This study aims to examine Korean notations and Korean phonemes in the early 20th century. This study investigated fifteen Japanese learning readers published between the 1900s and the 1930s, focusing on transitional characteristics from early modern Korean to late modern Korean, and discussing them. These Japanese learning readers were very practical and used so many conversational expressions that they tended to be colloquial. Thus, based on historical linguistics and generative phonology, this study analyzes the changes in notations and phonemes, and alternations in phonemes.
Chapter 2 of this study examines Korean notations in Japanese learning readers. These notations were divided into Korean notations of Korean and those of Japanese. Korean notations of Korean in Japanese learning readers had a strong tendency to be conservative like early modern Korean. All literature still much used the notations of compound consonants like ㅺ, ㅼ, ㅽ for tense sounds. For final consonants, no other consonants than ㄱ, ㄴ, ㄹ, ㅁ, ㅂ, ㅅ, ㅇ, ㄺ, ㄻ, and ㄼ were used. The notation of saisiot (inserted ㅅ) was primarily used as a marker of compounds, and secondarily as evidence of tensification. Between diphthongs, [ll] was still used as ㄹ and ㄴ.
Korean notations of Japanese were divided into three parts: voiceless sounds, voiced sounds, and aspirated sounds. For voiceless sounds, チ was marked as ‘디’ (di), and ツ as ‘두’(du) in some textbooks, helping learners easily memorize what they had learned. In other textbooks, チ was marked as ‘지’(?i), and ツ as ‘쓰(s), or 츠(?)’, which shows that these notations were close to their pronunciation. As for voiced sounds, it is examined how voiced obstruents were marked in Korean phonemes. Korean notations of voiced obstruents in literature were classified into five types (type 1: 가-ga, type 2: ㅺㅏ, type 3: ㅇ가, type 4: 가＂, and type 5: ?ㅏ). For aspirated sounds, Korean notations of voiceless bilabial stops were examined. There were also five types of these notations (type 1: 바-ba, type 2: 파-pa, type 3: ?ㅏ, type 4: ㅇ바, and type 5: ㅽㅏ). Among these notations, type 5 of voiced sounds and type 3 of aspirated sounds were marked closest to their pronunciation.
Chapter three of this study analyzes changes in phonemes. The following characteristics were found. ㅚ was pronounced as [o], [we], or [e], and ㅟ was pronounced as [u] or [wi], which showed transitional features. Vowel raising, vowel fronting, i-regressive assimilation, vowel rounding, and vowel unrounding extended more widely than in early modern Korean. With these phonological changes extending and widely used, speakers showed hypercorrection, intentionally refusing phonological changes for palatalization, ㅢ monophthongization, and vowel fronting. Word forms to which phonological changes in early modern Korean did nor apply appeared in Japanese leaning readers. These examples were confirmed in palatalization, vowel rounding, vowel unrounding, vowel fronting and i-regressive assimilation, which shows the conservativeness of the language.
Chapter 4 of this study analyzes phonological alterations. Phonological alterations such as tensification, nasal assimilation, place assimilation, glide formation and glide insertion occurred in the same phonological environments as phonological alterations in late modern Korean. The difference from contemporary Korean is that Japanese learning readers in the early 20th century reflected phonologically altered pronunciation in their notations. This confirms the transitional characteristics of Korean in the early 20th century.