Strong verbs

Strong verbs are verbs that form their past tense and past participle forms through a change in vowel; this change is called ablaut. The past participle also ends or formerly ended with the suffix -en.

All disused forms or forms rebuilt on the basis of regular sound changes are in bold. If verbs have regular forms as well, the regular forms will not be listed, since this is mainly about the irregular forms.

A few notes in regard to categorization:
 * The system herein is mainly historical; even if a strong verb's forms were influenced to some extent by those of another class of strong verbs, or the strong verb has become weak, the strong verb will still be listed in its original class. The new forms will be listed alongside it.
 * If an originally strong verb has irregular weak forms, it and the irregular weak forms are listed.
 * Regular verbs that have been made strong by analogy in New English are not included if these forms are not used in standard speech.
 * Derivatives such as understand and arise are not included.
 * A (*) put after a form shows that it is what the strong form would have likeliest become and is not attested in New English.

In each class, the principal parts (that is, the vowel pattern that the verb had in Old English) are given:


 * Infinitive - the vowel in the infinitive.
 * Past singular - the vowel in the first- and third-person singular past indicative.
 * Past plural - the vowel in all persons plural and second-person singular past indicative.
 * Past participle - the vowel in the past participle.

For a list of other irregular verbs, see here. For a list of reconstructed strong verbs, see here.

Class 1
The principal parts of the verb rīdan were:


 * Infinitive: ī - rīdan.
 * Past singular: ā - rād.
 * Past plural: i - ridon.
 * Past participle: i - riden.

The first became ride; the second rode; and the last two rid and ridden, respectively. As one can see, we have two ways of forming the past tense:


 * From the past singular, e.g., drove, shone.
 * From the past plural and the past participle, e.g., bit, slid.

This twofold way of forming the past tense explains that the surviving forms are not fully consistent. For example, the past tense forms for ride were once rode and rid, and rode was chosen over rid. However, in the case of bite, bit was chosen over bote, and in that of slide, slid was chosen over slode; undoubtedly, analogy with irregular weak verbs like light and meet helped establish this. It is reasonable to treat slide as an irregular weak verb now, but since bite still has bitten as its past participle, bite ought to still be treated as a strong verb.

Most of this class's verbs that have stayed strong use the original past participle forms. It can be seen that the past participle forms have a shortened form of the original infinitive vowel, e.g., rīdan (ride), riden (ridden).

This class has been preserved well; all the members that still have this class's strong forms are uniform in the infinitive vowel.

Notes on infinitive forms:


 * The verb shit is unattested in Old English, but it is attested in the form of besciten, the past participle of the unattested verb bescītan. In Middle English, the verb is found with strong forms. The infinitive vowel was originally long (and shite is still used in a few dialects), but likeliest from influence of the noun and the past participle shitten, the vowel was later shortened.

Notes on inflected forms:


 * Since shit (formerly shite) no longer has the same vowel as a usual Class 1 verb, the forms shote and shitten became archaic, and for the past tense and the past participle, shit has:
 * Shit - from the past plural and the past participle.
 * Shat - by analogy with the Class 5 verb sit.
 * Shitted - the regularized form.
 * The pronunciation of shone appears to differ in British English, in which it rhymes with gone instead. The American pronunciation, which has it rhyme with bone, follows this class's pattern.
 * As for why the past tense of strike shifted from stroke to struck, according to Hereward Thimbleby Price, it seems that it was influenced by stuck, since struck became widespread only around the same time that stuck popped up, and it explains why strike is the only Class 1 verb to have such a strong form.
 * The disused past tense form strick, which uses the vowel in the past participle, is seldom attested in the early 17th century; it was likely formed by analogy with verbs like bite and slide.

Notes on infinitive forms:


 * The current verb spew once had the same vowel as other members of this class, as it was spīwan in Old English.

Notes on inflected forms:


 * Abide (etymologically a derivative of bide) occasionally uses abode as the past tense and even the past participle (the original past participle would be abidden), but it is now generally conjugated regularly.

Class 2
The principal parts of the verbs crēopan and sūcan were:


 * Infinitive: ēo, ū - crēopan, sūcan.
 * Past singular: ēa - crēap, sēac.
 * Past plural: u - crupon, sucon.
 * Past participle: o - cropen, socen.

The past singular later came to have the same vowel as the lengthened past participle, whence came such past tense forms as chose and froze.

This class has more or less fallen apart, as no living Class 2 verb uses ea for the past tense. Instead, most Class 2 verbs that stayed strong ended up with forms indistinguishable from those of Class 4. Moreover, not only have most Class 2 verbs become regular, but sound changes have destroyed the original vowel pattern, and so surviving members of this class no longer show consistency in their infinitive vowels.

