An earlier article read the form yonde ita in Kokoro’s opening line for what it means — as the past tense of “-te iru.” This article steps back one stage further and looks at the mechanics themselves — why does yobu (“to call”) become yonde?
“Yonde ita” — bu becomes n, and de
Kokoro’s opening line reads:
Watashi wa sono hito o tsune ni sensei to yonde ita. — I always called that person Sensei. (working translation)
Yonde ita is built from the verb yobu (“to call”), plus -te (the te-form) and ita (the past tense of iru, “to be”). But yobu + te doesn’t simply become yobite — it changes to yonde. The sound bu shifts to n, and the following te voices to de — this is a regular sound change found in Japanese godan verbs (what this article will call “group 1” verbs).
“Kogonde iru” — mu becomes n, and de, too
The beach scene has another verb showing a similar change.
Soko ni kogonde iru nihonjin ni, hitokoto futakoto nanika itta. — He said something, a word or two, to the Japanese man crouching there. (working translation)
Kogomu (“to crouch”) belongs to the same group as bu-ending verbs: its mu shifts to n, and te voices to de, just as with yobu. Yobu and kogomu end in different sounds, but both follow exactly the same pattern: n, plus a voiced de.
“Ugoite ita” — ku becomes i, te
Let’s look at another pattern.
Suna no ue ga ugoite ita. — The sand’s surface was moving. (working translation)
Ugoku (“to move”) ends in ku. Here, ku shifts to i, and te stays as-is, unvoiced: ugoite. This is clearly a different change from the n + de pattern of yobu and kogomu.
“Nagamete ita,” “kite ita” — verbs that don’t change shape
Some verbs, on the other hand, don’t change shape at all.
Nagai aida seiyōjin no umi e hairu yōsu o nagamete ita. — He had been watching the Westerner go into the sea for a long while. (working translation)
Junsui no nihon no yukata o kite ita kare wa, sore o shōgi no ue ni suporito hōridashita mama… — He, wearing a plain Japanese yukata, had tossed it carelessly onto the bench… (working translation)
Nagameru (“to watch”) and kiru (“to wear”) only need to drop the final -ru and add -te. None of the sound changes seen in yobu, kogomu, or ugoku happen here. These belong to a different type of verb, called ichidan verbs (“group 2” verbs here).
Summing up three patterns
From this one passage alone, three representative patterns of the Japanese te-form can be confirmed.
| Verb | Te-form | Pattern |
|---|---|---|
| yobu, kogomu | yonde, kogonde | “bu/mu” → “nde” (voiced) |
| ugoku | ugoite | “ku” → “ite” (unvoiced) |
| nagameru, kiru | nagamete, kite | Drop “ru,” add “te” — nothing else changes |
Knowing a Japanese verb’s dictionary form doesn’t automatically tell you its te-form. You need to understand that the pattern of change depends on whether the verb is a godan verb (group 1) or an ichidan verb (group 2). This single passage from Kokoro happens to line up several of the representative patterns, one after another.
Worth checking as you read
First, godan-verb te-forms have other patterns beyond the “bu/mu → nde” and “ku → ite” covered here — for example, “su → shite” and “tsu/ru/u → tte.” This article confines itself to the patterns actually confirmable in the Kamakura beach passage.
Second, the verb iku (“to go”) is a godan verb, but it conjugates irregularly to itte, not ikite. This special case doesn’t appear in this passage and is not covered in this article.
Summary
| Pattern | Example | Rule of change |
|---|---|---|
| bu / mu | yobu → yonde, kogomu → kogonde | The ending shifts to “n,” and “te” voices |
| ku | ugoku → ugoite | The ending shifts to “i” |
| Ichidan verbs | nagameru → nagamete, kiru → kite | Just drop “ru” and add “te” |
In short — a Japanese verb’s te-form follows a pattern determined by the verb’s type, and this single passage from Kokoro happens to let us confirm several of the representative patterns with real examples.