Tense, Aspect, and Mood in Japanese: A Map
Tense, aspect, and mood in Japanese are three separate jobs that the language often packs into one verb ending. That is why calling た "past tense" is a useful shortcut, but one that breaks down fast.1 This page maps the three layers, shows where each form sits, and points you to the right drill-down for each.
Overview
Japanese marks aspect and mood far more richly than it marks tense. The verb has only two tense slots. It also has a broad aspect layer built on the te-form and the verb stem, plus an outer band of modal suffixes that report the speaker's stance.
The confusion that sends learners here comes from English glosses. English forces a tense contrast on every verb ("built" / "will build" / "was building"). A learner may then read the single Japanese suffix slot as "tense" and miss the aspect and mood riding alongside.1
This article treats tense, aspect, and mood as answers to three different questions: when did it happen, what shape does the event have, and what is the speaker's stance toward it.
The forms span roughly N5 through N3. The tense forms and ている sit around N5–N4. The modal, evidential, and volitional suffixes range N4–N3. The official JLPT publishes no grammar-point syllabus, so these placements follow standard graded references such as Makino & Tsutsui's basic and intermediate grammar dictionaries rather than a government list.2
Three Systems, One Verb Ending
What Tense, Aspect, and Mood Each Mean
Three cross-linguistic categories sit behind the Japanese verb ending. Defining them in language-neutral terms first keeps them from blurring once the forms appear.
Tense relates the time of a situation to some other time, usually the moment of speaking. A tense morpheme does not change what happened; it only locates the event on the time line.13
Aspect views the internal temporal shape of a situation: complete, ongoing, or leaving a result behind. An aspect morpheme looks inside the event and changes the proposition's content. For example, the English progressive in "John was building a house" does not imply a finished house.1
Mood (or modality) marks the speaker's stance toward the situation: how sure they are, how they learned it, whether they want it to happen. Modality lets a speaker talk about circumstances that need not be part of the actual course of events at all.4
The same event can be dressed in each layer in turn. The three blocks below change one layer at a time.
太郎は家を建てた。1
"Tarō built a house." (tense layer: locates the event in time)
太郎は家を建てている。1
"Tarō is building a house." (aspect layer: the same event mid-course; no finished house)
太郎は家を建てるだろう。4
"Tarō will probably build a house." (mood layer: the speaker's conjecture)
Why the Three Get Conflated in Japanese
One Japanese suffix can carry more than one of these jobs at once, and the English translation often collapses them into a single tense form. English marks tense obligatorily and overtly. Japanese can pack aspectual and modal information into the same slot a learner reads as "tense."1
た is the clearest case. Linguists disagree over whether た is a tense morpheme or an aspect morpheme. One influential position treats it as ambiguous between a past (過去, kako) reading and a perfective reading.1
The blur lives in the translation, not in the grammar. Japanese keeps tense, aspect, and mood in distinct slots. The English gloss must choose a tense, which fuses them and creates the learner's confusion.1
Tense: Only Two Slots (Non-Past vs. Past)
The Non-Past Form Covers Present and Future
Japanese has no overt future-tense marker on the verb. Future-oriented meanings are carried by the simple non-past form, with context or an adverb fixing the reading.1
The same 起きる covers a present habit and, with the right adverb, a future event.
毎朝七時に起きる。1
"I get up at seven every morning." (present/habitual reading of the non-past)
明日東京へ行く。1
"I will go to Tokyo tomorrow." (same non-past form, future forced by 明日)
When a speaker needs to point explicitly at a future event, Japanese reaches for future-oriented nouns such as つもり (tsumori, "intention") and 予定 (yotei, "plan") rather than a tense form.1 English has the auxiliary will, often treated as a future-tense morpheme. Japanese has no counterpart.1
新政権に働きかけるつもりだ。4
"I intend to make approaches to the new government." (future intention carried by the noun つもり, not by tense)
Relative vs. Absolute Tense
Absolute tense is oriented to the moment of speaking. Relative tense is oriented to some other reference time.1 This distinction is the hinge of the whole article, because Japanese た behaves as a relative-tense morpheme. It locates an event in relation to a tense that locally governs it, not necessarily in relation to the utterance time.1
The cleanest proof is an embedded clause, a clause inside another clause. In the sentence below, the watching is past relative to the studying, yet the whole sentence is about the future.
太郎はテレビを見た後で勉強する。1
"Tarō will study after watching TV." (the た-clause is future of speech time; た marks "before the studying")
The same relativity shows up under a verb of saying. The embedded た is anchored to Tarō's saying, not to now.