Notes on infinitive forms:


 * Choose and the now weak verb shoot represent atypical phonetic changes, as their OE forms cēosan and scēotan would have normally yielded cheese and sheet. It seemed that stress was transferred to the second part of the diphthong, perhaps because of the initial palatal consonant.
 * Cleave is rather strange, since OE clēofan would have normally yielded cleeve. Of course, there is no difference in sound now, but cleave and cleeve were formerly pronounced differently. The form cleave appears to have come from confusion with the unrelated weak verb cleave (adhere).

Notes on inflected forms:


 * The past tense and past participle forms of choose was were affected by Verner's law; the past plural and the past participle of cēosan (choose) were curon and coren, and those of frēosan were fruron and froren. In Middle English, these changes were gotten rid of, as the strong forms adopted the consonant in the infinitive and the present tense by analogy.
 * In addition, for choose, the initial consonant of curon and coren was /k/, not /tʃ/. The consonant in the infinitive and the present tense was later extended to those forms, which thus yielded chose and chosen.
 * Later writers such as Milton used frore (a reduced form of the original past participle) as a poetic word for frozen.
 * The past tense of fly lacks the same vowel as is in the past participle, which makes it an exception to the usual change in verbs of this class. It seems that because of the past participle flown, a past tense form flew was made by analogy with Class 7-1 verbs such as know and grow.

Notes on infinitive forms:


 * Bid was bēodan in Old English, and thus, its expected reflex would be beed. The current vowel seems to have been due to confusion with the Class 5 verb bid (the regular reflex of OE biddan).
 * The vowel in brook is somewhat puzzling. From the few examples of /uː/ before /k/, the vowel seemed to consistently shorten before /k/. Hence, OE sūcan and tūcian became suck and tuck, and if OE brūcan had followed the same line of development, it would be bruck instead. It is possible that /uː/ simply failed to shift before /k/. There is another explanation, however: in Middle English, there was apparently a byform broken showing /oː/, which would have naturally yielded brook. This then suggests that the modern word came from an unattested OE weak verb brucian (to which open-syllable lengthening could apply and yield the ME byform, as shown in OE wudu > ME wode > NE wood). This is probably the correct explanation, since it explains why the spelling is brook, and  suggests that it formerly had /oː/.
 * Shove, suck, and sup show shortening of /uː/.

Notes on inflected forms:


 * The past tense and past participle of bid is bid, but in the phrase bid fair, bade and bidden are often used because it has been confused for the Class 5 verb bid.
 * Forbid is not etymologically connected to the Class 5 verb bid (meaning command), but because of the similarity in form and meaning, forbid was later thought to be as a derivative thereof and thus became conjugated so; the past tense is now forbade, and the past participle forbidden.
 * The old past tense and old past participle were bod and bodden (variants bode and boden, wherein the vowels had not been shortened). The shortened vowel was likely due to the dental consonant (compare with shot, sodden, gotten, and trodden).
 * Flee and fly once shared the same past tense and past participle forms, and the two verbs continued to be confused for each other in Middle English. Only later on were they formally differentiated in this aspect. Flee became an irregular weak verb, whereas fly kept its strength.
 * The past participle of shoot has a short vowel, which is likely due to shortening before dental consonants (compare with gotten). The vowel in the past tense is from influence of the past participle or from shortening, and it was later strengthened by analogy with irregular weak verbs like meet and light. Because it is now indistinguishable from those verbs, it is best that it be deemed an irregular weak verb.
 * Sodden, the archaic past participle of seethe, lives on as an adjective that has diverged from the meaning of seethe, as it now means soaked, saturated instead of boiled.
 * That sodden has a d instead of the expected th is due to alterations brought about by Verner's law. In Middle English, an alternative form was sothen (included here to show a variant in which the alteration had been eliminated). A later form of the past tense was sod, which was formed by analogy with sodden.

Subclass 1
The principal parts of the verb singan were:


 * Infinitive: i - singan.
 * Past singular: a - sang.
 * Past plural: u - sungon.
 * Past participle: u - sungen.

Note that in every verb of this subclass, an m or an n follows the vowel.

In Middle English, for many verbs, the forms began to be mixed up, insofar as the past tense form came to have the same vowel as was in the past plural and the past participle. There was no consistency whether the original past tense form was kept, whence there are two ways of forming the past tense for verbs of this subclass:


 * From the past singular: began, sang, swam.
 * From the past plural and past participle: swung, stung.

In other words, it was once correct to say I swang and I wan, but now one says I swung and I won.

Note that for words ending in nd, sound changes have beclouded the original pattern. In the Middle English period, the vowel in the infinitive, the past tense, and the past participle were lengthened because vowels were lengthened before the consonant cluster nd. The past plural and the past participle of find later became found, and the singular form fand had died out. In other words, the modern past tense form draws its vowel from the past plural and the past participle.

The past participle forms originally ended in -en, but it has been dropped, which leaves the change in vowel as the lone marker of the past participle. Some of the -en forms survive as adjectives, e.g., drunken, shrunken, sunken.

