Clara Burgo
University of Illinois at Chicago
 
 
Introduction

“Child language acquisition and adult SLA involve different types of processing for language learning” (Doughty, 2003, p. 275). This claim is the logical inference resulting from different positions in the literature. The Fundamental Difference Hypothesis (Bley-Vroman, 1990) states that unlike child language learning, adult SLA involves explicit, general problem-solving strategies. DeKeyser (1995) agrees and adds that adults use analytical thinking to acquire a second language (L2). As Zobl (1989) explained, “the modular composition of the adult language faculty has been altered through the partial loss of the computational module” (p. 50). His findings suggested that modularity in adults must be considered as a source of interlanguage-first language (L1) difference. Likewise, the Competition Hypothesis (Felix and Hahn, 1985) proposed that initially implicit Universal Grammar (UG) and explicit problem-processes compete in adult SLA, but finally the latter imposes over the former. However, Liceras (1991) claims that the UG model is a key factor in SLA.

Child-adult differences in language acquisition are proof of the existence of maturational constraints (Doughty, 2003). One of the most obvious constraints on adult SLA is the accent. Adults are considered to be disabled L2 learners since there is an alteration of the speech-processing abilities through their experience with L1. Therefore, L2 adults usually process input with mechanisms “already attuned to their L1” (Doughty, 2003, p. 284). Cutler, Mehler, Norris and Segui (1986) investigated speech segmentation by adults in English speakers learning French as a L2. They discovered that adult English speakers do not use the syllabification strategy when listening to their L1 nor do they use it in French, even though the latter is considered easier to syllabify. It is also interesting that French native speakers who use the syllabification structure in French also use it when listening to English. Similarly, Cutler and Butterfield (1992) found that adult native English speakers use a stress-based segmentation strategy when listening to English. Similarly, Cutler and Norris (1988) observed that these speakers follow the same strategy when listening to Japanese (a language with a different rhythmic structure). The study of Cutler and Otake (1994) also emphasized that Japanese speakers use their mora-based strategy in English whereas native English speakers do not. Cutler et al. (1986) interpreted this evidence as a proof that segmentation strategies are not universal perceptual routines, but specific to a language. It seems rather, to be a transfer of the strategies used in the L1 to the L2: “during language acquisition, speakers adapt their perceptual routines so as to exploit with maximal efficiency the phonological properties of their native language” (Cutler et al., 1986, p. 397). Another important claim derived from these findings is that the segmentation strategy specific to a language is “in the listener, not in the speech signal” (Cutler, 2001, p. 11). Several studies of English-French bilinguals (Cutler, Mehler, Norris and Segui, 1992, 1989), English-Dutch and French-Dutch bilinguals (van Zon, 1997) showed that bilinguals commanded the strategies of their preferred language. However, the proficient bilinguals never misapplied the segmentation strategy of their preferred language when using their less-preferred language. This finding led Cutler (2001, p. 16) to conclude that “inappropriate language-specific segmentation is avoidable.” Interestingly, Doughty (2003) argued that discrimination ability remains throughout adulthood for phonemes that do not exist in the L1 and that are not pre-empted by any L1 contrast: for example, English speakers can discriminate Zulu clicks (Best, McRoberts, and Sithole, 1988).
 
Discussion

The question in L2 adult learners is whether they can be trained to use processing strategies other than those of their L1 (Doughty, 2003). Cutler (2002) proposed that “non-native listening skills are less flexible” (p. 3). In the native language, variations in pronunciation, unfamiliar voices or even external factors like background noises are not a real problem when listening. Nevertheless, they might be when listening to the L2. It appears that L2 learners use the same processing strategies for the L2 in the absence of proper instruction, although this is not very efficient (Doughty, 2003).

