Friday, August 7, 2020

Lexical Aspect

In grammar, actionality, or aktionsart, refers to a temporal property of a predicate. The term lexical aspect is used for the actionality of words (lexemes). Words that have different lexical aspects yield different meanings when used with the same syntax. For example:

  • "Tarou is running" means that "Tarou already ran" for a while, even if just for one second.
  • "Tarou is dying" doesn't mean that "Tarou already died" for a while, not even for one second.
  • Tarou wa hashitte-iru
    Tarou is running. (translates to progressive, "is ~ing.")
  • Tarou wa shinde-iru
    Tarou has died. (translates to perfect, "has ~ed.")

Above, we see that the verbs "to run," hashiru 走る, and "to die," shinu 死ぬ, have different temporal meanings when conjugated to the same forms, the progressive form in English, and the ~te-iru ~ている form in Japanese.

This happens because these words have different lexical aspects.

Note: "lexical aspect," goi-teki asupekuto 語彙的アスペクト, shouldn't be confused with "grammatical aspect," bunpou-teki asupekuto 文法的アスペクト. The progressive, perfect, perfective, imperfective, and so on are grammatical aspects, not lexical aspects.


There are two works regarding lexical aspect that are important for understanding Japanese.

The first is by the Japanese linguist Kindaichi Haruhiko 金田一春彦, published in 1950.

The second is by the German linguist Zeno Vendler, published in 1957.

These two theories provide different definitions for the term "stative verb," joutai-doushi 状態動詞. Consequently, sometimes you'll see this term based on Kindaichi's definition, other times you see it based on Vendler's definition.

In fact, there are several other definitions of state and stative verb in Japanese(Wu, 2017), but we won't be seeing all of them.

金田一 (1950)

In his work about aspect in Japanese, Kindaichi 金田一 (1950) divided Japanese verbs into four categories according to their relationship with the ~te-iru ~ている form:

  1. joutai-doushi
    Stative verbs. Verbs that don't accept the ~te-iru form.
    • The existence verbs aru ある (and its irregular negative form nai ない), and iru いる.
    • Copulas such as de aru である and de gozaru でござる.
    • Potentials such as the potential verb hanaseru 話せる, "to be able to speak," and dekiru できる when it expresses ability: the dekiru which is the potential form of suru する, and that appears in phrases like koto ga dekiru ことができる.
    • The perceptional verb mieru 見える, "seems."
    • The cognitive verb wakaru 分かる, "to know."
    • The verbs yousuru 要する, "to require," and ataisuru 値する, "to be worth."
    • The verb iu 言う of phrases like {Hideyoshi to iu} hito 秀吉という人, "a person {called HIyoshi}."
    • The auxiliary verb ~sugiru ~すぎる which attaches to the stem of adjectives, e.g. chiisa-sugiru 小さすぎる, "too small."
  2. keizoku-doushi
    Continuation verbs. Verbs that show an action has already started, but not finished when conjugated to ~te-iru.
    • Actions performed by someone, yomu 読む, "to read," kaku 書く, "to write, " warau 笑う, "to laugh," etc.
    • Natural phenomena like furu 降る, "to rain," physical movement like yureru 揺れる, "to shake," and moeru 燃える, "to burn."
  3. shunkan-doushi
    Moment verbs. Verbs that show an action has already occurred, and a resultant state remains when conjugated to ~te-iru.
    • The verbs "to die," shinu 死ぬ, "to disappear," kieru 消える, "to end," owaru 終わる, "to start," hajimaru 始まる, "to marry," kekkon suru 結婚する, along with many others.
    • The cognitive verb shiru 知る, "to come to know."
  4. dai-yon-shu no doushi
    Fourth-type verbs. Verbs that are never used without ~te-iru. (called so because naming things is hard.)
    • The comparisonal verbs niru 似る, "to resemble," sobieru 聳える, "to tower over," sugureru 優れる, "to surpass," arifureru ありふれる, "to be common."
    • The verb suru する in phrases like marugao wo shite-iru 丸顔をしている, "to have a round face."

Although not listed among the examples in the study, we can safely assume that the following are also stative verbs:

Among Kindaichi's observations is that you don't use an imperative with stative verbs(pp. 59–60). A stative verb merely expresses the way things are at the moment. There's no action. You can't order someone to do or to not do something that's not an action. For example:

  • yomu na!
    Don't read it!
  • *yomeru na!
    *Don't be able to read it! (this is wrong.)
  • *wakaru na!
    *Don't know it! (also wrong.)