This subclass is rather well preserved, as nearly every surviving verb of this subclass has kept at least a trace of the original vowel pattern.

Notes on infinitive forms:


 * The OE infinitive beginnan showed /j/ rather than /g/. The change to /g/ later happened from influence of inflected forms.
 * Run was rinnan or irnan (because of metathesis) in Old English. The expected development of the original form, rin, is now used only in some dialects, and how the standard form became run is somewhat unclear. One possible cause is that the metathesized form was the usual form in Old English (which was also the case for burn, OE beornan, originally brinnan). The metathesized form was passed down to Middle English, and it is guessed that after the vowel changed, metathesis happened again, which yielded run. That metathesis was undone may have been due to influence from Scandinavian forms as well. Another possible cause is that the past plural and the past participle influenced the infinitive.

Notes on infinitive forms:


 * Burn (Old English beornan) is the result of metathesis in prehistoric Old English (the expected OE infinitive is brinnan, which would have become brin), and the only attested Old English example of the original form is found in onbran (set fire to (past)).

Notes on inflected forms:


 * The past tense and past participle of climb in Old English were clamb and geclumben, and the vowel in the infinitive was short, whence the verb followed this subclass's pattern fully. But sometime later, the vowel was lengthened, and the strong forms clamb and clumb became disused in standard speech (though they stayed in use in some dialects). A strong form clomb was also used; it first arose as a result of vowel lengthening in ME clamb (as vowels underwent lengthening before mb), and later on, it was used alongside climb likely by analogy with Class 1 verbs such as drive and shine. In the end, however, climb was regularized.

Subclass 2
This class makes up the rest of Class 3 verbs, as these verbs do not have an m or an n following the vowel. The principal parts of the verb bregdan were:


 * Infinitive: e - bregdan.
 * Past singular: æ - brægd.
 * Past plural: u - brugdon.
 * Past participle: o - brogden.

The past plural and the past participle were kept throughout this subclass, but the infinitive and the past singular were highly inconsistent because of phonological developments such as breaking and diphthongization. For comparison, the principal parts of the verbs ceorfan, gieldan, and helpan were:


 * Infinitive: eo, ie, e - ceorfan, gieldan, helpan.
 * Past singular: ea - cearf, geald, healp.
 * Past plural: u - curfon, guldon, hulpon.
 * Past participle: o - corfen, golden, holpen.

The verbs fēolan, frignan, murnan, spurnan were exceptional in their infinitives, but followed the above patterns for the other principal parts.

In Middle English, there was a tendency for the past tense to adopt the past participle's vowel, but in the end, nearly every verb became regular or gained regular forms, and so in contrast to Class 3's other subclass, this subclass has nearly disappeared.

Notes on inflected forms:


 * The old past participle of fight is preserved in the archaic phrase foughten field (battleground).

Notes on infinitive forms:


 * The OE infinitive ceorfan (carve) showed /tʃ/ rather than /k/. The consonant in the infinitive later shifted to /k/ from influence of inflected forms such as curfon and the past participle corfen.
 * The vowel in milk appears to be from the OE weak verb milcian rather than the strong verb melcan.
 * Mourn (OE murnan) shows lengthening of /u/ before /rn/. On the other hand, lengthening is not seen in spurn (OE spurnan).
 * Swallow is somewhat unusual, as the OE form was swelgan, which would have normally yielded swellow. The verb appears to have been later influenced by the noun swallow referring to the bird.
 * Thresh later had thrash as a phonetic variant, and thrash is now the form generally used to mean beat, flog, though thresh, in addition to having the same meaning as thrash, also has the specialized meaning of separate grain from a plant.

Notes on inflected forms:


 * The verb burst (Old English berstan) underwent metathesis in Middle English from Old Norse influence, whence the past tense had brast as a variant. However, the verb was later reverted to its original form, and the vowel in the infinitive was later transferred to the past tense and past participle. The past participle ending was later dropped, which left burst to be used for all three functions.
 * Corven and storven became obsolete as past participles. Carven and starven were later formed by analogy with strong verbs that have the same vowel in the infinitive and the past participle, but these forms were, in the end, displaced by the regular forms.
 * Molten is still used as an adjective.

Class 4
The principal parts of the verb teran were:


 * Infinitive: e - teran.
 * Past singular: æ - tær.
 * Past plural: ǣ - tǣron.
 * Past participle: o - toren.

This class had two anomalous verbs in Old English: cuman (come) and niman (take).

Note that in a Class 4 verb, the vowel is followed by a liquid or a nasal. The two exceptions are brecan (break) and hlecan (cohere), both of which apparently had originally been Class 5, but had already gotten a past participle of this class by analogy.

The past tense forms developed into such forms as tare, brake, and bare, but they became influenced by the vowel in the past participle and so were changed into tore, broke, and bore.