Regarding phonological discrimination, once native phonemic categorization is established, it cannot be altered except for phonemes that are not part of the L1, as indicated previously (Best, Lafleur and McRoberts, 1995). Fortunately, not all segmentation strategies are so limited. Weber (2000) showed that proficient German-English bilinguals were sensitive to both native German and non-native English phonotactic sequence constraints. They were given nonsense words with German and English onsets and had to detect the word luck. English speakers first detected the word luck in moyshluck. This was attributed to the fact that shl- is not a possible onset in English, which made it more salient than the other possibilities (moysluck and moyfluck). German-English proficient bilinguals more easily detected the wordluck in the word moysluck. Moreover, they detected the word luck in the word moyshluck faster than in the word moysfluck. This study offered evidence that these listeners were sensitive to both native sequencing constraints and also acquired some sensitivity concerning English phonotactics. Unlike the phonemic categorization that seems to be fixed, Doughty (2003) claimed that there are other patterns of language structure that are not so fixed. Doughty proposed that bilinguals pay attention to the cues in the input of one language as well as they do with the other language.

Regarding the most evident maturational constraint, the accent, there are different views about the relationship between the age of L2 learning and the degree of foreign accent. Long (1990) claimed that a L2 is spoken without an accent if learned before the age of six, with a foreign accent if learned after the age of 12 and that there is variability if the language is learned between the ages of 6 and 12. Similarly, Patkowski (1990) predicted the critical period to be the explanation of why some individuals speak their L2 with a foreign accent. On the other hand, Bongaerts, Planken and Schils’ (1995) study presents a challenge for Patkowski’s prediction. They found some highly motivated individuals who started learning the L2 after the critical period and spoke the L2 without a foreign accent. Moreover, Flege (1999) argued that Patkowski’s judges were probably influenced by other factors other than pronunciation such as the speakers’ choice of words or grammar. He proposed three hypotheses that account for foreign accents. The first hypothesis is the exercise hypothesis, which states that “one’s ability to learn to produce and perceive speech remains intact across the life span, but only if one continues to learn speech uninterruptedly” (p. 105, see also Bever, 1981; Hurford, 1991 for an overview). The second hypothesis is the unfolding hypothesis, which states that “the more fully developed the L1 phonetic system is at the time L2 learning begins, the more foreign-accented the pronunciation of the L2 will be” (p. 105, see studies by Oyama, 1979; Elman, 1993; Marchmann, 1993; Flege, 1992a, 1992b; Best, 1995 for an overview). The third and the last hypothesis is the interaction hypothesis in which” bilinguals are unable to fully separate the L1 and L2 phonetic systems, which necessarily interact with each other” (p. 106, see Anisfeld & Semogas, 1969; Cutler et al; 1989, Flege, 1995; Ho, 1986; Macnamara, 1973).

So far, I have reviewed several cases of Adult SLA and adults’ limitations when acquiring the phonology of a L2. However, there have been numerous attempts to challenge the existence of maturational constraints as indicated by L2 learners who acquired native-like proficiency. Obviously, the most salient and, in some cases, misleading feature is the L2 learners’ apparent absence of a non-native accent when speaking in the L2. Let us have a look at these studies.

Historically, the first study to approach this issue was conducted by Coppieters (1987). He selected 21 highly successful and educated foreign learners of French and engaged them in a syntactic / semantic judgment task and follow-up interviews. They were selected for their absence of foreign accent. In spite of this, their overall performance was below that of native controls. However, Birdsong (1992) found that 15 of his 20 adult learners of French performed in the same range as native speakers on a difficult grammaticality judgment task.

In phonology, Bongaerts (1999), Bongaerts et al. (1995) and Bongaerts, van Summeren. Planken and Schils (1997) examined a group of highly proficient post-puberty Dutch learners of English and French. They were selected because of their native-like pronunciation and their proficiency in these languages. They were asked to pronounce aloud a series of sentences that contained words difficult to pronounce for Dutch native speakers. Significant proportions of these subjects passed as native speakers even in the upper range of native controls. Likewise, Bongaerts, Mennen and van der Slik. (2000) investigated the pronunciation of very advanced and naturalistic learners of Dutch as a L2. They selected 30 highly educated learners with different L1s and different age onsets (AOs) within the ages of 11 and 34. The results showed that two participants (age 21 and 14) passed as native speakers. More recently, Moyer (1999) investigated the pronunciation of 24 very proficient American adult learners of German. They were exposed to the language for several years in Germany. They were given a read-aloud task that included a word list, a paragraph and sentences. Four native German judges independently rated the speech samples. The results showed that the judges were able to distinguish between the non-native and native speakers. There was just one of participant who performed within the range of the natives across all pronunciation tasks. Moyer (1999) considered him/her an exceptional learner with a very high motivation for “sounding German”. In this study the investigator also found a trend toward more native accent ratings in the shorter tasks, like word list reading. Once the task became longer (sentences, paragraph) and less structured, the foreign accent was more evident. Hyltenstam and Abrahamsson (2003, p. 551) interpret this as “a risk that these studies highlight skills other than real pronunciation skills.” They wondered if these late L2 learners who passed as native speakers would have done so if they had been given longer tasks that required more spontaneous speech. However, Moyer (1999) claimed that in order to test the phonological fluency alone, it is not possible to use very long naturalistic speech since other linguistic features such as lexical, syntactic or pragmatic factors would influence in these L2 learners’ performance.

