/ to /?/ in a randomized order, and were required to judge whether a particular auditory stimuli sounded more like the first vowel inpereorpera. Catalan group was able to map the stimuli onto the two categories, whereas Spanish group performed around the chance level. In an AX discrimination task, both groups were presented with phoneme pairs and the phonemes were selected from the stimuli in the phoneme categorization task. Participants were required to judge whether the pair of the phonemes were same or different. Only the Catalan group showed a peak in their discrimination performance for the s3-s5pair, which suggested that the boundary that led them to divide the continuum into two phonemic categories fell between s3 and s5. In contrast, the Spanish group did not show a salient boundary.

Considering that explicit judgment task may not be able to reflect listeners’ potential unconscious sensitivity to the L2 phonological contrasts (Sebastia′n-Galle′setal. 2005), Navarraetal. (2005) contrasted native Catalan speakers’ and native Spanish speakers’ performance on Catalan /e/ and /?/ contrasts via an implicit task with the speeded classification paradigm. Participants were presented with two di-syllabic stimuli of which the second syllable contained the /e/ - /?/ contrasts or no variation, and were required to decide whether the two stimuli were same based on the first syllables of the two nonwords, and the first syllable were /pu/ or /ti/ in the task. Spanish group did not show significant difference no matter whether the 2ndsyllable differed from each other, whereas Catalan group responded in a significantly slower speed in the lists where the 2nd syllable contained the target contrast from trial to trial even if the task only required them to make the judgment on the first syllable. This result suggests that the /e/ - /?/ discrimination was detected by Catalan group and influenced their response implicitly.

NeuroimagingresearchonL2phonologicalperception

In addition to behavioral data, neuroimaging data also revealed participants’ ability to discriminate phonemic contrasts. For example, N??t?nen (2001) showed that listeners with different discriminating abilities showed different mismatch negativity (MMN) in a phonemic discrimination task. MMN is a negative component of the auditory event-related potential (ERP) which is usually elicited by a distinguishable change when a listener is exposed to a series of ongoing auditory stimulation. The better the listeners are able to detect phonemic contrasts, the larger MMN is elicited.

Recent research on phonemic discrimination examined the brain areas that may play a role and found that motor cortex is involved in phonological perception (D’Ausilioetal. 2009; M?tt?nen and Watkins 2009; Wilson 2009). Using repetitive transcranial magnetic stimulation (rTMS) technique, M?tt?nen and Matkins (2009) showed that motor cortex contributes to listeners’ phonological perception. Native English speakers were required to perform phoneme identification and discrimination tasks before and after the rTMS training, which temporarily disrupt the lip representation in the left primary motor cortex. The phoneme contrasts used in the study are the continua of /ba/-/da/, /ka-/ga/, /pa/-/ta/ and /da/-/ga/. After rTMS was conducted, participants’ performance on the contrasts which involve lips in their articulation (i.e. /ba/-/da/ and /pa/-/ta/) was significantly worse than that before rTMS was used. However, participants’ performance on the contrasts that do not require lips in the articulation (i.e. /ka/-/ga/ and /da/-/ka/) was not negatively influence. These results suggested that the motor circuits which are in charge of the mapping between sounds and motor representations are important in phonological perception. Using similar design and technique (i.e. TMS), D’Ausilioetal. (2009) achieved the similar results that listeners’ performance on discriminating lip-articulated phonemes was negative influenced by disrupting the lip representation in the left primary motor cortex, and that their performance on differentiating tongue-articulated phonemes was negatively influenced by disruption the tongue representation in their left primary motor cortex.

Navarra and Soto-Faraco (2007) showed that articulatory information also helps L2 learners to discriminate non-native phoneme contrasts. Native Spanish speakers showed difficulty in discriminating Catalan sounds /?/ and /e/ when only auditory stimuli were provided. However, combining visual information about the articulatory gestures (e.g. lip movements) did enhance their ability to discriminate this contrast. Perhaps the activation of motor cortex is involved in the articulatory gestures, which is consistent with M?tt?nen and Watkins (2009)’s finding that motor cortex plays a role in phonological perception in English.

SupragementalperceptioninL2

Studies focused on suprasegments showed similar findings that adults have difficulties in phonological perception in L2. For example, French is a language in which stress can only be on the last syllable of a word, whereas stress can be used to discriminate different lexical items in Spanish. Researchers found that native French speakers have difficulties differentiate auditory nonwords that differ only in stress location whereas Spanish speakers did not have from these difficulties (Dupouxetal. 1997; Tremblay 2008, 2009). In an ABX tasks, participants were provided two auditory stimuli (A and B) which differed only in stress location and were then presented a third sound. They were required to decide whether the third sound was same as A or B. Compared with Spanish speakers, French speakers encountered more difficulties in doing the task. In a phoneme-based ABX task, it was difficult for Spanish participants to ignore the different stress position of auditory stimuli, whereas French participants were not influenced by the stress position significantly. In summary, it has been acknowledged that, adult L2 learners have difficulties in phonological perception in their L2 when the phonological features do not occur in their native languages.

TrainingeffectsofL2phonologicalperception

Although L2 phonological perception is difficult for L2 learners sometimes, appropriate training is able to improve L2 learners’ performance (Werkeretal. 1981). For example, Logan, Lively, and Pisoni (1991) found that Japanese adults were able to learn to discriminate the English /r/-/l/ distinction after appropriate training. A group of Japanese listeners were trained to discriminate the English /r/-/l/ distinction over three weeks. During the training, they were asked to distinguish /r/ and /l/ via listening to multiple natural exemplars which contrast /r/ and /l/ from a variety of context. Specifically, the /r/ or the /l/ was in different positions in a syllable, and the stimuli were recorded by multiple speakers. Participants showed small but reliable improvement in discriminating /r/ and /l/ after the training. Bradlowetal. (1999) extend above findings to long-term retention of learning. Japanese trainees performed on a perceptual identification task on English /r/ and /l/ contrasts prior to, immediately after, and three months later after the training. Participants did not only show a reliable improvement after the training session, but also maintained the improvement in the 3-months follow-up test. Escuderoetal. (2011) showed that Spanish learners of Dutch performed significantly better on discriminating Dutch contrast /ɑ/ and /a/ after receiving the enhanced bimodal distributional training. During the training, a series of sounds along the acoustic continuum that comprises the two categories that are to be trained (e.g. /ɑ/ and /a/) were provided, among which the tokens near the end points of the continuum were provided most frequently. Listener’s performance on discriminating /ɑ/ and /a/ in ABX tasks was significantly improved. The results suggested that laboratory training is able to improve L2 learners’ performance on discriminating L2 phoneme contrasts and the training effect is long-lasting.

