Irregular verbs

English has many irregular verbs, and at first, there seems to be no rhyme or reason in these verbs. However, historically, irregular verbs can be put into different groups, and these verbs are listed here for those interested.

First things first, it is important to know that irregular verbs can be split mainly into two categories of conjugation: strong and weak. The difference is:


 * Strong verbs form their past tense and past participle forms with a change in vowel, and the past participle ends (or formerly ended) with -en. This change in vowel is called ablaut.
 * Weak verbs need an additional ending such as -ed for their past tense and past participle forms.

There are also two other categories: preterite-present and "anomalous". Verbs of those two categories are few, but are nonetheless among the speech's commonest verbs.

Strong verbs
See here.

Weak verbs
See here.

Preterite-present
In Proto-Germanic, certain verbs commonly thought to have come from the Proto-Indo-European perfect came to have present significance. For example, the archaic verb wit (meaning know) is akin to Latin videō (I see), and it is not hard to see how past sight of something implies present knowledge thereof. Hence, these verbs have the same form in the first and third-person singular present, unlike normal verbs. Most of these verbs came to be used as modal auxiliaries, which are now defective and lack infinitives and participles.

Note that the preterite-present verbs originally sprang from strong verbs, and it is usually clear from the forms which class they belonged to. For example, OE wāt has the same vowel as past tense forms of Class 1 strong verbs.

The preterite-present verbs, their Old English forms, and their modern descendants are (all reconstructed OE forms are in brackets):

Notes:


 * Can originally meant know, which later became know how to, whence its current meaning of be able to.
 * The infinitive would have become cun. Likewise, the present participle would have become cunning, which survives as an adjective meaning sly.
 * The past tense would have normally become couth (rhyming with mouth), but in Middle English, the th became d by analogy with the d used for the weak past tense. Even so, the word would have normally become coud (rhyming with loud), but in Early New English, it became could by analogy with should and would.
 * The etymologically original past participle was couth (OE cūþ), but by Old English, it had become an adjective meaning known. The adjective became archaic, but it is preserved in uncouth, which originally meant unknown and now means unrefined; the modern couth is a backformation from uncouth. The modern pronunciation of (un)couth, which has it rhyme with booth rather than mouth, is not regular; it appears to have come from a Northern variant of the word.
 * May originally meant have the power to do, whence its current meanings of possibility and permission. The original meaning of may is found in the related noun might and the adjective mighty.
 * In Middle English, the infinitive has a few different variants; the reflex of magan did not survive into Middle English, but would have yielded mawen (which would now be maw). One variant is traced back to an unattested OE variant mugan, which yielded muwen (later written as mowen), and this variant would now be mow rhyming with cow. The vowel was later transferred to the past tense, which then yielded the variant mought (now used only in some nonstandard dialects and rhyming with out).
 * The OE past tense form meahte would have yielded maught (rhyming with taught); the current form might is from the OE variant mihte, an umlauted variant that may have come from influence of the related noun miht (now might).
 * Shall originally meant owe, whence its current meaning of obligation and current use as a future tense auxiliary.
 * The modern pronunciation of shall is exceptional, since the normal reflex of OE sceal would have it rhyme with fall. The modern pronunciation comes from a weak form with /æ/.
 * Should, the past tense form, shows lengthening of o before ld, which yielded the Early New English pronunciation /ʃuːld/. The modern pronunciation, /ʃʊd/, comes from a variant that shows loss of /l/ and vowel shortening. The modern spelling with ou seems to come from an Early New English convention of representing /uː/ with ou.
 * Must was originally the past tense form of mōtan, but it was used so often as a past subjunctive that it became a present-tense auxiliary, whence the lack of any distinct past tense form. The word would have normally become moost, but the vowel was shortened in late Middle English because of the st cluster, whence the modern form must.
 * The original present tense would have normally become moot, though the vowel in late Middle English was often shortened, which yielded forms such as mut (by analogy with must).
 * Owe is now a regular verb and is no longer used as an auxiliary.
 * The present tense would have become ough (rhyming with dough). Of course, this makes no difference now, but in Middle English, ough and owe represented two different pronunciations, and the original present tense was replaced with owe (found in such forms as the infinitive).
 * Owe originally meant have, possess. Own, originally the past participle, became an adjective. The weak verb āgnian was formed in Old English, but seemed to become obsolete in Middle English; the current verb own is apparently a backformation from owner.
 * Owe later shifted in meaning from possession to obligation, whence the current meaning of ought. Ought was the past tense of owe, but it was used so often as a past subjunctive that it became its own verb, which is why it lacks a distinct past tense form.
 * OE āgan, notwithstanding that the present tense has ā (a characteristic sign of the past tense of Class 1 strong verbs), is now commonly thought to have been from Class 7 instead. The reason is that āgan showed a lack of vowel distinction between the singular and the plural. In contrast, OE witan showed a singular-plural distinction in the present tense (singular wāt and plural witon).
 * Dare is now a regular verb when not used as an auxiliary. The original infinitive is unattested in Old English, but begins to be attested in Middle English and would have become dur.
 * The present tense would have become dar, but the vowel was later lengthened by analogy with past tense forms of Class 4 and 5 strong verbs such as spake and bare (which would have otherwise become spack and bar).
 * The past tense form would have become dorst, but in Middle English, it was influenced by forms with u such as the present plural, whence the past tense became durst (which later became archaic).
 * Wit, meaning know, has a full conjugation as well while keeping the irregularity in its forms, much like be, but it is not used as an auxiliary. The verb is now archaic, but it survives in (un)witting and to wit. Related words to the verb include the noun wit and witness.
 * The present tense would have normally become wote, but the vowel was later shortened because of the dental consonant.
 * The past participle would have become witten, but later on, the past tense form wist became used for the past participle. The archaic adverb iwis meaning surely (OE gewiss) shows a trace of the etymologically original past participle, but by Old English, it had become an adjective meaning sure, certain.
 * Dow is a dialectal verb meaning thrive, prosper.
 * Its past tense form is dought, and the verb is related to the adjective doughty, which, notwithstanding its spelling, rhymes with pouty (the expected reflex of OE dohtig would have it rhyme with naughty). The vowel in doughty and dought looks to have been transferred from the infinitive dow.
 * OE munan (generally used in the derivative gemunan) meant remember. The original infinitive (attested in gemunan) would have become mun.
 * OE nugan is found only in the derivatives benugan and genugan (both of which have unattested infinitives). Benugan means require, and genugan means suffice and is related to the adjective enough. The original infinitive would have become now.
 * OE þurfan means need, require, and in the negative, it means not have to, which makes it the opposite of must not. The original infinitive would have normally become thurve, but in Middle English, the infinitive and inflected forms commonly showed loss of f, and there was a strong tendency to mix this word up with durren (now dare).
 * OE unnan means grant, allow. The original infinitive would have become unn.
 * Sometimes, the verb need is conjugated as if it were a preterite-present, so one may see such phrases as he need not go and need I say any more?. This use is confined to questions and grammatically negative phrases, and it is still usual to treat need like a regular verb, e.g., he does not need to go, do I need to say any more?.