Class 4, though small, has been preserved rather well; nearly every normal Class 4 verb is uniform in their past tense and past participle forms.

Notes on infinitive forms:


 * Break is an atypical example of Early New English /eː/ not shifting to /iː/. If it had shifted, the word would now rhyme with speak.

Notes on inflected forms:


 * Derivatives of bear use borne as the past participle, e.g., forbear - forborne.
 * The past tense of OE cuman was cōm, which would have normally yielded coom. However, the a form began to be used in late Old English and later overthrew the old form in the standard speech. The a form was apparently formed by analogy with Class 4 and 5 verbs, and the vowel was later lengthened, which yielded came.
 * Shore appears to be used mainly in Australian English, other areas preferring sheared as the standard past tense.

Class 5
The principal parts of the verb metan were:


 * Infinitive: e - metan.
 * Past singular: æ - mæt.
 * Past plural: ǣ - mǣton
 * Past participle: e - meten.

A few Class 5 verbs have i for the infinitive instead, e.g., biddan (bid, as in bid farewell), sittan (sit), and licgan (lie, as in lie down).

In a Class 5 verb, the vowel is followed by a consonant that is neither a liquid nor a nasal (which separates it from a Class 4 verb).

In Middle English, the past tense forms were lengthened, whence the old past tense of speak became spake. However, as for sit, the unlengthened variant, the expected reflex of the Old English form, survived, whence we use sat.

As one can see, the principal parts are nearly identical to those of Class 4, the one exception being the past participle. It is thus not surprising to see why there is a strong tendency for Class 5 verbs to gain strong forms modeled on those of Class 4 verbs; that is, the vowel in the past participle is substituted with o, after which the vowel in the past tense is substituted therewith.

Class 5 has not been preserved well; there is no longer any uniformity in the infinitive vowel, and many surviving verbs have become regular or gained Class 4 strong forms.

Notes on infinitive forms:


 * The current forms of fret and tread are somewhat atypical, since the expected reflex of OE fretan is freat, just as OE etan became eat, and likewise, tread would rhyme with mead rather than head. The shortened vowel may have been due to the final dental consonant.
 * Lie is not gotten from the OE infinitive licgan, which would have yielded lidge instead. Instad, it is gotten from such inflected forms as ligst and ligþ.
 * Mete uses an archaic spelling representing Early New English /eː/; the spelling expected would be meat.
 * Speak and the related noun speech both had r in OE, e.g., sprecan, sprēc. The loss of r happened in late Old English.

Notes on inflected forms:


 * The past participles of OE biddan and sittan were beden and seten, which might have become beaden and seaten. However, later on, they were altered to bidden and sitten by analogy with the vowel in the infinitive.
 * Bid is sometimes conjugated like the Class 2 verb bid (meaning offer) because both verbs have been confused for each other from the similarity in form and meaning.
 * The past tense traditionally rhymes with glad and is occasionally spelled bad (but bade is now the spelling generally used), but an alternative pronunciation has it rhyme with made.
 * Forbid is not connected to bid of this class, being a derivative of bid of Class 2, but because of the similarity in form and meaning, forbid is now treated as a derivative of bid of Class 5; the past tense is forbade, and the past participle forbidden.
 * Curiously, eat (and its etymological derivative fret) in Old English had lengthened forms for the past tense, even in the singular. However, there must have also been a shortened form (whether inherited or formed by analogy with other Class 5 verbs), as the shortened form was later lengthened and became ate.
 * The American pronunciation of ate has it rhyme with eight and matches this class's pattern. The British pronunciation, however, traditionally has it rhyme with bet; it seems that this pronunciation rose from vowel shortening, which was applied to weak verbs like lead and read.
 * The past participle of lie, lain, naturally developed from the Old English form; a past participle lien was later formed by analogy, but it is now archaic.
 * The inflected forms of see underwent a few different analogical changes.
 * The OE past singular was seah, which became saugh in ME. The modern form saw appears to have been partly influenced from an ME past plural form sowen from OE sāwon.
 * The usual past participle form in OE was sewen or segen. The past participle form seen later arose in ME probably by analogy with the infinitive and the present tense and from influence of the ME adjective isene (visible).

Notes on infinitive forms:


 * Bequeath is the lone remnant of the obsolete verb queath meaning say.

Notes on inflected forms:


 * The past tense of obsolete queath became quoth (which lives on as a literary word for said) from a few different phonetic causes. However, bequeath (the only surviving derivative) was later regularized, and so the past tense is no longer bequoth.
 * The current pronunciation of quoth is artificial, since it is a literary word, and the spelling suggests that it is pronounced /kwoʊθ/. However, since it is the past tense of a verb inherited from Old English, a more natural pronunciation would be /kwoʊð/.
 * The rare past participle quothen is attested only in bequothen, which did not live past Middle English.