Nevertheless, training is not always successful. For example, Wanrooijetal. (2012) found the enhanced bimodal distributional training present no effect on improving native Dutch listeners’ ability to discriminate English contrast // - /ε/. MacKainetal. (1981) suggested that effective training has to be extensive and naturalistic. For example, Lively, Logan, and Pisoni (1993) found that rich phonetic environment and talker variability is important in the training. Consistent with above statement, Strange and Dittman (1984) found that extensive training without rich phonetic environment and talker variability only improved Japanese speakers’ ability to discriminate /r/ from /l/ in a laboratory task but not in natural speech. In addition, some studies suggested that adult L2 learners may still fail to achieve native-like levels of performance on discriminating L2 speech sounds even if the training is quite helpful and successful (see Polka 1991; Werker 1994). Loganetal’s(1991) results have supported this statement-the improvement was small (less that 10%) even after three weeks training, despite the fact that the enhancement was reliable.

In summary, L2 adult learners have difficulties in phonological perception in their L2. It is difficult for listeners to discriminate phonemic contrasts in L2 if the contrasts do not occur in their L1. In terms of suprasegments, it is difficult for them to discriminate minimal pairs which only differ in stress location if such pairs do not occur in their L1. Although appropriate training may be able to help with adult learners’ L2 phonological perception, it is challenging to reach native-like level. Recent studies have focused more on investigating the underlying brain activities corresponding to L2 phonological perception. Future neuro-imaging data may help us learn more about underlying mechanism of adults’ poor phonological perception in L2.

The influence of L1 on L2 phonological perception

Previous research has shown that adults’ performance on L2 phonological perception is strongly influenced by experience with the phonological system of their L1 (Abramson and Lisker 1970). A traditional explanation of this phenomenon is that a native-language “phonological filter” influences the perception of non-native phonological information (Polivanov 1931). This filter makes the discrimination between non-native contrasts which do not occur in native language difficult for mature listeners. In other words, non-native speakers’ poor performance on perception of L2 phonemic contrasts can be attributed to their previous linguistic experience of their native language. That native Japanese speakers’ poor performance on discriminating English /r/ and /l/, that native English speakers’ poor performance on discriminating Hindi /d/ and /t/, and that native Spanish speakers’ poor performance on discriminating Catalan /e/ and /?/ can all be explained by L1 phonological filter. Interestingly, Polka (1992) found that it was difficult for L2 learners to discriminate an unfamiliar L2 phonemic contrast even if an analogous contrast exists in their native language. Native English speakers were not be able to discriminate Salish uvular versus velar glottalized stop contrast because they were not exposed to this contrast in their native language. However, native Farsi speakers did not performed better on discriminating this contrast even if an analogous contrast occurs in Farsi. Conversely, Best and McRoberts (1988) found that adult non-native speakers were able to discriminate some non-native contrasts quite well, even without prior exposure to these contrasts or to their analogous contrasts- American English listeners were able to successfully discriminate Zulu click consonants, despite the fact that they had never been trained to do so. Werkeretal. (1981) showed that, compared with the place-of-articulation contrast, the Hindi voicing contrast is easier for native English listeners to discriminate. In summary, adult L2 learners do not have the difficulties to the same degree when discriminating non-native phonetic contrasts which do not occur in their L1.

TheoreticalmodelsofL1effectsonL2phonologicalperception

Some researchers proposed different models to explain the variability. For example, Best explained the variability via the Perceptual Assimilation Model (PAM) (Best 1994a, 1994b, 1995), Flege proposed Speech Learning Model (SLM) (Flege 1995), and Kuhl proposed Native Language Magnet model (NLM) (Iverson and Kuhl 1995; Kuhl 1991). All of these models agreed that the experience of native language contributes to L2 learners’ perception of non-native phoneme contrasts, but they differ in how the L2 learners build perceptual framework of their native languages. PAM focused on na?ve non-native listeners and explained L1 influence on non-native phonemic perception based on perceptual learning of phonetic-articulatory pattern (Best and Tyler 2007). The model suggests that non-native speakers will perceptuallyassimilatethe nonnative phoneme to the most articulatorily similar native phoneme. L1 influence occurs at both phonological and phonetic level. SLM is more focused on L2 learners and discusses L1 influence at phonetic level. The model suggests that the more dissimilar L2 phonemes are from the closest native phonemes, the more easily they can be acquired and produced accurately. NLM discusses the issue of L2 phonemic acquisition at acoustic level instead of phonetic level. According to NLM, L1 phonemes serve as a prototype, just like a magnet, and the nearby perceptual space is “shrunk”. As a result, it is difficult for listeners to discriminate phonemes which fall around the prototypes than to discriminate those that are around the non-prototypes. It is difficult for L2 learners to form prototypes of the categories of the L2 phonemes because of the lack of acoustic experience, thus making it difficult for them to discriminate between the non-native phonemes. There is no agreement in the literature as to which model can best explain L2 learners’ various performance on non-native phonological perception.

EffectsofL1orthographyonL2phonologicalperception

Recent researchers examined other factors that may play a role in L2 phonological perception in addition to L1 phonology. One important factor, L1 orthography, has been shown to influence the perception of L2 phonology (Weber and Cutler 2004; Escudero and Wanrooij 2010). Cutleretal. (2006) suggested that orthographic knowledge influences lexical representation, which ultimately influences phonological perception. For example, Dutch listeners have difficulties discriminating English // and /?/ contrast, however, Weber and Cutler (2004) still showed that they performed differently when hear these two sounds, and the difference was largely due to their L1 orthographic knowledge. Using the eye-tracking paradigm, they found native Dutch speakers tend to look at the picture of a pencil when they hear the first syllable of the English word panda (p//nda whereas they are less likely to look at the picture of a panda when they hear the first syllable of the word pencil (p/?/ncil). According to Cutleretal.’s suggestion, Dutch participants may create lexical representations that only letterebut not letteracan be pronounced as /?/, perhaps due to the fact that the lettereis pronounced similarly in Dutch L1 and English L2 whereas the letterais pronounced quite different in these two languages. Escudero and Wanrooij (2010) supported above interpretation by indicating that when discriminating Dutch phoneme contrasts, the pairing of auditory stimuli with orthographic labels can influence Spanish listeners’ performance. Native Spanish speakers were asked to discriminate six Dutch vowels, which were /ɑ/, /a/, /y/, /y/, // and /i/ in an auditory task with an ABX paradigm. They were also required to match a sound with one of the following six orthographic presentationsaa,a,ie,i,uuandu. Spanish participants’ performance differed across the two tasks: the most difficult phonemes in the auditory task were the easiest in the orthographic task and the phonemes which are more likely to be perceived accurately in the auditory task were relative poorly judged in the orthographic task.

