Is it possible to split an Ateji reading, that is assigned to a compound of Kanji? For example 貴女 (anata). Can it be split into 貴(ana) and 女(ta) or into 貴(a) and 女(nata)?

I'm not asking if it makes sence to assign the parts of the reading to a corresponding individual Kanji, I just want to know if there is a correct way to do it theoretically as in 私達 (watashi·tachi).

Thank you for your input!

No. 熟字訓 (jukujikun) type of ateji are assigned to the whole word; it's not possible to split it up. So if you choose to split it up for some reason, there's no way to do it "correctly" since it's just not done in the first place.

Some can be, like 大丈夫, but not the one you gave. I think 今日　（きょう）is in the same category.

大丈夫 read as だいじょうぶ is not ateji.

Really? I thought ateji referred to all forms of kanji misuse! Is it only for uses that don't use the standard pronunciation?

「大丈夫」has three On readings. I assume you're referring to the fact that the「夫」reading is voiced? That happens in every(?) compound unless the attached word in question is either conspicuously foreign or has a voiced second consonant (e.g. 日影, hi-kage).

大丈夫 　（だいじょうぶ）
大 = だい (#1 most common reading)
丈 = じょう (#1 most common reading)
夫 = ぶ (#3 most common reading)

Unlike say:
今日（きょう）
今 != きょ (doesn't have any readings "kyo")
日 != う (no such reading)

私達 is not ateji either.

harhol Wrote:That happens in every(?) compound unless the attached word in question is either conspicuously foreign or has a voiced second consonant (e.g. 日影, hi-kage).

Those are two cases where it usually doesn't happen, but there's no way to predict where it will happen.

And yeah, there's no "misuse" in 大丈夫.

So if I understand this correctly, Ateji readings and meanings are not found in the dictionary entries of the contained Kanji.
Would you say it is a benefit if a digital dictionary offered them (listing the whole compound of course)? As a beginner I would say so.

Thanks for your opinions!

jcdietz03 Wrote:私達 is not ateji either.

That's why I used it as an example for a splittable compound.

Interesting topic, thanks for the information!

So, to complete (or is there more?) my information on this: can name readings (what are they called again.. not that I have an ime activated anyway) always be split into Kanji portions? Or is this a moot question, because some name readings are just so rare that one might as well consider them to be ateji-like, i.e. not deductible from the single components, but rather from the compound as a whole?

Some can, some can't. Name readings are generally a kind of kun-yomi; once you get beyond the Jouyou list, the difference between a kun-yomi and a meaning is not entirely clear and it starts to get difficult to tell what's actually a "real reading" and what's not.

re: names being used with kun reading a lot.
This is true for a lot of things, except like 藤本 vs. 佐藤, the 藤 changes.

Now...how about using 戯 for the ざ, as in ざれる? Unfortunately, ざ isn't a very common syllable in Japanese names, so I've had quite a time trying to kanjify my name...

I think it's pretty unusual to use a kanji like that -- it would have to be ざれ, not just ざ.

There's always 座, that's about the only one I know...

I think that　大丈夫 would count as misuse in some sense because the characters are being used solely phonetically, without any regard to their underlying connotations. Interesting, I thought "ateji" covered that. I apparently was wrong.

That is one type of ateji, but 大丈夫 is not an example of characters being used solely phonetically. They are being used for meaning as well. The word 丈夫 means "strong" or "healthy". 丈 is used in other words with that connotation as well (頑丈, for instance). 大 just intensifies the meaning, and the current meaning of "OK" is a semantic shift.

True examples of this kind of ateji are rather rare in contemporary Japanese because a lot of them have fallen out of use and the words are just written in hiragana (or katakana). 馬鹿（ばか） and 野暮（やぼ） are two common examples, though.

Surely plains (野) and livelihood (暮) for foolishness/boorishness (野暮) can't just be a coincidence... seems like the Japanese equivalent of "redneck".

Ateji don't necessarily have to be coincidental; there's no etymological link between the word やぼ and the morphemes underlying the characters 野暮. It just happened that they were able to pick a good set of characters for the ateji.

Interesting! Thanks to all for the information!