Japanese Verb Conjugation

Japanese calls verbs 動詞 (doushi), or movement words. Most verbs in Japanese fall into one of two types: 一段 (ichidan), often called -ru verbs, and 五段 (godan), often called -u verbs. The category affects how the verb conjugates, as explained below. Each verb form gets its own description, conjugation tables where appropriate, generic conjugation instructions, and examples for each verb type.

To be more accurate, there are two types of ichidan verbs: 上一段 (kami ichidan) and 下一段 (shimo ichidan). However, these two subgroups conjugate the same way in modern Japanese, so it makes sense to group them together. The difference is that the inflected ends of 上一段 verbs carry the イ sound (the row above the central ウ row in the traditional kana chart, hence the name, "upper one-row"), like 生きる (ikiru), while those of 下一段 ("lower one-row") verbs carry the エ sound (the row below the middle), like 食べる (taberu). For whatever reason, the kana before the る (き and べ for these examples) is regarded as part of the ending rather than the base, even though it's always there and never changes regardless of the verb form being used. Maybe for historical reasons.

And on that note, 五段 ("five-row") verbs are called that since they use all five of the vowels in their various inflected endings.

There are also what I call "old-style -aru verbs," a subtype of godan verbs that are holdovers from an earlier form of Japanese and conjugate slightly differently. As far as I know, only five remain in use: いらっしゃる (irassharu), to be, come, or go (honorific); ござる (gozaru), to be (polite); 為さる (nasaru), to do (honorific); おっしゃる (ossharu), to say (honorific); and 下さる (kudasaru), to give (honorific).

Another way verbs can be classified is by transitivity. Transitive verbs (those that normally have direct objects), involve a subject that takes an action, and are called 他動詞 (tadoushi), or move-other-words. Intransitive verbs (those that cannot have a direct object) basically indicate an action that changes the state of the subject, and are called 自動詞 (jidoushi), or move-self-words. Which of these categories a verb falls under affects the meaning of some conjugations.

Additionally, although I sometimes refer to forms as "casual" and "formal", the "casual" form is grammatically required in certain situations regardless of formality. These situations are noted where relevant. Formal form is also called long form, polite form, and distal style. Casual form is also called short form, plain form, and direct style.

The Copula

So what's a "copula" anyway? It's just a word that links (couples) the subject and predicate of the sentence in a way that indicates equivalence. In short, it's a linguistic equal sign. The copula in English is "be", though it also has other uses including acting as the existence verb, as in "there is a(n) X", and functioning as an auxiliary verb in several tenses (such as the progressive tense—"is doing"—and the passive tense—"was done"). The Japanese copula doesn't have that ambiguity.

The Japanese copula isn't even technically a verb (being formed from a particle and an existence verb), but it does conjugate much like one, so here are its various forms. I'll explain what the different forms are used for below.

Do not assume that a form is at all common just because it exists. でして, じゃありませんで, でしたら, and でしたり in particular are uncommon and too formal for normal usage.

Conjugation table:

FormLongShortLiterary
Stem N/A, but で (de) is comparable であり (de ari)
Nonpast affirmative です (desu) だ (da) である (de aru)
Nonpast negative じゃありません (ja arimasen) じゃない (ja nai) ではない (de wa nai)
Past affirmative でした (deshita) だった (datta) であった (de atta)
Past negative じゃありませんでした
(ja arimasen deshita)		 じゃなかった
(ja nakatta)		 ではなかった
(de wa nakatta)
Tentative * でしょう (deshou) だろう (darou) であろう (de arou)
Volitional / Hortative N/A, but でいよう (de iyou) is close
Positive -te form でして (deshite) で (de) であって (de atte)
Negative -te form じゃありませんで (ja arimasen de) じゃなくて (ja nakute) ではなくて (de wa nakute)
Potential N/A, but でいられる (de irareru) is close
Passive N/A, but にされている (ni sarete iru) is close
Causative N/A, but にさせている (ni sasete iru) is close
Causative-Passive N/A, but にさせられている (ni saserarete iru) is close
Command N/A, but になれ (ni nare) is close
Conditional でしたら (deshitara) だったら (dattara) であったら (de attara)
Provisional N/A † であれば (de areba)
Alternative でしたり (deshitari) だったり (dattari) であったり (de attari)

* This form is the equivalent to the volitional/hortative form, but is used differently. Refer to the tentative section.

† Though there are no non-literary provisional forms of the copula, ならば (naraba) and なら (nara) mean nearly the same thing.

Contractions, etc.:

じゃ (ja) is a contraction of では (de wa). The uncontracted form is used more often in writing, and hardly ever when spoken, except when being very proper or for effect.

です is a contraction of であります (or でございます, depending on your source), and similarly with だ from である, でした from でありました, and so on. The longer versions are not often used, but this at least explains why the various forms of ある keep showing up.

Variations:

In somewhat casual speech, です (desu) may turn into っす (ssu), as in 面白いっす (omoshiroi ssu = it's interesting). This is generally a more laid-back and masculine tendency.

The highly respectful version is でいらっしゃる (de irassharu), which conjugates as いらっしゃる (irassharu) does. The humble one, でござる (de gozaru), similarly conjugates like ござる (gozaru). Both いらっしゃる and ござる are old-style -aru verbs, and conjugate somewhat differently than most verbs.