Class 6
The principal parts of the verb faran were:


 * Infinitive: a - faran.
 * Past singular: ō - fōr.
 * Past plural: ō - fōron.
 * Past participle: a - faren.

There were some exceptions, as they differed in the infinitive: hebban, hliehhan, sceaþan, scieppan, steppan, and swerian.

On the whole, this class no longer shows any uniformity.

Notes on infinitive forms:


 * The Old English verb sacan (meaning disagree) survives only in the derivative forsake.
 * Slay and the now weak flay are not gotten from the OE infinitives slēan and flēan. Instead, the infinitive forms were later influenced by such past participle forms as slægen and flægen, which yielded slain and now disused flain.

Notes on inflected forms:


 * The past tense forms of draw and slay (and formerly flay and gnaw) appear to have come about from influence of Class 7 verbs. Since the past participle of those verbs is formed simply by adding the past participle suffix to the infinitive, analogy with Class 7 verbs such as know and grow was then applied.
 * Stand is noteworthy in that the past tense, stood, has no n in it. The reason is that the n is historically a nasal infix that marks the present tense, whence stood naturally lacks an n. Stand is the only English verb that shows such a feature.
 * Swear looks like a modern Class 4 verb, but in Old English, the past tense was swōr, which matched the past tense forms of other Class 6 verbs. The past participle, interestingly enough, was sworen, which appears to have been made by analogy with Class 4.

Notes on infinitive forms:


 * Heave is not from the OE infinitive form hebban. Rather, the infinitive was newly shaped by such inflected forms as hefest and hefeþ, as well as the past participle form hefen.
 * Shape is not from the OE infinitive form sceppan. Rather, the infinitive was influenced by the past participle form scapen, which became now disused shapen.

Notes on inflected forms:


 * Though hove, the strong form of heave, is no longer found in common speech, it is still used in nautical contexts and the mainly nautical phrase heave in(to) sight/view, e.g., we waited until a ship hove in sight.
 * Hove is not the usual development of the Old English form, as it would have become hoove instead. As the past participle became hoven at one point, it suggests that hove and hoven are forms made by analogy with verbs that had Class 4 forms.
 * Heft was an irregular form made by analogy with verbs like leave and cleave, but it is no longer used.
 * The standard pronunciation of laugh comes from a variant using /f/. If the past tense lough (OE hlōh) had not become obsolete and had changed likewise, it would rhyme with tough (OE tōh).
 * Shapen and shaven are still used as adjectives.
 * Wash was conjugated in Old English mainly like a Class 6 verb, though one instance of a Class 7 past tense form (wēox, the last consonant due to metathesis) appeared.

Class 7
This is not so much a class as it is a category for all the verbs whose past tense was formed by reduplication in Proto-Germanic. Therefore, there was no uniformity in the infinitive vowel. The past tense, however, had ē or ēo, and the past participle had the same vowel as was in the infinitive, the lone exception being wēpan (weep).

Subclass 1
There is an obvious pattern in these surviving Class 7 verbs, so these are classified as a subclass. This subclass has suffered a few losses, but it has held up rather well.

Notes on inflected forms:


 * Blow in the usual meaning of move through air has the same form as the verb blow meaning produce flowers. Etymologically, in Old English, both verbs belonged to Class 7 and shared the same past tense forms, though their past participle forms differed.

Notes on inflected forms:


 * The past tense crew is occasionally used when the verb is used to describe a cock's characteristic cry, but crow is now generally conjugated regularly.

Subclass 2
This subclass consists of the remaining Class 7 verbs. Unsurprisingly, there is no uniformity among them at all.

Strong verbs currently used in standard speech

Notes on infinitive forms:


 * Hang comes from two OE verbs: the strong verb hōn and the weak verb hangian. The infinitive became hangen/hongen in ME from influence of the weak verb and the strong forms with ng such as the past tense hēng and the past participle hangen. The same process happened to the now dialectal verb fang meaning catch.
 * The expected reflex of the ME verb would be hong because of lengthening of a before ng, as seen in long (OE lang) and strong (OE strang). It seems that the vowel was shortened from influence of trisyllabic forms such as hangede.

Notes on inflected forms:


 * Oddly enough, the past tense of beat is spelled not as beet (from OE bēot) but as beat (beat and beet were formerly pronounced differently); it may be that that the infinitive and the past participle (which had the expected form, beaten) influenced the past tense form.
 * The past tense of hang was hēng in OE, and it naturally became hing later (e regularly changes to i before ng). However, the Old Norse verb hengja came into northern dialects, whence those dialects used hing as the infinitive instead of hang. By analogy with Class 3 verbs like sing, hang was used as a past tense, and hung was formed by analogy. Hung later entered standard speech, whence the current conjugation is hang, hung, hung.
 * The old past participle of hold is found in the adjective beholden, which has a different meaning from that of behold now.
 * Fell and held show vowel shortening, as the OE forms had ēo.

