1. Classification of the Buyeo Languages

The ancient Koreanic languages are divided into two branches: the Buyeo languages and the Han languages.^[1] According to ancient Chinese records, the Goguryeo languages appear to have been different from ancient Tungusic languages like Mohe.^[2][3][4] The Japonic languages are divided into the Japanese branch and the Ryukyuan branch.

A number of linguists such as Kim Banghan, Vovin and Unger classify Goguryeo as Old Korean. They note that the toponyms which resemble Japanese are mostly found in the central part of the Korean peninsula, and theorize that they do not reflect the Goguryeo language but rather the language of the pre-Goguryeo population in the central and southern parts of the Korean peninsula. Since a number of Japanese-like toponyms found in the historical homeland of Silla^[5] were also distributed in the southern part of the Korean peninsula, these linguists propose that there was once a Japonic language spoken on the Korean peninsula, which forms a substratum of the Silla language; Unger suggests that the ancestors of the Yayoi people would have settled in Japan and came from the central or southern part of Korea. None of the Japanese cognates have been found in the historical homeland of Goguryeo in the northern part of the Korean peninsula or south-western Manchuria. Koreanic toponyms, on the other hand, are distributed across the entire territory of the Three Kingdoms, from Manchuria to the southern Korean peninsula.

1.1. Japanese–Koguryoic Hypothesis

The linguist Beckwith suggests that Japanese is closely related to the language of Goguryeo and both are part of the Buyeo languages. The Korean state of Baekje (18 BCE — 660 CE) was founded by Goguryeo (37 BCE – 668 CE) princes, and considered itself descended from the kingdom Buyeo (2nd century BC — 494 AD). Baekje subsequently had close relations with Yamato period Japan; Christopher Beckwith suggests that at that point the Japanese may have still recognized a relationship to Buyeo. Beckwith reconstructs about 140 Goguryeo words, mostly from ancient place names, including Gaya.^[6] Many include grammatical morphemes which appear to be cognate with morphemes of similar function in Japanese, such as genitive -no and attributive -si. In a review of Beckwith's book for Korean Studies, Thomas Pellard criticized Beckwith's work for serious methodological flaws, such as rejecting mainstream reconstruction of Chinese and Japanese and using his own instead. He suggests that the language of Goguryeo is most closely to Koreanic but both Koreanic and Japonic are both related and from one common proto-language.^[7]

1.2. Problems and Contradictions of Japanese-Koguryoic Hhypothesis

The footsteps of the Goguryeo language versus the Ancient Koreanic footsteps found in Manchurian and Japanese

Many of the existing studies have been done in the way of finding the "Japanese footprints" characteristic in the Goguryo language. If this is the case, we should point out that the "Japanese footprints" of the Goguryeo and Manchu languages, which were deeply related to the Goguryeo language during the Goguryeo-Balhae (Bohai) period, should be found. However, Alexander Vovin's research has not found such findings. (This is a theory that denies Goguryeo's claim that it is a systematic language like the Japanese languages) Instead, it says that Korean-related characteristics can be found in many ways - in vocabulary and grammar. These characteristics are not found in other Tungusian languages but are found only in the Jurchen and Manchu languages. It is clear that this is a foreign element rather than a characteristic of Tungusu. Vovin interpret it as having received.

Korean morphological characteristics found in Goguryeo inscription

Vovin discusses the morphological characteristics of Korean revealed in the analysis of Koguryo inscriptions. The use of '之' is similar to that of the Silla inscription, and it is said that such use is not found in Japanese. It also says that the usage of '伊' is related to ancient Korean.

The usage of '伊' is also found in Ancient Western-Japanese, but it is not found in the whole of the Japanese archipelago. This interprets ancient Western-Japanese as the result of Korean influence. (An analysis of the fact that the ancient Western-Japanese is related to the Korean language is analyzed in detail in the following.) (The specific usage is consecutively found in the Goguryeo inscription, Ancient Korean, Ancient Japanese, Can be seen as an indication of homogeneous language usage on the peninsula.)

Baekje's "bilingual system" and Baekje vocabulary found in Ancient Western-Japanese

Based on the two expressions of '鞬吉支 (geon-gil-ji)' and '於羅瑕 (eoraha)', the claim that Baekje was a bilingual state is a well-known hypothesis.

However, Vovin says that these claims are poorly grounded. In the meantime, we analyze the Baekje language found in Nihonshoki. It is also analyzed that the Baekje language was introduced into Ancient Western-Japanese. But Vovin insists these vocabularies are related to ancient Korean. Vovin has three circumstantial evidence that these vocabularies are related to ancient Korean, apart from specific linguistic analysis. First, some of these vocabulary flowed into Japanese in antiquity, and this distinctive vocabulary was found only in Ancient Western-Japanese. Second, there is a vocabulary used throughout the Japanese archipelago that corresponds to a semantic pair in these vocabularies. (This means that the vocabulary was introduced as a foreign word in Ancient Western-Japanese, with the Japanese already spreading widely.) Third, these vocabularies are etymologically 'too obvious' to Old/Middle Korean.

Text in Baekje language

Vovin claims that Seodongyo, a Hyangga of the Silla period, was used as a Baekje language.

Seodongyo uses '乙' (ur) as a cover letter. "Gyunyeo`s Hyangga"(均如鄕歌) and "Doijangga"(悼二將歌), except for all the other same thing is used as a cover letter '肸' (gur) is. Vovin interprets this as 'g' in the Baekje language (the dialect of ancient Korean) as 'g'. (g is easy to drop out as a researcher). However, there is no use of '肸' in the proper noun of the Baekje language of the Samguksagi, based on the fact that '乙' is used several times. This is the phonological characteristic of the Baekje dialect, This is also reflected in Seodongyo.

1.3. Korean-Japanese

The theory that Korean is related to Japanese has supporters due to similar grammatical features that have been elaborated upon by such researchers as Samuel E. Martin^[8] and Roy Andrew Miller.^[9]

Some of the common features are:^[10]

• the languages are agglutinative,
• follow the SOV order,
• nouns and adjectives follow the same syntax,
• particles are post-positional,
• modifiers always precede modified words.

The two languages have previously been thought to not share any cognates (other than loanwords),^[11] for their vocabularies do not phonetically resemble each other. However, a recent 2016 paper proposing a common lineage between Korean and Japanese traces around 500 core words that show a common origin. ^[12]

Some previous linguists have listed a tiny amount of translations which are perceived as phonetic parallels and cognates. This is exemplified in the following table:

2. Languages of the Buyeo Family

• Buyeo languages
  • Ye-Maek language
  • Gojoseon language (It is also associated with Silla language)
  • Okjeo language
  • Buyeo state language
  • Goguryeo language
    • Old Baekje
  • New Baekje
  • Sillan language
    • Korean language
    • Jeju language
  • Gaya language
    • Japonic languages
      • Japanese language
      • Ryukyuan languages