Note that all negative forms of the copula generally add は (wa) after で (de). This can be omitted, leaving forms like でない (de nai), but only rarely is. Refer to the negative sentences subheading under the topic particle は (wa) for more related information.

Some dialects use じゃ (ja) or や (ya) instead of だ (da), and there may be other variants as well.

Positive negative:

じゃない isn't always really a negative. Occasionally, it's added to the end of a fact when the fact is unexpected. It's sort of like "Isn't this..." or "It can't be...", but doesn't really translate. I'm not sure whether じゃありません is used similarly in formal speech, or if this is something that only happens in casual. As an example, when Crono finds the missing queen's hairpin in Chrono Trigger, Lucca, after looking at it closely, says:

Although the sentence literally says, "This isn't the Guardia royal family's coat of arms!" it's obvious in context that she means that it is the royal family's coat of arms. The じゃない just means she never expected to find something like that there. The translation can safely be rephrased as: "Isn't this the Guardia royal family's coat of arms!?"

The Verb Stem

The verb stem usually has something attached to it, rather than being used alone, but is useful enough to deserve its own heading. For those keeping track of such things, this is properly called the 連用形 (ren'you-kei), or conjunctive form, and is one of six basic conjugated forms.

Conjugation table:

DictionaryStem
する (suru)し (shi, not su)
来る (kuru)来 (ki, not ku)
Regular -u verbschange the final -u to -i
Old-style -aru verbschange the final -aru to -ai
-ru verbsremove ～る (-ru) from the end
Examples:
飲む (nomu)飲み (nomi)
下さる (kudasaru)下さい (kudasai)
食べる (taberu)食べ (tabe)

Sample uses:

Pronunciation shifts:

I think this is clear enough, but just in case, changing the -u in -u verbs to -i means changing the final kana in the dictionary form (which always ends in the 'u' vowel) to the corresponding kana with an 'i' vowel. This causes a change in the romanization of the consonant, and arguably some shift in pronunciation, for verbs that end in つ (tsu) or す (su), which become ち (chi) and し (shi), respectively.

Nonpast Affirmative (-masu and dictionary forms)

The most basic form. It's a lot like a normal present tense verb in English, for example, "does", except that it can refer to the future ("will do") as well as the present. There is no special future form in Japanese. When talking specifically about the future, time adverbs are often used, or other forms that refer to intentions, decisions, or something similar.

For intransitive verbs, this form is also essentially the same as the English present progressive ('-ing') form, since those verbs are more concerned with the result of the action than the action itself. Transitive verbs use a different syntax involving the -te form to indicate actions in progress. When there's a reason to indicate an action in progress with an intransitive verb, most often a movement verb such as 行く (iku, "go"), add 途中 (tochuu, "in the middle of doing") after the short nonpast affirmative form.

Conjugation table:

DictionaryLongShort
する (suru) します (shimasu) する (suru)
来る (kuru) 来ます (kimasu) 来る (kuru)
-u verbs add ます (masu) to the stem no change—dictionary form
is the same as short form
-ru verbs
Examples:
話す (hanasu) 話します (hanashimasu) 話す (hanasu)
いらっしゃる (irassharu) いらっしゃいます (irasshaimasu) いらっしゃる (irassharu)
食べる (taberu) 食べます (tabemasu) 食べる (taberu)

The ～ます is actually an auxiliary verb that attaches to the conjunctive form and indicates respect for the listener. For many long forms, ～ます conjugates rather than the base verb, though for others (such as the passive form), the base verb conjugates and ～ます is then added to that if needed.

Dictionary form:

Since the short nonpast affirmative is the same as the dictionary form, verbs listed in a dictionary can be used as-is, at least when the casual nonpast affirmative is appropriate. Most dictionaries, anyway. There have been rare cases of dictionaries that list a different form.

Nonpast Negative (-masen and -nai forms)

This is the same as the nonpast affirmative except that it means the action does not occur. It's like "does not do" or the future form "will not do" in English. As with the positive, time adverbs or indications of things like intentions can make the future explicit.

Conjugation table:

DictionaryLongShort
する (suru) しません (shimasen) しない (shinai)
来る (kuru) 来ません (kimasen) 来ない (konai)
-u verbs add ～ません
(-masen) to stem		 change -u to -a and add ～ない (-nai)
Exception: ある (aru) ない (nai), also written 無い
-ru verbs drop ～る (-ru), add ～ない (-nai)
Examples:
会う (au) 会いません (aimasen) 会わない (awanai)
ござる (gozaru) ございません (gozaimasen) ござらない (gozaranai)
見る (miru) 見ません (mimasen) 見ない (minai)

More accurately, the short form nonpast negative is in all cases just adding ～ない to the 未然形 (mizenkei), or irrealis form, one of the six basic conjugated forms.

Pronunciation shifts:

There are, of course, some romanization changes in some casual forms. つ (tsu) becomes た (ta), but more significantly and confusingly, う (u) becomes わ (wa). Apparently, such verbs once ended in ふ (hu) and conjugated to は (ha), but pronounced these as 'u' and 'wa' even then.