Vendler (1957)

The work of Vendler (1957) serves as basis for countless theories formulated about tense and grammatical aspect. In "Vendlerian terms," there are four distinct types of actionalities:

  1. Activities.
    Which have a duration, but no completion point (they're atelic).
    • "To run" is an activity. If you say "I'm running," you've already run a bit, so it's durative.
  2. Accomplishments.
    Which have a duration, and a completion point (they're telic).
    • "To run a mile" is an accomplishment.
    • The phrases "I finished running" and "I ran in 30 minutes" don't make sense alone because running has no completion point, it's an activity.
    • However, the phrases "I finished running a mile" and "I ran a mile in 30 minutes" make sense, as the action finishes when we run a whole mile., so it's an accomplishment.
    • If I say "I'm drawing a circle," that means I've already drawn part of the circle. If I say "I'm reading a manga," that means I've already read a page or so. So accomplishments have durations.
  3. Achievements
    Which have no duration.
    • "To die" is an achievement.
    • If you say "I'm dying" that doesn't mean I died yet. If I say "I'm reaching the top of the mountain" that doesn't mean I reached the top yet. If I say "I'm marrying with the love of my life," that doesn't mean I married yet.
  4. States.
    Which are durative, but don't involve actions at all, they're static, instead of dynamic.
    • "To know" is state.
    • If I say "I know French," that means "knowing French" is state that's true about me right now. On the other rand, if I say "I watch anime," that doesn't meaning "watching anime" is something I'm doing right now. Even though they're both in the simple present in English.
    • Similarly: "I think so," "I hate spinach," "I love Emilia," "I don't feel so good," etc. are all used in the simple present to state something that's true right now.

The Japanese translations would be(陳, 2014:32):

  1. katsudou-doushi
    Activity verb.
  2. tassei-doushi
    Accomplishment verb.
  3. toutatsu-doushi
    Achievement verb.
  4. joutai-doushi
    State verb.

Vendler's theory is about predicates, about whole phrases and sentences that describe subjects. However, it's typically used to refer to the lexical aspect of verbs, which would be about words instead.

For instance, "to run" is lexically an activity, it's the infinitive form of "run," a lexeme. Meanwhile, "to run a mile" is a whole phrase, not a single lexeme.

Adding the modifier "a mile" to "to run" changes its actionality to accomplishment. Hence why the theory deals with whole predicates, not just individual words.

One of Vendler's observations about state verbs is that they don't make sense in the progressive(p. 25).

  • *I'm knowing french. (this is wrong.)

Telic predicates—achievements and accomplishments—have a feature called the "imperfective paradox," which is when they're in the progressive form and don't entail the telos (the completion) has actually happened(Bertinetto, 2001:6). Observe:

  • If I'm dying, and I stop.
    • I haven't died.
  • If I'm running a mile, and I stop.
    • I haven't run a mile. Maybe I've run half of a mile, but not a whole mile.
  • If I'm running, and I stop.
    • I have run. Maybe I haven't run a lot, maybe I just ran a bit, but I HAVE run.

English verbs that involve movement are not stative, so some people call them "dynamic verbs," which would contrast with "static verbs."

In Japanese, some non-stative verbs are cognitive, so it would be weird to call them dynamic—the term eventive is used instead.

An "eventive verb" is a verb that expresses an event, while a "stative verb" expresses an state. Eventive verbs include activities, accomplishments and achievements. A lot of research is primarily done on the distinction between stative and eventive verbs.


The theories we've seen above define lexical aspect mostly according to the usage of words with the English progressive and the Japanese ~te-iru ~ている form.

Unfortunately, this creates a grave weakness: if we can find instances of words being used in a way that contradicts the usage predicted by the theories, then the theories lose strength overall.

Both theories forbid the usage of stative verbs with the progressive and the ~te-iru form.

In English, "stative verbs can be exceptionally used in the progressive form to indicate temporary state," for example(Freund, 2016:123, citing Quirk et al.,1985; Biber et al., 1999; Leech et al., 2009):

  • George is loving all the attention he is getting.
    • Similarly "I'm thinking" is also normal.
    • However, "I'm knowing" is still odd.

In Japanese, stative verbs can also exceptionally be used in ~te-iru form to express a temporary state(adapted from 加藤, 2010:134–135):

  • Tarou wa, ima no tokoro, chanto setsumei dekite-iru
    Tarou, for now, has been able to explain it properly.
    • This phrase implies that, while Tarou has been able to explain it properly so far, it's uncertain whether he will able to explain the whole thing properly.

Naturally, there must be something very bizarre must be going on here for both predictions to be broken, and for both of them result in the same "express a temporary state" effect.

Fortunately, there's a way to explain how this all works.

Unfortunately, it's too complicated.

The short version is that the progressive form, perfect form, ~te-iru ~ている, and ~te-aru ~てある have the function to instantiate the event expressed in the verb, somehow making it relevant in the present. This function ends up resulting in several usages and meanings.

The long version is that if we apply genericity(Krifka, 1995), and assume that states are inherently gnomic for they lack a spatiotemporal property, and that the progressive form has the function to bind them to a temporal episode through evidential coercion, whereas the simple present abstracts events into gnomic habitual by removing their temporal anchoring(Fernald, 1999), and group the progressive and the perfect into an episodicalizer category(or we could just dump them all into the imperfective, whatever works(Smith, 1997, as cited in Shirai, 2000:330n3)), then we can come up with an unifying theory that explains both English syntax and ~te-iru ~ている and ~te-aru ~てある, distinguishing the Japanese episodicalizers by giving them an actualization requirement that English progressives lack with achievements, and this would account even for the distinction between gnomic habituals and episodic iteratives(Bertinetto & Lenci, 2010), the latter of which is expressed through ~te-iru as predicted by the claim that it functions as an episodicalizer(see Sugita, 2009, although she calls iteratives habituals because different authors have different definitions for these two things), so long as we redefine stage-level predicates as assertions with the stage as the topic(Erteschik-Shir, 1997, as cited in Cohen & Erteschik-Shir, 2002:133) in order to explain why certain statives, which should be syntactically gnomic, are ultimately bound to episodes in SLPs, which would fit neatly with how double subject constructions work in Japanese(Shibatani, 1999), given that we'd be considering stages to be a covert topicalized large subjects that can't be overtly marked simply because there's no word to refer to it so it can't be uttered, and that would also even explain why the ga が particle changes behavior when used in SLPs compared to ILPs(鈴木, 2014).

