Japanese speakers learning r and l

Japanese has only one liquid consonant, a lateral flap IPA|/ɺ/, while English has an alveolar lateral approximant IPA|/l/ and an alveolar or retroflex central approximant IPA|/r/. When Japanese speakers learn English later than childhood, they typically have difficulty in hearing and producing the difference between IPA|/l/ and IPA|/r/.

Perception

Harvcoltxt|Goto|1971 reports that native speakers of Japanese who have learned English as adults have difficulty perceiving the acoustic differences between English IPA|/r/ and IPA|/l/, even if the speakers are comfortable with conversational English, have lived in the U.S. for extended periods, and can articulate the two sounds when speaking English.Harvard reference
last=Goto
first=Hiromu
year=1971
title=Auditory perception by normal Japanese adults of the sounds "l" and "r""
journal=Neuropsychologia
volume=9
pages=317-323]

Japanese speakers are, however, able to perceive the difference between English IPA|/r/ and IPA|/l/ when these sounds are not mentally processed as speech sounds. Harvcoltxt|Miyawaki|Strange|Verbrugge|Liberman|1975 found that Japanese speakers could distinguish IPA|/r/ and IPA|/l/ just as well as native English speakers if the sounds were acoustically manipulated in a way that made them sound less like speech (by removal of all acoustic information except the F3 component).Harvard reference
last=Miyawaki
first=Miyawaki
last2=Strange
first2=W.
last3=Verbrugge
first3=R.R.
last4=Liberman
first4=A.M
last5=Jenkins
first5=J.J.
last6=Fujimura
first6=O.
year=1975
title=An effect of linguistic experience: the discrimination of [r] and [l] by native speakers of Japanese and English
journal=Perception and Psychophysics
volume=18
issue=5
pages=331-340]

Harvcoltxt|Lively|Pisoni|Yamada|Tohkura|1994 found that speakers' ability to distinguish between the two sounds depended on where the sound occurred. Word-final IPA|/l/ and IPA|/r/ with a preceding vowel were distinguished the best, followed by word-initial IPA|/r/ and IPA|/l/. Those that occurred in initial consonant clusters or between vowels were the most difficult to distinguish accurately.Harvard reference
last=Lively
first=S.E.
last2=Pisoni
first2=D.B.
last3=Yamada,
first3=R.A.
last4=Tohkura
first4=Y.I.
last5=Yamada
first5=T
year=1994
title=Training Japanese listeners to identify English /r/ and /l/: III. Long-term retention of new phonetic categories.
journal=Journal of the Acoustical Society of America
volume=96
issue=4
pages=2076–2087 ]

Harvcoltxt|Bradley|Pisoni|Yamada|Tohkura|1997 provide evidence that there is a link between perception and production to the extent that perceptual learning generally transferred to improved production. However, there may be little correlation between degrees of learning in perception and production after training in perception, due to the wide range of individual variation in learning strategies.Harvard reference
last=Bradlow
first=A.
last2=Pisoni
first2=D
last3=Yamada
first3=R.A.
last4=Tohkura
first4=Y
year=1997
title=Training Japanese listeners to identify English /r/ and /l/: IV. Some effects of perceptual learning on speech production
journal=Journal of the Acoustical Society of America
volume=101pages=2299-2310]

Production

Harvcoltxt|Goto|1971 reports that Japanese speakers who are unable to hear the difference between IPA|/r/ and IPA|/l/ may still learn to produce the difference, presumably through articulatory training in which they learn the correct places and manners of articulation required for the production of the two sounds. In this sense, they learn to produce IPA|/r/ and IPA|/l/ in much the same way a deaf person would. Although they have only a single acoustic image corresponding to a single phoneme intermediary between IPA|/r/ and IPA|/l/, they are able to know they are producing the correct sound based on the tactile sensations of the speech articulators (i.e. tongue, alveolar ridge, etc.) coming into contact with each other without any auditory feedback or confirmation that they are indeed producing the sound correctly.

Variations in acquisition

There is some indication that Japanese speakers tend to improve more on the perception and production of IPA|/r/ than IPA|/l/.

Harvcoltxt|Aoyama|Flege|Guion|Akahane-Yamada|2004 conducted a longitudinal study examined the perception and production of English /l/, /r/, and /w/ by adults and children who were native speakers of Japanese but living in the US. Over time, the children improved more on English /r/ than English /l/.Harvard reference
last=Aoyama
first=Katsura
last2=Flege
first2=James Emil
last3=Guion
first3=Susan
last4=Akahane-Yamada
first4=Reiko
last5=Yamada
first5=Tsuneo
year=2004
title=Perceived phonetic dissimilarity and L2 speech learning: the case of Japanese /r/ and English /l/ and /r/
journal=Journal of Phonetics
volume=32
pages=233–250 ]

