Korean Language Stack Exchange is a question and answer site for linguists, teachers and students of the Korean language. It only takes a minute to sign up.
Sign up to join this community
I am trying to make sense of the grammar behind ~ㄹ/을 때 for past sentences.
As far as I know, ~ㄹ/을 때 is making use of the future tense modifier, used for example for the common construction ~(으)ㄹ 거예요, and it feels a bit strange to use the future tense modifier with the past verb stem (먹었을 때) in order to say "when I ate" instead of the past tense modifier (먹은 때) that you would normally use for something like 먹은 음식 (the food that I ate).
In addition to the perfectly scoped answer from @user13229973 , I just wanted to add that from linguists' perspective, using past tense in expressing conditional (or non-/counter-factual) modality is actually quite common in many languages, for example specifically in English, there is the past subjunctive "were" (often called "irreal/irrealis" mood by linguists).
As a quick note on mood (서법) and modality (양태) in Korean: many Korean linguists agree that the modern Korean language generally lacks grammatical markers (mood) for expressing modality. However, some argue that complementary usage of -ㅁ vs -기 (in nominalization) and -ㄴ vs -ㄹ (in attribution) show they are "weak" mood markers for epistemic modalities. Some of them also argue they are residue of middle-age Korean which is believed to have more explicit mood markers in a broader conjugational range.
For what it's worth, here are some random examples of -ㄴ, -ㄹ (or even -ㅆ- + -ㄹ) used as more mood markers than tense markers.
- 먹을 때에는 식사예절을 지키세요 (temoporal overlap)
- *먹은 때에는 식사예절을 지키세요
- ?먹은 때에는 양치를 하세요 (temporal "after")
- 먹은 후에는 양치를 하세요 (temporal "after")
- 먹었을 때에는 양치를 하세요 (temporatal "after")
- 조용히 먹을 것 (deontic) (also as see https://korean.stackexchange.com/a/3430/358)
- 지금쯤이면 먹은 줄 알았지 (epistemic necessity)
- 지금쯤이면 먹을 줄 알았지 (epistemic possibility)
- 젊을 때 다쳤어요 (temporal overlap)
- *젊은 때 다쳤어요
- ?젊었던/젊던 때 다쳤어요
- 젋었을 때 다쳤어요 (temporal overlap)
- 젊을 때는 다 그래 (generic youth, that's the way it is for the youth)
- ?젊었을 때는 다 그래
- 젊었을 때 얘기해 주세요 (your specific youth, tell me about your old days)
- ?젊을 때 얘기해 주세요
- 시간될 때 얘기해 주세요 (conditional, talk to me when you're available)
- 시간되는 때 얘기해 주세요 (attribution, tell me the time you're available)
(*: unacceptable, ?: marginal/unnatural, subjectively judged by me)
I think it'd be more helpful that we talk about the semantics of the phrase (1) 뭘 했었을 때 that 먹었을 때 is based on, versus the phrase (2) 뭘 했던 that 먹은 is based on.
(1) 뭘 했었을 때
You're right in saying that 뭘 했었 "다" is strictly the past tense of an action, but when you conjugate it to 뭘 했었 "을 때", then it becomes the conditional premise A of the proposition if A then B. And the same conditional premise 뭘 했었을 때 is used to complete the proposition that could take place in either the past or future. So in English, the conditional premise 뭘 했었을 때 depending on its tense is equivalent to saying either when you did that OR when you do that.
i. Example of conditional premise 뭘 먹었을 때 in the past tense: 너 김치 먹었을 때 배 많이 아팠잖아. When you ate kimchi, you had a bad stomachache.
ii. Example of conditional premise 뭘 먹었을 때 in the future tense: 너 김치 먹었을 때 배 많이 아플거다. When you eat kimchi, you'll have a bad stomachache.
(1) 뭘 했던
뭘 했던 is also the conjugation of an action in the past tense, but it behaves more like an adjective to describe the subject that carried out that action.
i. Example of adjective 뭘 먹은: 밥 많이 먹은 강아지 똥 많이 싼다. The puppy that ate a lot poops a lot
Because 먹은 is an adjective, adding 때 would be redundant as 때 more or less describes the state of something. So 먹은 때 would sound unnatural to native speakers. At least to me it does.
