-ㅅ-

Usage notes

This morpheme can surface only in the following environments:

• After a vowel and before a consonant or /i/
• After a sonorant consonant, i.e. a nasal or liquid, and before an obstruent

If it is followed by /i/ or /j/, the interfix /n/ intervenes between the morpheme and the second element: hence 뒷일 (dwinnil, “later things”) is realized as [twinnil] rather than [twisil]. While the morpheme may theoretically also be present between obstruent consonants, its tensing effects are indistinguishable from the regular tensing of an obstruent when preceded by another obstruent. Therefore, its existence cannot be ascertained in those compound nouns.

The existence of the interfix in specific words varies greatly depending on age and location. Two compounds involving the same morpheme and with similar semantics may still differ in their use of the interfix: there is tensing in 비빔밥 (bibimbap, “bibimbap, a rice dish with mixed ingredients”) but not in 비빔국수 (bibimguksu, “bibim-guksu, a noodle dish with mixed ingredients”), despite effectively identical semantics. In Sino-Korean, compare:

• 불법 (佛法, bulbeop, “Buddha's law”) without tensing but 신법 (神法, sinppeop, “god's law”) with tensing
• 소수 (小數, sosu, “decimal number”) without tensing but 소수 (素數, sossu, “prime number”) with tensing

With compound Sino-Korean words, the interfix appears (in the form of tensing) only before certain hanja, sometimes even in non-genitive or non-attributive constructions. A leading hypothesis is that the interfix has a tendency to appear in Sino-Korean compounds which are still transparent compounds in modern Korean. For instance, 병 (病, byeong, “disease”) is tensed in 화병 (火病, hwappyeong, “sickness from frustration”) because Koreans perceive the word as a compound of 화 (火, hwa, “anger, frustration”) and 병 (病, byeong, “disease”). It is not tensed in 나병 (癩病, nabyeong, “leprosy”) because 나 (癩, na) in isolation is not a valid morpheme in Korean, and the word is therefore not perceived as a compound by Korean speakers. Nonetheless, a recent study (Yu 2019) notes that it appears impossible to posit any satisfactory semantic explanation that explains all cases of Sino-Korean tensing.

The extremely common word 거 (geo, “thing”) is always preceded by the interfix in genitive constructions, and 속 (sok, “inside”) is also very commonly preceded by it:

• 니 꺼 (ni kkeo, “your thing”) - from 니 (ni, “you”) + ㅅ + 거 (geo, “thing”)
• 일본 꺼 (ilbon kkeo, “something from Japan”) - from 일본 (日本, ilbon, “Japan”) + ㅅ + 거 (geo, “thing”)
• 컴퓨터 쏙 (keompyuteo ssok, “inside the computer”) - from 컴퓨터 (keompyuteo, “computer”) + ㅅ + 속 (sok, “inside”)

Contrast with non-genitive forms, with no tensing:

• 차가운 거 (chagaun geo, “cold thing, something which is cold”)
• 속이 안 좋아 ― sok-i an joa ― I have a stomachache

However, these are usually analyzed as special allomorphs of these common words, rather than as the morpheme continuing to function as a true particle.

In North Korea, the morpheme is only pronounced and not written. In the South Korean prescriptive standard, it is only written under the following conditions:

• The first component of the compound word ends in a vowel
• At least one of the components is a native word, with six exceptions in which it occurs in entirely Sino-Korean words.

It is unwritten in other conditions. In practice, it is often omitted even when it should prescriptively be written.

The latter criterion did not exist in South Korea until 1988, and Sino-Korean words were also prescriptively written with the morpheme if the first element ended with a vowel:

• 잇과 (理科, itgwa), 홋수 (戶數, hotsu) instead of current 이과 (理科, igwa), 호수 (戶數, hosu)

This was deprecated in a 1988 spelling reform, except for six words which were excluded from the reform for unclear reasons, but certain officially deprecated spellings such as 갯수 (gaetsu) for 개수 (個數, gaesu, “number”) and 댓가 (daetga) for 대가 (代價, daega, “price”) remain widespread if nonstandard.