In Classical Japanese, the auxiliary verb "-つ" ("-tsu") has a perfective function, indicating the completion of an action or process. According to Haruo Shirane's Classical Japanese: A Grammar, "-て" ("-te"), the ren'youkei or continuitive conjugation of this auxiliary verb, later developed into a conjunctive particle with several different meanings: (1) temporal/sequential connection, (2) parallel connection, (3) causal connection, and (4) existing condition.

I can see how a usage of a marker for perfective aspect could develop meanings 1 and 3, as these both rely on posterior or definite qualities, but my question is how do meanings 2 and 4 derive from such a form? 1 and 2 in particular seem to have conflicting definitions, the first meaning "and then", the second meaning "at the same time". Are there any explanations for how the meaning could shift in this way? And are there any papers that discuss the semantics behind the development of this kind of change (whether specific to Japanese or not)?


1 Answer 1


While Shirane follows the traditional grammarian's account (in this as in everything else), according to linguist Bjarke Frellesvig's A History of the Japanese Language, the non-perfective, connective -te already existed in Old Japanese. In this analysis, there is then no evidence to say that the connective -te (also called a "gerund") is a later derivation of the perfective -tu/te; both of these usages go as far back as we have records. The connective has a wide range of uses comparable to English "and" or "-ing", among which are included the things about parallelism, conditions etc.

Consider for example the grammatical role of -te in the following passage of the Kojiki (ca. 712 – one of the oldest sources). The goddess Suseri-bimye (Suseri-hime) is talking to her husband Opokuninusi (Ōkuninushi) about his polygamy:

賣迩斯阿禮婆 那遠岐弖 遠波那志 那遠岐弖 都麻波那斯
めにしあれば なをきて をぱなし なをきて つまぱなし

1 mye ni si areba
2 na wo kite
3 wo pa na-si
4 na wo kite
5 tuma pa na-si.

1 Being a woman [as I am],
2 excluding you,
3 [I can] have no other man;
4 excluding you,
5 [I can] have no other husband.

Similarly wide uses of -te can be found elsewhere in the Kojiki, the Man'yōshū etc., meaning they go as far back as we have data.

Pefective -tu/te makes a pair with perfective -nu; they occur in Old Japanese more or less with the same function, but with -tu/te preferred for transitive verbs and -nu for intransitives. They're thought to be both derived from copulas, -te from to and -nu from ni. Frellesvig thinks connective -te, perfective -te, copulas -to, -tu, genitive -tu and continuative -tutu all trace back to some proto-element with a -t-; but this is reconstruction, we don't have documentation of this common ancestor.

  • Also of note regarding grammatical roles of -te is the -te aru form, which dates to antiquity (later shortened to -taru > modern -ta); and sound changes like kakite > kaite, which are found in the record as far back as Heian (though they were excluded from formal, classical writing). Commented Jul 25, 2017 at 16:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.