In another study, Ioup (1995) investigated the performance of two talented L2 learners of Arabic. The subjects were assigned very demanding tasks: apart from production tasks, they had dialect differentiation and grammatical competence tests. Both performed within the range of native controls on the dialect differentiation test. They were rated as natives by 13 judges. However, even though both scored high on tests of grammatical intuition, they performed below native controls. Therefore, Ioup, Boustagui, Tigi and Moselle (1994) concluded that the assumed neurocognitive changes that occur after the critical period do not happen in the usual way. However, it is not clear whether the ordinary acquisition system keeps working or whether there is an alternative system.

White and Genessee (1996) considered that even individuals who seem to have acquired a native-like proficiency differ from natives “in subtle ways” (p. 234). In their study they selected the subjects in a very strict way in order to properly separate the native speakers from the non-natives. Most of the subjects that were considered to be near native started learning English as a L2 before the age of 12 unlike most of the non-native speakers who started learning English after the age of 12. The results showed significant differences between the native and the non-native group, but no significant differences between the near native and the native group on any of the measures assessed. There were no effects on the AOs either. Therefore, they concluded that UG is unaffected by age and even late L2 learners are able to achieve native-like proficiency. Hyltenstam and Abrahamsson (2003) argued that, in spite of having been very rigorous in their selection criterion, White and Genesee (1996) had not taken into account that their subjects spoke French as their L1, a language that works similarly to English in subjacency and regarding the Empty Category Principle (ECP). Consequently, it is completely normal that the speakers had not encountered any difficulties with sentences involving these aspects. Thus, this could be interpreted as continued access to L1 rules (Eubank and Gregg, 1999).

In other studies, researchers examined younger learners to discover whether it was possible to find L2 learners that achieved native-like proficiency at earlier ages. Hyltenstam (1992, 1988) tested the grammatical and lexical performance of 24 highly proficient Spanish and Finnish L2 learners of Swedish with AO of six years or earlier (16 subjects) and seven years or later (eight subjects). He found that “the age 6 or 7 does seem to be an important period in distinguishing between near-native and native-like ultimate attainment” (p. 364). Not all early learners achieved native-like proficiency though. This is why he finally concluded that an early AO is not a sufficient requirement for ultimate native-like attainment. Similarly, Ekberg (1998) found significant differences in frequencies of the use of specific structures, like sentence connectors or complex predicates, between two groups of native and non-native early L2 learners. Finally, Hyltenstam and Abrahamsson (forthcoming) concluded that there were significant differences between L1 speakers and L2 speakers regardless of their AOs. These studies and others, like Butler (2000), seem to indicate that “even a short delay in onset has effects on the ultimate level of language proficiency” (Hyltemstam and Abrahamssom, 2003, p. 555).

Comparing the studies that examine late L2 learners and those that examine early L2 learners, it seems that both failed to find learners who achieved overall native-like proficiency. Hyltemstam and Abrahamssom (2003) examined late and highly advanced L2 learners and found several limitations. Other studies also found that advanced L2 learners could not achieve full proficiency (Coppieters, 1987; Hyltenstam and Abrahamsson, forthcoming; Ioup et al., 1994; Moyer, 1999). Baralo (1996) mentioned that native-like speakers have different grammaticality intuitions than native speakers. According to her, errors are unavoidable but are also necessary in learning a language. Regarding the errors that are typical of the fossilization processes, the learner has to be aware of his own learning strategies in order to monitor his production in the L2. Other studies instead showed that subjects performed like non-native speakers in other areas that were not specifically investigated (Birdsong, 1999). Specifically, the investigator found that those speakers who passed as native speakers “deviated very little from native forms” (p. 9).  However, a small deviation is still deviation. Bongaerts (1999) observed some non-native features, beyond pronunciation, in his advanced subjects during free oral production. White and Genesse (1996) suggested that late L2 learners are able to achieve native-like proficiency at least in UG principles. However, it is still not clear “whether this is true of all domains and, if not, in which areas native-like success is not attainable and why not” (p. 262).