Interation with "negative" words:

A note on negative verbs and their interaction with other words: Some words are typically or always used only with negative forms, or are generally listed as having opposite meanings when used with a negative. These aren't double negatives. For example, いつも (itsumo) is often defined as "always, never (with negative)". What this actually means is that 「いつも待つ」 (itsumo matsu) means "always waits" as expected, but 「いつも待たない」 (itsumo matanai) is not "never doesn't wait". It's "always doesn't wait", more naturally written in English as "never waits". This is just moving the negative from the verb, where it goes in Japanese, into an adverb, where it tends to go in English. Remember that the negative in Japanese is part of the verb conjugation and not a separate word.

Variations:

In spoken informal language, the ～ない (-nai) in most short nonpast negatives may be shortened to ～ん (-n). From what I can tell, there must be at least two kana before ～ない for this to apply; otherwise the result is too short. As a result, あわない ⇒ あわん is okay, but I've never seen or heard みん (from みない). This shortened form tends to be a more masculine usage, and is apparently also more common in western dialects.

Be careful with this, as it's identical to an archaic version of the positive volitional form. On the other hand, the archaic volitional is far less common in the modern language. The archaic volitional version is shortened from the archaic ～む (-mu) verb ending, while this is derived from the archaic ～ぬ (-nu) ending (see the Other negatives section below).

(Thanks to Samir Alam for noting that ～ん is more common in western dialects and pointing out the connection to the archaic ～ぬ ending.)

Another fairly common shortening is replacing ～らない (-ranai) at the end of forms that contain it with ～んない (-nnai). This tends to be a more girlish usage, and perhaps sounds a bit petulant. As with the ～ない to ～ん shift, this only seems to happen when there are at least two kana before ～らない.

The same shortening may also occur in other words ending with ～らない (the -i adjective つまらない ⇒ つまんない comes to mind), but short nonpast negatives are the most common occurance of the sound.

Invitations:

Adding か to a sentence with a nonpast negative verb makes it an invitation to do that activity. In casual usage, か is optional, and the question is indicated with a simple rising intonation (in speech) or question mark (in writing).

Other negatives:

In Kansai dialects, the ～ない and ～ません of nonpast negative forms are typically replaced with ～へん (-hen) and ～まへん (-mahen). ある in short form becomes あらへん (arahen).

Short form nonpast negative verbs may end with ～ぬ (-nu) or ～ず (-zu) instead of ～ない (-nai). For most verbs, the only difference is the changed ending, but する becomes せぬ (senu) and せず (sezu) rather than しぬ (shinu) and しず (shizu), while ある stops being an exception and becomes あらぬ (aranu) and あらず (arazu).

～ぬ is archaic and very uncommon in the modern language, but may be used for poetic or other effect, or appear in idioms and other set phrases. Don't jump to conclusions if you see a ～ぬ, though, since a few uninflected verbs, most often 死ぬ (shinu, to die), end in ぬ to begin with.

～ず is also archaic, and is usually followed with に (ni) in modern usage, though it can be used without it. The ～ずに (-zu ni) ending appears primarily in writing, and is equivalent to the connective uses of the negative -te form. ～ず without に also occurs in some idioms and other set phrases. When modifying a noun, ～ず becomes ～ざる (-zaru). (Or ～ぬ, which also should only be used before a noun).

Past Affirmative (-mashita and -ta forms)

The past equivalent of the nonpast affirmative form. The action already happened, like "did". The short form conjugations for -u verbs vary here depending on the final consonant. It's technically using an alternate conjunctive form, but that's likely to be more confusing than helpful, so don't worry about it.

Conjugation table:

DictionaryLongShort
する (suru) しました (shimashita) した (shita)
来る (kuru) 来ました (kimashita) 来た (kita)
-u verbs ending in う (u), つ (tsu),
or る (ru), including old-style -aru verbs		 add ～ました
(-mashita) to stem		 drop last kana, add ～った (-tta)
-u verbs ending in む (mu),
ぶ (bu), or ぬ (nu)		 drop last kana, add ～んだ (-nda)
-u verbs ending in く (ku) drop ～く, add ～いた (-ita)
Exception: 行く (iku) 行った (itta)
-u verbs ending in ぐ (gu) drop ～ぐ, add ～いだ (-ida)
-u verbs ending in す (su) drop ～す, add ～した (-shita)
-ru verbs add ～た (-ta) to stem
Examples:
やる (yaru) やりました (yarimashita) やった (yatta)
おっしゃる (ossharu) おっしゃいました (osshaimashita) おっしゃった (osshatta)
飲む (nomu) 飲みました (nomimashita) 飲んだ (nonda)
書く (kaku) 書きました (kakimashita) 書いた (kaita)
泳ぐ (oyogu) 泳ぎました (oyogimashita) 泳いだ (oyoida)
話す (hanasu) 話しました (hanashimashita) 話した (hanashita)
寝る (neru) 寝ました (nemashita) 寝た (neta)

In Kansai dialects, the っ found in the past affirmative form of some verbs tends to be replaced with a lengthening of the previous vowel. A typical example is that 言う (iu, often pronouned as yuu) ends up as ゆうた (yuuta) instead of the usual いった (itta).