本文基于模糊控制原理，將Elman網(wǎng)絡(luò)與小波變換相結(jié)合，針對(duì)隧道開(kāi)挖過(guò)程中位移的動(dòng)態(tài)變化和非線性，提出新的預(yù)測(cè)模型。通過(guò)小波將監(jiān)測(cè)隧道位移變形序列分解成具有不同屬性和頻率的子序列，采用Elman神經(jīng)網(wǎng)絡(luò)對(duì)子序列進(jìn)行預(yù)測(cè)，建立起網(wǎng)絡(luò)，疊加預(yù)測(cè)結(jié)果、計(jì)算殘差并測(cè)試網(wǎng)絡(luò)精度，利用模糊控制方法對(duì)殘差進(jìn)行控制，考慮前向已檢測(cè)樣本的步驟及誤差值，控制輸出后續(xù)待檢測(cè)樣本的預(yù)測(cè)值和誤差，修正后得到小波變換下的模糊控制Elman網(wǎng)絡(luò)位移預(yù)測(cè)模型。

In summary, the absence of some phonemic contrasts in L1s is an important cause for adult learners’ difficulties in phonological perception in L2. In addition, L1 orthographic knowledge also plays an important role because it influences the formation of phonological representation. Several models have been put forth to explain how L2 learners acquire their L2 phonological representation taking into account L1 phonological knowledge. Furthermore, L1 orthography is an important factor in studying L2 phonological perception. Nevertheless, previous models have not taken this into consideration.

L2 phonological processing in adults

The investigation of phonological awareness is an important aspect in the research about phonological processing. Phonological awareness has been shown to be a good predictor of reading performance in English L1 studies (Jueletal. 1986; Bryantetal. 1990; Ziegler and Goswami 2005). Research has shown that levels of phonological awareness in L1 and L2 are different for the same individuals (Saiegh-Haddad and Geva 2010; Russak and Saiegh-Haddad 2011). For example, Russak and Saiegh-Haddad (2011) showed that among Hebrew-English adult bilinguals, phonological awareness of their L1 is better than that of their L2. Three tasks were conducted in the study. In the phoneme isolation task, participants were required to isolate a particular phoneme from an auditory word, and articulate it. In the full segmentation task, participants were asked to segment a word into its individual phonemes and articulate the phonemes correctly. In the phoneme deletions task, participants were required to remove a particular phoneme from an auditory word and articulate the remaining part of the word. In all of these three tasks targeting phonological awareness, the Hebrew-English bilinguals performed better in Hebrew (L1) than in English (L2), although the difference in full segmentation task did not reach significance. Lexical Restructuring Model (Metsala and Walley 1998) suggested that phonological awareness is related to individuals’ vocabulary knowledge. L1 words are better represented than L2 words given learner’ familiarity and more practice with the L1 words.

EffectsofL1orthographyonL2phonologicalprocessing

Similar to the influence of L1 orthography on L2 phonological perception, L1 orthography also influences L2 phonological processing. Koda (1989) suggested that L1 orthographic representation has impact on L2 phonological coding strategies. Four groups of participants with different L1 backgrounds were included: Arabic, Spanish, Japanese and English. Unlike the other three groups whose native languages are alphabetic, Japanese group’s L1 is logographic (i.e. Kanji). The former three groups were all English L2 learners. A short-term memory recall task was conducted in which participants were require to recall pronounceable letter string (e.g. CACE) and unpronounceable letter string which violated English orthographic rules (e.g. XJWZ). Unlike other three groups, Japanese participants performed better in recalling unpronounceable strings than pronounceable strings. It seems that the violation of English orthographic rules produced less negative impact on Japanese participants than the other three groups when encoding the phonological information of the stimuli. Doctor and Klein (1992) showed that spelling-sound correspondences of bilingual speaker’s both languages contribute to the recognition of a letter string. A group of balanced English-Afrikaans bilinguals were administered an inclusive lexical decision task (i.e. press the “yes” button if a letter string is a real word in either language), and it took them longer to make judgment on interlingual English-Afrikaans homophones (e.g. LAKE-LYK) than interlingual homographs or words exclusive to either English or Afrikaans.

Wangetal. (2003) investigated the influence of L1 orthographic knowledge on L2 reading performance among native Chinese and Korean readers. English proficiency was matched between the two groups. This is an important control which was not given enough attention in previous research. The ESL learners’ relative reliance on phonological and orthographic information was examined. Van Orden’s (1987) semantic category judgment task was selected to test the involvement of orthography and phonology in reading for meaning. In the semantic category judgment task, the participants were asked to judge whether the target word is a member of a category, for example, whetherrowsis aflower, or whethertowistheendofyourfeet). By varying the phonological and spelling similarity of the target words to the category exemplars, the use of phonological and visual-orthographic information in second language learners with a nonalphabetic L1 background was examined. The second task was a phoneme deletion task. This task requires phoneme deletion in an English word, followed by a spelling of the new word that results from the deletion. The uniqueness of this task is that the deletion of the required phoneme in the word leads to a new word with a different spelling form from the original one (e.g. removing the /t/ sound frommightcreate a wordmywhich has a distinct spelling frommight. This feature requires the participants not only to manipulate the individual phonemes in the word, but also to accurately access their spelling knowledge of the new word.

In the semantic category judgment task, the Korean ESL learners made more false positive errors in judging stimuli that were homophones to category exemplars than they did in judging spelling controls. However, there were no significant differences in responses to stimuli in these two conditions for Chinese ESL learners. Chinese ESL learners, on the other hand, made more accurate responses to stimuli that were less similar in spelling to category exemplars than those that were more similar. Chinese ESL learners may rely less on phonological information and more on orthographic information in identifying English words than their Korean counterparts. Further evidence supporting this argument came from the phoneme deletion task, in which Chinese subjects performed more poorly overall than their Korean counterparts and made more errors that were phonologically incorrect but orthographically acceptable. The researchers suggest that cross writing system differences in first languages and first language reading skills influence could be responsible for these ESL performance differences.

L2 phonological processing can also be influenced by some other factors such as L2 proficiency and whether the speakers had just used their L1 before carrying out the L2 tasks (Jared and Kroll 2001). In a naming task, participants named a block of English experimental words, a block of French filler words, and then a second block of English experimental words. The target English words included three groups: the no-enemies group, the French-enemies group, and the English-enemies group. The non-enemies group were words whose word bodies are consistent in English but do not exist in French (e.g.bump). For the French-enemies words, their word bodies are consistent in English (e.g.bait) but are pronounced in multiple ways in French (e.g.faitandlait). For the English enemies group, their word bodies are pronounced in multiple ways in English (e.g.beadanddead) and do not exist in French. French-English bilingual speakers’ performance on English naming blocks before and after the French block was compared. For participants whose dominant language was English, the influence of French enemies on their performance was smaller than that on participants whose dominant language was French. For the participants whose dominant language was French, their naming latency on French enemies group was only longer than that on no enemies group to a very small degree before the French filler block. However, after the French filler block, their naming latency on French enemies group was much longer than that on no enemies group. These results suggested that participants who were more dominant in their L1 and those who have used their L1 recently are more likely to be influenced by their L1 phonological knowledge in L2 phonological processing.

