Why do some loanword sounds get "contracted" when katakanized, but not others?

Question

Not sure the best way to word this. Why is it that some sounds from a loanword get combined down to a shorter or different sound while other do not? Here are some examples:

Shortened, but sound weird to me/many Japanese learners

• Pineapple　→　パイナップル　not　パイン・アップル　→　The 'n' and 'a' sound combine to ナ instead of remaining separate
• Menu　→　メニュー　not　メン・ユー　→　'n' and 'y' combine to ニュ instead of remaining separate
• Double-"you"　→　ダブリュー　not　ダブル・ユー　→　'r' and 'y' combine to リュ instead of remaining separate

Separate, seemingly breaking the above patterns

• Time-out　→　タイムアウト　not　タイマウト　→　'mau' becomes 'mu' + 'au' seemingly unnecessarily
• Backup　→　バックアップ　not　バッカップ　→　why not similar to the "pineapple" pattern?
• Piano　→　ピアノ　not ピャノ　→　Similar to "menu" in my eyes

What gives?

Answer 1

When words are borrowed in speech, they're generally "repaired" to match the phonology of the target language. In Japanese, that usually means picking the nearest consonant and vowel sounds and adding epenthetic vowels to avoid consonant clusters that aren't allowed (like /str/ → /sutor/), although other methods of repair are occasionally used (e.g. deletion).

But this isn't an exact process. It depends on the pronunciation of the original word (which may vary) and the way the word is perceived in the target language (which may also vary). Sometimes more than one repair is possible, and sometimes words are borrowed more than one way: consider strike ストライク and strike ストライキ, or Hepburn ヘボン and Hepburn ヘプバーン, as in James Curtis Hepburn and Audrey Hepburn.

And of course, there are further complications. Many words are borrowed via writing rather than speech, and words have been borrowed at different points in time, resulting in different pronunciations than you might expect. Most speakers no longer say the /h/ in white /hwaɪt/, but it was relatively common back when ホワイト was borrowed.

---

Let's look at your examples.

In the case of time-out and backup, we can see that each word corresponds rather transparently to a pair of words, time out and back up. I suspect that this correspondence influenced both words, whether they were borrowed via speech or writing. What's more, both time-out and backup can be pronounced like a pair of words in careful speech. Given that, the way they're borrowed makes sense.

I think that pineapple is less transparent (despite the spelling) and that most speakers pronounce the /n/ as a syllable-initial consonant (following the maximum onset principle), like "pie napple" rather than "pine apple". Honestly, the reason this word in Japanese doesn't match my expectations as an English speaker is the location of the pitch accent, not because it has ナ in the middle.

A couple of your examples are implausible. Menu simply doesn't sound like メンユー, it sounds like メニュー. Try saying it both ways! And there is no separation to "remain separate". Piano doesn't sound like ピャノ. The vowels are fairly distinct in English and I see no reason why a Japanese speaker would perceive them as a glide or interpret the spelling as representing a glide.

But I'm not sure it's really possible to answer a question like this rigorously. I see no reason why 'W' double-you couldn't be ダブル・ユー as well, but whoever borrowed it first must have perceived it the other way and ダブリュー became more popular. For that matter, I don't see why time-out couldn't have been borrowed as タイマウト. It's not an exact science and I'm afraid there are no absolute rules to explain it―all we can really do is try to find tendencies and patterns.

Answer 2

I think that along with snailboat's answer, some of these can be explained by nasalization occurring in Japanese.

Where you perceive a syllable break, e.g. in [men.ju], trying to convert the phonetics to kana doesn't work out, because a final ン is usually nasal, i.e.

[menjuː] menu
[meɴjuː] メンユー
[menʲuː] メニュー

Similarly,

[pain.æpl̩] / [painæpl̩]          pine apple
[painap̚.puɽu]                   パイナップル
[paiɴ.ap̚.puɽu] / [paiɴjap̚.puɽu] パイン・アップル

So メンユー would usually be pronounced with a nasal ン, i.e. [meɴjuː], which is further than [menʲuː] from the English [menjuː] and similarly for pine apple.

(Note that in パイン・アップル we see the [j] of 千円 [seɴjeɴ] appear, which gives yen.)

The result is that the syllable border in English may fall within a single mora in Japanese to avoid nasalization, and thus stay closer to the English pronunciation.

ピアノ may be explained either

• by being a case of romanization by spelling, or
• by romanization by sound, since native English speaker do pronounce piano [pi'ænoʊ], not just ['pʲænoʊ]. (Wiktionary in fact only lists the former.)

As for タイムアウト and バックアップ, my theory is that the syllable border has not been welded into a single mora in Japanese is that in this case, the syllable border is in fact a word border and both words on either side have probably existed longer than the compound word.

Answer 3

I basically agree with the 1st 3 paras of Snailboat's answer but to put it differently:

If an English word is not pronounced consistently throughout the English speaking world then, even if the spelling is consistent, you cannot predict how it will be pronounced as a loan word.

"Schedule" is one example: In the States the "sch" is pronounced to sound like "sk", in Britain it usually sounds like "sh". To my ears the Japanese スケジュール shows the influence of an American drawl, but I don't really know, and FWIW many in the UK now say the word with a "sk" sound, and some regional accents give the "d" a "j" sound.

If the adoption of loan words is organic then you can't really expect to find rules but you might expect some consistency. To identify the consistency you probably need know how the foreign word (and the pronunciation of its source) sounds to native Japanese ears, not yours or mine.

Answer 4

Japanese is full of strange pronunciation of foreign words(most English).
Although ability in English is increasing in Japan, it has been quite low especially for a developed exporting nation, and even though there has and still is a lot of interest in other cultures, things are often taken into Japan and used in a way to fit Japanese people.
I don't think a lot of these words enter Japanese through an official process, and I don't think many Japanese people question the authenticity of most of these words, so I think this is the most probable reason for these differences in pronunciation.