Verbs in the Korean language come in last place in a clause. Verbs are the most complex part of speech, and a properly conjugated verb may stand on its own as a complete sentence. This article uses the Yale romanization in bold to show morphology.
Classification Korean verbs are typically classified into four categories: action, state (or description), existential, and the copulas.
Action or processive verbs involve some action or internal movement. For a list of Korean action verbs, see wikt:Category:Korean verbs. Stative or descriptive verbs are sometimes called adjectives. For a list of Korean stative verbs, see wikt:Category:Korean adjectives. Existential verbs convey the existence of something, or its presence in a particular location or a particular being's possession. This category was created for the verb itda "to exist" and its opposite, eopda "not to exist." Copulative verbs allow a non-verb to take verbal endings. In Korean this category was created for the affirmative and negative copula. The affirmative copula is ida "to be," and the negative copula anida "not to be." However, there are many other verbs in Korean that also serve to attach verb endings to nouns, most notably hada "to do."
The distinction between action verbs and descriptive verbs is visible in verb conjugation in a few places. The copulas conjugate like stative verbs, but the existential verbs conjugate like action verbs. Some verbs can be either stative or active, depending on meaning.
Forms Korean verbs are conjugated. Every verb form in Korean has two parts: a verb stem, simple or expanded, plus a sequence of inflectional suffixes. Verbs can be quite long because of all the suffixes that mark grammatical contrasts.
A Korean verb root is bound, meaning that it never occurs without at least one suffix. These suffixes are numerous but regular and ordered. There are over 40 basic endings, but over 400 when the combinations of these endings are counted. Grammatical categories of verb suffixes include voice (passive or causative), tense (past, present, or future), aspect (of an action – complete, experienced, repeated, or continuing), honorification (appropriate choice of suffix following language protocol), and clause-final conjunctives or sentence enders chosen from various speech styles and types of sentences such as interrogative, declarative, imperative, and suggestive.
Sound changes A great many verbs change the pronunciation of the final consonant of the root after the addition of a suffix. Some of these changes are the result of regular consonant assimilation or cluster simplification, but some of them are irregular. The irregular verbs contain root-final consonants that were historically lenited and which, as a result disappeared or mutated between vowels but remained next to a consonant.
Citation form The lemma or citation form of a Korean verb is the form that ends in ta da without a tense-aspect marker. For verbs, this form was used as a past perfect declarative form in Middle Korean, but is no longer used in Modern Korean. For adjectives, this form is the non-past declarative form.
Infinitive form Besides a verbal root itself that precedes ta in the citation form, there is also a long stem with an additional harmonic vowel, called by linguist Samuel E. Martin the "infinitive" form. This tense-neutral form also does not express any honorifics and speech levels. Thus they are often used for literary titles, subtitles and chapter titles, since they are not specifically directed toward an individual or a group.
This so-called infinitive, however, must not be confused with the citation form mentioned above. It is formed by attaching eo/a to the root, according to vowel harmony. If the verbal root ends in a vowel, the two vowels may merge or contract.
Without vowel contraction al al "know" + → al.a ara mek meok "eat" + → mek.e meogeo
With vowel contraction ka ga "go" + → ka ga o o "come" + → wa wa se seo "stand" + → se seo i i (copula) + → ye yeo ssu sseu "use" + → sse sseo
toy ta 되다 doeda "to become" may or may not undergo contraction. ha ta 하다 hada "to do" is irregular.
This infinitive form is not used as a noun, but it can be used in compound verbs, serial verb constructions, and before certain (not all) verb endings. It may be compared to the conjunctive in Japanese.
Finite verb endings Verbs are the most complex part of speech in Korean. Their structure when used as the predicate of a clause is prefix + root + up to seven suffixes, and can be illustrated with a template:
+Finite verb template Derivational suffixes Sentence-final endingsPrefix0IIIIIIIVVVIVIInegative*ROOTvalencyhonorifictense-aspectformalitysyntactic moodspragmatic moodspoliteness suffix
*The negative prefix is an "not"; the word mos mot "cannot" also occurs in this position.
I Valency may be passive or causative. These often involve a stem change, followed by the suffix i (the spelling of this suffix may change, depending on the stem change of the verb).
II The honorific suffix is -usi -eusi- after a consonant, -si after a vowel. The i is reduced to a glide before another vowel. For example, with a following past tense, sie-ss -si-eoss- reduces to sye-ss -syeoss-.
This shows deference towards the topic of the conversation, for example when speaking of one's elders.
III If there is no suffix in this slot, the verb is in present or gnomic tense. Future tense & prospective aspect is key-ss -get-, past perfective is --ss -eot-/-at but with vowel harmony. If there is no intervening consonant, this reduces, both in pronunciation and in writing: a-ss to at-, and wa-ss to wat-. The verb o "to come" is therefore wa-ss wat- in the perfective. The verb ha ha "to do" is an irregular hay haet- in the perfective.