太郎は花子が本を読んだと言った。1
"Tarō said that Hanako had read a book." (embedded た = before Tarō's saying, not before now)
Because the point that fixes its interpretation need not be the utterance time, た in this respect resembles the English perfect.1 After surveying the alternatives, Ogihara settles on the relative-tense analysis.1
One residual puzzle remains for the relative-tense theory: in a relative clause, a clause that modifies a noun, た can mark a current state rather than a past event.
太郎は底の抜けた桶を持っている。1
"Tarō has a pail that has no bottom." (relative-clause た marks a present state, which Ogihara flags as a residual problem)
Aspect: The Heart of the System
た as Perfective, Not Past
A major view in the literature treats た not as tense but as an aspect morpheme, specifically a perfect.1 た can carry a result-state meaning similar to the English perfect. It also co-occurs with もう ("already"), another perfect-like behavior.1
バスはもう出た。1
"The bus has already left." (perfect-like reading: もう with た marks a completed event whose result holds now)
Beyond past reference, the た-form marks the discovery or recall of a present situation. Standard reference grammars list this "have just realized / found it" use alongside the past and present-perfect uses of た.2
あ、あった！2
"Oh, there it is! Found it!" (discovery of a present situation; the thing exists now, yet た is used)
The same non-past behavior drives a brusque spoken command, where no past time is involved at all.
どいた、どいた！2
"Move it, move it! Out of the way!" (a command in the た-form with no past reference)
ている: Progressive and Resultant State
ている is the single most important aspect form. The modern study of it starts with Kindaichi's 1950 work.15 Its meaning splits in two, and the split is decided by the verb, not by the suffix.
With a continuative (durative) verb, ている gives an in-progress reading: part of the action is done, part remains.15
太郎は今木を倒している。1
"Tarō is now felling a tree." (continuative verb + ている = action in progress; 今 forces the progressive)
With an instantaneous verb, the action has already ended and ている reports that its result persists.15
太郎は今あそこで倒れている。1
"Having fallen down, Tarō is now lying over there." (instantaneous verb + ている = resultant state, not an ongoing fall)
A third reading exists. With a past-time adverbial, ている can report that the subject has the experience of having done something.1
太郎は1970年に結婚している。1
"Tarō got married in 1970 (and that is a fact about him)." (experiential reading; the 1970 adverbial blocks both progressive and plain result-state)
Verb Class Decides the Reading
Kindaichi (1950) divides Japanese verbs into four classes, using ている itself as the diagnostic that produces the split.15 A verb's class decides whether ている reads as progressive or as resultant state.
| Class | Kindaichi term | Behavior with ている | Examples |
|---|---|---|---|
| 1 Stative | 状態動詞 | Never takes ている | ある (exist), できる (be able) |
| 2 Continuative | 継続動詞 | In-progress reading | 読む (read), 書く (write), 降る (fall) |
| 3 Instantaneous | 瞬間動詞 | Resultant-state reading | 死ぬ (die), 結婚する (marry), 知る (come to know) |
| 4 Fourth class | 第四種の動詞 | Occurs only in ている | そびえる (tower above) |
Stative verbs denote states and never take ている; the existence verbs ある and いる are the canonical members.
太郎は東京にいる。1
"Tarō is in Tokyo." (stative; いる takes no ている)
The instantaneous class explains Kindaichi's opening puzzle. 知る names the instant of coming to know, so the resultant-state 知っている, not the bare 知る, is how Japanese says "I know."5
父はそのことを知っている。5
"My father knows that." (instantaneous 知る + ている = resultant state "knows")
The fourth class is tiny and needs only a brief look: verbs like そびえる appear only in the ている form to express a state, never in a bare past form for an action.15
山が高くそびえている。15
"A mountain stands tall." (fourth class; そびえる occurs only in the ている form)
Where English treats have, live, and know as statives, their Japanese counterparts 持つ (have), 住む (live), and 知る (know) fall into the instantaneous class instead.5 That is why each needs ている to express the ongoing state a learner expects: 持っている, 住んでいる, 知っている.
The Wider Aspect Layer
ている is only one of many aspect morphemes in Japanese; the aspect layer is broad.1 Two attachment sites host the rest of it: the te-form and the verb stem (連用形).
The diagram below sorts the wider aspect layer by where each construction attaches.
The te-form combines with auxiliary verbs. てある marks the resultant state of a deliberate action, ておく marks something done in advance, and てくる / ていく marks movement and temporal change toward or away from a reference point.26
窓が開けてある。2
"The window has been left open (on purpose)." (てある: resultant state of a purposeful action)
The verb stem hosts a set of phasal and manner-of-completion suffixes. 始める / 終わる / 続ける mark begin, finish, and continue. 出す / 切る / 込む / 直す mark onset, completion to the end, inward or thorough action, and redoing.6
雨が降り始めた。6
"It started to rain." (phasal 始める on the stem 降り: onset of the process)
The ところ family and ばかり mark aspect-of-timing. They show where the speaker places the moment in an event's run-up or aftermath.2
今出かけるところだ。2
"I'm just about to go out." (ところ family: on the verge of the event)
Mood and Modality: The Speaker's Stance Layer
Evidential Suffixes: How Do You Know?