Past Negative (-masen deshita and -nakatta forms)

It's past and it's negative; not much else to say. It just means that the action didn't happen. Comparable to "did not do".

Conjugation table:

DictionaryLongShort
する (suru) しませんでした (shimasen deshita) しなかった (shinakatta)
来る (kuru) 来ませんでした (kimasen deshita) 来なかった (konakatta)
-u verbs add ～ませんでした
(-masen deshita) to stem		 change ～ない (-nai) in casual nonpast
negative to ～なかった (-nakatta)
-ru verbs
Examples:
分かる (wakaru) 分かりませんでした (wakarimasen deshita) 分からなかった (wakaranakatta)
起きる (okiru) 起きませんでした (okimasen deshita) 起きなかった (okinakatta)

The short form negative is again built on the irrealis form, just as the nonpast equivalent is.

Volitional / Hortative (-mashou and -ou forms)

There's an actual meaning behind the fancy name. "Volitional" because it can describe decision (volition) and "hortative" because it can urge (exhort) an action. In short, the volitional function indicates intention to do something, while the hortative function encourages someone else to do something.

Conjugation table:

DictionaryLongShort
する (suru) しましょう (shimashou) しよう (shiyou)
来る (kuru) 来ましょう (kimashou) 来よう (koyou)
-u verbs add ～ましょう (-mashou) to stem change -u to -o and add う (u)
-ru verbs add ～よう (-you) to stem
Examples:
泳ぐ (oyogu) 泳ぎましょう (oyogimashou) 泳ごう (oyogou)
出かける (dekakeru) 出かけましょう (dekakemashou) 出かけよう (dekakeyou)

For the linguistically inclined out there, the short form is built on an alternate irrealis form.

Refer to the "Nix that" subheading below for information on the uncommon negative volitional ending ～まい (-mai).

Let's...

In the hortative function, this verb ending resembles "Let's..." in English.

In casual speech, the dictionary form is sometimes used similarly, but that's more of a "we'll do this" than a "let's do this."

Shall we...?

As a question, the hortative function resembles "Shall we ... ?" or "Shall I ... for you?"

This also works in the volitional sense, particularly when people are talking to themselves.

Note that か (ka), which is often dropped in casual questions, is more likely to remain when using this form.

Making an effort (-ou to suru)

Add とする (to suru) to a volitional form verb and conjugate する (suru) as appropriate.

This is trying in the sense of making an effort to do something. For trying in the sense of experimentation to see what happens, see ～てみる (-te miru). There's also a clause-ending ようにする (you ni suru) that has essentially the same meaning as [volitional]とする.

This also works, when used with the -te form and a second action, in the sense of making an effort to accomplish something by doing the second thing.

Decisions:

A sentence in volitional form, followed by the appropriate conjugation of と思っている (to omotte iru), means that the action has been decided on. Using と思う (to omou) instead suggests that the decision is made on the spot.

Casual shortening:

As often happens in words with long vowels, ～ましょう is sometimes pronounced ～ましょ when the situation isn't too formal.

Nix that:

The usual way of making a negative volitional in the modern language is to add ようにしましょう (you ni shimashou), ことにしましょう (koto ni shimashou), ようにしよう (you ni shiyou), or ことにしよう (koto ni shiyou) to the short nonpast negative form.

Another common way of saying "let's not" is to noun the verb and add the object particle を (wo) and 止めよう (yameyou) or 止めましょう (yamemashou), the volitional form of 止める (yameru, to stop or cease).

There is a negative volitional form, though it's more literary and uncommon outside of certain set phrases. Simply add ～まい (-mai) to the short nonpast affirmative form, 取るまい (torumai) for example. Apparently, this can only be used in the volitional sense, not the hortative (it can express intention to avoid doing something, but not trying to dissuade someone else from doing it).

One of the few phrases where this remains comparatively common is [positive volitional]が[negative volitional]が, which is roughly equivalent to "it doesn't matter whether or not [action occurs]."

This ～まい ending may also be used as a negative tentative form, though this is even rarer in the modern language.

Archaic volitional:

Though this is now obsolete, old-style Japanese has an alternate volitional form. Replace the ～ない (-nai) in the short nonpast negative with ～む (-mu), often (usually?) shortened to ～ん (-n). する (suru) and 来る (kuru) become せん (sen) and 来ん (kon), respectively, while ある (aru) becomes あらん (aran). This form occasionally appears for effect, but is mostly significant because it remains used in the modern language in certain set phrases such as 言わんばかりに (iwan bakari ni).

Be careful, though, since this ～ん ending form is identical to an informal variant of the short nonpast negative, where that variant applies, but has a much different meaning.


The -te Form

This form is used for many different things and has no single English equivalent. It usually appears as a connective or part of a larger phrase, but may appear by itself as an informal request or an incomplete thought. Refer to the verb auxiliaries section for more on the many uses of the -te form.

