^{page 474 note 1} A form of Japanese poems, consisting of five lines and thirty-one syllables.

^{page 474 note 2} A generic term for the particles and the flexional endings of verbs and adjectives. Its origin and meaning are explained in Sansom's, G. B.An Historical Grammar of Japanese, 1928, pp. 8–9.Google Scholar

^{page 474 note 3} It is very similar to the tanka, but is so called because the first hemistich was comj by one man and the second by another.

^{page 475 note 1} They are, besides Motoori, Kada-Azumaro (, 1669–1736), Kamo-Mabuti, and Hirata-Atutane ( , 1776–1842).

^{page 475 note 2} A full description of this rule will be given later.

^{page 476 note 1} Although this was completed in 1782 it was not published until 1886.

^{page 476 note 2} The following abbreviations will be used in this article:—

(a) HK = (Teniwoha) Himokagami.

(6) TW = (Kotoba no) Tamanowo.

(e) KS = (Mikunikotoba) Katuyosyo.

^{page 476 note 3} In this article the term “ noun ” usually includes the pronoun. In modern Japanese grammar these two classes are treated as one under the term taigen (, substantive words).

^{page 476 note 4} In the case of inu, actually in is the uninflected part and u the flexional ending, but in Japanese writing nu cannot be divided into a consonant and a vowel, and the whole of it is therefore generally regarded as a flexional ending.

^{page 476 note 5} Kono humi no na yo, tamanowo to simo tukeruru yosi wa, hito no mi no yosoi ni mo, yorozu no mono no kazari ni mo, agareru yo ni wa, takaki iyasiki hodohodo ni, mina tama wonamu mono site, imiziki takara no oya to wa sikeru;… saru wa ito kagirinaku medetaM mono no kazari naramu ni mo, nukituranetaran sama ni sitagaite namu, ima hitokiwa no hikari mo sowarinubeku, mata haenaku kiete mo mienubekereba, kono wo koao ito nanome narumaziki mono ni wa arikere; masite suzinaku midare mo si tae mo si namu ni wa ika ni medetaku to mo sono kazari itazura narazi ya wa; kotonoha no tama no yosoi hata kono nukituranuru teniwoha kara nan tomokaku mo an'meru waza nareba mata yoaoete namu (TW vol. i, pp. 1–2).

^{page 477 note 1} zo no ya nani no teniwoha aru toki…. (TW vol. i, p. 13).

^{page 477 note 2} Ue no kotodomo wo mote ayakotoba () no teniwoha mo subete uta to mataku onaziki koto wo satorubesi (TW vol. vii, p. 41).

^{page 477 note 3} Teniwoha wa kamiyo yori… sono motosue ( ) wo kanaeawasuru sadamari nan aru (TW vol. i, p. 9).

^{page 477 note 4} This style of Japanese was commonly used in writing until the end of the nineteenth century.

^{page 477 note 5} In this article underlining signifies emphasis.

^{page 479 note 1} Himokagami no migi no kudari no dandan wa wa, mo, tada, no musubi-kotoba nari; naka no kudari no dandan wa zo, no, ya, nani, no musubi-kotoba nari; hidari no kudari no dandan wa hoso no musubi-kotoba nari (TW vol. i, p. 12).

^{page 479 note 2} Musubi to wa kotoba no tozime wo iu… kotoba no kiraru tokoro wa mina sono tozime nite ue ni okeru teniwoha no musubi nari (TW vol. i, p. 11).

^{page 480 note 1} Kakaru is an intransitive verb with various meanings, and there is no complete accord among grammarians about the interpretation of the term used by Motoori. Aston interprets it as “ to affect ” (Grammar of the Japanese Written Language, 1877, p. 186), Sansom gives “ to connect or join ” (An Historical Grammar of Japanese, 1928, p. 258), and Yositake suggests “ to bind ” (“ The Japanese Particles, Wa, Ga, and Mo ”, Bulletin of the School of Oriental Studies, Vol. VIII, part I, p. 29).

^{page 480 note 2} Motoori does not use the term kakari-kotoba (, kakari- words) for the particles given in HK, although he seems to make a distinction between musubi and the musubi-kotoba as abstract and concrete notions respectively. However, as the kakari and the musubi are generally applied to the kakari-kotoba and the musubi-kotoba, the kakari and the musubi will be used in this article in the sense of the kakari- words and the musubi-words respectively.