Notes on infinitive forms:


 * Let shows vowel shortening, likely because of the final dental consonant. Though this is not shown in read, it is present in dread, etymologically a derivative of read.
 * Read comes from OE rǣdan, the West Saxon variant, which is somewhat interesting as the modern form of words generally comes from the Anglian variant instead. In this case, the Anglian variant was rēdan, which would have yielded reed.
 * Shed is not the expected reflex of OE sceādan since the e was apparently only orthographic and was there to show that sc represented /ʃ/. The expected reflex would then be shode, which is attested in ME as shoden. One possible explanation is that the past tense form (OE sceād and scēd) became weak and yielded Middle English forms such as shadde/shedde. This would resemble past tense forms such as ladde/ledde, the past tense forms of ME leden (lead), and by analogy, a new infinitive sheden was formed. This would have become shead (rhyming with mead), but the vowel was shortened later on.
 * Swoop was swāpan in OE, and in ME, because the preceding consonant cluster had w, the vowel was further shifted such that the word became swoop. The same change can be seen in two (OE twā) and who (OE hwā).

Notes on inflected forms:


 * The past tense of ban had a long vowel in OE, but it probably would have been shortened later from influence of forms showing gemination.
 * Even in Old English, read had already come to be used as a weak verb; seldom did the strong forms appear.
 * The past tense of let and shed show vowel shortening, probably because of the final dental consonant and by analogy with weak verbs with the same vowel in both the infinitive and the past tense, e.g., spread, set, shut.
 * Wax was, in Proto-Germanic, a strong verb of Class 6, but in the West Saxon dialect of Old English, it came to be conjugated like a Class 7 verb; the old past tense form, wōx (found in awōx, the past tense of aweaxan), was confined to the Northumbrian dialect.

Weak verbs made strong
There are some weak verbs that have become strong by analogy.

Notes:


 * The past tense of Old English hȳdan was hȳdde, which naturally became hid. But hide was later felt to be a strong verb by analogy with Class 1 verbs like slide and bite, so the strong past participle hidden was formed, overthrowing the weak past participle.
 * The same process appears to have happened for chide. Interestingly enough, even though chide formerly had chode as an analogical past tense, hide does not seem to have ever developed hode by analogy.
 * Dig and stick are grouped with Class 3-1 verbs, though they do not end with a nasal consonant. It is noteworthy that they both end with velar consonants.
 * The history of the strong forms of stick is somewhat complicated; originally, in Middle English, it formed strong forms by analogy with Class 4 verbs, and these forms are the ones used today. But there was also another verb, ME steken, which had strong forms based on Class 5 and with which stick had been often confused. In the end, ME steken became disused in standard speech, and stick ended up using stuck as the past tense and the past participle.
 * Stuck seems like a form analogical to Class 3-1 forms such as stung and clung, and it is certainly now interpreted as such, but according to Hereward Thimbleby Price, it seems likelier that it was a natural development of variants of the older Class 4 forms, stoke and stoken, and such development can be found in blood and flood.
 * Get itself is a borrowing from Old Norse geta, which was also a Class 5 verb, but the native cognate is attested in derivative words in Old English and would have become yet, whence the uninfluenced forms of forget and beget would be foryet and beyet.
 * Give is from Old Norse, since the native form would have become yeave.
 * The archaic past tense of show, shew, is not to be confused with shew, an archaic spelling of show.
 * Though showed is used for the past participle by some people, the much preferred form is now shown.
 * String gained strung as its past tense and past participle; the past tense form strang was formed much later and is used only in some dialects.
 * Wake is a rather complicated case. In OE, it is attested only in the form of the past tense form wōc; the expected infinitive, wacan, is unattested. Hence, it is sometimes said that it serves as the past tense of the otherwise weak verb wæcnan (which later became the modern verb waken). In Middle English, there is a verb waken (which later became the modern verb wake), and it has the strong past participle waken. These forms, if not reflexes of unattested OE forms, were probably made by analogy with Class 6 verbs such as shake. As reflexes of the infinitive are also from OE wacian, for a long while, the verb could be conjugated weakly, but nowadays, the strong forms are generally used.
 * In OE, the verb was used mainly in the sense of come into being (but the sense of become awake is found in the past tense form awōc). The related weak verb wacian meant stay awake. The two verbs merged into one in Middle English, and the meaning of stay awake was later lost, but it is still found in phrases such as waking hours.
 * The original past participle, waken, survives in the adjective awake, in which the n has been dropped from awaken.

