Umlaut

This page lists words gotten from others through a sound change called umlaut (also known as i-mutation). An example of a word gotten from umlaut is feed, which is clearly linked to food. Another example is mice, the plural of mouse.

What exactly caused this change in vowel? To put it simply, the change happened whenever the following syllable had /i/ or /j/. For example, mice in Proto-Germanic was mūsiz. In the transition to Old English, the z was lost, and later, because of the following i, the main vowel was changed, which led to mȳsi. The final i was later dropped, which led to the Old English form mȳs.

Many remnants of umlaut can be found in the present speech. Because sound changes have beclouded many of the original links, the Old English forms are given as well. All forms that are no longer used in current standard speech are put in bold. A (*) shows that a form has been reconstructed and is not attested in New English.

By the way, the word umlaut is from German; a calque thereof would be umbloud, wherein umb means around, and loud means sound.

Vowel changes
The changes gotten from umlaut can be summarized thus:

Notes:


 * OE a before the nasal consonants (m and n) became e instead, e.g., mann (human) - menn.
 * Theoretically, the change of non-nasal a to æ should not have happened since a sound change called Anglo-Frisian brightening caused all instances of a to change to æ (including forms where umlaut would later happen). However, a was often analogically restored, whence all instances of non-nasal a to æ are analogical. Note that this applied only to the short vowel; the change of ā to ǣ was not unusual.
 * Theoretically, the change of e to i should not have happened since in Proto-Germanic, e was changed to i whenever it occurred before i or j (in other words, an early form of umlaut). It is possible that e was analogically reintroduced in forms where umlaut would later apply, but there is no clear evidence of changes of e to i that happened in OE.
 * Examples of the change of o to e are rare since Proto-Germanic did not have o; instead, o came from Proto-Germanic u. However, Proto-Germanic u was kept in forms where umlaut would later apply, which explains OE words with o had y from umlaut, e.g., OE gold and gyldan (gild).
 * One rare case of o to e was in OE dehter, the dative singular of dohtor (daughter). This is quite unexpected, since the Proto-Germanic forebear of dohtor had u, so the vowel in Old English would have changed into y, which would have yielded the form dyhter. However, sometime before umlaut, the o in dohtor was analogically introduced to the dative singular form, which is why the vowel in that form mutated into e.
 * Note that this applied only to short o; the change of ō to ē was not unusual.
 * The change of ea to ie was present only in the West Saxon dialect; in other dialects, the umlauted vowel became e instead. Since our modern standard speech is mainly based on the Anglian dialect, the mutated forms that we have inherited come from the Anglian variants. For example, the West Saxon verbal derivative of tēam (offspring) was tīeman (give birth), but the Anglian variant was tēman, whence our modern word teem.
 * The change of eo to ie is slightly trickier since there are two different sources of eo: original eo and io.
 * Theoretically, the change of original eo to io should not have happened for the same reason that there is no clear evidence of the change of e to i in OE. However, it is possible that eo was analogically reintroduced in forms where umlaut would later apply. Nonetheless, io became eo in most later variants of Old English, which thus undid the change.
 * Original io changed to ie through umlaut. However, the original vowel io changed to eo in most later variants of Old English, which created alternations such as frēond (friend) and the plural form frīend.
 * The mutated vowels ie and īe later changed into y (or i) and ȳ (or ī), respectively. The later forms using y will be used here.
 * Derived forms characteristic of West Saxon are labeled WS. The Anglian variants can be inferred from those. In this instance, NE forms (be they reconstructed or not) are based on the Anglian variants (unless noted otherwise).
 * In a few cases, the form of either the umlauted derivative or the original word ended up taking over the other. For example, kiss was originally the verb gotten from a noun that would have become coss, but the form later became used for the noun as well. Here, the normal reflexes are recorded instead of the analogous ones.

Plural forms
Technically, the vowel change was originally not only a characteristic sign of the plural. In Old English, the vowel change was also found in the dative singular, and not all case forms in the plural showed such change. Nonetheless, this vowel change is now seen as a sign of the plural, so for convenience' sake, it is treated as such for the Old English forms.