Taking into account all this evidence, Hyltestam and Abrahamssom (2003) suggested that there are a few exceptional late learners who, although not at all levels, are able to achieve a native-like proficiency in some areas of the L2. The possible reasons for this could be found in their high motivation (Moyer, 1999), their high aptitude for learning languages (DeKeyser, 2000; Harley and Hart, 1997; Ioup et al., 1994) or their intensive L2 instruction (Bongaerts, 1999; Moyer, 1999).

There have been numerous attempts to give a theoretical framework for the existence of these maturational constraints in a broad sense. The two most popular formulations have been the controversial Critical Period Hypothesis (CPH) and the not-so-radical Sensitive Period Hypothesis (SPH). I used the adjective radical to describe the CPH since it considers maturation as a phase in the life span, which is “abruptly set off from the rest at a specific age (puberty or earlier)” (Hyltenstam and Abrahamsson, 2003, p. 556). Thus I consider the SPH more reasonable and flexible. The SPH proposes that the sensitivity disappears progressively throughout the lifespan (late childhood, puberty and adolescence) instead of in such an abrupt way, as the result of inefficiency of the organism or of the peripheral input (Eubank and Gregg, 1999). However some researchers considered the SPH a revised version of the CPH (Obler and Hannigan, 1996) and use both terms interchangeably (Hyltenstam and Abrahamsson, 2003).

Johnson and Newport (1989) proposed two possible versions of the CPH: the exercise version and the maturational state version. The exercise hypothesis suggests that if a person does not exercise his/her capacity to learn a L1 in early childhood, it will disappear. However, if they exercise this capacity during this time, it will be intact. Therefore, they predicted that late learners of L1 will never acquire the same proficiency in their L1 than the rest of L1 learners. In this hypothesis, however, it is suggested that this is not the same for late L2 learners. They might be able to acquire native-like proficiency in some levels in their L2, but not all. I argue that this hypothesis is contradictory to the maturational state version, since the latter predicts that maturation has effects on any language acquisition (L1, L2, etc). That is, if learners do not acquire the language during early childhood, their capacity to learn a language will deteriorate with maturation, whether exercised or not. Thus, learners will not be able to achieve native-like proficiency when learning a language after the critical period. However, this version does not deny the positive effects of exercise (see study by Mayberry, 1993). There is a clear difference between late learners of a L1, like the famous cases of Genie or Chelsea who achieved rudimentary levels in their L1, and late learners of a L2 who acquired their L1 before the critical period. In late L2 acquisition, the consequences of learning a language so late are not as drastic as late L1 acquisition (Genie or Chelsea). As Eubank and Gregg (1999) explained in adult L2 learners “the neural architecture is already developed” (p. 78).

According to Harley and Wang (1997) the characteristics that play a role in the CPH are the onset, the terminus, the intrinsic component, the extrinsic component, the affected system and the ultimate causes. Among these the ones that have arisen the most interest among researchers are the onset and offset. However, there is not a consensus on what age corresponds to these characteristics. Lenneberg (1967) proposed an onset at the age of two, when learners start developing syntactic complexity. Other proposals suggested earlier ages such as six months, when babies are sensitive to phonetic categories or at birth, when infants are sensitive to segmental and prosodic distinctions and turn taking (Singleton, 1989). Singleton (1989) discussed the existence of different onsets in different domains of a language.¹ This observation is captured in the notion of multiple critical periods (Seliger, 1978).