We have discussed L2 phonological processing in adults in terms of segments and the factors that may play a role such as L1 phonology and L1 orthography. In terms of suprasegmental processing among L2 learners, their L1s also play a role. For example, Linetal. (2012) found native Mandarin speakers may have an advantage over Korean speakers in English stress processing because of the suprasegmntal features of their L1s. In Mandarin, contrastive stress exists and is realized in the same way as in English (Duanmu 2007). In both Mandarin and English, the same word that carried stress in different positions may differ in meaning. In contrast, lexical stress and minimal word pairs that only differ in stress do not exist in Korean (Jun 2005). In a sequence recall task, monolingual English speakers, native Mandarin speakers and native Korean speakers heard a series of non-word strings which consist of minimal pairs that only differ in stress (e.g. /mipa/ and /mipa/), and were asked to recall the sequence of the string. Mandarin English L2 speakers performed better than Korean English L2 speakers after other factors such as English proficiency had been controlled. In an auditory lexical decision task, compared with Mandarin speakers, it was more difficult for Korean speakers to reject non-words that were created by shifting the stress of real English words (e.g. the non-word /hyumn/ was created by shift the stress location of the wordhuman). These results suggested that Mandarin speakers utilize their stress encoding strategies from L1 to English L2, so that they outperform Korean speakers on stress processing in English. This is consistent with previous studies that focused on suprasegmental processing in L2 among Romance-language speakers (Dupouxetal. 2001, 2010).

L2 PHONOLOGICAL PERCEPTION AND PROCESSING IN CHILDREN

Although L2 phonological perception and processing is difficult for non-native speaking adults, research has shown that infants are able to discriminate speech sounds according to phonetic category without prior specific language experience in a non-native language (Werkeretal. 1981). Unfortunately, for adults and children as young as age one (Eimas 1975; Werker and Tees 1983), this ability may be lost or declined due to the specific language experience with their native language and lack of experience with the non-native language. The current section reviews L2 phonological perception and processing in a developmental perspective.

Perception of non-native sounds in infants

As early as 1971, Eimasetal. found that 1-4 months old infants are able to discriminate auditory syllables /ba/ and /pa/, which differ in only the initial phoneme. Later studies further investigated infants’ ability to discriminate non-native auditory contrasts. For example, Streeter (1976) found that young Kikuyu-learning infants are able to discriminate phonemic contrasts that are used in English but not in Kikuyu; Werkeretal. (1981) found that young English-learning infants can discriminate Hindi phoneme contrasts which do not occur in English; Tsushimaetal. (1994) showed that young Japanese -learning infants can discriminate contrasts that are used in English but not in Japanese. Considering the different performance children and adults have on non-native phonemic discrimination, researchers suggested that infants are born with the ability to discriminate phonetic contrasts in all languages, but unfortunately, this ability may be lost or declined due to the lack of particular linguistic experience to the non-native sounds (Eimas 1975; Strange and Jenkins 1978).

Studies that aimed to find the mechanism of infants’ ability to discriminate universal sets of phonetic contrasts have been conducted, and as mentioned earlier in the current paper, several models have been put forward to describe how learners acquire the phonetic knowledge (e.g. PAM, SLM and NLM). Early studies (MacKain 1982; Werker and Tees 1984) suggested that the learning of phonetic categories is minimal pair-based, infants learn phonetic distinction as they acquire the lexical knowledge of the language. For example, when an infant learns that the wordspearandbearhave different meanings, they also learns that /p/ and /b/ are different phonemes. The semantic information of words is critical for infants to acquire phonemic categories. However, Mayeetal. (2002) provided evidence that infants’ acquisition of the category of the sounds may be due to distributional learning. In their study, two groups of infants aged six to eight months were exposed to an 8-step continuum of /da/ to /ta/ auditory phonemes synthesized via a computer program. One group heard more sounds of the two center points of the continuum (i.e. steps 4 and 5, corresponding to unimodal frequency distribution) which did not provide salient cue for them to distinguish the /da/-/ta/ contrast, while the other group heard more sounds of steps 2 and 7, corresponding to bimodal frequency distribution, which was more helpful in training to distinguish the /da/-/ta/ contrast. Both groups heard same numbers of the remaining sound in the continuum. After a familiarization period, only the second group showed reliable ability to discriminate steps 1 from 8. Yoshidaetal. (2006) found a similar result with a non-native contrast. Werkeretal. (2007) further demonstrated this distributional learning mechanism.among Japanese and English learning infants.

Unfortunately, infants’ ability to discriminate non-native minimal contrasts cannot be sustained until their adulthood. On one hand, as mentioned earlier, this ability declines within the first year of one’s life because of the lack of specific linguistic experience, and this is directly related to the L1 linguistic environment where infants are exposed to (Kuhletal. 2001). On the other hand, the acquisition of non-native phonetic features does not only require simple exposure to sounds, but also requires appropriate contextual support. For example, Kuhletal. (2003) trained English learning infants at the age of 9-10 months to study a Mandarin contrast which does not exist in English: an alveolo-palatal affricate /th/ and an alveolo-palatal fricative //. The critical finding is that only short-term exposure to the Mandarin contrast, without interpersonal interaction, did not significantly help infants’ acquisition of the Mandarin feature. A more recent study (Yoshidaetal. 2009) also showed that 14-month-old infants are not able to learn two similar-sounding words unless appropriate contextual support (e.g. a picture that matches the new word) is provided. There are two possible explanations: one is that the acquisition of the new features depends on infants’ development of their general cognitive abilities. The second is that the contextual support facilitates the learners’ lexical representation of the phonological information.

In summary, infants are born with the ability to discriminate universal sets of phonemic contrasts. Unfortunately, their abilities to discriminate non-native phonemic contrasts may be lost or declined in the first year of their life, largely due to the lack of corresponding linguistic experience. Recent research has explored the mechanism as to how infants acquire phonemic categories. Compared with traditional minimal pair-based learning hypothesis, distributional learning hypothesis seems to better capture the process how phonemic categories are acquired.

Perception and processing of L2 phonology in young children

When young children learn L2, their ability to discriminate universal sets of phonetic contrasts may have declined or have been lost. Instead, the phonological skills in their native languages may begin to influence those in their L2s. For example, Cisero and Royer (1995) showed that among young Spanish-English bilingual children, kindergarten children’ scores on their Spanish L1 phonological awareness predicted the level of their English L2 phonological awareness in their first grade. Durgunoglu (1998) also showed that first-grade Spanish-English bilingual children’s phonological awareness of Spanish (L1) is highly correlated with the phonological awareness of English (L2). Dickinsonetal. (2004) replicated these results among low-income Spanish-English bilingual preschool children.