Conjugation table:

DictionaryAffirmative -te FormNegative -te Forms
する (suru) して (shite) しないで (shinaide)
しなくて (shinakute)
来る (kuru) 来て (kite) 来ないで (konaide)
来なくて (konakute)
-u verbs change -a in short
past affirmative to -e		 add ～で (-de) to short nonpast negative
change ～ない (-nai) in short nonpast negative to ～なくて (-nakute)
Exception: ある (aru) becomes なくて (nakute), never ないで (naide) *
-ru verbs
Examples:
乗る (noru) 乗って (notte) 乗らないで (noranaide)
乗らなくて (noranakute)
読む (yomu) 読んで (yonde) 読まないで (yomanaide)
読まなくて (yomanakute)
聞く (kiku) 聞いて (kiite) 聞かないで (kikanaide)
聞かなくて (kikanakute)
行く (iku) 行って (itte) 行かないで (ikanaide)
行かなくて (ikanakute)
泳ぐ (oyogu) 泳いで (oyoide) 泳がないで (oyoganaide)
泳がなくて (oyoganakute)
話す (hanasu) 話して (hanashite) 話さないで (hanasanaide)
話さなくて (hanasanakute)
出かける (dekakeru) 出かけて (dekakete) 出かけないで (dekakenaide)
出かけなくて (dekakenakute)

The affirmative is built on an alternate conjuctive form, and the negatives are built on the irrealis form.

* I'm fairly certain of this, though I haven't seen it officially stated anywhere. Apparently it's because the ない (nai) is technically an adjective rather than a verb form. In any case, since ある (aru) is also the basis for the default copula, this means that the negative -te form of です (desu) is similarly always じゃなくて (ja nakute) rather than じゃないで (ja naide).

In Kansai dialects, the っ in the -te form of some verbs tends to be replaced with a lengthening of the previous vowel. A typical example is that 言う (iu, often pronouned as yuu) ends up as ゆうて (yuute) instead of the usual いって (itte).

Superformal:

It's possible to create an extremely formal -te form by changing the nonpast affirmative ～ます (-masu) ending to ～まして (-mashite) and the nonpast negative ～ません (-masen) ending to ～ませんで (-masende). This is overdoing it in most situations, but there are times when it's useful.

Chaining actions together:

When X is a clause ending in a -te form verb and Y is another clause, X + Y means "X and (then) Y". The tense (and formality) is determined by the verb in Y, which conjugates as normal. Either the ～ないで or the ～なくて negative can be used to indicate that the action is not taken, though their connotations are slightly different—more on that below.

There may be a relatively weak sense of "X in order to Y" ("I went home so I could sleep" in the first example) or "Y because of X" ("I cry since I can't do anything" in the last example), depending on context.

Adding ために (tame ni) to X in the dictionary form can more explicitly express the first meaning:

Using から (kara) or ので (no de) can express the second meaning more directly:

Actually, ため can also express cause-effect, but that tends to be a more literary usage, and usually omits the に anyway.

Differences between the two negative forms:

Many practical usage differences are noted on the verb auxiliaries page. For instance, it's normally ～なくてもいい (-nakute mo ii) and ～ないでください (-naide kudasai), rather than ～ないでもいい (-naide mo ii) or ～なくてください (-nakute kudasai). While that's always nice to know, what about in more general usage?

It seems that while either one is techincally usable in essentially any case, there are certain nuances that, often strongly, prefer one above the other. These include:

Circumstances

When the negated first phrase gives background information or otherwise describes the situation under which the second phrase occurs, use ～ないで (-naide). This often translates fairly cleanly to "without", as the first action remains undone when the second part happens.

In this example, not eating breakfast describes the condition that I went to school in, and it would be awkward to use ～なくて (-nakute) instead.

The exceptions are that the existence verb ある (aru) and the copula (basically the equivalence verb) only have ～なくて forms, so will necessarily use ～なくて. However, I can't think of any sentence using either in this sense that wouldn't be phrased in a way that avoids them. Possible alternatives include using まま (mama), which indicates an unchanged state, or choosing a different verb entirely:

Contrast

When the first phrase is avoided or neglected in favor of the second, use ～ないで (-naide). This typically translates to "instead of".

It's not uncommon to see this contrasting sense used in requests or commands.

When the two phrases are contrasted against each other like this, it's awkward to use ～なくて (-nakute).

The exceptions are that the existence verb ある (aru) and the copula (basically the equivalence verb) only have ～なくて forms, so will necessarily use these. In fact, it's so common to use the the negative -te form of the copula for contrast that I gave じゃなくて (ja nakute) its own entry.

Explanation

When the second phrase comes as a result of the first, ～なくて (-nakute) is typically used.

Followed by a particle

Though I haven't confirmed this, experience suggests that when a negative -te form is followed by a particle, typically は (wa) or も (mo), the ～なくて (-nakute) form is always used rather than the ～ないで (-naide) form. See, for instance, ～なくてはいけない (-nakute wa ikenai) and ～なくてもいい (-nakute mo ii) on the verb auxiliaries page.

Potential (-eru, -rareru, and -reru forms)

Another useful and common form, and not very complicated either. A verb in potential form just means that the action is possible, or can be done.