For the acquisition of phonology, Long (1990) proposed an upper limit AO of 6 and 12 years, similar to the AOs of other sub-components of language, like morphology or syntax (see studies by Long, 1990; Johnson and Newport, 1989; Hyltemstam, 1992). It is interesting to notice that a low AO does not guarantee a native-like acquisition of a language. There are extrinsic factors that might influence this process. Hyltenstam (1992) suggested that frequency, quality of input and identity issues “interact with maturational constraints for the outcome even at a low age” (p. 364).

Biological factors seem to play a role in explaining why the acquisition of phonology (and also syntax) is easy early in life but difficult with age. Penfield and Roberts (1959) and later, Lenneberg (1967) introduced the idea that there is loss of flexibility or plasticity of the brain with age. Eubank and Gregg (1999) defined the notion of cerebral plasticity as “the ability of neurons to make new connections, and varied connections depending on the stimulus” (p. 69). In fact, the “strengthening of connections between neurons probably represents the neurological basis for learning” (Pulvermueller and Schumann, 1994, p. 691). As Hyltenstam and Abrahamsson (2003) pointed out, this includes language acquisition. They summarized this process, called myelination, as one that plays an important role in the ability of neurons to make connections. This is a physical-chemical process in the brain necessary for providing neurons with nutrition so that they can extend the information transfer to larger distances in the brain. The maturation of the brain is considered to take place when most of the cortical areas have completed the process of myelination. There are two systems of cortical connections between neurons: the A-system and the B-system². The first one uses apical dendrites and axons that connect different cortical areas and the second one uses basal dendrites that are “close to the cell body and local branches of the axons” (Hyltenstam and Abrahamsson, 2003, p. 562).

A strengthening of connections occurs first locally through the B-system and then extends to neighboring cells in the A-system . This could explain how maturation affects the acquisition of phonology. The phonological, as well as the syntactic acquisition, relies on connections within the B-level. However, other aspects of language, like pragmatics or semantics, are less affected since the acquisition of these aspects relies on connections within the A-level (for a more detailed explanation, see Pulvermueller and Schumann, 1994; Hyltenstam and Abrahamsson, 2003).

On the other hand, other factors such as motivational, affective-attitudinal, input or social-psychological factors offer tremendous advantages for late L2 learners (Hyltenstam and Abrahamsson, 2003). Hyltenstam and Abrahamsson (2003) argued that it is undeniable that maturation is responsible for age-related declines in learning. However, the differences between highly proficient and other L2 learners with the same AOs rely on the factors just mentioned. Moyer (1999) demonstrated that the amount and type of instruction plus the degree of motivation are correlated with success in late L2 learners. In the same way, Bongaerts (1999) suggested that intensive training in pronunciation, in combination with high motivation and continued access to L2 input, were crucial for native-like pronunciation in late L2 learners. DeKeyser (2000) proposed that high verbal analytical abilities are essential for acquiring high levels of proficiency in adult L2 learners. In addition to all these factors, Ioup et al. (1994) mentioned the language learning talent as an explanation for the apparent exceptions to the CPH.

Among the critics of CPH, Bialystok and Hakuta (1999) and Hakuta (2003) claimed that there is a linear decline of ultimate proficiency in L2 English with age. They deny that this is due to the existence of a critical period (this decline is constant across age), but rather attribute it to certain cognitive mechanisms such as “working memory capacity, cognitive processing speed and attention” (Hakuta, 2003, p. 32). However, Hyltenstam and Abrahamsson (2003) claimed that there is no co-variation between problem solving, metalinguistic abilities and language proficiency. If these cognitive explanations accounted for age-related differences, there would be different learning processes for children and adults, but research has shown otherwise. “In fact, we see these recent research findings of a linear decline in ultimate attainment as even stronger evidence for the role of maturation than the typical, non-linear pattern” (p. 571). They summarized the studies we saw so far in the following two statements:

1-    Native-like L2 proficiency is only possible in early learners.

2-    Native-like L2 proficiency is possible in early learners and a few late learners.

However, they proposed a third statement to offer a more accurate interpretation of these studies:

3-    Native-like L2 proficiency is not possible in either early or late learners.