Other studies showed that there is even cross-language transfer of phonological awareness across non-alphabetic and alphabetic writing systems (Chiang 2002; Wangetal. 2005). Chiang (2002) reported that Chinese-speaking kindergarten-aged children’s phonological awareness in Chinese was positively correlated with their level of phonological awareness in English. Wangetal. (2005) showed similar findings among Chinese immigrant children at the age of Grade 2 and 3 in the US. Onset, rime, and tone matching tasks in Chinese were administered, in which an auditory stimuli was provided first, and children were asked to judge which of two candidates share the same onset, rime or tone with the target sound. Onset, rime matching and phoneme deletion tasks in English were administered. Results showed that the bilingual children’s performance on Chinese onset matching task was significantly correlated with that on English onset and rime matching tasks.

Nevertheless, the evidence of cross-language transfer in phonological awareness does not mean that the phonological representations of L1 and L2 are same. For example, Wang and Cheng (2008) found that Chinese-English bilingual children in the US preferred CV unit in a CVC (consonant-vowel-consonant) spoken Chinese (L1) syllable. In contrast, the same children revealed a preference for VC over CV unit in a spoken English (L2) syllable. This difference can be explained by the different syllabic structures across Chinese and English. There are only two coda /n/ and /?/ in Chinese and a large number of Chinese syllables are open (i.e. CV) syllables. This linguistic experience may encourage children to pay more attention to CV unit than VC unit. Therefore, it is likely that bilingual children’s subsyllabic unit preference in their L2 is more determined by the spoken properties of the L2 than the influence of the L1. Despite that L1 phonological processing skills can be transferred from L1 to L2, cross-language differences may still exist as a result of different properties of target spoken languages.

Kuo and Anderson (2010) investigated the effects of early bilingualism on phonological awareness beyond L1-L2 transfer. They argued that learning L2 has an effect on the formation of phonological representation at an abstract level by compared monolingual and bilingual children’s phonological awareness. In their study, the monolingual groups were kindergarten, first-grade and second-grade children in Taiwan whose native language was Mandarin. The bilingual participants were children at the same age but had exposure to an additional language-Southern-Min. Children’s onset, rime and tone awareness were measured in oddity tasks. For each trial, three auditory syllables were provided, and children were required to select the syllable that have the different onset, or rime, or tone with the other two. In addition to the syllables that exist in Mandarin and Southern-Min, novel syllables that do not conform to the phonotactics of either of the two languages were also constructed. For kindergarten and first grade children, bilingual participants outperformed their monolingual peers on these novel syllables in all of the three oddity tasks. These results suggested that early exposure to L2 allows children to be exposed to phonological segments in more contexts, thus facilitating the abstraction of the phonological segments. In addition, bilingual children’s better performance on tone awareness suggested that bilingual experience enable them to segment novel syllables into subsyllabic units and detect the tonal information more efficiently than their monolingual peers. To sum up, learning L2 has a positive effect on the formation of phonological representation.

In summary, L1-L2 cross-language transfer on phonological perception and processing skills occurs not only among languages that share similar scripts, but also among languages that have different writing systems. Learning L2 is able to facilitate the formation of phonological representation among bilingual children.

Contribution of L1 phonological awareness to L2 reading skills

L1 does not only affect L2 phonological awareness, but also contribute to L2 reading performance in bilingual children. Durgunogluetal. (1993) showed that Spanish beginning readers’ performance on English visual word recognition is predicted by the level of their phonological awareness in both English L2 and Spanish L1. Carlisleetal. (1999) also showed that L1 phonological awareness contributed to L2 reading achivement among unbalanced Spanish-English bilingual children from Grade 1, Grade 2 to Grade 3. English L1 phoneme and onset-rime awareness is an important predictor of bilingual children’s reading performance in French L2. Kim and Stoel-Gammon (2009) found that both onset-rime and phoneme awareness in Korean were positively correlated with English reading skill among Korean-English bilingual children. However, only Korean phoneme awareness was significantly correlated with word reading performance in Korean. This may be due to the different orthographic depth of the two languages. In contrast to English, Korean is a language with more consistent grapheme-to-phoneme correspondences (GPCs). Readers may develop phonological recoding strategies based on small unit only in Korean, but instead rely on reading strategies based on both small unit (i.e. phoneme) and large unit (e.g. onset-rime) in English.

Some studies investigated that whether non-alphabetic L1 phonological skill contributes unique variance to alphabetic L2 reading. Gottardoetal. (2001) showed that Cantonese-English children’s phonological skills in Cantonese L1, such as their performance on rhyme detection and tone detection tasks in Cantonese, contributed unique variance to English L2 reading skill. Wangetal. (2005) also showed that Mandarin L1 phonological awareness influences English L2 reading performance among Chinese-English bilingual children. Children’s performance on both real Pinyin and pseudo-Pinyin naming tasks were significantly correlated with their performance on English pseudoword reading. Children’ scores of Chinese tone matching tasks contributed to English pseudoword reading after English phoneme processing skill was taken into consideration.

LIMITATION AND FUTURE RESEARCH

Since phonological information encompasses not only segmental information but also suprasegmental information, it is important to address both segmental and suprasegmental perception and processing in further research. Although it is well-documented that L2 learners have poor perception of non-native segmental contrasts, L2 perception of non-native suprasegmental contrasts has received less attention. More research is certainly needed in understanding how L2 suprasegmental information is perceived and processed. As one of the first few attempts to address L2 suprasegmental perception and processing, Linetal. (2012) used sequence recall task and lexical decision task to examine the encoding of stress information in two groups of L2 learners that differ in their L1 backgrounds, Chinese vs. Korean English L2 learners. Future research needs to include more L1 backgrounds, and different research paradigms in this line of research.

In the line of research addressing the cross-language transfer issue in L2 phonological awareness and L2 reading, most current studies have reported correlational data where the language- and reading-related tasks were administered simultaneously. Therefore we cannot make any suggestion on directionality of the relationships between phonological awareness in the two languages and between phonological awareness in L1 and reading in L2. To better understand these relationships, future research needs to address the issue of bidirectional relationships. In line with monolingual research on the reading processes and reading acquisition, future studies should examine the predictive power of phonological awareness in L2 reading acquisition over time. It would be especially interesting to test the relationship between phonological awareness in one language at early time points and reading skills in another language at later time points, as well as to explore the reciprocal developmental relationship between reading skill and phonological awareness. Furthermore, future research also should consider phonological training studies on L2 or bilingual children to see if phonological awareness in one language would produce substantive gains in students’ word reading and reading comprehension in the other language. Combining longitudinal studies and training studies will allow researchers to establish the causal relationship between phonological awareness and reading skill in L2 acquisition.

Finally, L2 proficiency is an important factor in L2 phonological perception and processing. Highly proficient L2 learners will overall have better L2 phonological perception and processing skills. However, how L2 proficiency interacts with different types of phonetic contrasts and phonological units is interesting and worthy of more extensive research.