Conjugation table:

DictionaryPotential
する (suru) 出来る (dekiru)
来る (kuru) 来られる (korareru)
-u verbs change -u to -e and add ～る (-ru)
-ru verbs drop ～る (-ru), add ～られる (-rareru)
Examples:
話す (hanasu) 話せる (hanaseru)
食べる (taberu) 食べられる (taberareru)

Technically, only godan verbs have a potential form, which is created by converting to a shimo ichidan as described above. する is obviously replaced with a different verb entirely. For 来る and ichidan verbs, this is actually just using one sense of the helper verb られる. It attaches to the irrealis form, which is identical to the conjuctive form (our familiar stem) for ichidan verbs, but is こ rather than き for 来る. I'm not sure why the irrealis, since it usually shows up when expressing that which is not (negative forms, most often), but perhaps it's because potential indicates that it could happen rather than that it actually is happening?

Usage:

The resulting potential-form verb conjugates as a standard -ru verb, so you can just as easily say things like "can't do", "want to be able to do", and so on, in both long and short forms. The only complication is that the particle を (wo) is often replaced with が (ga) in a potential sentence, especially when using 出来る (dekiru).

It's also not uncommon to see potentials expressed as [verb] + 事が出来る (koto ga dekiru) or [verb] + のが出来る (no ga dekiru). This doesn't seem to be any more or less formal than the plain potential form as far as I can tell. My best guess, based on context, is that the plain potential form is more personal ("I can do it") and this alternate form is more universal ("It is possible to do it"), but I'm far from certain, and there seems to be little if any difference in practice.

Potential for confusion:

Potential form is identical to passive form for -ru verbs (because they're using different senses of the same auxiliary verb, which can also express respect or spontaneity, though the last two are rare in modern Japanese). Use context to tell which meaning is intended.

Pronunciation shifts:

As usual with changing a vowel, there's a minor romanization change. つ (tsu) becomes て (te), which shouldn't be surprising.

Alternate form:

The ～られる (-rareru) ending of -ru verbs is longish, so the unofficial variant ～れる (-reru) is sometimes used in casual speech.

Passive (-areru)

Like English, Japanese has a passive form that focuses emphasis on the object of an action rather than the actor. Unlike in English, the passive form is comparatively common, and is not mindlessly assumed to be poor style just by being used.

Conjugation table:

DictionaryPassive
する (suru) される (sareru)
来る (kuru) 来られる (korareru)
-u verbs change -u to -a and add ～れる (-reru)
-ru verbs change ～る (-ru) to ～られる (-rareru)
Examples:
呼ぶ (yobu) 呼ばれる (yobareru)
捨てる (suteru) 捨てられる (suterareru)

This is from built the irrealis form. I'm not sure why the irrealis, since it usually shows up when expressing that which is not (negative forms, most often), but perhaps it's because the subject is not doing anything, instead being affected by the action of another.

Usage:

The resulting passive form verb conjugates as a regular -ru verb. Use the topic marker は (wa) or subject marker が (ga) to indicate the object of the action, and the particle に (ni) to mark the one performing the action. By emphasizing the one affected, this form often suggests victimization...

...but not necessarily.

The important point is that the focus falls on the target of the action, or on the action itself, rather than on the performer.

The passive form also works with intransitive verbs (those with no object), with the meaning that the action is an inconvenience or annoyance. English doesn't have an equivalent usage, so this generally takes some creativity to translate effectively.

"They" did something:

One common usage in English is to say that "they" did something, when it's unclear or irrelevant who actually did it, and all that really matters is that it happened. Although it's not unusual to omit a subject in Japanese, that's normally done when the subject is implied, not unknown or unimportant. However, this is a perfect situation for using a passive verb, which inherently takes the focus away from the subject.

Shortening:

The passive form of -ru verbs is sometimes shortened by leaving out the ら (ra), such as using 食べれる (tabereru) rather than 食べられる (taberareru). This is considered nonstandard, but does occur.

Potential for confusion:

Passive form is identical to potential form for -ru verbs (because they're using different senses of the same auxiliary verb, which can also express respect or spontaneity, though the last two are rare in modern Japanese). Use context to tell which meaning is intended.

Looks passive, actually respectful:

A verb in passive form can be used to express the non-passive action of someone for whom you want to show respect (the helper verb involved can express either meaning, and can additionally be used to indicate potential or spontaneity instead). Though this is uncommon in modern Japanese, I've seen it often enough in written text to make a note of it.

Causative (-aseru, -saseru, -asu, -sasu)

A causative verb indicates that someone makes or allows someone else to take action (in order words, causes it to happen).

Conjugation table:

DictionaryCausativeCausative (shorter variant)
する (suru) させる (saseru) さす (sasu)
来る (kuru) 来させる (kosaseru) 来さす (kosasu)
-u verbs change -u to -a and add せる (seru) change -u to -a and add す (su)
-ru verbs change ～る (-ru) to ～させる (-saseru) change ～る (-ru) to ～さす (-sasu)
Examples:
飲む (nomu) 飲ませる (nomaseru) 飲ます (nomasu)
迎える (mukaeru) 迎えさせる (mukaesaseru) 迎えさす (mukaesasu)

This is built on the irrealis form. I'm not sure why the irrealis, since it usually shows up when expressing that which is not (negative forms, most often), but perhaps it's because the subject is not actively doing anything, instead being compelled to by another.