They argued that the data of studies with both early and late L2 proficient learners is under analyzed. They claimed that there is not a single subject who achieved native-like L2 proficiency. They assumed that these subjects who passed as natives did so because “they have reached proficiency levels above the limit of perceivable non-nativeness” (pp. 571-572). These are exceptional learners who are not easily recognized as non-natives. In their view, the learning mechanism has to be stimulated from birth in order to avoid its deterioration. They argued that maturation accounts for the decline in learning of subjects with increasing AOs. Nevertheless, it cannot account for the difference between exceptional and non-exceptional late L2 learners within the same AOs³. They suggested that what accounts for these cases are social / psychological factors instead. In this way, they positioned themselves for the existence of maturational constraints.
 
Conclusion

To conclude, I found that there is evidence for the existence of maturational constraints. However, there have been attempts to challenge this view. It is obvious that there is not a consensus regarding whether it is possible to achieve a native-like proficiency in L2 learners. I support Hyltenstam and Abrahamsson’s (2003) perspective that, so far, none of these learners can be considered as natives without additional evidence (i.e. testing all levels of proficiency). Further research and more complete studies are needed on this topic, especially focusing on more levels that would be indicators of language proficiency. Until then I remain skeptical. Regarding the existence or not of maturational constraints in L2, I would argue for their existence.
 
Notes

 
Bibliography

Anisfeld, M., Anisfeld, E., & Semogas, R. (1969). Cross-influences between the phonological systems of Lithuanian- English bilinguals. Journal of verbal learning and verbal behavior, 8, 257-261.

Baralo, M. (1994). Errores y fosilización. Coleccion Aula de Español. Madrid. Fund. Antonio de Nebrija.

Best, C. (1995). A direct realist view of cross-language speech perception. In W. Strange (Ed.), Speech perception and linguistic experience: Theoretical and methodological issues (pp. 171-206). Timonium, MD: York Press.

Best, C. T., Lafleur, R., & McRoberts, G. W. (1995). Divergent developmental patterns for infants’ perception of two non-native contrasts. Infant behavior and development, 18, 339-350.

Best, C. T., McRoberts, G. W., & Sithole, N. M. (1988). Examination of perceptual reorganization for non-native speech contrasts: Zulu click discrimination by English-speaking adults and infants. Journal of experimental psychology: human perception and performance, 14, 345-360.

Bever, T. (1981). Normal acquisition processes explain the critical period for language learning. In K. Diller (Ed.), Individual differences and universals in language learning aptitude (pp. 176-198). Rowley, MA: Newbury House.

Bialystok, E. & Hakuta, K. (1999). Confounded age: linguistic and cognitive factors in age differences for second language acquisition. In D. Birdsong (Ed.),Second language acquisition and the critical period hypothesis. Mahwah, NJ: Lawrence Erlbaum Associates, 161-181.

Birdsong, D. (1992). Ultimate attainment in second language acquisition. Language, 68, 706-755.

Birdsong, D. (Ed.). (1999) Second language acquisition and the critical period hypothesis. Mahwah, NJ: Lawrence Erlbaum Associates.

Bley-Vroman, R. (1990). The logical problem of foreign language learning. Linguistic analysis, 201-202, 3-49.

Bongaerts, T. (1999). Ultimate attainment in L2 pronunciation: The case of very advanced late L2 learners. In D. Birdsong (Ed.), Second language acquisition and the critical period hypothesis. Mahwah, NJ: Lawrence Erlbaum Associates, 133-159.

Bongaerts, T., Mennen, S., & van der Slik, S. (2000). Authenticity of pronunciation in naturalistic second language acquisition: The case of very advanced late learners of Dutch as a second language. Studia linguistica, 54, 298-308.

Bongaerts, T., Planken, B., & Schils, E. (1995). Can late learners attain a native accent in a foreign language? A test of the critical period hypothesis. In D. Singleton & Z. Lengyel (Eds.), The age factor in second language acquisition (pp. 30-50). Clevedon, England: Multilingual Matters.

Bongaerts, T., van Summeren, C., Planken, B., & Schils, E. (1997). Age and ultimate attainment in the pronunciation of a foreign language. Studies in second language acquisition, 19, 447-465.

Butler, Y. G. (2000). The age effect in second language acquisition: Is it too late to acquire native-level competence in a second language after the age of seven? In Y. Oshima-Takane, Y. Shirai, & H. Sirai (Eds.), Studies in language sciences 1,( 159-169). Tokyo: Japanese Society for Language Sciences.