I think. I'm not quite sure, to be honest.


  1. 金田一春彦, 1950. 國語動詞の一分頬. 言語研究, 1950(15), pp.48-63.
  2. Vendler, Z., 1957. Verbs and times. The philosophical review, 66(2), pp.143-160.
  3. Krifka, M., Pelletier, F.J., Carlson, G., Ter Meulen, A., Chierchia, G. and Link, G., 1995. Genericity: an introduction.
  4. Fernald, T.B., 1999. Evidential coercion: Using individual-level predicates in stage-level environments.
  5. Shibatani, M., 1999. Dative subject constructions twenty-two years later.
  6. Shirai, Y., 2000. The semantics of the Japanese imperfective-teiru: An integrative approach. Journal of pragmatics, 32(3), pp.327-361.
  7. Bertinetto, P.M., 2001. On a frequent misunderstanding in the temporal-aspectual domain: The ‘perfective-telic confusion’. Semantic Interfaces: Reference, Anaphora and Aspect, Stanford: CSLI Publications, pp.177-210.
  8. 陳建瑋, 2014. 日本語のアスペクト形式 「テイル」 の習得に関する横断研究―動詞の語彙的アスペクトによる影響について―. 言葉と文化, 15, pp.31-47.
  9. Cohen, A. and Erteschik-Shir, N., 2002. Topic, focus, and the interpretation of bare plurals. Natural Language Semantics, 10(2), pp.125-165.
  10. Sugita, M., 2009. Japanese-TE IRU and-TE ARU: The aspectual implications of the stage-level and individual-level distinction. City University of New York.
  11. Bertinetto, P.M. and Lenci, A., 2010. Iterativity vs. habituality (and gnomic imperfectivity). Quaderni del laboratorio di linguistica, 9(1), pp.1-46.
  12. 加藤重広, 2010. 北奥方言のモダリティ辞. 北海道大学文学研究科紀要, 130, pp.125-左.
  13. 鈴木彩香, 2014. ガ格の総記/中立叙述用法と裸名詞句の総称/存在解釈の統一的説明. 言語学論叢 オンライン版, (7).
  14. Freund, N., 2016. Recent Change in the Use of Stative Verbs in the Progressive Form in British English: I’m loving it. University of Reading Language Studies Working Papers, 7, pp.50-61.
  15. Wu, Y., 2017. 日本語の< 状態>< 状態動詞> 再考. Journal of Humanities and Social Sciences, 43, pp.101-118.


Leave your komento コメント in this posuto ポスト of this burogu ブログ with your questions about Japanese, doubts or whatever!

All comments are moderated and won't show up until approved. Spam, links to illegal websites, and inappropriate content won't be published.

  1. I have this idea that in the case that certain verb types are always transitive or intransitive or direct or indirect or something that it would be helpful to add them and to clarify what is not.

    1. They aren't related.

      Stative verbs are often transitive, but not always. E.g. "aliens exist" is intransitive and stative.

      Some ergative verbs are activities intransitively but accomplishments transitively: "I'm reading" is an activity while "I'm reading a book" is an accomplishment. However, you can also make them into accomplishments by adding a time duration: "I'm reading for 1 hour" is intransitive and an accomplishment predicate.

      There are intransitive achievements, "I died," as well as transitive ones, "I reached the top of the mountain."

      The same is true for Japanese. Transitivity and lexical aspect are mostly unrelated.

  2. Would you suppose the temporary state thing is the only exception?And at least "gnomic" and "genericity" don't have language related definitions on their front page search results.

    1. I think so. From what I understand, some predicates are gnomic by default, and some sentences can't be gnomic, so you have to convert them to episodic (i.e. temporary) predicates through the English progressive and Japanese ~te-iru forms. Typically this only happens when you have a specific, explicit time, e.g. I was being able to do this YESTERDAY, but sometimes this time is implicit instead, e.g. I was being able to do this.

  3. When it says (Japanese has non stative verbs that are like English "dynamic verbs" but are cognitive?),it might be helpful to give an example.At the time of reading this I am personally having difficulty trying to imagine such a verb.

    1. それを後悔する = [I] will regret that.
      それを後悔している = [I] regret that.

      "Regret" is a stative verb in English, but its Japanese translation, 後悔する, is an eventive verb in Japanese, since it becomes future-tensed when used in nonpast form, and requires ~te-iru to become present-tensed.