Notes:


 * Dive and sneak are the best known examples of recent irregularization of verbs.
 * Around the 19th century, dive became strong by analogy with verbs of this class and gained the past tense form dove, which is now widespread in American English (though dived is still common and always acceptable). Those who use dove as the past tense of dive keep dived as the past participle, however.
 * Sneak has snuck as an inflected form in American English (though sneaked is still common and always acceptable), but the root thereof is not certain, since there is no clear analogy for the shift from sneaked to snuck.
 * Proven was formed from preve, an alternate form of prove that was used in Middle English. Preve died out in standard speech, but survived in Scotland, and proven first appeared in Scottish legal contexts. The Scottish pronunciation has proven rhyme with woven, but in all other areas, because prove is used as infinitive, the vowel was changed to match the infinitive's. Hence, it looks as if it had been formed by analogy with verbs like hew.
 * The past participle is used only for the base verb and disprove; approve, though etymologically related to prove, is conjugated regularly.
 * In Old English, there was a weak verb spǣtan, which meant the same thing as spit and whose past tense was spǣtte. The past tense naturally became spat, which was later reinterpreted to be the past tense of spit by analogy with the Class 5 verb sit, and the original verb whence spat was gotten became obsolete. A past participle spitten was later formed, but it is now dialectal.
 * Strive may be found as a regular verb, but traditionally, the strong forms are preferred.

Notes:


 * The Old English weak verb rignan (meaning rain) had rīnan as a variant, whence came the analogous past tense rān (which is attested only once; the analogous past participle is unattested). In Middle English, the verb became rinen, and the past tense rone (the weak past tense died off in early Middle English); it is a rare instance of an Old English weak verb's conversion to a strong verb. As for the verb, it died off in the 15th century.
 * Reap comes from Old English repan, a variant of the Class 1 verb rīpan. It came to have Class 5 (and later Class 4) forms by analogy, though in the end, it became a regular verb. The original strong verb became ripe and is now used only in a few dialects as a regular verb. In Old English, repan is used far less than rīpan, attested mainly as the inflected form reopaþ, and strong forms appear found only once as rǣpon, the past plural. The strong past participle begins to appear only in Middle English.
 * OE swīþan (strengthen) was generally conjugated as a weak verb, but a strong form swāþ is attested once.
 * The following regular verbs gained strong forms in Middle English, but these forms are no longer used (except possibly in some dialects):
 * Arrive (French) - past tense: arrove, past participle: arriven.
 * Ding (as in strike, possibly Norse) - past tense: dang, past participle: dung.
 * Quake - past tense: quook. The analogous past participle, quaken, is unattested.
 * Snow - past tense: snew, past participle: snown.
 * Occasionally, in poetry and older usage, -en is used instead of ed to form a past participle with weak verbs, e.g., wreathen, paven, loaden (by analogy with laden), rotten (truly a Scandinavian loanword and only an adjective in modern standard speech).

Modern grouping of strong verbs
Obviously, it is no longer apparent that all these different strong verbs once belonged to seven different classes. From a modern point of view, the strong verbs can be grouped differently. Standard American pronunciation is used.

Mixed verbs (verbs that have either a past tense or a past participle that is clearly not strong):

The overall changes are:


 * Consonantal differences between the infinitive and inflected forms have been thoroughly wiped out.
 * OE drīfan had [v], but drāf, the past singular, had [f] instead. The reason is that [v] was simply an allophone of /f/ in Old English. Later on, [f] and [v] became treated as separate phonemes, and so the consonant in the past tense form was changed to conform with that in the infinitive.
 * OE frēosan shows a seemingly odd consonantal difference in frure (the past plural) and froren (the past participle). The reason for this is due to Verner's law. Of course, later on, this difference was done away with, and so we now use frozen rather than frorn as the past participle.
 * The initial consonant for some strong verbs differed because of palatalization. For example, OE beginnan showed /j/, but the past tense (begann and begunnon) and the past participle begunnen showed /g/ instead. Later on, the unpalatalized forms came to influence the infinitive, whence the verb is now begin rather than beyin. For OE cēosan (choose), however, the old strong forms were influenced by the consonant in the infinitive, and so the strong forms became chose and chosen.
 * The past tense no longer shows a difference in number. Hence, sang (previously the past singular only) came to be used for the past plural as well.
 * For the second-person singular (thou), the past tense formerly did not use the characteristic second-person singular suffix -est; in OE, the past tense of singan for the second-person singular was sunge (sharing the same vowel as that in the past plural). Later on, however, because of the strong association between thou and -est, the past tense of all strong verbs came to use -est. Hence, the past tense form became sangest.
 * Alterations to inflected forms due to other sound changes were often reverted by analogy. For example, for the Class 3-1 verb sing, the OE past tense form sang yielded song in ME because of vowel lengthening before ng (and indeed, the related OE noun sang became modern song). However, analogy with other Class 3-1 verbs caused sang to overthrow song.
 * The strong past participle no longer consistently has the -en ending.
 * Class 3-1 verbs and a few other strong verbs no longer use -en for the past participle, but some -en forms survive as adjectives, e.g., sunken, shrunken, beholden.
 * For many other strong verbs, past participle forms without -en were formerly used as well, but the -en forms eventually became the standard forms. Some forms without -en such as broke and bespoke survive as adjectives.