Coppieters, R. (1987). Competence differences between natives and near-native speakers Language, 63, 544-573.

Cutler, A. (2001). Listening to a second language through the ears of a first. Interpreting, 5, 1-18.

Cutler, A. (2002). Native listeners. European Review, 10, 27-41.

Cutler, A. & Butterfield, S. (1992). Rhythmic cues to speech segmentation: evidence from juncture misperception. Journal of Memory and Language, 31, 218-236.

Cutler, A. & Norris, D. G. (1988). The role of strong syllables in segmentation for lexical access. Journal of experimental psychology: Human perception and performance, 14, 113-121.

Cutler, A. & Otake, T. (1994). Mora or phoneme? Further evidence for language-specific listening. Journal of memory and language, 33, 824-844.

Cutler, A., Mehler, J., Norris, D. G., & Segui, J. (1986). The syllable’s differing role in the segmentation of French and English. Journal of memory and language, 25, 385-400.

Cutler, A., Mehler, J., Norris, D., & Segui, J. (1989). Limits on bilingualism. Nature, 340, 229-230.

Cutler, A., Mehler, J., Norris, D., & Segui, J. (1992). The monolingual nature of speech segmentation by bilinguals. Cognitive psychology, 24, 381-410.

DeKeyser, R. M. (1995). Learning second language grammar rules: An experiment with a miniature linguistic system. Studies in second language acquisition, 17, 379-410.

DeKeyser, R. M. (2000). The robustness of critical period effects in second language acquisition. Studies in second language acquisition, 22 (4), 493-533.

Doughty, C. (2003). Instructed SLA. In Doughty, C., & Long, M. (Eds.) The handbook of second language acquisition (pp. 256-310). Blackwell.

Ekberg, L. (1998). Regeltillaempning kontra lexicokunskap I svenskan hos invandrarbarn I Malmoe. In J. Moller, P. Quist, A. Holmen, & J. N. Jorgensen (Eds.),Nordiske Sprog som andetsprog (pp. 99-116). Kovenhavn: Institut for humanistiske fag, Danmarks Laererhojskole.

Elman, J. (1993). Learning and development in neural networks: The importance of starting small. Cognition, 48, 71-99.

Eubank, L. & Gregg, K. R. (1999). Critical periods and (second) language acquisition: divide et impera. In D. Birdsong (Ed.), Second language acquisition and the critical period hypothesis. Mahwah, NJ: Lawrence Erlbaum Associates, 65-99.

Felix, S. & Hahn, A. (1985). Natural processes in classroom second-language learning. Applied linguistics, 63, 223-238.

Flege, J. E. (1992a). The intelligibility of English vowels spoken by British and Dutch talkers. In R. Kent (Ed.), Intelligibility in speech disorders: Theory, measurement and management (pp. 157-232). Amsterdam: John Benjamins.

Flege, J. E. (1992b). Speech learning in a second language. In C. Ferguson, L. Menn, &  C. Stoel-Gammon (Eds.), Phonological development, models, research, and applications (pp. 565-604). Parkton, MD: York Press.

Flege, J. E. (1995). Second-language speech learning: Findings, and problems. In W. Strange (Ed.), Speech perception and linguistic experience: Theoretical and methodological issues (pp. 233-273). Timonium, MD: York Press.

Flege, J. E. (1999). Age of Learning and Second Language Speech. In D. Birdsong (Ed.), Second language acquisition and the critical period hypothesis (pp. 101-131). Mahwah, NJ: Lawrence Erlbaum Associates.

Hakuta, K. (2003). Critical evidence: A test of the critical-period hypothesis for second language acquisition. American psychological society, 14, 31-38.

Harley, B. & Hart, D. (1997). Language aptitude and second language proficiency in classroom learners of different starting ages. Studies in second language acquisition, 19, 379-400.

Harley, B. & Wang, W. (1997). The critical period hypothesis: where are we now? In M. B. De Groot and J. F. Kroll (Eds.), Tutorials in bilingualism: Psycholinguistic perspectives. London: Lawrence Erlbaum Associates.