Notes:


 * The umlauted plurals of oak, book, and borough had undergone palatalization in OE. Of course, it is possible that if they had survived, they would have been influenced by the consonant in the singular later on; for example, the modern umlauted plural of oak might be eak instead.
 * OE brēc (which meant breeches) was the plural of the unattested brōc and later became breech. The form survived for some time, but is now wholly displaced by breeches, a double plural.
 * OE brōþor (meaning brother) generally showed mutation only in the dative singular; the form brēþer was seldom found in the plural and appears to have been formed by analogy with nouns with vowel mutation in both the plural nominative and the dative singular. The mutated plural, brether, later was used with the -en plural ending, which yielded brethren, a double plural, and it is now used as the plural of brother only in some meanings.
 * OE byrg lives on as bury in place names such as Canterbury. In these names, however, byrg represented not the plural but the dative singular, as many place names were often used in the dative.
 * The mutated plural of cow became kye, which is still used in some dialects. In standard speech, the -en plural ending was later added, which yielded kine, a double plural, but it is now archaic.
 * OE ding, a variant of dyng, was the dative singular of the unattested dung (meaning dungeon).
 * OE fyrh and grȳt were the dative singular forms of furh (furrow) and grūt (coarse meal; the word later became grout).
 * OE þrūh (meaning pipe or tomb) showed the expected vowel change in þrȳh, the dative singular, though it is not attested in the nominative plural. The word later became through and is found in some dialects with the meaning of horizontal gravestone.
 * Woman (OE wīfmann) is etymologically a compound of wife (originally meaning woman) and man (originally meaning human), but the present form is special, as the plural shows a vowel change in the first syllable. This change, however, is not an example of umlaut. Originally, one variant of woman was influenced by the foregoing w, and the vowel later became /ʊ/, whereas another variant remained uninfluenced, and the vowel later became /ɪ/. The two variants were used together for some time before the /ʊ/ variant became restricted to the singular, and the /ɪ/ variant to the plural (perhaps by analogy with pairs such as mouse-mice).

Verbs
Verbs gotten from nouns:

Notes:


 * The OE verb ǣlan has not survived in its base form, but it lives on in the derivative anneal (OE onǣlan).
 * Boot meaning remedy is now archaic, but it is still found in the phrase to boot.
 * Kemb, the verbal derivative of comb, is still found in the past participle adjective unkempt (untidy, disheveled); modern kempt is a backformation from unkempt.
 * The modern word dot seems to have come from Old English dott (meaning head of a boil), though it is not attested at all in Middle English and begins to be attested again in the late 16th century. The dialectal verb dit meaning close is from Old English dyttan and may be derived from dott.
 * It is unclear whether OE drēam is the source of the modern word dream, as the meaning of dream is unattested in OE (though it is still possible that it was present, but simply unattested). The meaning of sleeping vision begins to be attested in Middle English and was due to or strengthened by Old Norse influence.
 * OE flēdan is attested in the derivative oferflēdan (overflow) and the adjective fiþerflēdende (flowing in four parts). The base verb is attested in Middle English as fleden.
 * OE ǣþan, hendan, myldan, and myrþran are attested only in geǣþan, gehendan, bemyldan, and formyrþran.
 * The modern word murder shows a consonantal shift that was likely strengthened by Anglo-Norman forms; the shift may still be native, however, as seen in burden, OE byrþen.
 * The word sale corresponds to sell as tale does to tell, but sale is a borrowing from Old Norse rather than an Old English word inherited from Proto-Germanic. However, it seems that if the word had been passed on to Old English and had survived up to New English, it would have still become sale.
 * OE þeccan would have normally yielded thetch (which variant lived up to Early New English), but the verb's vowel was later influenced by the noun's vowel, so the verb became thatch. The noun also became thatch from influence of the verb.
 * The vowel in tellan is based on an older form that had æ.
 * The OE verb wyrcan would have normally become worch, but inflected forms showing lack of palatalization and the noun work later influenced the verb.

