In Plain-language explainer 1 we saw that Matsushita Daizaburō divides language into three steps (danku, nenshi, genji). In this article we check the lowest step, the genji, in the general remarks of Part 2, “Genji-ron” (pp. 61–62). (English renderings of the quotations are ToL translations.)
What does the theory of genji discuss?
The opening of Part 2 is this sentence.
原辞論は文法学の二分科の一つであって、念詞の材料たる原辞の、念詞中に於ける文法的法則を論ずるものである。
(The theory of genji is one of the two branches of grammar, and discusses the grammatical rules of the genji, the material of the nenshi (word), within the nenshi.)
What matters here is the phrase “within the word.” What genji-ron deals with are rules inside a word. Rules between words (how a sentence is put together) are a separate matter.
Matsushita’s way of fixing a unit
Matsushita takes out the genji in roughly this order. If you cut language by the condition “can express an idea by its own power,” everything you get is a nenshi (word). But in the construction of language, below the word there is a further unit, the genji. Words are made of genji, he says.
A genji is what you get by dividing as far as it will go. As examples of genji, Matsushita gives “yama” (mountain), “san” (mountain, Sino-Japanese), “umi” (sea), “kai” (sea, Sino-Japanese), “yuku” (go), “kuru” (come), “o,” “ni,” “nari,” “tari,” and so on. (The readings are according to the furigana in the original.)
Here something interesting happens. Genji are of two kinds.
- Those that can be a word even alone — the native readings, like “yama” and “umi”
- Those that cannot be a word alone — the Sino-Japanese readings “san” and “kai,” and o, ni, and so on
That is, the character 山, if read “yama,” can be a word by itself. But if read “san,” it cannot. Only by joining with other genji, as in “kōzan” (high mountain) or “sanga” (mountains and rivers), does it become a word.
Checking with “gakusei-rashii”
Matsushita shows the relation between word and genji with a concrete example.
例へば「学生らしい」は一つの念詞であるが、「らしい」は単独の力では一観念を表はさないから「学生」と「らしい」との二念詞とすることは出来ない。併し自己単独の力であつてもなくても一つの意義を表はすとしてならば、分解して「学」「生」「らしい」の三つとすることが出来る。
(For example, “gakusei-rashii” is one nenshi, but since rashii does not express an idea by its own power, it cannot be made into the two nenshi “gakusei” and “rashii.” But if, whether by its own power or not, we take each as expressing one meaning, it can be divided into three: “gaku,” “sei,” “rashii.”)
Matsushita’s way of thinking shows well here.
- “Gakusei-rashii” is one word.
- It is not divided into two words, “gakusei” and “rashii.” Rashii cannot express an idea alone.
- But if we divide by the looser criterion “has meaning,” it can be broken into three genji: “gaku,” “sei,” “rashii.”
Renji and renshi
What happens when a genji joins other genji? Matsushita says: those that cannot become a word by themselves — “san,” “kai,” o, ni — join other genji and together become one genji, as in “sanga,” “kōzan,” “hana o,” “tsuki ni,” and only then become one word. This joined genji he calls a renji.
Those that can become a word by themselves, like “yama” and “umi,” too, often become a word only after joining other genji and becoming a renji, as in “takayama,” “yamakawa,” “aruumi,” “uchiumi,” “yama o,” “umi ni.” These too are renji.
On the other hand, what is made by joining words is called a renshi. Matsushita gives an example.
「春風ぞ吹く」は一つの連詞であるが、「春風ぞ」といふ念詞と「吹く」といふ念詞から成立して居る。
(Harukaze zo fuku is one renshi, but it consists of the nenshi “harukaze zo” and the nenshi “fuku.”)
Note that “harukaze zo” is counted here as one word. “Harukaze” with zo attached is one word. Zo is a part.
Two kinds of rules that act on genji
Finally, Matsushita divides the rules that genji-ron deals with into two. One is rules that exist in that genji alone (independent rules); the other is rules that exist in relation to other genji (correlative rules). For that reason genji-ron divides into “genji tandoku-ron” and “genji sōkan-ron.”
The examples Matsushita gives are of the following kind.
| Kind | Example |
|---|---|
| Independent rule | The genji “yuku” changes into “yuka,” “yuki,” “yuke.” “Shiro” (white) also becomes “shira” |
| Correlative rule | One says 人殺し (hitogoroshi) but not 人殺, and says 殺人 (satsujin); one says 帰り道 (kaerimichi) but not 帰れ道 (kaeremichi) |
Conjugation (the change of “yuku” into “yuka,” “yuki,” “yuke”) is placed on the side of genji-ron, as a rule belonging to that genji itself. This too is Matsushita’s way of organizing, different from school grammar.
What to keep in mind when reading
First, the genji is close to a unit like the “morpheme,” but not the same. Matsushita thinks of a genji as something with meaning, divided as far as it can be divided. The relation to the modern “morpheme” we have not checked in this article.
Second, these general remarks are the entrance to genji-ron. How the individual rules of genji (conjugation, etc.) are actually organized cannot be known without reading the body of Part 2. We have not gone that far in this feature.
Summary
| Point | Detail |
|---|---|
| Genji-ron | Deals with the rules the material of words (genji) follows within a word |
| Genji | What has meaning and is divided as far as it can be divided |
| Two kinds of genji | Those that can be a word alone (yama) / those that cannot (san, o, ni) |
| “Gakusei-rashii” | One word. Can be broken into three genji: “gaku,” “sei,” “rashii” |
| Renji, renshi | Joining of genji is renji; joining of words is renshi |
| Two rules | Rules of a genji alone (conjugation, etc.), and rules in relation to other genji |
In a nutshell — Matsushita Daizaburō divided words into parts called genji, and drew the line of units so fine that yama is a full member but san is only a part.