Thread by @ZevHandel

So here’s a cool thing about the Korean word “bibimbap” 비빔밥—aside from the fact that it’s the name of a delicious, healthy, and versatile food.

The letter ㅂ <b> appears four times in the written form of the word, 비빔밥. And each one of those occurrences represents a different sound!

That’s it, that’s the tweet.

Well, okay, we can talk about this in more detail. Follow on if you want to get into the weeds. Let’s look at each of the four ㅂ sounds in order.

1. The first ㅂ, at the beginning of the word, represents a voiceless, slightly aspirated bilabial [p]. To English speakers, it sounds intermediate between /b/ and /p/. That’s why some romanizations of Korean render it as <p> and others as <b>.

2. The second ㅂ, at the start of the second syllable, represents a fully voiced sound [b].

3. The third ㅂ, at the start of the third syllable, represents a “tense” consonant. There isn’t a widely accepted International Phonetic Alphabet (IPA) representation, but [p͈] will do for our purposes.

This is a sound that can be, and often is, written ㅃ <bb>. But it isn’t written that way in this word, for reasons I’ll get into a bit later.

4. The fourth ㅂ, at the very end of the word, represents an unreleased bilabial stop [p̚ ].

Three of these sounds—the ones labeled #1, #2, and #4 above—are allophones of the Korean phoneme /b/. This means that they are variant pronunciations that are automatically and unconsciously produced by native Korean speakers in a predictable way.

When Korean /b/ occurs at the beginning of a word it is automatically pronounced [p]. When between two vowels, it is automatically pronounced [b]. And at the end of a word, it is pronounced [p̚ ].

Korean speakers not only do this automatically, they don’t perceive these sounds as different. But if, like me, you are a foreign learner of the language, you have to make a conscious effort to master these positionally-determined variant pronunciations.

An analogous phenomenon is found with the English phoneme /p/. Most English speakers aren’t aware that the actual pronunciation of this sound at the beginning of a word, as in “peek”, is quite different from its pronunciation after /s/, as in “speak”.

The /p/ of “peak” has quite a bit of aspiration, meaning there is a significant, audible puff of air blown out after the lips are separated: [pʰ]. The latter has no puff of air at all: [p]. So the actual pronunciations are [pʰik] and [spik].

(If you are a native speaker, you can test this yourself by holding the palm of your hand in front of your lips at a distance of 3 or 4 centimeters while you say each word. You’ll feel the puff of air on “peak”.)

English speakers make these distinct pronunciations of /p/ automatically and unconsciously; they perceive the two sounds as being the same; and they write them with the same letter, <p>.

But if you switch the pronunciations, for example by pronouncing “speak” as [spʰik], it sounds really weird!

As for pronunciation #4, “unreleased” means that after the [p] is articulated, the lips remain closed. This is a pretty common way that word-final consonant sounds are articulated in many languages, but not in English.

If you look at a native English speaker saying the word “sup”, you’ll notice their lips parting at the end of the /p/ sound, which provides a brief but audible release burst.

But if you ask a Korean speaker to say “sip” 십 ‘ten’, you can observe that their lips stay closed at the end of the word. The resulting sound is a bit quieter than the English word-final /p/.

It's the same with Cantonese “sap” 十 ‘ten’. The lips stay closed. That’s one reason English “sup” and Cantonese “sap” sound different.

Sound #3, the “tense b”, is perhaps the most interesting. It’s not an allophone, but a distinctive Korean sound that is commonly written differently from ㅂ /b/. Like the other tense consonants of Korean, it’s written with a double letter: ㅃ <bb>, as in “bbul” 뿔 ‘horn’.

This tense sound is perceived as distinct from /b/ by native speakers, and contrasts with it. There are plenty of examples of pairs of words like “bul” 불 ‘fire’ and “bbul” 뿔 ‘horn’ where the tense pronunciation distinguishes one word from the other.

So why isn’t the tense b sound in the word “bibimbap” written as ㅃ <bb>? Why is the written form of the word 비빔밥, not 비빔빱?

I'm glad you asked!

The answer is connected to two interesting Korean-language phenomena: one is a word-formation change called “compound tensing”, and one is a feature of the writing system—a feature that is coincidentally shared by English orthography.

Compound tensing is a phenomenon that happens in Korean when a compound noun is formed out of two component nouns, like the English words “mailbox” (“mail” + “box”) and “classroom” (“class” + “room”).

The word “bibimbap” is a compound formed from two nouns: “bibim” 비빔 ‘a mix, a hash’ and “bap” 밥 ‘cooked rice’.

When two nouns combine into a single word in Korean, and the second of those nouns starts with a plain consonant like /b/, more often than not that plain sound changes into a tense sound. This is the reason that “bibim” + “bap” is pronounced as /bibim/ + /bbap/.

So why is it *written* as “bibimbap” 비빔밥 and not “bibimbbap” 비빔빱?

It's because Korean writing is decidedly *not* "phonetic", despite what is often claimed about it.

Korean spelling is the result of a conscious decision that was made by language reformers in the early 20th century as they were debating how to standardize Korean orthography using the Hangul alphabet.

The question was: Should the spelling reflect the actual pronunciation? Or should it represent in a consistent way the underlying components that make up the words?

They ended up deciding on the latter. This means that an underlying component—like “bap” 밥 ‘cooked rice’ — gets spelled the same way in every word that it occurs in, even when its pronunciation shifts to /bbap/.

Now here’s the part that is really interesting to me: This is actually a lot like English spelling, although in English it’s the result of historical accident and conservative orthographic habits, not deliberate language policy.

Think about a word like “photograph”. When we add the suffix “-y” to get the word “photography”, the pronunciation of the “photograph” part changes dramatically.

Among these changes: the second vowel of “photograph” is schwa [ə], and the second vowel of “photography” is “ah” [ɑ], yet both are spelled identically with the letter <o>.

A more striking example from English is “electric” and “electricity”. The spelling of “electricity” reflects its underlying structure as “electric” + “-ity”. But if English spellings indicated actually sounds, they would be written something like “elektrik” and “elektrisity”.

To a second-language learner trying to pronounce unfamiliar written words, “elektrik” and “elektrisity” would make life a lot easier. But there are advantages, especially to native speakers, to having the same word component spelled consistently, in Korean as in English.

One more example from English. The plural “-s” suffix in “kits” is pronounced [s], while that same plural suffix in the word “kids” it is pronounced [z].

But we don’t write “kidz”; we English speakers prefer our plural suffix to be written consistently as “s”, and as native speakers we automatically apply the rule that yields [z] after certain sounds.

Just the same way that a Korean speaker automatically applies the rule that changes “bap” to “bbap” in “bibimbap”

That brings us to the end of the thread. Hopefully I haven’t left you as mixed up as a bowl of 비빔밥!

Side note: Good dictionaries will tell you where compound tensing exists.

https://twitter.com/zevhandel/status/1357909886583984129