Is 給う an irregular verb?

While I was writing the section on the -た form in my grammar book, I started to look at what others had done when introducing it. After all, it’s a pretty complex thing, usually more so than people assume at first glance. As usual, my first instinct was to visit the guide I had used myself when I first tried (and failed) to learn Japanese: Tae Kim.

Of course, Tae Kim does not call it the -た form. He simply calls it the past form. Honestly, I do not really mind this simplification, but since I try to avoid saying half-truths too often, I will stick with the name -た form. After all, the -た form has another very important use, the perfect aspect. Anyway, this discussion is for another post, so let’s go back to what interests us: How to conjugate verbs to the -た form.

As is usual, Tae Kim gives conjugation tables for all three major categories of verbs in Japanese: 1-grade verbs (what he calls “る-verbs”; terrible name, oh well), 5-grade verbs (what he calls “う-verbs”; literally every verb, oh well) and exceptions. And to my surprise, the exceptions include more than the usual two suspects. Tae Kim lists する, くる and 行く. But 行く is noted 行く*, and the asterisk mentions “exceptions particular to this conjugation”.

Of course, this is not wrong. Still, the adjectives “exception” and “irregular” are often used in Japanese grammar to discuss specific kinds of verbs (する and くる in Modern Japanese; more in Classical Japanese, e.g. 死ぬ). But it’s undeniable that 行く does not conjugate like you’d assume it would. The issue is that, well, it’s not the only one. Take a look at this example sentence for 給う in Jitendex:

我々に生命を与えた神は我々に同時に自由を与え給うた。

Now, I’m no expert, but this seems to be a translation of that famous Thomas Jefferson quote: “God who gave us life gave us liberty.”

Clearly, this is the -た form, so why is it 給うた and not 給った? After all, the rules of conjugation are pretty clear: 5-grade verbs that end in う change to った, not た. If at least it attached to the 連用形, the conjunctive form of the verb (e.g. 読み for 読む), then I could chalk it up to a more archaic form. But that’s not what’s happening here. It’s literally just adding た to the verb (or so it seems). So what’s going on?

The key here is not actually the -た form itself, but rather what we call euphony, in Japanese 音便 (onbin).

As per Wikipedia:

Onbin […] is a set of sound changes that occurred in Early Middle Japanese around the end of the eighth century to the beginning of the tenth century, first attested in written texts of the Heian period.

So what does this have to do with our initial question? Well it turns out that onbin is the reason why the -た (and -て) forms are so messy. But that was not originally the case. For instance, what we call the -て form of a verb is not really a form at all, rather it is the conjunctive particle て that attaches to the 連用形 (conjunctive form) of the verb.

At least originally, this is how it worked. So, for instance, you would have 読みて, 死にて, etc. Similarly, you would have 読みた, 死にた, etc. Through onbin, these became 読んだ, 死んだ, etc.

There are many different kind of onbin in Japanese. It turns out that some very common words were formed through onbin. For instance, take the words 素人, 弟 and 妹. See a pattern here? They all end the same way, in -ōto. That is no coincidence.

One of the more common type of onbin is ウ音便 whereby the first part of the second element of a compound changes to -u-. In old Japanese, pito (modern Japanese, 人) gave us imopito, otopito, and siropito. Through ウ音便, these became 妹 (imouto), 弟 (otouto), and 素人 (shirouto). These words come from the same sound change, the pi from pito became u, which in modern Japanese gave us the ō we see in all three of these words. This is the same process that gave us 尊い, 小路 and even 神戸.

Other kinds of onbin exist though, for instance イ音便. This often turned ki into i, one of the most famous example of this being 刃, from 焼き刃 (forged blade). Similarly, 撥音便 had various sounds (such as ni, mu or gu) turn into ん. For instance, kamunusi became kannusi, modern 神主. Some words had different potential outcomes, a good example being ka-kupasi becoming either 香ばしい (こうばしい) or 芳しい (かんばしい). In the first one, ウ穏便 occured; in the second, 撥音便 occured.

Cool. So what about 給う and 行く?

So we’re expecting 給った, in other words tamatta. It turns out that the 促音 (the little つ) can also also go through onbin. Through this sound change, tamatta became tamouta, which is exactly what you see in 給うた. So even though it looks like tamauta, it really is tamouta. And it’s not the only verb that does this. You might also know 問う (probably the best example of this phenomenon) and maybe also 請う・乞う (to beg).

So to answer the question in the title, yeah, in a sense, 給う is irregular, but it’s irregular in a highly regular way. It’s not so much that it doesn’t follow the rules, it’s that it follows the rules of Japanese phonology rather than those of Japanese grammar, and why wouldn’t it?

Grammar is a formalization, it always bows down to the actual language, and one of the aims of phonology is to describe sound changes, i.e., how the language as it was actually spoken changed over time. It makes sense that phonological change would overrule Japanese grammatical “rules”.

And what about 行く? In what way is it irregular? What is happening here is another phonological phenomenon.

In Japanese, we tend to avoid vowel-doubling, by which I mean saying the same vowel twice in a row. A good example is the adjective 大きい which is pronounced ōkī and not ookii. This extends even to other vowels at times, e.g., 栄光 pronounced ēkō and not eikou.

Well what would happen if we conjugated 行く as one would expect? We’d get 行いた (iita), doubling the i vowel. To avoid this, we say 行った (itta). Can you think of any other verb that does something similar, maybe in a different tense or mood?