CONCLUSION

The current paper reviewed and discussed the literature on perception and processing of L2 phonology among infants, children and adults. Adult L2 learners have difficulties discriminating some non-native phonetic contrasts that do not occur in their native languages, however, the degree of the difficulties may differ across contrasts. Infants are born with the ability to discriminate universal sets of phonetic distinctions, but this ability may decline or be lost later on due to lack of corresponding linguistic experience. The acquisition or learning of the L2 phonological knowledge does not only require simple linguistic exposure, but also need contextual support and social interaction. The mechanism underlying the L2 phonological learning has been proposed including PAM, SLM, NLM and distributional learning. For both adults and children, their L2 phonological processing skill is influenced by the knowledge of their L1. Furthermore, not only L1 phonology but also the orthography affects the phonological processing skill in L2. Finally, L1 phonological skills contribute to the development of L2 reading skills over and above the within-L2 phonological skills.

REFERENCES

Abramson, A. S. and L. Lisker. 1970. ‘Discriminability along the voicing continuum: Cross language tests,’Proceedingsofthe6thInternationalCongressofPhoneticSciences: 569-73.

Best, C. T. 1994a. ‘The emergence of native-language phonological influences in infants: A perceptual assimilation model’ in J. C. Goodman and H. C. Nusbaum (eds):TheDevelopmentofSpeechPerception:TheTransitionfromSpeechSoundstoSpokenWords. Cambridge, MA: MIT Press, pp. 167-224.

Best, C. T. 1994b. ‘Learning to perceive the sound pattern of English’ in C. Rovee-Collier and L. Lipsitt (eds):AdvancesinInfancyResearch. Hillsdale, NJ: Ablex Publishers, pp. 217-304.

Best, C. T. 1995. ‘A direct realist perspective on cross-language speech perception’ in W. Strange (ed.):SpeechPerceptionandLinguisticExperience:IssuesinCross-languageResearch. Timonium, MD: York Press, pp. 171-204.

Best, C. T. and G. W. McRoberts. 1988. ‘Examination of perceptual reorganization for nonnative speech contrasts: Zulu click discrimination by English-speaking adults and infants,’JournalofExperimentalPsychology:HumanPerceptionandPerformance14/3: 345-60.

Best, C. T. and W. Strange. 1992. ‘Effects of phonological and phonetic factors on cross-language perception of approximants,’JournalofPhonetics20: 305-30.

Best, C. T. and M. D. Tyler. 2007. ‘Nonnative and second language speech perception: commonalities and complementaries’ in M. J. Munro and O. -S. Bohn (eds):SecondLanguageSpeechLearning:TheRoleofLanguageExperienceinSpeechPerceptionandProduction. Amsterdam: John Benjamins, pp. 13-34.

Bosch, L., A. Costa, and N. Sebastia′n-Galle′s. 2000. ‘First and second language vowel perception in early bilinguals,’EuropeanJournalofCognitivePsychology12: 189-221.

Bradlow, A. R., R. Akahane-Yamada, D. B. Pisoni, and Y. Tohkura. 1999. ‘Training Japanese listeners to identify English /r/ and / l/: long-term retention of learning in speech perception and production,’Perception&Psychophysics61: 977-85.

Bryant, P., M. Maclean, and L. Bradley. 1990. ‘Rhyme, language, and children’s reading,’AppliedPsycholinguistics11: 237-52.

Carlisle, J. F., M. M. Beeman, L. H. Davis, and G. Spharim. 1999. ‘Relationship of metalinguistic capabilities and reading achievement for children who are becoming bilingual,’AppliedPsycholinguistics20: 459-78.

Chiang, P. Y. 2002.PhonologicalawarenessinChinese-Englishbilingualchildren:Effectsofarticulationtraining. Master’s thesis, National Taiwan University, Taipei, Taiwan.

Cisero, C. A. and J. M. Royer. 1995. ‘The development and cross- language transfer of phonological awareness,’ContemporaryEducationalPsychology20: 275-303.

Crystal, D. 2008.ADictionaryofLinguisticsandPhonetics,SixthEdition. Malden, MA: Blackwell Publishing.

Cutler, A., A. Weber, and T. Otake. 2006. ‘Asymmetric mapping from phonetic to lexical representations in second-language listening,’JournalofPhonetics34: 269-84.

D’Ausilio, A., F. Pulvermu¨ller, P. Salmas, I. Bufalari, C. Begliomini, and L. Fadiga. 2009. ‘The motor somatotopy of speech perception,’CurrentBiology19: 381-85.

Dehaene-Lambertz, G. 1997. ‘Electrophysiological correlates of categorical phoneme perception in adults,’NeuroReport8/4: 919-24.

Dickinson, D. K., A. McCabe, N. Clark-Chiarelli, and A. Wolf. 2004. ‘Cross-language transfer of phonological awareness in low-income Spanish and English bilingual preschool children,’AppliedPsycholinguistics25: 323-47.

Doctor, E. A. and D. Klein. 1992. ‘Phonological processing in bilingual word recognition’ in R. J. Harris (ed.):CognitiveProcessinginBilinguals. Amsterdam: Elsevier, pp. 237-52.

Duanmu, S. 2007.ThePhonologyofStandardChinese,SecondEdition. New York, NY: Oxford University Press.

Dupoux, E., C. Pallier, N. Sebastian, and J. Mehler. 1997. ‘A distressing “deafness” in French?’JournalofMemoryandLanguage36: 406-21.

Dupoux, E., S. Peperkamp, and N. Sebastia′n-Galle′s. 2001. ‘A robust method to study stress “deafness”,’JournalofAcousticalSocietyofAmerica110: 1606-18.

Dupoux, E., S. Peperkmap, and N. Sebastia′n-Galle′s. 2010. ‘Limits on bilingualism revisited: stress “deafness” in simultaneous French-Spanish bilinguals,’Cognition114: 266-75.

Durgunoglu, A. Y. 1998. ‘Acquiring literacy in English and Spanish in the United States’ in A.Y. Durgunoglu and L. Verhoeven (eds):LiteracyDevelopmentinAMultilingualContext. Mahwah, NJ: Erlbaum, pp. 135-45.

Durgunoglu, A. Y., W. E. Nagy, and B. J. Hancin-Bhatt. 1993. ‘Cross-language transfer of phonological awareness,’JournalofEducationalPsychology85/3: 453-65.

Ehri, L., S. Nunes, D. Willows, B. Schuster, Z. Yaghoub-Zadeh, and T. Shanahan. 2001. ‘Phonemic awareness instruction helps children learn to read: evidence from the National Reading Panel’s meta-analysis,’ReadingResearchQuarterly36: 250-87.

