One mouse, two mice; a layman's intro to Germanic umlaut
While speaking English, you may have noticed a strange set of irregular plurals. One mouse, but two mice? One foot, two feet? This is the result of a phenomenon known as Germanic umlaut; when vowels are either raised, fronted or both. But what does this mean?
Oversimplified introduction to the IPA
The IPA (International Phonetic Alphabet) is what linguists use to notate sounds in human languages accurately. This article will use IPA notation surrounded in slashes. Here’s what the whole table looks like:
The columns, on the contrary show the place of articulation, where the sound is produced. The first column consists of the bilabials; consonants pronounced by using both lips. In cells with two consonants, the left is voiceless and the right is voiced, but for the sake of this article we’ll skip over that distinction. Let’s compare two sounds:
- /p/: voiceless bilabial plosive
- /m/: voiced bilabial nasal Both are pronounced by closing both lips together, but /p/ is created by almost hitting your lips together, whereas /m/, a nasal is created by closing your lips and allowing airflow through the nose.
But what we’re here for are the vowels.
Proto-Germanic..?
In linguistics, linguists have theorized that the ancestor of English, among many other languages is Proto-Indo-European.
Out of these languages, there is a branch known as the Germanic languages, seen in red on this map. By comparing Germanic languages, such as English, German and Norwegian, linguists have created a theoretical model of what they believed the Proto-Germanic language — ancestor of all Germanic languages — would’ve sounded like.
There is a major misconception amongst people that English is descended from Latin or French, but this is not the case. Its grammar and core vocabulary is inherited from Proto-Germanic, which can be seen by its similarities to other Germanic languages in that regard. It has a large portion of vocabulary borrowed from Latin or French — nearly 60%, in fact — particularly borrowed during the Norman Conquest throughout the 10th to 12th centuries. This can be seen how vocabulary considered to represnt more “basic” or “core” concepts tend to be Germanic (i.e. breathe vs respirer, eat vs manger), whereas more “advanced” concepts tend to be borrowed (i.e. language vs langue, calculate vs calculer).
Note that throughout this article, Proto-Germanic words will be prefixed with an asterisk to denote that these are unattested (no records of it existing, and are pure reconstructions).
Back to plurals and vowel raising.
Proto-Germanic (BC 500 - AD 200)
A commonality between all these vowel-swapping irregular plurals; mouse/mice, foot/feet and tooth/teeth; these all end in consonant sounds. In Proto-Germanic, nouns ending in consonants other than /n/, /r/ and /z/ were considered to be consonant stems, and took the nominative plural suffix of *-iz. (Nominative here refers to when a word is the subject/actor of a verb.)
| English | Proto-Germanic Singular | Proto-Germanic Plural |
|---|---|---|
| mouse/mice | *mūs /muːs/ | *mūsiz /muːsiz/ |
| foot/feet | *fōts /ɸɔːts/ | *fōtiz /ɸɔːtiz/ |
| tooth/teeth | *tanþs /tɑnθs/ | *tundiz /tundiz/ |
Note that þ (/θ/) is pronounced the same as the English th, as in thin, and /ː/ represents a long vowel.
In Proto-West Germanic, the ancestor of languages including English, the final *-z on polysyallbic (more than one syllable) words was lost. By Proto-West-Germanic, the three words had shifted to these:
| English | Proto-Germanic | Proto-West Germanic |
|---|---|---|
| mouse, mice | *mūs, *mūsiz | *mūs, *mūsi |
| foot, feet | *fōts, *fōtiz | *fōt, *fōti |
| tooth, teeth | *tanþs, *tundiz | *tanþ, *tanþi1 |
(1 *tundiz is an irregular inflection of *tanþs; somehow it re-regularizes into *tanþiz > *tanþi??)
Old English (AD 475-900)
Across languages, vowels and consonants are often influenced by the syllables surrounding it, a phenomenon known as assimilation. We saw this in the previous article with palatalization; English, along with most Germanic languages underwent umlaut; or the fronting and raising of vowels.
In Late Proto-West Germanic (with the exception of the Gothic branch), the vowel of the penultimate, appearing directly before the *-iz suffix undergoes umlaut, influenced by the succeeding /i/ sound, in a phenomenon known as i-mutation. In Old English, the vowels shifted as such:
The final -i was then lost after the vowel was weakened. At this point, the /ø/ vowel unrounds to form /e/. In Old English, the words have now evolved to form:
| English | Proto-West Germanic | Old English |
|---|---|---|
| mouse, mice | *mūs, *mūsi | mūs, mȳs |
| foot, feet | *fōt, *fōti | fōt, fēt (ø > e) |
| tooth, teeth | *tanþ, *tanþi | tōþ, tēþ (ø > e) |
Middle English (AD 1400-1700) to Modern English (1700-)
Finally, we reach the Great Vowel Shift in Middle English. In Middle English, /y/ was unrounded to /i/. Futhermore, the distinction between long and short vowels are lost in some cases due to diphthongization. This is demonstrated by the below diagram (each vowel only shifts once):
- /yː/ > /iː/ > /aɪ/
- /eː/ > /iː/
- /uː/ > /aʊ/
- /oː/ > /uː/ > (in some cases) /ʊ/
Therefore creating:
| Old English | Middle English2 | Modern English |
|---|---|---|
| mūs /muːs/, mȳs /myːs/ | mus /muːs/, mys /myːs/ | mouse /maʊs/, mice /maɪs/ |
| fōt /foːt/, fēt /feːt/ | fot /foːt/, feet /feːt/ | foot /fʊt/, feet /fiːt/ |
| tōþ /toːθ/, tēþ /teːθ/ | tooth /toːθ/, teth /teːθ/ | tooth /tuθ/, teeth /tiːθ/ |
2 There’s a lot of variation in spelling for Middle English. The word fot has also showed up as foot, foit, fod, fut, foȝt and so on, due to a lack of standardization.
And that’s how the irregular plurals were formed!
Wait, then wouldn’t we have a lot more words with irregular vowel plurals..?
If every consonant resulted in this kind of vowel shift, we’d have a lot more plurals that alternate based on its vowel, right? Well, that was exactly the case in Old English! Take the word book for instance, bōc in Old English, with the plural bēċ (the ċ is pronounced as the Modern English ch), with the final c palatalized to form ċ by the preceeding e. If we follow these steps, in theory we should have the pair book and beech. The same with the word cow; with the pair cū and cȳ, which should yield cow and cee.
The main theory is that although these words attained an irregular vowel plural, in Modern English it was converted back into a regular -s plural through following the pattern of the majority of other nouns. A pair like “book” and “beech” would seem strange to native speakers, and they would prefer a regular plural like “books”.
How about verbs like sing, sang sung?
That’s the result of another phenomenon; Indo-European Ablaut, which I would like to cover in another article, but does require more explanation of Proto-Indo-European, so I’ll leave it here.
No comments yet!