Anomalous
There are some verbs that are truly irregular, as they do not belong to any of the aforesaid classes. There are only four: be, do, go, and will.

Be, the English tongue's most irregular verb, is historically a blend of two different verbs: bēon and wesan. From the bēon conjugation are:


 * The infinitive, imperative, and present subjunctive, be.
 * The present participle, being.
 * The past participle, been.

And from the wesan conjugation are:


 * The first-person singular present indicative, am.
 * The second-person singular present indicative, are (the form for thou is art).
 * The third-person singular present indicative, is.
 * The plural present indicative, are.
 * The first and third-person singular past indicative, was.
 * The second-person singular past indicative, were (the forms for thou are wast and wert).
 * The plural past indicative and subjunctive, were.
 * The past singular subjunctive, were (the form for thou is wert).

These forms came from originally distinct verbs.

Do is somewhat irregular in that the present tense is not formed regularly. The inflected forms does (as well as archaic dost and doth) all show /ʌ/ instead of /u(ː)/. The reason is that in Early New English, the vowel was shortened in these forms. This shortening is also found in other words such as blood and flood, which formerly had /uː/.

The past tense of do, did, was originally formed by a reduplicated form, which makes do the only living verb whose past tense came from reduplication. Other than that, it is straightforwardly gotten from OE dyde.

The past participle, done, is straightforward, as it has the past participle ending -en; the OE form was dōn, and the vowel later underwent shortening like the inflected present tense forms.

The past tense of go in OE was ēode (which itself was a suppletive form), which would have naturally become eed. In ME, forms such as yede and yode appeared, the y being the result of shift of stress in the diphthong, and these would now be yeed and yood. But by Early New English, they had already been replaced with went, a new suppletive form that had originally been the past tense of the weak verb wend (which now uses wended instead).

The past participle of go is gone, which has the usual past participle ending; the OE form was gān, and the vowel later underwent shortening in Early New English. The same shortening is found in the British pronunciation of shone.

Will (distinct from the regular verb will as in God willed it so) was originally gotten from the Proto-Germanic subjunctive. Hence, unlike the other modal verbs, it is not a preterite-present verb. However, it is easy to mistake it as such, since the present tense is now the same in both the first-person singular and the third-person singular, and no personal ending is added to the latter; the OE form for the first-person singular was wille, and that for the third-person singular wil(l)e.

The past tense of will is would, but the verb has no past participle (both in Old English and in New English), nor has it now an infinitive or present participle (which survives as the adjective willing). The verb originally was a normal verb meaning desire (like the German cognate wollen), but later came to be used as an auxiliary not only to show desire but also to refer to the future. The pronunciation of would followed the same development of should. First, the o was lengthened before ld, so after the Great Vowel Shift, it was naturally pronounced /wuːld/. The modern pronunciation is from an Early New English variant that shows loss of /l/ and vowel shortening.