Japanese has been genetically linked to several other languages. We can conveniently divide the various theories into basically two kinds according to how scholars, linguists in particular, have responded in terms of interest and research. The first category, the inner circle, refers to studies conducted by several persons over many decades and commanding a wide range of support; the outer circle encompasses those small-scale studies sometimes associated with the work of just one person. Let us outline these two categories very briefly.
The Outer Circle
1. Tibeto-Burman and Japanese
Percival Lowell first hypothesized that Burmese was related to Japanese in a paper before the Asiatic Society of Japan in 1891. The relation was developed by C. K. Parker in the Dictionary of Japanese Compound Verbs in 1939 and later in the Japanese language edition entitled Japanese and Tibeto-Burman are Sister Languages. Parallels were made between grammatical structures. However, Parker, and before him G. Taylor (Environment, Race and Migration 1937), suggested that migratory groups of Tibeto-Burmese, Thai and Mon-Khmer over several generations expanded along the coast of the China Sea and down the west coast of Korea finishing in Japan. Parker's theories were later expanded in the work Various Tribes of southern Asia and the Japanese and have received some support from scholars such as Susumu Ono (1970).
2. Mon-Khmer
Based on his study of the lexicon for parts of the body in the Semang and Sakai dialects of Mon and Khmer, Matsumoto (1942) noted similarities with Old Japanese. The formation of pronouns also seemed to show common characteristics.
3. Sumerian
Junichiro Takakusu (1944) made passing references to various ethnographic similarities between Sumerian and other languages of Asia notably Mon-Khmer, Melanesian and Japanese. This comparison was taken up by R.Yoshiwara (1991) who pointed out the following shared characteristics between Sumerian and Japanese: agglutination, SOV word- order, pre- and post-positioning, vowel harmony, no morphological distinction for gender, asyndetic usage of conjunctions and other features.
The Inner Circle
1. Austronesian
Japanese vocabulary seems to contain many elements of Malayo-Polynesian or Austronesian (a proto-language for which we have no records). Research began in this area with the important paper by Matsumoto Nobuhiro in 1928 on "Le Japonais et les langues austro-asiatiques" and following this the work of Otto Dempwolff (1934-38) and lzui Hisanosuke (1952). Murayama (1974) pointed out the large number of similarities between the morphology of Japanese and Oceanic languages and in a seminal study (1975) succeeded in explaining the etymologies of the entire Japanese numeral system (1-1,000) by means of Proto-Austronesian morphemes.
Kawamoto Takao (1985) is currently the leading proponent of the Austronesian connection. Based upon fieldwork in the Pacific, Kawamoto (1976) reconstructed the proto-system of all the Japanese verb combinations based on the incomplete reconstructions of Susumu Ono (1953) and indicated the following shared features: presence of phonemic accent, tendency to disyllabism and canonical morphemic shapes in Old Japanese and Proto-Austronesian, syllables closed with special phonemes only, vowel harmony, agglutination, SVO and adjective-noun order, question forms made by adding a particle to a statement, derivation by vowel mutation, vocalization or nasalization, plurality expressed by affixation and reduplication and other shared features.
In the vocabulary stock, particularly those items dealing with marine life, many convincing comparisons have been made suggesting lexical contact between Japanese (J) and proto-Austronesian (PAN): J ika 'cuttlefish' and PAN ikan 'fish'; J hana 'flower' and PAN buna 'flower'. Additionally, many of the so-called 'vulgarisms' of the Fudoki texts have been linked with Austronesian, in particular: OJ (Old Japanese) isa 'whale' from PAN i'ti ; OJ fisi 'sandbar' from PAN pat'iy.
2. The Ural-Altaic Theory
The Ural-Altaic language comprises a wide range. Ural includes present-day Finnish and Hungarian. Altaic indicates Turkish, Mongolian, Samoyed, Tungus and Korean. H. J. Klaproth, the German traveller and scholar is credited with first suggesting the relation between Japanese and the Ural-Altaic languages (Asia Polyglotta 1823). Castren (1857) further elaborated the category of 'Altaisch' and ongoing research on language typology led subsequent investigators to place Japanese and Korean in that group: Rosny (1864), Winkler (1884, 1894), Grunzel (1895). The Austrian Anton Boller (1857) emphasised the connection thus: "Nachweiss, dass das Japanische zum ural-altaischen Stamme gehort". J. Hoffman, Professor of Japanese at Leiden University voiced the same opinion (1867). By the turn of the century, evidence for the Altaic relationship led at least one scholar of Old Japanese to describe it as: "the primal extant deliverances of the whole Ural-Altaic stock" (1906: 2xxv).
Much attention has been focused on vowel harmony in the establishment of the Japanese- Altaic connection. K. Fujioka in 1908 isolated fourteen essential features of Ural-Altaic languages and identified thirteen of those features shared by Japanese: all except vowel harmony. Shinkichi Hashimoto's discovery that 8th century Japanese had 3 more vowels than had previously been thought led to the discovery that Japanese also exhibited vowel, harmony as in Ural-Altaic. Roy Miller (1971) in his Japanese and Other Altaic Languages, Shibata Takeshi (1949) and a substantial number of other scholars now favour the Altaic connection.
3. Japanese and Korean
The British diplomat W. G. Aston published his pioneering 'A Comparative Study of the Japanese and Korean Languages' in 1879. From his analysis of 100 pairs of words, Aston concluded that Korean and Japanese were related to the same extent as two distant languages in the Indo-European family. Meiji scholars continued to research the etymology of Korean elements in Old Japanese. Hattori (1898) supported this view with a study of 200 lexical items. Kanazawa (1910) as much with eye towards political harmony as towards linguistics concluded in 'The Common Origin of the Japanese and Korean Languages' that both Korean and Ryukyuan were "branches of our Great Japanese Empire". The Finnish scholar Rarnstedt was guardedly affirmative regarding the connection between Japanese and Korean using historical-comparative linguistics to compare *sima 'island' and *pai 'boat' in his paper 'Two Words of Korean and Japanese' (1926). The Koreanist Ogura Shinpei in his 1920 study remarked that between these languages "in some points greater or lesser different, but anyone could see the predominance of similarities" (1920: 31-32 quoted in Lewin 1976). This statement is probably still an accurate summary of today's state of opinion regarding the two languages.