Usage:

The resulting causative form verb conjugates as a regular -ru verb, while the shorter variant shown in the rightmost column conjugates as a regular -u verb ending in -su. Use the topic marker は (wa) or subject marker が (ga) to indicate who or what causes the action, and the particle に (ni) to mark who or what actually carries out the action. If the verb is intransitive (has no object), the object marker を (wo) can be used instead of に.

It's unclear without proper context whether the causer is allowing an action (that the actor wants to do) or forcing one (that the actor doesn't want to do). For example, the sentences above could also be read "Joyce's friends allowed her to watch a movie" and "The parents make their children play outside."

Appropriate descriptive words can make the intended meaning clearer, for example using adverbs such as 無理に (muri ni) when the action is forced. Another way to make the "allow" sense unambiguous is to use the -te form of the causative verb and follow with one of the "give" or "receive" verbs (あげる, くれる, etc.):

This usage is also one way to ask permission.

Shortening:

There's a shorter variant of the causative form, shown in the conjugation chart above. According to conjugation notes on Jim Breen's WWWJDIC, this form has a connotation of a more direct causation. Assuming he's right, it would mean that, for example...

Japanese Wikipedia indicates that this alternate form is ungrammatical for する (suru), 来る (kuru), and ichidan (-ru) verbs, but a simple Google search shows that it's fairly common regardless.

Causative-Passive (-asareru, -saserareru)

Causative-passive is just what it sounds like... a causative sentence in passive form. That is, instead of saying that X makes Y do Z as in a normal causative sentence, it says that Y is made to do Z by X.

Conjugation table:

DictionaryCausative-Passive
する (suru) させられる (saserareru) or さされる (sasareru)
来る (kuru) 来させられる (kosaserareru) or 来さされる (kosasareru)
-u verbs change -u to -a and add ～せられる (-serareru) or ～される (-sareru)
-ru verbs change ～る (-ru) to ～させられる (-saserareru) or ～さされる (-sasareru)
Examples:
渡す (watasu) 渡させられる (watasaserareru) or 渡さされる (watasasareru)
転ぶ (korobu) 転ばせられる (korobaserareru) or 転ばされる (korobasareru)
落ちる (ochiru) 落ちさせられる (ochisaserareru) or 落ちさされる (ochisasareru)

Base verb in irrealis form plus causative suffix in irrealis form plus passive suffix.

Usage:

The resulting causative-passive form verb conjugates as a regular -ru verb. Like the simple passive, it changes the emphasis and tends to add a sense of victimization. The "victim" is marked with the topic marker は (wa) or subject marker が (ga), and the "slave driver" with the particle に (ni).

The causative-passive is just the causative form also put into passive form, except that the trailing ～せられる (-serareru) is often shortened to ～される (-sareru), unless that would cause a double さ (sa) and sometimes even then. This comes from the shorter variant of the causative form, and, properly speaking, should only be used with godan (-u) verbs but is used with others often enough regardless.

Imperative (-nasai, -e, -yo, -ro, -na)

Though it's normally better to make requests instead of giving orders, there are times and situations when commands are appropriate. That, and they're used constantly in any sort of fiction with any amount of action or drama.

(Relatively) polite commands:

First off, the nicer sort of command: Add ～なさい (-nasai) to a verb stem to create a command with the connotation that the commanding person knows better and is, generally speaking, giving the order for the commanded person's own good. Parents talking to children and teachers talking to students often use this form.

There is no common negative form, but ～てはいけない (-te wa ikenai) ("must not do") is comparable. The uncommon negative form is to add ～なさるな (-nasaru na) to the verb stem, but I can't recall ever encountering it outside of conjugation charts.

Harsher commands:

Of course, there are also not-so-nice commands. These are a bit more complicated, but they're also briefer and sound more like orders. These blunt commands are the kind most often used in the action genre, and are the way a drill sergeant might talk.

Conjugation table (blunt commands):

DictionaryDo it!Don't!
する (suru) しろ (shiro), or the more literary せよ (se yo) add ～な (-na) to the dictionary form
来る (kuru) 来い (koi)
Regular -u verbs change final -u to -e
Old-style -aru verbs change final ～る (-ru) to ～い (-i)
-ru verbs change final ～る (-ru) to ～ろ (-ro),
or the less common older and somewhat milder ～よ (-yo)
Examples:
取る (toru) 取れ (tore) 取るな (toru na)
いらっしゃる (irassharu) いらっしゃい (irasshai) いらっしゃるな (irassharu na)
見る (miru) 見ろ (miro) or 見よ (miyo) 見るな (miru na)

Appropriately, the positive ones use 命令形 (meirei-kei), or imperative form.

There are situations in which this form isn't rude, such as when cheering people on by telling them to do what they want to do anyway, or in an emergency situation when the urgency is appropriate. In general, however, it should be avoided.

In addition, since honorific verbs are, after all, honorific, their command forms are not rude either. In particular, いらっしゃい (irasshai, to come or go) is a common welcome in stores, ください (kudasai, to give) is often used in requests, and なさい (nasai, to do) is used in the gentler commands explained above, as well as in idiomatic expressions such as お休みなさい (o-yasuminasai, "good night").

Other:

～でない (-de nai) may be added to the dictionary form as a negative imperative, but that's more archaic.

