My friend Amanda, who is studying Japanese in order to teach English in Japan, writes:
Bayes (if it’s pronounced how I think it would be pronounced) would actually translate to beizu in katakana. Otherwise, Beizu-tsukai would be perfect, I think. The person is probably thinking of “mahou-tsukai” which is technically a “magic user” but always translates to “witch” or “wizard” in English.
For comparison, beisu would be pronounced roughly like “base”. If the sibilant in “Bayes” is voiced, as I would expect, then beizu would be more accurate (this actually bothered me a bit when reading Eliezer’s earlier stuff, but it seemed too petty to bring up at the time). However, I have seen a few badly-rendered loanwords imported into Japanese in a way that matches the romaji letters to the original spelling, rather than matching the actual phonetics.
For extra fun, these would be written as ベイズ使い or ベイス使い (the latter using the unvoiced consonant).
In hindsight it would have made more sense for me to just look that up in the first place rather than pontificating about transliteration in my previous post. Ah, well.
Could always handwave something about the fictional world in which the beisutsukai exist as having drifted norms of English pronunciation such that a word-terminal “s” is never voiced, making “Bayes” sound like “base”.
Other than that, though, beizu remains more accurate, unfortunately.
Also assorted words that are actually in dictionaries, e.g.: kozukai, janitor; zoutsukai, elephant trainer; ryoutoutsukai, two-sword fencer/expert in two fields; &c. Sounds like mahoutsukai and other related uses in fiction are much more common, though.
Also note that, as with kozukai, sometimes compound words in Japanese will voice the consonant of the first syllable of the second word. My knowledge of Japanese is not extensive enough to recall when this is done, however, especially in the case of mashing something onto the end of a loanword like ベイズ.
The voicing thing is known as rendaku. Generally it’s a bit of a mystery when it will and will not happen. This thesis lists a bunch of proposed rules, two of which seem relevant:
Rendaku is favoured if the compound words are native-Japanese (yamatokotoba). This might be the reason for kozukai vs mahoutsukai, ko is native-Japanese and mahou is sino-Japanese. So by analogy, one would not expect voicing for beizutsukai.
Noun+Verb compounds exhibit rendaku if the noun is an “adverbial modifier” but none if it’s a direct object. In “using magic” 魔法を使う magic is a direct object, so no voicing. On the other hand kozukai (‘little servant’?) is an Adjective+Verb, which explains the voicing.
In any case, I guess the upshot is that we should expect beizutsukai, without rendaku,
My friend Amanda, who is studying Japanese in order to teach English in Japan, writes:
For comparison, beisu would be pronounced roughly like “base”. If the sibilant in “Bayes” is voiced, as I would expect, then beizu would be more accurate (this actually bothered me a bit when reading Eliezer’s earlier stuff, but it seemed too petty to bring up at the time). However, I have seen a few badly-rendered loanwords imported into Japanese in a way that matches the romaji letters to the original spelling, rather than matching the actual phonetics.
For extra fun, these would be written as ベイズ使い or ベイス使い (the latter using the unvoiced consonant).
Huh. The way I mentally pronounce this seems to be closer to bei-su-tzkai than bei-zu-tzkai, but when I say it out loud, it can come out either way.
Does anyone know how “Bayes” itself is standardly written in Japanese?
Apparently it’s written as beizu.
In hindsight it would have made more sense for me to just look that up in the first place rather than pontificating about transliteration in my previous post. Ah, well.
Well now I’m torn. Damn it, in writing, “beisutsukai” looks far better than “beizutsukai” and it may even sound better.
Could always handwave something about the fictional world in which the beisutsukai exist as having drifted norms of English pronunciation such that a word-terminal “s” is never voiced, making “Bayes” sound like “base”.
Other than that, though, beizu remains more accurate, unfortunately.
You’ve already coined the word. Too late to change it!
Not to me. z’s are inherently cooler than s’s.
The suffix applies beyond mahou, though; e.g. mizutsukai for watercaster.
Also assorted words that are actually in dictionaries, e.g.: kozukai, janitor; zoutsukai, elephant trainer; ryoutoutsukai, two-sword fencer/expert in two fields; &c. Sounds like mahoutsukai and other related uses in fiction are much more common, though.
Also note that, as with kozukai, sometimes compound words in Japanese will voice the consonant of the first syllable of the second word. My knowledge of Japanese is not extensive enough to recall when this is done, however, especially in the case of mashing something onto the end of a loanword like ベイズ.
The voicing thing is known as rendaku. Generally it’s a bit of a mystery when it will and will not happen. This thesis lists a bunch of proposed rules, two of which seem relevant:
Rendaku is favoured if the compound words are native-Japanese (yamatokotoba). This might be the reason for kozukai vs mahoutsukai, ko is native-Japanese and mahou is sino-Japanese. So by analogy, one would not expect voicing for beizutsukai.
Noun+Verb compounds exhibit rendaku if the noun is an “adverbial modifier” but none if it’s a direct object. In “using magic” 魔法を使う magic is a direct object, so no voicing. On the other hand kozukai (‘little servant’?) is an Adjective+Verb, which explains the voicing.
In any case, I guess the upshot is that we should expect beizutsukai, without rendaku,