There are also compound tenses: remote past --ss-e-ss -eosseot-/-asseot-, past-future --ss-key-ss -eotkket-/-atkket-, remote past-future (An action that should have been completed in the past but has not actually been done) --sse-ss-key-ss -eosseotkket-/-asseotkket-
IV The formal suffix is -p after a vowel (it is normally written in the same block as that vowel), -sup -seup after a consonant in a declarative or interrogative verb, -up -eup after a consonant in a proposition. (After a consonant s or ss the letter in the suffix drops.)
This shows deference towards the audience of the conversation, for example when speaking to one's elders. If speaking both to and of one's elders, one would use both the formal and the honorific suffixes.
V The syntactic moods, for lack of a better term, are the indicative -nun -neun, -ni , or n ; the retrospective (imperfective) -ten -deon, ti -di, or t -d-; and the subjunctive si -si or s . None of these are used in the casual or intimate styles, and the formal plain indicative declarative can only occur in the gnomic tense.
-nun -neun and -ten -deon are used in the formal plain and familiar interrogative styles. After a vowel, -nun -neun reduces to n . Before declarative la ra, -ten -deon reduces to te -deo.
-ni , -ti -di, and -si are used in the formal polite style.
-n , t -d-, and s are used in the familiar declarative and subjunctive styles.
VI The pragmatic moods, for lack of a better term, are the declarative -ta -da (formal polite), -la -ra (formal plain), and ey -e (familiar); interrogative kka , ya (formal) and -ka -ga (familiar); propositive -ta -da (formal polite), -ca -ja (formal plain), and ey -e (familiar); and the imperative o (formal polite), - la -eola/-ala (formal plain), and -key -ge (familiar).
Style: These distinctions are not made in the intimate and casual styles. Instead, this slot is taken by the intimate suffix - -eo (a after an a or o ) or the casual suffix -ci -ji.
VII The polite suffix yo (-i yo after a consonant) appears in the informal styles. It expresses one's relationship to the audience.
Negation A verb is typically negated in Korean by using a suppletive negative form, if it exists, or by putting a negative adverb in front of it.
There are two possible negative adverbs: mos mot, and an . mos is used for when a person or animate being subject tries to accomplish an action, that is, begins and is unable to finish it successfully. an is a more common negator which is used in all other instances. The two prefixes are mutually exclusive.
Derivational suffixes Derivational endings are attached directly to the verb root, and are followed by the tense suffixes. These derivational suffixes end with the high vowels i or wu which is reduced to a glide in the long stem form. For example, with a following past tense, -(u)si -(eu)si reduces to -(u)sye-ss -(eu)syeot.
Valency Valency in Korean is partly lexical and partly derivational. Many forms can change their valency by the addition of the passive or causative derivational suffixes, -i , -hi , -li -ri, -ki , -wu -u, -kwu -gu, or -chwu -chu, sometimes with additional changes to the stem.
Subject honorific The subject honorific suffix -(u)si derives an honorific verb, that is, a verb which is used when the subject of a sentence is higher in social status than the speaker. Such verbs are used, for example, when speaking of one's elders, one's social superiors (parents, teachers, bosses), or strangers.
The full form -usi is only used after a consonant. Otherwise, the initial vowel is absorbed, becoming -si.
While the honorific suffix is necessary, some verbs have honorific alternatives which must be used in addition to -(u)si. For instance, iss ta itda becomes kyey'si ta gyesida.
Tense and aspect Following the derivational endings, Korean verbs can contain up to three suffixes in a row which represent a combination of tense, aspect, and mood.
Past This suffix is an enclitic consonant 'ss after the infinitive form of the verb (ending in ), forming 'ss (the final consonant is pronounced before a vowel and before a consonant). This suffix, which is conventionally called "past" or "perfective" by various linguists, has many different meanings, depending on the semantics of the verb that it is attached to and the context; it may be a simple past or a present perfect.
Etymologically, 'ss is a contraction of the existential verb iss via vowel absorption. The contracted form - iss, was originally a present perfect.
Pluperfect A verb can superficially have two copies of the above-mentioned suffix, the second of which, however, is always -ess -eot and represents a true past tense. This results in the combination 'ss.ess -eosseot/-asseot. This combination communicates a past perfect or a more remote past.
Future The future suffix is -keyss -get, often used to describe future events. It is used, when the speaker has valid reasons to believe something will be certain to happen. For instance, the suffix is used in broadcasting contexts in Korean such as weather forecasts.
But it may be used together with the perfective and pluperfect suffixes, or in a present tense context. If used with the perfective suffix, this makes an inferential or conditional past -'ss-keyss -eotget/-atget "should have, would have, must have." If used with the remote past suffix it makes an inferential or conditional remote past -'ss-ess-keyss -eosseotget/-asseotget, though this is rare. Because this infix is occasionally used for a conditional, or inferential tense, depending on context it is sometimes called irrealis.
Etymologically, the future suffix is the result of the merger of a resultative verb ending -key and the existential root iss , via vowel absorption, as mentioned above. This contraction and change in meaning has its parallel in the future tense of Vulgar Latin.
Sentence-final endings +Sentence-final endingsIVVVIVIIformalitysyntactic moodspragmatic moodspoliteness suffix