It's also possible to conjugate the respectful ～ます (-masu) ending into the imperative form as ～ませ (-mase), but this essentially never happens outside of a few uses that have become semi-idiomatic, like いらっしゃいませ (irasshaimase), the longer version of the common service-sector greeting.

See also:

The んだ (n da) sentence ending may also be used in a commanding sense.

Though more archaic in usage, adding お～ (o-) to a verb stem also functions as a command form.

Conditional (-tara)

To turn a verb into the conditional form, just add ～ら (-ra) to the end of the past affirmative form (to make a positive condition) or past negative form (to make a negative condition), for all verbs. Note that the short past form is typically used as the base. Using the long forms is not incorrect, but sounds very formal, excessively so for normal usage.

A sentence in the form "P (positive conditional verb) Q" means that Q will happen once P does. If P is fairly likely to happen or almost definitely will, it's essentially the same as "Q when P". If P may or may not happen, it's closer to "Q if P", but is still only used when P happens before Q does. P can even be entirely false, which is used to describe hypothetical situations—the entire sentence is usually followed by が (ga), けど (kedo), のに (no ni), でしょう (deshou), or something similar in this case to make it more tentative, speculative, or contrastive. In any case, P happening means that Q will happen too, but only after P does and not before.

A negative conditional works the other way. In a sentence P (negative conditional verb) Q, P not happening means that Q will happen. Negative conditionals frequently appear when describing what might have happened if the situation had been different.

Note that this indicates absolutely nothing about whether the condition is true, or about what might happen if the condition is false.

The key difference between this form and other constructs, such as the provisional form, that roughly translate to "if P, then Q" is the emphasis on time. Q does not happen until after P does. Note particularly that in the first example above, there is little if any cause-effect relationship between P and Q, just the fact that I read my e-mail only after I go home. This lends itself to making hypothetical statements in the sense that, since P failed to happen (if past tense) or is unlikely to ever happen, the timeframe for Q to happen (after P) simply doesn't exist, so Q is necessarily pure speculation.

Suggestions:

The conditional form is also one way to make suggestions. Express the suggestion with a conditional verb at the end and follow with どうですか (dou desu ka). Casual language may use just どう (dou) or どうだ (dou da), or simply end with the conditional verb and a rising intonation (question mark if written). The literal translation is roughly, "what would happen if...?" or "how would it be if...?" The time sense applies here too, though it's unnatural to phrase it that way in English: "If you did P, what would things be like after that?"

Provisional (-eba)

To put a verb in the provisional form, just change the ending -u in the dictionary form to -eba, or change the ～ない (-nai) in the short nonpast negative form to ～なければ (-nakereba) to make the negative provisional form.

This is built on the 仮定形 (katei-kei), or conditional form, of the verb itself when positive or of the ～ない suffix when negative.

A sentence in the form "P (provisional verb) Q" means that Q is true if P is true, or, to use the fancy name for this tense, Q will be true provided that P is true. Using this form says that situation P leads to result Q, and has some implication that Q is unlikely to be true unless P is true, though it doesn't go as far as demanding that P must be true for Q to be true. It is less often used when Q is an undesirable outcome, or when the cause-effect relationship between P and Q is fairly weak (try the conditional form instead for those).

Putting a verb in provisional form and adding いい (ii, good) or some other, often equally vague, desirable result is one way to advise an action.

Must do:

The negative provisional form followed by one of several generic negatives is one way to say that an action must be taken (i.e., if don't do, not good). Refer to the -te form + ikenai subsection above for further details.

Regret:

To say that something that didn't happen would have been better off happening, or that something that did happen would have been better off not, use a positive result in the past tense, usually よかった (yokatta, was good), after the provisional form. This usage is often followed with のに (no ni) or other contrasting forms to highlight the difference between reality and what might have been.

Contrast this with the similar phrase -te form + yokatta that expresses satisfaction with a favorable outcome:

Shortening:

It seems that many people find ～なければ (-nakereba) to be cumbersome, as there are at least three shorter variants. ～ねば (-neba) seems to be more literary, while the ～なきゃ (-nakya) and ～なけりゃ (-nakerya) contractions commonly occur in casual speech. I've also seen ～なくば (-nakuba) once or twice, but that's an archaic form, and furthermore should properly be used only when the condition is at least suspected to be false.

Similar to:

There are some differences in nuance, but the provisional form is very similar in function to the particle なら (nara) as used after a verb.

Alternative Form (-tari)

To list actions as examples and indicate that it's not a complete list, attach ～り (-ri) to the short past affirmative of each action. You then usually end the chain with a する (suru) after the last ～り. This gives an "and so on" sort of meaning, or alternately, a "some of this, some of that" meaning. The phrase formed can be used anywhere a simple verb can be, and only する changes its conjugation.

Speculation:

The above isn't the only use of the form; I've seen a number of situations in which it's followed by して (shite), often as だったりして (dattari shite), at the end of a sentence. This roughly equates to "what if" or "could be", and is used to make speculations. This function may also use negatives by adding ～り to the past negative form, such as じゃなかったりして (ja nakattari shite).

I think the nuance has to do with this being only one of numerous possibilities (because of the alternative form). You're throwing it out there as something that is conceivable, but not at all certain or even necessarily likely.