Evidential markers indicate the source of the information. Epistemic markers, by contrast, qualify how well a proposition fits the speaker's beliefs.4 The three core evidentials らしい, そうだ, and ようだ all signal that the speaker is not in complete possession of the facts, and that a report or conjecture is being made.78
そうだ is the trap, because it is really two forms. Reportive そうだ (hearsay) attaches to the plain, sentence-final form. Conjectural そうだ ("looks like") attaches to the verb or adjective stem.78 The minimal pair turns on that single difference of attachment.
雨が降るそうだ。7
"I hear it's going to rain." (reportive そうだ on the plain form 降る)
雨が降りそうだ。7
"It looks like it's going to rain." (conjectural そうだ on the stem 降り)
The other two core evidentials differ by where the evidence comes from. ようだ expresses a conjecture from the speaker's own firsthand sensory evidence. らしい draws on non-firsthand information such as hearsay and tends to mark less commitment to the conjecture.789 みたい is the casual counterpart of ようだ. ようだ is the formal, written register for the same resemblance-and-inference function.2
雨が降っているらしい。7
"It seems, from what I gather, to be raining." (らしい: conjecture from non-firsthand information)
The だろう-class and the ようだ-class also differ in how they reason: だろう-type elements deduce forward to results, while evidential ようだ-type elements reason back to causes.4
Epistemic Suffixes: How Sure Are You?
Where evidentials mark the source of information, epistemic suffixes mark the degree of belief. だろう, はずだ, にちがいない, and かもしれない all belong to the domain of knowledge and belief.4 They line up, roughly, on a scale of certainty.
だろう (polite でしょう) expresses the outcome of an inferential process, that is, a conjecture the speaker has reasoned to.4
ケンは試験に受かるだろう。4
"Ken will probably pass the exam." (だろう: the speaker's inferred conjecture)
はず, originally a noun, follows a tense-inflected copula. It marks the prejacent, the proposition it attaches to, as what ought to be the case given the speaker's premises.4
ビールは今ごろ冷えているはずだ。4
"The beer ought to be cold by now." (はず: expectation following from premises)
At the ends of the scale sit にちがいない and かもしれない. にちがいない marks its prejacent as entailed ("there is no mistake in"). かもしれない marks it as merely compatible with what is known ("can't know whether").4
明日は雨が降るかもしれない。4
"It may rain tomorrow." (かもしれない: low-certainty possibility)
The scale is a guide, not a clean partition. The standard four-criteria split between epistemic and evidential markers does not sort them neatly. にちがいない is normally epistemic, yet it patterns with the evidentials in allowing abductive inference. Embeddability under 思う, one of the tests meant to separate the two classes, also divides them only roughly. (Epistemic だろう embeds under 思う where the evidential ようだ-class resists it.)47
Volitional and Intention
A small cluster carries the will, intention, and proposal corner of modality. The volitional ending ～(よ)う expresses prioritizing modality and groups with the imperative among the markers that set a sentence's clause type.4
寿司を食べよう。4
"Let's eat sushi." (plain volitional / cohortative)
つもり states a plan, and its reading pivots cleanly on the tense of what comes before it. With a non-past prejacent, it reads as a plan or intention.4
新政権に働きかけるつもりだ。4
"I intend to make approaches to the new government." (non-past prejacent → plan/intention)
With a past prejacent, the same つもり shifts to an epistemic reading, a belief about the past rather than a plan.4
精一杯やったつもりです。4
"I believe I did the best I could." (past prejacent → epistemic "I think/believe")
The plain volitional pairs with 思う to state an intention the speaker is forming: ～(よ)うと思う. The volitional supplies the will, 思う reports the speaker entertaining it, and the "I'm thinking of doing X" reading is built from the two parts.4
How to Read a Japanese Verb Ending
The Stacking Order
Japanese is agglutinative: the verb is built by stringing suffixes onto a stem. Voice, aspect, polarity, tense, and modality each surface as a successive suffix, and several can chain in one word.6 The order is fixed by scope, with each layer answering a different question.
