Anyone who enjoyed my earlier post on the way Japanese conjugates verbs made from borrowed words, based on one at No-Sword, will want to read his new entry “More unusual Japanese verbs.” One of his tidbits:

gomakasu — means “misrepresent (in a deceptive way)”, and again, has ateji that mean “mis-bewitch-style” + s + u. The origin story is great, though:
gomadouran (“sesame seed bags [or cases]” — not ateji here, these characters reflect etymology) were a kind of, uh, sesame seed candy, hollow on the inside. (Hence “bag” or “case”, I suppose.) They quickly became known by the more direct name gomakashi (“sesame seed candy”).
This word then came to have a figurative meaning: something which has an appetising exterior, but nothing inside; and then, the abstract idea of misrepresenting something in this way. gomakasu the verb was a back-formation, because gomakashi sounds the nominalised form of such a verb.
Once this verb had been born, ateji were used to write it — maybe because of a desire to pun, maybe because the first person to write it down wasn’t aware of the etymology.
So, to summarise: we have inaccurate S-J ateji used to write a verb which is a back-formation from a legitimate S-J word.
Best half-S-J verb ever.

He explains ateji thus:

ateji might be translated as “characters that hit the target”. It’s kind of hard to pin down a definition, but it basically refers to kanji used to write words which are not, etymologically, related to those kanji. Sometimes the kanji are used for their phonetic value, and sometimes the kanji are used for their meanings (and given entirely new phonetic values). What is important is the absence of an etymological link.
For example, the ten of tempura (<- tenpura) is often written with the kanji for heaven, ten, even though the word itself comes from Portuguese and obviously has nothing to do with the Chinese word for heaven. That’s ateji in action.
An English analogy might be spelling television “tell-a-vision”: the sounds are there, and the meaning is kind of acceptable, but Greek tele– has nothing to do with Germanic tell.
Incidentally, hiragana and katakana (the two Japanese syllabic character sets) both derive from standardised ateji sets, representing sound only. Their present forms are simplifications (hiragana) or selected parts (katakana) of the original kanji.

He ends with the following teaser: “Tune in next time for the final instalment: Japanese stems + non-Japanese endings!”


  1. I’m already regretting that definition of ateji, you know, because it doesn’t really explain why Japanese (not SJ) words written with kanji aren’t ateji. As far as I can tell, the only answer is “because those kanji were assigned to those words long enough ago that they get a free non-ateji pass.”
    I guess ateji are like porn: you know them when you see them.

  2. I’m not sure about the long ago argument. Plenty of ancient Buddhist writings use ateji, and it is recognized as such. Probably because the terminology is so specialized.
    The most direct experience I (and I suspect many others) have had with ateji is in parlor games, making up ateji for people’s names. Best done drunk.

  3. It might be worth noting that the word goma, though it may have come to Japanese through Chinese, was actually a loan word in Chinese itself, according to this web page (modern Mandarin no longer uses this word). However, the story is a little more complicated, because according to the 漢字源, the character for go in goma is in several (originally Chinese) words that indicate goods or plants that were grown in areas to the north and west of China. So by that reasoning, goma could be an original Chinese word.
    But, how the word is treated in modern Japanese is probably a separate deal. You could argue that compounds like goma-abura, shiro-goma and kuro-goma indicate its ‘native-ness,’ similar to the (originally loaned) word sara ‘plate.’ Unfortunately, the goma starts with a voiced obstruent, and so we can’t use a rendaku (consecutive voicing) test for nativeness.

  4. Minor point: it would be rare to see “gomakasu” written with its ateji–hiragana is the norm.
    My etymology dictionary gives a secondary possibility, that “gomakasu” derives from the term “goma-no-hae”, a kind of Edo-period pickpocket.

Speak Your Mind