Ho, D. (1986). Two contrasting positions on second-language acquisition: A proposed solution. International review of applied linguistics, 24, 35-47.

Hurford, J. (1991). The evolution of the critical period for language acquisition. Cognition, 40, 159-201.

Hyltenstam, K. (1988). Lexical characteristics of near-native second-language learners of Swedish. Journal of multilingual and multicultural development, 9, 67-84.

Hyltenstam, K. (1992). Non-native features of near-native speakers: On the ultimate attainment of childhood L2 learners. In R. J. Harris (Ed.), Cognitive processing bilinguals. Amsterdam: Elsevier Science, 351-368.

Hyltenstam, K. & Abrahamsson, N. (2003). Maturational constraints on SLA. The handbook of second language acquisition (pp. 539-588). Blackwell.

Hyltenstam, K. & Abrahamsson, N. (forthcoming). Age of onset and ultimate attainment in near-native speakers of Swedish. In K. Fraurud & K. Hyltenstam (Edss),Multilingualism in global and local perspectives: Papers from the 8th Nordic Conference on Bilingualism, November 1-3, 2001, Stockholm-Rinkeby. Stockholm: Rinkeby Institute of Multilingual Research.

Ioup, G. (1995). Evaluating the need for input enhacement in post-critical period language acquisition. In D. Singleton & Z. Lengyel (Edss), The age factor in second language acquisition. Clevendon: Multilingual Matters, 95-123.

Ioup, G., Boustagui, E., Tigi, M., & Moselle, M. (1994). Reexamining the critical period hypothesis: a case study in a naturalistic environment. Studies in second language acquisition, 16, 73-98.

Johnson, J. S. & Newport, E. L. (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, 60-99.

Lenneberg, E. (1967). Biological Foundations of Language. New York: John Wiley.

Liceras, J. M. (1992). La adquisición de lenguas extranjeras: hacia un modelo de análisis de la interlengua. Madrid. Visor.

Long, M. (1990). Maturational constraints on language development. Studies in second language acquisition, 12, 251-285.

Macnamara, J. (1973). Nurseries, streets, and classrooms. The modern language journal, 57, 250-254.

Marchman, V. A. (1993). Constraints on plasticity in a connectionist model of English past tense. Journal of cognitive neuroscience, 5, 215-234.

Mayberry, R. I. (1993). First-language acquisition after childhood differs from second-language acquisition: The case of American sign language. Journal of speech and hearing research, 36, 1258-1270.

Moyer, A. (1999). Ultimate attainment in L2 phonology: The critical factors of age, motivation, and instruction. Studies in second language acquisition, 21, 81-108.

Obler, L. K. & Hannigan, S. (1996). Neurolinguistics of second language acquisition and use. In W. C. Ritchie & T. K. Bhatia (Edss), Handbook of second language acquisition. San Diego: Academic Press, 509-523.

Oyama, S. (1979). The concept of the critical period in developmental studies. Merrill-palmer quarterly, 25, 83-103.

Patkowski, M. (1990). Age and accent in a second language: A reply to James Emil Flege. Applied linguistics, 11, 73-89.

Penfield, W. & Roberts, L. (1959). Speech and brain mechanisms. New York: Athenaeum.

Pulvermueller, F. & Schumann, J. H. (1994). Neurobiological mechanisms of language acquisition. Language learning, 44, 681-734.

Seliger, H. W. (1978). Implications of a multiple critical periods hypothesis for second language learning. In W. Ritchie (Ed.), Second language acquisition research. New York: Academic Press, 11-19.

Singleton, D. (1989). Language acquisition. The age factor. Clevendon: Multilingual Matters.

Weber, A. (2000). The role of phonotactics in the segmentation of native and non-native continuous speech. In A. Cutler & J.M. McQueen (Eds), Proceedings of the workshop on spoken word access processes. Nijmegen: MPI, 143-146.

White, L. & Genesee, F. (1996). How native is near-native? The issue of ultimate attainment in adult second language acquisition. Second language research, 12, 233-265.

Zobl, H. (1989) Modularity in adult L2 acquisition. Language learning, 49-79.

van Zon, M. D. C. M. (1997). Speech processing in Dutch: A cross-linguistic approach.  Unpublished doctoral dissertation. Tilburg University.