Eimas, P. D. 1975. ‘Developmental studies in speech perception’ in L. B. Cohen and P. Salapatek (eds):InfantPerception:FromSensationtoPerception(vol. 2). New York: Academic Press.

Escudero, P. and K. Wanrooij. 2010. ‘The effect of L1 orthography on non-native vowel perception,’LanguageandSpeech54/3: 343-65.

Escudero, P., B. Titia, and W. Karin. 2011. ‘Enhanced bimodal distributions facilitate the learning of second language vowels,’JournaloftheAcousticalSocietyofAmerica130 /4: EL206-EL212.

Flege, J. E. 1992. ‘Speech learning in a second-language’ in C. A. Ferguson, L. Menn and C. Stoel-Gammon (eds):PhonologicalDevelopment:Models,Research,andImplications. Parkton, Md.: York Press.

Flege, J. E. 1995. ‘Second language speech learning: theory, findings, and problems’ in W. Strange (ed.):SpeechPerceptionandLinguisticExperience:IssuesinCross-languageResearch. Timonium, MD: York Press, pp. 233-76.

Flege, J. E. and W. Eefting. 1986. ‘Linguistic and developmental effects on the production and perception of stop consonants,’Phonetica43: 155-71.

Gillon, G. 2004.PhonologicalAwareness:FromResearchtoPractice. New York: Guilford Press.

Goto, H. 1971. ‘Auditory perception by normal Japanese adults of the sounds “l(fā)” and “r”,’Neuropsychologia9: 317-23.

Gottardo, A., B. Yan, L. S. Siegel, and W. L. Wade. 2001. ‘Factors related to English reading performance in children with Chinese as a first language: More evidence of cross-language transfer of phonological processing,’JournalofEducationalPsychology93: 530-42.

Iverson, P. and P. K. Kuhl. 1995. ‘Mapping the perceptual magnet effect for speech using signal detection theory and multidimensional scaling,’JournaloftheAcousticalSocietyofAmerica97: 553-62.

Jared, D. and J. F. Kroll. 2001. ‘Do bilinguals activate phonological representations in one or both of their languages when naming words?’JournalofMemoryandLanguage44: 2-31.

Juel, C., P. Griffith, and P. Gough. 1986. ‘Acquisition of literacy: A longitudinal study of children in first and second grade,’JournalofEducationalPsychology78: 243-55.

Jun, S.-A. 2005. ‘Korean intonational phonology and prosodic transcription’ in S.-A. Jun (ed.):ProsodicTypology:ThePhonologyofIntonationandPhrasing. Oxford: University Press, pp. 201-99.

Kim, M. and C. Stoel-Gammon. 2009. ‘The acquisition of Koreanword-initial stops,’JournaloftheAcousticalSocietyofAmerica125: 3950-61.

Koda, K. 1989. ‘Effects of L1 orthographic representation on L2 phonological coding strategies,’JournalofPsycholinguisticResearch18: 201-22.

Kuhl, P. K. 1991. ‘Human adults and human infants show a “perceptual magnet effect” for the prototypes of speech categories, monkeys do not,’Perception&Psychophysics50: 93-107.

Kuhl, P. K., F. M. Tsao, H. M. Liu, Y. Zhang, and B. de Boer. 2001. ‘Language/culture/mind/ brain: progress at the Margins between Disciplines’ in A. Domasioetal. (eds):UnityofKnowledge:TheConvergenceofNaturalandHumanScience. New York: The New York Academy of Sciences, pp. 136-74.

Kuhl P. K, F.-M. Tsao, and H. -M Liu. 2003. ‘Foreign-language experience in infancy: Effects of short-term exposure and social interaction on phonetic learning,’ProceedingsoftheNationalAcademyofSciences,USA, 100.

Kuo, L. and J. R. C. Anderson. 2010. ‘Beyond cross-language transfer: reconceptualizing the impact of early bilingualism on phonological awareness,’ScientificStudiesofReading14/4: 265-380.

Lin, C. Y., M. Wang, W. J. Idsardi, and Y. Xu. 2012. under review.StressProcessinginMandarin-EnglishandKorean-EnglishBilinguals.

Lively, S. E., J. S. Logan, and D. B. Pisoni. 1993. ‘Training Japanese listeners to identify English /r/ and /l/: II. The role of phonetic environment and talker variability in learning new perceptual categories,’JournaloftheAcousticalSocietyofAmerica94: 1242-55.

Logan, J. S., S. E. Lively, and D. B. Pisoni. 1991. ‘Training Japanese listeners to identify English /r/ and /l/: A first report,’JournaloftheAcousticalSocietyofAmerica89: 874-86.

MacKain, K. S. 1982. ‘Assessing the role of experience on infants’ speech discrimination,’JournalofChildLanguage9: 527-42.

MacKain, K. S., C. T. Best, and W. Strange. 1981. ‘Categorical perception of English /r/ and /l/ by Japanese bilinguals,’AppliedPsycholinguistics2: 369-90.

Metsala, J. L. and A. C. Walley. 1998. ‘Spoken vocabulary growth and the segmental restructuring of lexical representations: precursors to phonemic awareness and early reading ability’ in J. L. Metsala and L. C. Ehri (eds):WordRecognitioninBeginningLiteracy. New York: Erlbaum, pp. 89-120.

Maye, J., J. F. Werker, and L. Gerken. 2002. ‘Infant sensitivity to distributional information can affect phonetic discrimination,’Cognition82: B101-B111.

M?tt?nen, R. and K. E. Watkins. 2009. ‘Motor representations of articulators contribute to categorical perception of speech sounds,’TheJournalofNeuroscience:TheOfficialJournaloftheSocietyforNeuroscience29/31: 9819-25.

Moyer, A. 1999. ‘Ultimate attainment in L2 phonology: The critical factors of age, motivation, and instruction,’StudiesinSecondLanguageAcquisition21: 81-108.

N??t?nen, R. 2001. ‘The perception of speech sounds by the human brain as reflected by the mismatch negativity (MMN) and its magnetic equivalent (MMNm),’Psychophysiology38: 1-21.

Navarra, J. and S. Soto-Faraco. 2007. ‘Hearing lips in a second language: Visual articulatory information enables the perception of second language sounds,’PsychologicalResearch71: 4-12.

Navarra, J., N. Sebastian-Galles, and S. Soto-Faraco. 2005. ‘The perception of second language sounds in early bilinguals: New evidence from an implicit measure,’JournalofExperimentalPsychology:HumanPerceptionandPerformance31/5: 912-18.

Pallier, C., L. Bosch, and N. Sebastian-Gallés. 1997. ‘A limit on behavioral plasticity in speech perception,’Cognition64: B9-B17.

Polivanov, G. and E. Dmitrievic. 1931. ‘La perception des sons d.une langue étrangère,’TravauxduCercleLinguistiquedePrague4: 79-96.