^{page 480 note 3} Subete teniwoha no kotoba nite tomarite ue e kaeru kokoro no uta wa, izure mo mina sono tomari no teniwoha no kanarazu ue no kotoba no kiruru tokoro made e kakaru y6 ni yomu ii, p. 6).

^{page 481 note 1} Keru wa ue ni zo no ya nani nado no kotoba nakereba keri to iubeki kaku () naru wo, keri to iwazu site keru to iu wa ika ni to iu ni, subete uta no hasigaki wa sono uta e kakarite iu mono yue ni tune no kotoba no tozime to wa koto ni site, kotoba wa kirete mo kokoro wa kirezu site yagate uta e tuzuku nari (TW vol. vii, p. 42).

^{page 481 note 2} Zo no musubi wa Himokagami no naka no kudari no dandan no kotoba nari (the musubi for zo are those words shown under column II of HK) (TW vol. ii, p. 12).

^{page 482 note 1} It should be noticed that the adverbial particles sometimes come before a case particle: e.g. otome NOMI to omowaru (it is thought that there are only girls), which has a different meaning from otome to NOMI omowaru (one is taken for nothing but a girl).

^{page 482 note 2} Occasionally we find no coming immediately after a case particle: e.g. miyako YORI NO tayori (the message from the capital), where no follows a case particle yori (from), the function of which, as stated above (on this page), is to indicate the relation between the word preceding it (miyako, the capital) and a predicate word, which is not expressed, while no indicates the relation of one noun, which is not expressed, to another (tayori, message). This implies that there is the omission of a word which has the faculty of being both the predicate word for miyako and the noun (or word in the substantive form) to be linked with tayori: e.g. kitarisi, a predicate word meaning “ came ” and in the form to join on to a noun; and the sentence is equivalent to miyako yori kitarisi tayori (the message came from the capital).

^{page 483 note 1} Motoori mentions that the particle ga has a similar meaning and behaviour to no (TW vol. iii. p. 29);ga was another particle indicating the relation of one noun to another, but in the modern language it points out prominently something to which it is attached, rejecting all other relevant matters.

^{page 483 note 2} Tadasi zo ya nani ni kurabureba no wa yaya karuki yue ni kano sadamareru kaku ni hazurete musubu koto naki ni arazu (TW vol. iii, pp. 21–2).

^{page 484 note 1} Nani nado no kotoba wo okite sono aita no musubi to no aida ni ka-mozi wo hasamu koto tune ni dsi (TW vol. iv, p. 33).

^{page 484 note 2} Mo to ukuru kaku musubi ni kakawarazu (TW vol. iv, p. 37).

^{page 485 note 1} Simo is supposed to be a compound of si and mo, and si was once extensively used as an emphatic particle, but its use became less popular in the Heian period.

^{page 485 note 2} Tada to wa ioa mo zo no ya nani koso nado iu kotoba no naki wo ima kari ni kaku iu nari (TW vol. i, p. 11).

^{page 487 note 1} Wa mo to zo no ya nani koso to kasanaru koto tune ni dsi. Sono toki wa mo wa karoku zonoya nani koso wa omoki yue ni omoki kata no kaku ni yorite musubu nari (TW vol. ii, p. 7).

^{page 488 note 1} In this connexion, wa has the peculiarity of pointing out a particular something (fact, act, state, or property) to which it ia attached, without referring to other relevant matters of which one is consciously or unconsciously thinking; and mo points out a particular something placing it among other relevant matters of which one is consciously or unconsciously thinking. For example, when one says ha otu (leaves fall), one simply expresses “ the falling of leaves ” and the falling of other things is not in one's mind, but when it is said ha wa otu, one expresses “ the falling of leaves ” apart from “ the falling of other things ” of which one is consciously or unconsciously thinking, e.g. the falling of flowers or branches of a tree;ha mo otu expresses “ the falling of leaves ” as well as “ the falling of other things ”, such as flowers or branches of a tree.