Similarly, Harvcoltxt|Guion|Flege|Akahane-Yamada|Pruitt|2000 found that Japanese speakers who received training in distinguishing English sounds improved more on IPA|/r/ than on IPA|/l/. They suggest that English IPA|/l/ is perceived as more similar to Japanese IPA|/ɺ/ than English IPA|/r/ is, and hence it is harder for Japanese speakers to distinguish IPA|/ɾ/ from IPA|/l/ than IPA|/ɺ/ from IPA|/r/.Harvard reference
last=Guion
first=Susan
last2=Flege
first2=James Emil
last3=Akahane-Yamada
first3=Reiko
last4=Pruitt
first4=JC
year=2000
title=An investigation of current models of second language speech perception: The case of Japanese adults’ perception of English consonants
journal=Journal of the Acoustical Society of America
volume=107
issue=5 Pt. 1
pages=2711-2724]

Harvcoltxt|Kuzniak|Zapf|2004 found differences between the second and third formants in IPA|/r/ and IPA|/l/ of a native Japanese speaker and a native English speaker. The results showed that the Japanese speaker had a hard time producing an English-like third formant, especially that which is required to produce an /l/.Harvard reference
last=Kuzniak
first=Kinnaird
last2=Zapf
first2=Jennifer
year=2004
title= [https://www.indiana.edu/~iulcwp/pdfs/04-kinnaird.pdf An acoustical analysis of a Japanese speaker's production of English /r/ and /l/] ]

Effects of training

There have been a number of experiments in training Japanese subjects to improve their perception of IPA|/r/ and IPA|/l/.Harvard reference
last=Logan
first=John
last2=Lively
first2=Scott
last3=Pisoni
first3=David
year=1991
title=Training Japanese listeners to identify English /r/ and /l/: afirst report
journal=Journal of the Acoustical Society of America
volume=89
pages=874–886]

Harvcoltxt|Lively|Pisoni|Yamada|Tohkura|1994 found that monolingual Japanese speakers in Japan were able to increase their ability to distinguish between /l/ and /r/ after a 3 week training period, which involved hearing minimal pairs (such as 'rock' and 'lock') produced by five speakers, and being asked to identify which word was which. Feedback was provided during training, and participants had to listen to the minimal pairs until the correct answer was given. Participants performed significantly better immediately after the 3-week training, and retained some improvements when retested after 3 months and after 6 months (although there was a decrease in recognition ability at the 6 month test). Reaction time decreased during the training period as the accuracy went up. Participants were able to "generalize" their learning somewhat: when tested they could distinguish between new /l/ and /r/ minimal pairs, but performed better when the pairs were said by one of the 5 speakers they had heard before than a new speaker.

Harvcoltxt|Lively|Logan|Pisoni|1993 also found that subjects who were trained by listening to multiple speakers' production of IPA|/r/ and IPA|/l/ in only a few phonetic environments improved more than subjects who were trained with a single talker using wider range of phonetic environments. [Harvard reference
last=Lively
first=Scott
last2=Logan
first2=John
last3=Pisoni
first3=David
year=1993
title=Training Japanese listeners to identify English /r/ and /l/: II. The role of phonetic environment and talker variability in new perceptual categories
journal=Journal of the Acoustical Society of America
volume=94
pages=1242-1255]

Harvcoltxt|McClelland|Fiez|McCandliss|2002 argue that it's possible to train Japanese adults to distinguish speech sounds they found difficult to differentiate at first. They found that speech training results in outcomes indicating a real change in the perception of the sounds as speech, rather than simply in auditory perception.Harvard reference
last=McClelland
first=J.L.
last2=Fiez
first2=J.A.
last2=McCandliss
first2=B.D.
title=Teaching the /r/-/l/ Discrimination to Japanese Adults: Behavioral and Neural Aspects
volume=77
pages=657-662 ]

However, it is not clear whether adult learners can ever fully overcome their difficulties with IPA|/r/ and IPA|/l/. Harvcoltxt|Tagaki|Mann|1995 found that even Japanese speakers who have lived 12 or more years in the US have more trouble identifying IPA|/r/ and IPA|/l/ than native English speakers do.Harvard reference
last=Tagaki
first=Naoyuki
last2=Mann
first2=Virginia
year=1995
title=The limits of extended naturalistic exposure on the perceptual mastery of English /r/ and /l/ by adult Japanese learners of English
journal=Applied Psycholinguistics
volume=16
issue=4
pages=379-405]

Examples

There are numerous minimal pairs of words distinguishing only IPA|/r/ and IPA|/l/. For their study, Harvcoltxt|Kuzniak|Zapf|2004 used::Right/light:Red/Led:Load/Road:Climb/Crime:Bled/Bread:Flows/Froze:Alive/Arrive:Collect/Correct:File/Fire:Mole/More

References

See also

* English phonology
* Non-native pronunciations of English