Polka, L. 1991. ‘Cross-Language speech perception in adults: Phonemic, phonetic, and acoustic contributions,’JournaloftheAcousticalSocietyofAmerica89: 2961- 77.

Polka, L. 1992. ‘Characterizing the influence of native experience on adult speech perception,’PerceptionandPsychophysics52: 37-52.

Russak, S. and E. Saiegh-Haddad. 2011. ‘Phonological awareness in Hebrew (L1) and English (L2) in normal and disabled readers,’ReadingandWriting24: 427-42.

Saiegh-Haddad, E. and E. Geva. 2010. ‘Acquiring reading in two languages: An introduction to the special issue,’ReadingandWriting23: 263-67.

Sebastia′n-Galle′s, N., S. Echeverria, and L. Bosch. 2005. ‘The influence of initial exposure on lexical representation: Comparing early and simultaneous bilinguals,’JournalofMemoryandLanguage52: 240-55.

Sebastia′n-Galle′s, N. and S. Soto-Faraco. 1999. ‘Online processing of native and non-native phonemic contrasts in early bilinguals,’Cognition72: 111-23.

Strange, W. and S. Dittmann. 1984. ‘Effects of discrimination training on the perception of /r/-/l/ by Japanese adults learning English,’Perception&Psychophysics36: 131-45.

Strange, W. and J. Jenkins. 1978. ‘Role of linguistic experience in the perception of speech’ in R. D. Walk and H. L. Pick (eds):PerceptionandExperience. New York: Plenum Press.

Streeter, L. A. 1976. ‘Language perception of two-month old infants shows effects of both innate mechanism and experience,’Nature259: 39-41.

Takagi, N. and V. Mann. 1995. ‘The limits of extended naturalistic exposure on the perceptual mastery of English /r/ and /l/ by adult Japanese learners of English,’AppliedPsycholinguistics16: 379-405.

Tan, L. H., J. A. Spinks, J. H. Gao, H. L. Liu, C. A. Perfetti, and J. Xiong. 2003. ‘Brain activation in the processing of Chinese characters and words: A functional MRI study,’HumanBrainMapping10: 16-27.

Treiman, R., J. Mullennix, R. Bijeljac-Babic, and E. D. Richmond-Welty. 1995. ‘The special role of rimes in the description, use, and acquisition of English orthography,’JournalofExperimentalPsychology:General124: 107-36.

Tremblay, A. 2008. ‘Is second language lexical access prosodically constrained? Processing of word stress by French Canadian second language learners of English,’AppliedPsycholinguistics29: 553-84.

Tremblay, A. 2009. ‘Phonetic variability and the variable perception of L2 word stress by French Canadian listeners,’InternationalJournalofBilingualism13: 35-62.

Tsushima, T., O. Takizawa, M. Sasaki, S. Shiraki, K. Nishi, M. Kohno, P. Menyuk, and C. Best. 1994. ‘DiscriminationofEnglish/r-l/and/w-y/byJapaneseinfants6-12months:Language-specificdevelopmentalchangesinSpeechPerceptionAbilities.’ Paper presented at the International Conference on Spoken Language. Yokohama, Japan.

Van Orden, G. C. 1987. ‘A rows is a rose: Spelling, sound and reading，’MemoryandCognition15: 181-98.

Vellutino, F. R. and D. M. Scanlon. 2001. ‘Emergent literacy skills, early instruction, and individual differences as determinants of difficulties in learning to read: the case for early intervention’ in S. Neuman and D. Dickson (eds):HandbookforResearchonEarlyLiteracy. New York: Guilford press, pp. 295-321.

Wagner, R. K. and J. K. Torgesen. 1987. ‘The nature of phonological processing and its causal role in the acquisition of reading skills,’PsychologicalBulletin101: 192-212.

Wang, M. and C. Cheng. 2008. ‘Subsyllabic unit preference in young Chinese children,’AppliedPsycholinguistics29: 291-314.

Wang, M., K. Koda, and C. A. Perfetti. 2003. ‘Alphabetic and non-alphabetic L1 effects in English word identification: A comparison of Korean and Chinese English L2 learners,’Cognition87: 129-49.

Wang, M., C. A. Perfetti, and Y. Liu. 2005. ‘Chinese-English biliteracy acquisition: Cross-language and writing system transfer,’Cognition97: 67-88.

Wanrooij, Karin, Titia van Zuijen, and Paul Boersma. 2012. ‘MMN declines after distributional vowel training.’ Poster at MMN 2012.TheSixthConferenceonMismatchNegativityanditsClinicalandScientificApplication, New York, May 4, 2012.

Weber, A. and A. Cutler. 2004. ‘Lexical competition in non-native spoken-word recognition,’JournalofMemoryandLanguage50: 1-25

Werker, J. 1994. ‘Cross-language speech perception: Developmental change does not involve loss’ in Goodman and Nusbaum (eds):TheDevelopmentofSpeechPerception. Cambridge, MA: MIT Press, pp. 93-120.

Werker J. F., J. H. V. Gilbert, K. Humphrey, and R. C. Tees. 1981. ‘Developmental aspects of cross-language speech perception,’ChildDevelopment52: 349-53.

Werker J. F., F. Pons, C. Dietrich, S. Kajikawa, L. Fais, and S. Amano. 2007. ‘Infant-directed speech supports phonetic category learning in English and Japanese,’Cognition103: 147-62.

Werker, J. F. and R. C. Tees. 1983. ‘Developmental changes across childhood in the perception of non-native speech sounds,’CanadianJournalofPsychology37: 278- 86.

Werker, J. F. and R. C. Tees. 1984. ‘Cross-language speech perception: Evidence for perceptual reorganization during the first year of life,’InfantBehaviorandDevelopment7: 49-63.

Wilson, M. 1987. ‘Brain mechanisms in categorical perception’ in S. Harnand (ed.):CategoricalPerception:TheGroundworkofCognition. New York, NY, US: Cambridge University Press, pp. 387-417.

Wilson, S. M. 2009. ‘Speech perception when the motor system is compromised,’CognitiveScience13: 329-30.

Yeung, H. H. and J. F. Werker. 2009. ‘Learning words’ sounds before learning how words sound: 9-Month-olds use distinct objects as cues to categorize speech information,’Cognition113: 234-43.

Yoshida, K. A., F. Pons, and J. F. Werker. 2006. ‘Does distributional learning affect perception after phonemes are established?’ Poster presented at theInternationalConferenceonInfantStudies(ICIS), Kyoto, Japan.

Yoshida, K. A, C. T. Fennell, D. Swingley, and J. F. Werker. 2009. ‘Fourteen-month-old infants learn similar sounding words,’DevelopmentalScience12/3: 412-18.

Ziegler, J. C. and U. Goswami. 2005. ‘Reaing acquisition, developmental dyslexia, and skilled reading across languages: A psycholinguistic grain size theory,’PsychologicalBulletin131: 3-29.