^{page 488 note 2} Yamada considers a kakari to be “ a particle which is related to the force of statement ” (tinzitu no seiryoku ni kanyo suru zyosi), that is to say, it causes the word or group of words to which it is attached to “ call for a certain statement ” (ittei no tinzitu wo yd-su) (Yamada: “ Nippon Bunpôgaku Gairon ”, p. 476). But, when one wants to express the conception nusubito tama wo toriki (the thief took a jewel), does one not call for a certain statement toriki (took) after saying nusubito tama wo (the thief, a jewel) ? Therefore, Yamada's definition of kakari seems to apply also to the other classes of particles which indicate the relation between a noun and a predicate word.

^{page 490 note 1} Unless we understand “ changes ” as “ inflexions ” there are actually only two “ changes ” in what Motoori calls “ the musubi of three changes ”, but I shall follow his terminology in this article. It should also be noticed that, as Motoori points out in TW (vol. i, p. 12), some of the musubi have only two inflexions and in this case there is only one “ change ”.

^{page 491 note 1} Of course, when there is no supplementary verb-ending, hi of yuki is a musubi.

^{page 491 note 2} Actually masi inflects as mase, masi, masika, and it is one of the supplementary verb, endings:rasi is usually uninflected (though it once had inflexions), but its function is very similar to that of supplementary verb-endings;tutu is a repetition of tu on line 20, and together with Icana and gana it should be regarded as a particle.

^{page 492 note 1} For this reason Yamada recognizes mo as a kakari. but whether it has any special emphatio force apart from the mild prohibitive value is doubtful.

^{page 494 note 1} Subete no kotobazukai ni kiruru tokoro to tuzuku tokoro to no kezime aru koto wo mazu wakimae okubesi…. Kiruru kotoba wa kotogotokn Himokagatni no migi no kudari no dandan ni ari. Tuzuku kotoba wa naka no kudari no dandan ni aru nari (TW vol. i, p. 12).

^{page 494 note 1} However, those musubi on lines 33 to 43 in HK take the same form for “ breaking ” and “ joining ”. (Sate mata kiruru tokoro mo tuzuku tokoro mo onaziki kotoba ari; Himokagami no dai sanzyusandan yori owari made awasete zyiiitidan no kotoba kore nari (TW vol. i, p. 12).)

^{page 494 note 2} It is interesting to note that Motoori's conception of kiruru kaku and tuzuku kaku has something in common with Jespersen's idea of nexus and junction. See the following Japanese versions for the examples given by Jespersen (The Philosophy of Grammar, p. 114):

the dog barks. inu ho YU.

the rose is red. bara aka SI.

a barking dog. ho YURU inu.

a red rose. akaKI bara. where hoyu (to bark) and akasi (to be red) are in the “ breaking ” form and hoyuru and akaki are in the “ joining ” form; in other words, a combination with the “ breaking ” form is nexus and that with the “ joining ” form is junction.

^{page 495 note 3} Subete koso no musubi to naru kotoba wa ue ni koso nakute kiruru toki wa ôku ohosuru kotoba naru wo… ue ni koso nakute kotoba no tuzuku tokoro ni aru toki wa mina kanarazu ba to do to no hutatu no kotoba e tuzuku nari (TW vol. v, p. 1).

^{page 495 note 1} Adjectives have no imperative form, and some verbs have a different form for the imperative.

^{page 495 note 2} Naka no kudari wa tune ni wa tuzuku kotoba naru wo ue ni zo no ya nani no teniwoha aru toki wa sono musubi to narite kiruru nari (TW vol. i, p. 13).

^{page 495 note 3} Koso no musubi to naru kotoba wa ue ni koso nakute kiruru toki wa oku wa ohsuru kotoba naru wokoso to kakareba ohosuru kotoba ni narazu (TW vol. v, p. 1).

^{page 496 note 1} Oyoso utagai no ya ni kotoba no nakaba ni okite sita wo sono musubi-kotoba nite musubu to, kotoba no tozime ni okite kiru to no hutatu ari;… nakaba ni oku ya no ue wa kanarazu tuzuku kaku no kotoba yori uke, tozime ni okite kiruru ya no kanarazu kiruru kaku no kotoba yori ukuru ga sadamari nari (TW vol. iv, p. 4). Here Motoori's argument becomes illogical as, according to him, the “ breaking ” form signifies the conclusion of a sentence, but in this instance he says ya follows the “ breaking ” form at the conclusion of a sentence.

