The Language Teacher
November 2001

Romazi, Roomazi, Rohmaji, and Romaji: Psycholinguistic Implications of Shared L1-L2 Orthography in L2 Processing

Laura Fukunishi

Kobe City University of Foreign Affairs


The Problem

Documentation of a bewildering variety of mistakes in children's classroom work led to the formulation of an hypothesis: the introduction of romazi in the 4th grade of elementary school interrupted normal foreign language acquisition and caused serious confusion in the children's minds. Interference occurred in their confident acquisition and categorization of a foreign language, in this case, English.

Elementary school children's English orthographic and, subsequently, oral language processing, was being damaged, as it remained based on the visually similar but phonologically bi-valent systems of the English alphabet and the romazi tuzuri (writing/spelling). Not only by its study in the educational system but also in its use in the environment, romazi (romaji, roomazi, and rohmaji) was interfering with normal foreign language learning, students' speech comprehension and production, and even the psychological process of learning itself.

The Experiment

Subjects

Once the idea of romazi interference was postulated, it was decided to test both a control group and a subject group consisting of students aged 9, 10, and 18, in grades 3, 4, and 11, respectively. It was thought that this selection would show the range and demarcation of no romazi effect in a pre-learning year, great disturbance in the romazi class year, and the tapering off of the effect some years later.

The control group consisted of students at an international school who never studied romazi, but when studying Japanese, were told to transcribe new words using their own languages' alphabets. To further increase the validity of the romazi-effect research, this investigator included two other Japanese control groups: kindergarteners and college students.

Methods

Listening tests were conducted using both English and nonsense words, and the answers were recorded on an English distinctive features chart. All answers on the chart showed the sound spoken and the sound perceived.

Results

The kindergarteners' mistakes were normal, developmental ones with no romazi influence. College student answers showed that the longer the word, the better the chance to understand or produce the same sound. They appeared to hear the English, change the pronunciation to katakana and then write romazi. Romazi's influence had decreased only very slowly. The control groups did not show a statistically satisfying contrast, but the phonic mapping showed a combination of environmental and AFL factors in their mistakes or misperceptions concentrated in certain expected areas of the feature chart grid.

Discussion

The 4th graders' decisions showed they believed the visually learnt romazi orthography was actually a fair representation of spoken English, or even English itself. Learning their L1 Japanese phonemes by activating new neural responses appropriate to L2 graphemes, the 4th graders, reversing the procedure, used the assembly method. Ravid, 1996, posited that the Hebrew nikud orthography interfered not only with native speakers' reading, but that its ÅghelpÅh also had a deleterious effect on immigrant speakers' ability to process the language quickly.) The 4th graders were using a L2 learning method as their basis, but were putting into effect the already-empowered romazi phoneme-grapheme transfer, illustrating Jacobvits' ideas on transfer theory (Language Learning, No. 19, pp.55-86) on assuming equivalency of form and function of one system to another.

For these Japanese L2 learners, the normal holistic identification of the full orthographic form of the foreign word, which gives direct access to its phonological form in the lexicon, is being interfered with by the letter-by-letter method of assemblage, the analytic mapping of misapplied phonological elements onto the transparently supported orthographic segments: looking for an English word in an all-Japanese dictionary.

For them, romazi, a totally transparent orthography, the basis of which is stored in long-term memory, gives two possible lexicons. However, it is evident that the phonological information of Japanese is usually processed first and most quickly by sub-lexical, spelling-sound correspondence, and that it influences and interferes with L2 visual word recognition by prompting L1 retrieval instead of the target, L2. Grainger, (1993), doing work on visual word recognition, says that isolating and fixating the correct orthographical and phonological description in long-term memory leads to correct word meaning identification. In addition, Garman's (1996) work on perception adds this information on the human learning habit: visual perception involves integration of what we already know and what we expect to know.

Language learning is based on three points:

However, romazi contributes to 3 types of psychological stress which are pernicious to language learning: stress caused by the learner's using a dysfunctional code; stress caused because the code has been authoritatively provided and will be used to judge his ability; and stress leading to anxiety that he will most probably make a mistake, one which he almost cannot help making. This last step in the Ministry of Education's program is most deleterious because the learner believes that he is the cause of the error. Subsequently, with the seeds of failure thought to be within, the student is more unwilling to try to speak or express himself orally or on paper. This reaction would seem to credit Titone's Holodynamic Model (Titone, 1989, 1994) which posits a hierarchical structure of Tactics, Strategy, and Ego language-learning layers, all based on and rooted in the deepest layer of the Ego, its experience, and its basic characteristic of desiring to communicate if it has had unimpeded, normal development.

The results of the experiment, although carried out in the classroom rather than in the laboratory, show that when the acquisition of English is still in an unresolved stage, competing (romazi) stimulation has a deleterious effect on the acquisition of English on the morpho-phonological level. The results also show that, unfortunately, the psychological conditioning which occurred in the 4th grade, which now includes all Japanese under the age of 67, continues to produce feelings of discomfort. Their discomfort and/or apathy towards school are in large part based upon dislike and fear of the English language in particular, and upon fear of their inability to learn in general.

Solution

The solution to the problem is very simple: remove romazi from the 4th grade curriculum and reschedule it for junior high school or high school. There is no need for any Japanese to write anything in romazi except, perhaps, names and addresses. The second recommendation is for the Ministries of Communications, Transportation, Education, and Foreign Affairs to comply with the 1954 law signed into effect by the then Prime Minister Yoshida, which states that all different romazi's must consolidate and conform to the international, already agreed-upon 1934 ISO romazi alphabet. This means that all street signs, all train station signs, all passports, and all national and international documents using romazi must conform to a single romazi system.

References & Bibliography

Fukunishi, L., & Kohno, M. (1998). [Renzo Titone and bilingual education and the development of metalinguistic abilities: A research project]. IJPL, 10[1]. 5-14 Vol. 3, .

Garman, M. (1996). Psycholinguistics, Cambridge: CUP.

Grainger, D. (1993). Visual word recognition in bilinguals. Linguistics, 22, 11.

Kohno, M. (1998). Language Specific Developmental Changes in Speech Perception Abilities, and their Neuropsychological Reason ­ A Hypothesis. In Proceedings of the Int. Confer. of ISAPL Portugal.

Kohno, M. (1998). The Learning and Psychology of Language. Osaka: Kinseido.

Paradis, M. (1993). Linguistic, psycholinguistic, and neurolinguistic aspects of ÅginterferenceÅh in bilingual speakers: The Activation Threshold Hypothesis. IJPL, Vol. 9, No.2 [25], 133-145.

Ravid, D. (1995). Language Change in Child and Adult Hebrew: A Psycholinguistic Perspective. New York: OUP.

Ravid, D. (1996). Accessing the mental lexicon: evidence from incompatibility between representation of spoken and written morphology. Linguistics, 34, 1219-1246.

Saeki, K. & Yamada, H. (1977). The Romanization of Japanese writing: Hepburn vs. Kunrei system controversies. Tokyo: Zaidan hozin Nippon no romazi Sya.

ISO/TC46SC2, Paris, France. Resolution of the 16th Plenary Meeting of ISO/TC46 Documentation (Including ISO/TC46N1170E/F(1976/5). Brussels: Author.

Tatuoka, H. & Yamada, H. (1996). On the Theory and practice of the romanized writing system for Japanese. Gakuzitu Zyoho Senta Kiyo, Vol. 8

Titone, R. (1989). On the Bilingual Person. Ottawa: Univ. of Toronto Press.


All materials on this site are copyright © by JALT and their respective authors.
For more information on JALT, visit the JALT National Website