Archaic, dialectal, or obsolete verbs
Notes:


 * The dialectal verbs sie, thee, and wry were, in Old English, sēon, þēon, and wrēon, all of which were contracted verbs that originally had the same vowel as this class's other verbs and were often conjugated like Class 2 verbs because of the new infinitive vowel. OE sēon and wrēon would have normally become see and wree, but they were influenced by their other forms.
 * Sike, now a dialectal word meaning sigh, gained the irregular weak form sight by analogy with verbs like teach and buy. Sigh is likely a back-formation from sight (which form is now obsolete).

Notes:


 * The obsolete verb leese (OE lēosan) is now found only in the adjective forlorn, which is the past participle of forlese (forsake) and was affected by Verner's law. If the verb had stayed in regular use, the strong forms would have likely been changed by analogy and would have become loze and lozen.
 * The irregular verb lose is generally derived from Old English losian, but some trace it back to OE lēosan instead. The explanation is that as cēosan and scēotan became choose and shoot from altered stress of the diphthong, so did lēosan become lose; normally, cēosan and scēotan would have yielded cheese and sheet. This shift in stress is also seen in words such as show (OE scēawian), troth, originally a variant of truth (OE trēowþ), and hoo, a dialectal variant of she from OE hēo.

Notes:


 * The dialectal verb frain is from Old English frignan, which also had byforms because the g often disappeared. One byform was frīnan, which was partly conjugated like a Class 1 verb.
 * Though swelt is no longer used in the standard speech, it lives on in the frequentative verb swelter.
 * The archaic verb worth can still be found in the literary phrase woe worth (wherein worth is subjunctive, and the phrase thus means woe be to). The pronunciation /wərθ/ is a spelling pronunciation; as a verb passed down from Old English, the dental fricative would naturally be voiced, and so if the verb had stayed in general use, the expected pronunciation would be /wərð/ (barring other possible sound changes such as /ð/ sometimes shifting to /d/ after /r/, as seen in burden and afford).

Notes:


 * The dialectal verb heal began to be conjugated regularly in Middle English, mainly because the strong verb helan had merged with the weak verb helian.
 * The past participle of nim lives on in the adjective numb (the b is unetymological and not pronounced, and the past participle ending -en has been dropped). The old past participle (OE numen, often spelled as nomen in Middle English) had the same vowel as in come (OE cuman, past participle cumen).
 * In OE, niman had a variety of forms for the past tense. One form was nam, which was etymologically the original form. Another form was nōm, which came about by analogy with the plural form nōmon.
 * Nim, by some later writers, was conjugated like a weak verb, since by that point, the verb had fallen out of common use. If the strong forms had survived, they would have likely become nam and num, since it would have also been strengthened by analogy with Class 3-1 verbs.

Notes:


 * Yet was used only in derivatives in Old English, and the unprefixed verb get came from Old Norse. There are a few instances of yet used without a prefix in Old English, but they are likely shortened forms of prefixed verbs rather than a genuine survival of the base verb. In Middle English, there are also a few instances of lone yet, but they are likely the Norse verb influenced by derivatives. Perhaps even without the introduction of the Norse verb, the base verb yet would have been coined as a backformation from a prefixed verb such as beyet, which formerly meant obtain. This kind of backformation has happened before, as seen in gin, an obsolete variant of begin.
 * On the whole, yet shows Class 5 forms; seldom were there forms based on Class 4 forms, and even in those instances, they were likely from influence of the Class 4 forms of get.
 * The Old English word for be had wesan as one of its infinitives. As wesan was originally a Class 5 verb, the current past tense forms (was and were) are based on Class 5's pattern.
 * That the past indicative forms show an s/r difference is due to Verner's law, and be is the only current verb that shows a noticeable consonantal difference in the past tense with regard to number. Verner's law also explains why were is used for the singular in the past subjunctive.
 * In Middle English, the verb steken (which meant pierce) was conjugated like a Class 5 verb. It is attested no earlier than that, and so it may have come from an unattested Old English verb stecan; at the very least, it is Germanic and akin to German stechen and Dutch steken. It was often confused with the verb stick, and so the strong forms were also mixed up with the Class 4 strong forms of stick.

Notes:


 * Graven and laden are still used as adjectives.

Notes:


 * The latest form of the strong past tense of the dialectal verb fang was feng, which would have become fing later. Fang and hang were also parallel in conjugation in Old English, both verbs being contracted verbs.
 * The same would have also happened for the dialectal verb gang; the latest past tense form was geng (from which we would have gotten ging), though gang was not parallel in conjugation with hang and fang in Old English.