Compound Tensification & the 사이시옷 (물고기 → 물꼬기)

There is a third tensification, and it is the most meaning-driven of the three. When two nouns fuse into a compound, the plain initial consonant of the second noun often comes out tense — even when there is no stop batchim anywhere, sometimes no batchim at all. This is the 사잇소리 현상 ("linking-sound phenomenon"), and unlike the automatic after-stop rule, it is lexical and semantic: the tensing is a fossilized trace of a genitive "of" relation between the two nouns. 물고기 is literally "water-of-fish," and that hidden "of" is what you hear as the tense [꼬]. You learn these compounds one at a time, but the logic is consistent enough that you can usually predict them.

The pattern: the second noun's initial consonant tenses

Take two nouns, glue them into a compound, and the front consonant of the second one frequently tenses.

물고기 (water + fish) → [물꼬기]; 눈동자 (eye + pupil) → [눈똥자]; 길가 (road + edge) → [길까]; 밤길 (night + road) → [밤낄]. Notice the batchim in front is ㄹ, ㄴ, or ㅁ — sonorants that trigger nothing on their own. What licenses the tensing is not the batchim but the morpheme seam between the two nouns.

Why: a hidden "of," a compressed genitive

Historically these compounds carried a genitive marker — the medieval Korean "of" — between the nouns, and that marker has decayed into a purely phonetic residue: the tensing (and, in writing, sometimes an inserted ㅅ). So the tense onset is doing semantic work. It signals that the two nouns stand in a possessive or "belongs-to / made-of / is-for" relationship, not a loose sequence. This is why a genuine compound tenses but a mere two-word phrase does not:

• 물고기 [물꼬기] = one word, "fish" (a kind of thing that lives in water) → tense
• 물 고기 [물 고기] = two words, "water('s) meat / meat in water" (a description) → plain

Same syllables, different structure, different sound. The tensing is the audible signature of compoundhood.

When the first noun ends in a vowel: the 사이시옷 (inserted ㅅ)

Often the first noun of a compound ends in a vowel, with no batchim to write the linking sound on. Korean orthography solves this by inserting a ㅅ — the 사이시옷 — into the spelling, which then surfaces as a [t]-batchim and/or triggers the tensing.

바다 + 가 is written 바닷가 and said [바다까] (or, articulating the ㅅ, [바닫까]); 초 + 불 is written 촛불 [초뿔]; 코 + 등 is written 콧등 [코뜽]. The ㅅ you see in these spellings is not part of either root word — it is the written form of the linking sound. Recognizing a 사이시옷 tells you instantly that you are looking at a native compound and that the second element's onset is tense.

Two honest caveats. First, the 사이시옷 is only written when at least one of the two nouns is native Korean; pure Sino-Korean compounds normally don't get the ㅅ even when they tense — with six memorized exceptions (곳간, 셋방, 숫자, 찻간, 툇간, 횟수). Second, whether a given compound tenses at all is partly lexical: 김밥 is officially [김ː밥] but almost everyone says [김ː빱]; 고추장 stays plain [고추장]. Where the standard and real usage diverge, a dictionary's bracketed pronunciation is the authority.

The grammatical twin: the future ending -(으)ㄹ + a noun

The exact same tensing appears in a very high-frequency grammatical place: after the future/prospective modifier ending -(으)ㄹ, the following bound noun's consonant tenses. This is the ancestor of the compound linking sound showing up at a clause boundary, and it is one of the first tensings a beginner meets — inside 할 수 있어요.

할 수 있어요 is said [할쑤이써요] — the 수 tenses to 쑤 after the modifier ㄹ. Likewise 갈 데 → [갈떼], 할 것 → [할껃], 갈 곳 → [갈꼳], 먹을 게 → [머글께]. This is not the Sino-Korean ㄹ rule from the previous page — it is triggered specifically by the future adnominal ending, and it fires on native and Sino nouns alike. Because 할 수 있다 ("can do") is such a workhorse, drilling [할쑤] early pays off immediately.

Distinguishing it from the other two tensings

You now have three tensifications, and it helps to line them up:

The takeaway: the after-stop rule is pure phonetics; the other two are about structure and meaning. Compound tensing in particular is the one you have to think about semantically — ask whether the two nouns form a single "X-of-Y" unit.

Common Mistakes

1. Pronouncing a compound with a plain second consonant. The tensing is the compound's signature.

• ✗ 물고기 said [물고기] → ✓ [물꼬기]
• ✗ 눈동자 said [눈동자] → ✓ [눈똥자]

2. Missing the tensing in 할 수 있어요. One of the most common early errors.

• ✗ 할 수 있어요 said [할수이써요] → ✓ [할쑤이써요]

3. Ignoring the 사이시옷 as a pronunciation cue. The inserted ㅅ is telling you to tense.

• ✗ 바닷가 said [바다가] → ✓ [바다까]
• ✗ 촛불 said [초불] → ✓ [초뿔]

4. Over-applying it to loose phrases. Only genuine compounds tense; a two-word phrase stays plain.

• ✗ the phrase 물 고기 said [물 꼬기] → ✓ [물 고기] (plain — it's not the compound)

5. Expecting the romanization to show it. Revised Romanization never writes tensing — 물고기 is mulgogi, 할 수 is hal su. Read the bracket.

Key Takeaways

• Fusing two nouns into a compound often tenses the second noun's initial consonant: 물고기 [물꼬기], 눈동자 [눈똥자], 길가 [길까].
• This is semantic — the tense onset is a fossilized "of," so a real compound tenses where a loose phrase would not.
• When the first noun ends in a vowel, the linking sound is spelled with a 사이시옷 (ㅅ): 바닷가 [바다까], 촛불 [초뿔], 콧등 [코뜽].
• The future ending -(으)ㄹ triggers the same tensing on a following noun: 할 수 있어요 [할쑤이써요], 갈 데 [갈떼], 할 것 [할껃].
• It is lexical/grammatical, not automatic — and, like every tensing, invisible in Revised Romanization.