^{page 496 note 2} Subete kotoba no tozime ni okite kiruru ka no ue wa tuzuku kaku no kotoba yori ukuru sadamari nari; mata kiruru ya no ue wa kiruru kaku no kotoba yori ukuru sadamari nari (TW vol. iv, p. 27).

^{page 496 note 3} Ba ni sude ni sikaru koto wo iu to, imada sikarazaru koto wo iu to no hutatu ari (TW vol. iii, p. 4).

^{page 496 note 4} Tune no teniwoha no to wa ue wo kiruru kaku no kotoba yori ukuru wo, aimukaite iu to wa tuzuku kaku no kotoba yori ukuru sadamari nari (TW vol. v, p. 15).

^{page 497 note 1} Sude ni sikaru koto wo iu ni wa do domo to nigori, imada sikarazaru koto wo iu ni wa to tomo to sumite iu. Sono sumi nigori ni yorite ue wo ukuru kotoba no kaku mo kotonarn nari (TW vol. v, p. 17).

^{page 497 note 2} Nu has two more forms, i.e. naru and nure, the musubi under columns II and III respectively.

^{page 497 note 3} Nu wa sono kotoba no tuzuki ni sitagaite na to mo ni to mo nu to mo ne to mo naru nari (TW vol. vi, p. 4).

^{page 498 note 1} The following is the table of the fifty sounds of the syllabary:—

^{page 501 note 1} Teniwoha wa… agareru yo wa sara ni mo iwazu, nakamukasi no hodo made mo onozukara yoku totonoite tagaeru husi mo osaosa nakarikeru wo, yo kudarite wa uta ni mo saranu kotoba ni mo kono totonoe wo ayamarite… iniaie no yoki utadomo no wa kanarazu sikarade wa e aranu onozukara no kotowari arite sadamarituru mono ni nan arikereba karisome ni mo sore wo hanarete watakusi no sakasira wo ba tuyu bakari mo kuwaubeki ni arazu. Tada izuku made mo izuku made mo hurnki miti wo tazunete sono ato ni nan sitagai yorubeki waza narikeru (TW vol. i, pp. 9–10).

^{page 502 note 1} Akabori-Mataziro, : “ Kokugogaku Syomoku Kaidai ” an account of books on Japanese linguistics), 1902.Google Scholar

^{page 502 note 2} Ando-Masatuou, : “ Kokugogaku Tuko ” , outline of Japanese linguistics), 1931.Google Scholar

^{page 502 note 3} Aston, W. G:A Grammar of the Japanese Written Language, second edition, 1877.Google Scholar

^{page 502 note 4} Huzimtoa-Tukurtt, (editor): “ Nippon Bungaku Daiziten ” , 1932–5.Google Scholar

^{page 502 note 5} Hosina-Kdm, : “ Kokugogaku Seigi ”, comprehensive Japanese linguistics), 1909.Google Scholar

^{page 502 note 6} MOTOORI-NOEINAGA:

(a) “ Teniwoha Himokagami ”, 1771, reproduced in “ Motoori-Norinaga Zensyii ” complete works of Motoori-Norinaga), vol. ix, 1927 edition.

(b) “ Kotoba no Tamanowo ”, 1829 edition.

(c) “ Mikunikotoba Katuyosyd ”, from “ Motoori-Norinaga Zensyii ”, vol. x, 1927 edition.

^{page 502 note 7} Muraoka-Tensi, : “ Motoori-Norinaga ” (the life of Motoori-Norinaga) third edition, 1937.Google Scholar

^{page 502 note 8} Sansom, G. B.:An Historical Grammar of Japanese, 1928.Google Scholar

^{page 502 note 9} Tokieda-Seiki, : “ Motoori-Norinaga oyobi Huzitani-Sigeaya no teniwoha kenkyfi ni tuite” on the studies of the teniwoha by Motoori-Norinaga and HuzitaniSigeaya), “ Kokugo to Kokubungaku ” , Japanese language and literature), No. 46.Google Scholar

^{page 503 note 10} YAMADA-TAKAO:

(a) “ Nippon Bunp6ron ” (treatise on Japanese grammar), 1938 edition.

(b) “ Nippon Bunpdgaku Gairon ” (outline of Japanese grammar), 1936.

^{page 503 note 11} Yositake S.: “ The Japanese Particles, Wa, Ga, and Mo ”, Bulletin of the School of Oriental Studies, Vol. VIII, part 1.