Verbs gotten from adjectives:

Notes:


 * The modern word blank is not a survival of the Old English word but a borrowing from Old French, but the French borrowing is from the same Germanic source.
 * The derived blench now is an intransitive verb meaning flinch, which seems to be due to influence from blink.
 * The word couth used in standard speech is not a survival of the original adjective but a backformation from uncouth, which originally meant unknown and now means unrefined.
 * The verb file meaning corrupt is no longer used, but it is commonly found in the derivative defile.
 * OE hlēow is attested only in derivatives such as unhlēow (chill). The adjective later came to have the meaning of tepid, and it may be related to the first element of lukewarm.
 * OE hlencan is attested only in gehlencan.
 * OE sōm is attested only in gesōm (unified).
 * The modern verb seem is a borrowing from Old Norse, but is ultimately connected to the adjective, as it originally meant be suitable and gradually came to mean appear.
 * OE strengan is attested only in gestrengan and ætstrengan (deforce).

Nouns
Nouns gotten from adjectives and nouns with the addition of the -th suffix.

Notes:


 * The OE noun fǣhþu would have become faught; the cluster hþ became ht in Middle English, as seen in height (OE hēhþo) and sight (OE sihþ), and the vowel would have changed to yield faught, as seen in the Middle English word aughte (OE ǣht meaning possession).
 * Sleuth was later replaced by sloth, which was formed directly from slow in Middle English.

Other nouns:

Notes:


 * Bread was later replaced with breadth by analogy with nouns such as depth, strength, and length. It is still found in waybread, a name for the Eurasian plantain.
 * Proud and pride are clearly related, and their OE forms are prūd and prȳde (more commonly prūt and prȳte), but since proud is said to have been a late Old English borrowing from Old French, it seems that pride was formed by analogy with umlauted words inherited from Proto-Germanic.

Feminine nouns made with the -en suffix:

Notes:


 * The only remnant of the Germanic feminine ending is vixen, which is the feminine of fox and is gotten from the unattested OE fyxen. The word would have normally become fixen, but in the southern dialects, the initial consonant was voiced (which is also seen in vat and vane), and for whatever reason, vixen ended up becoming the standard form.
 * The feminine of wearg is attested only in grundwyrgen, which is used in Beowulf to describe Grendel's mother. Also, wyrgen shows palatalization, whence the consonant in the feminine differs from that in wearg.

Nouns (instrumental or diminutive) made with the -le suffix.

Notes:


 * OE bydel was later influenced in form by Old French bedel, whence the current word beadle.

Adjectives
Comparative and superlative forms:

Notes:


 * The eo in geong appears to have been not a true diphthong but an orthographic convention that stood for u and showed that the consonant was a palatal; had it been a true diphthong, the word would have later become yeng instead. The mutated forms showed i instead of y since i apparently was commonly substituted for y after palatal consonants.
 * Likewise, the e in sceort appears to have been orthographic.
 * The comparative OE grēat is attested as grȳtra (the West Saxon variant), but the superlative only begins to be attested in Middle English. Presumably, the OE superlative was grȳtest.
 * The mutated comparative of OE brād (broad) is attested as brǣdra, but the mutated superlative brǣdest is unattested (but the regular superlative is).
 * OE nēah (nigh) had a mutated superlative nȳhst, and the Anglian variant later became next. The comparative was nēarra and not nȳrra, but the comparative form of the adverb is attested as nȳr, however.

Adjectives made with the -en suffix:

Adjectives made with the -ish suffix:

Notes:


 * OE Angel is attested only as a prefix meaning Anglo-.
 * OE Frencisc would have normally yielded Frenchish, but the ending was later dropped.
 * OE myrcisc (Mercian) might have later undergone shortening and become something like Merch instead. Likewise, OE Scyttisc could have become Skitch, just as Scottish has Scotch as a (now dated) contracted form.

Adjectives made with the -y suffix:

Ablaut derivatives
See here.