The inner layers are about the event. The outer layers are about the speaker. The causative suffix sits inside the passive (causative ～させ then passive ～られ), giving ～させられる. This is a clear case of the layering principle.6
食べさせられた。6
"I was made to eat." (causative ～させ inside passive ～られ inside tense ～た)
Modal and evidential material sits at the outer, rightmost edge. Evidentials and epistemic markers attach only after tense and aspect have been spelled out. This fits the idea that modality scopes over the whole proposition the inner layers build.47
雨が降っているらしい。7
"It seems to be raining." (aspect ている spelled out first; evidential らしい attaches outside the tensed clause)
Evidentials can even stack on each other by scope. Reportive そうだ can follow a conjectural evidential. A form like 「降っているそうらしい」 ("it apparently seems to be raining") is attested, though stacked evidentials are marked and rare in everyday speech.78 The principle matters more than the rare form: a reportive evidential can embed a conjectural one. That is direct evidence that these suffixes layer outward by scope.
The order stem → voice → aspect → polarity → tense → modality is a pedagogical sketch of the agglutinative system, not a rigid slot grammar. Read it as "suffixes stack outward, and each layer answers a different question," with Shibatani's survey as the anchor for the agglutinative-morphology claim.6
A Worked Decode
A fully stacked verb can be peeled apart layer by layer. Each suffix answers one of the three questions: when, what shape, whose stance. The blueprint example exercises voice, aspect, tense, and modality in one word.
食べさせられていたらしい。146
"It seems he had been made to eat." (one word carrying voice, aspect, tense, and modality)
Peeling it from the inside out shows the three systems in distinct slots:
| Layer | Piece | Job |
|---|---|---|
| Stem | 食べ | the lexical core, "eat"6 |
| Voice (causative) | させ | "make/let eat"6 |
| Voice (passive) | られ | "be made to eat"6 |
| Aspect | てい | the resultant/progressive state1 |
| Tense | た | anchors the state to a past reference time1 |
| Modality | らしい | the speaker reports it as an inference47 |
The peel makes the headline point clear. Tense (た), aspect (てい), and mood (らしい) each occupy a different slot. Voice (させられ) is tucked inside aspect, and modality surfaces last because it scopes over everything else.146
Good to know
The "Past Tense" Trap
Reading every た as English past tense is the most common way the shorthand misleads. The textbook label "past tense" is correct for the ordinary main-clause case. But it breaks down in embedded clauses and in the discovery, recall, and soft-command uses, none of which are past at all.12
Glossing 「あった！」 as "it existed" treats a present discovery as a past event. The intended meaning is present-time: the thing exists now and the speaker has just located it.
あ、あった！2
"Oh, there it is! Found it!"
The deeper reason is that た is best analyzed as a relative-tense or perfective morpheme rather than an absolute past marker. Textbooks teach "past tense" because it fits the most frequent case, not because it describes the form.1
ている Is Not the English Present Continuous
Mapping ている onto English "-ing" for every verb produces the classic resultant-state error. Verb class, not the suffix, fixes the reading. Continuative verbs (読む, 書く) give a progressive, but instantaneous verbs (知る, 結婚する, 死ぬ) give a resultant state.15
Using bare 知る as a stative present to mean "I know him" is wrong, because 知る names the instant of coming to know. The resultant-state form carries the meaning a learner wants.
彼を知っている。5
"I know him."
By the same logic, 「結婚している」 means "is married," the persisting result, not "is in the act of marrying."5
Mood Lives at the End
A reliable mnemonic: the further right a suffix sits, the more it is about the speaker rather than the event. Voice and aspect cling to the stem and describe the event. Tense sits in the middle, and evidential or epistemic modality lands at the outer edge.46
Modality scopes over the proposition the inner layers build, so it surfaces last. When a verb ending feels far from the root, expect it to be reporting how the speaker knows or how sure they are.
A Note on Terminology
The Japanese terms are 時制 (jisei, "tense"), 相 or アスペクト (sō / asupekuto, "aspect"), and モダリティ or 法 (modariti / hō, "mood/modality"). "Mood" and "modality" are often used loosely, and their boundaries are genuinely contested.4
In the native Japanese-linguistics tradition, モダリティ is defined as the expression of the speaker's current attitude to a proposition. That scope can fold in politeness, negation, and even tense, and it differs from the formal-semantic definition.4
The disagreement reaches the individual forms. Linguists defend た as a tense morpheme, as a perfect (aspect) morpheme, and as a relative-tense morpheme. Even the epistemic-versus-evidential line is contested, since the standard diagnostics do not cleanly sort the markers, with にちがいない straddling the two classes.14
See also
- Inferential Suffixes in Japanese: ～そう, ～よう, ～らしい, ～みたい Compared
- The Ta-Form in Japanese: Construction Rules
- The Te-Form in Japanese: Uses (Linking, Cause, Light Imperative, Continuation)
- Passive Voice (受身形): Direct and Indirect Passives
- The Polite Volitional ～ましょう: How to Say "Let's" in Japanese
- The Japanese Imperative Form (命令形): Plain Commands and Prohibitions