7

I saw a spelling mistake on an SO question: submittion. That got me wondering, is there a name for the shift of ‑mit‑ to ‑miss‑ in submission, permission, admission and so on? Are there other patterns like this in English words? (Not simple stuff like dropping an e for confuse/confusion) and if so, are these patterns related to the language we took the word from?

5

Shifts of this sort are very common in words of Latin derivation. In the case you cite, missus is the participle of mittere, and the nominal forms of admit, commit, emit, permit, remit all take the terminal -ssion form. (Promise represents the same thing, though in this case the verb is a back-formation from the participle used as a noun, promissum, and there is no promit, at least today.)

Latin underwent many sound changes which caused stems to have different terminal vowels in different conjugational and declensional contexts. Here's just a handful of examples from what must be hundreds of words represented in English:

  • from claudere, p.ppl. clausum ... include, inclusion (and clause)
  • from miscere, p.ppl. mixtum ... promiscuous, mixture
  • from videre, p.ppl. visum ... video, visa
  • from adolescere, p.ppl. adultus ... adolescent, adult
  • from corpus, gen. corporis ... corpse, corporal
  • from genus, gen. generis ... genus, general
  • from magnus, comp. maior ... magnitude, major

In other cases, such as those pointed out by jlovegren, the change took place after the word was borrowed into another language and involved only one of these consonants — but that one took on two different pronunciations, depending on the phonological environment: regal/regicide, for instance. Sometimes English has borrowed a stem twice, once from Latin and once from its French daughter: alongside regal and regicide we have royal — and in that particular case we also have a descendant from an entirely different line of descent, rajah.

And similar pairings occur in “native” (Germanic) English words: was/were is the most common example, but consider also life/live, think/thought, lose/forlorn.

  • nice example of rhotacization in was/were. – jlovegren Dec 22 '12 at 17:17
  • @jlovegren my all-time favourite. – StoneyB Dec 22 '12 at 17:18
  • And pairs like star and stellar are related because they both came from Indo-European, one through Old English and the other through Latin. – Peter Shor Dec 22 '12 at 23:57
5

For background, see the Wikipedia entry on sound change. For explanation of the terms referring to consonant types, see the entry for obstruent.

Compare the latin admittere and admittionem. In the latter form the coronal stop [t] is followed by a high front vowel. A stop--high vowel sequence is one of the most common sources for a consonant mutation (generally, a sound change where a consonant's sound value is altered due to effects of a neighboring vowel). Well-known examples of consonant mutation are found in Celtic and in Bantu languages.

When a stop is changed to a fricative, the change is usually referred to as lenition (weakening). More specifically, when a stop changes to an affricate or fricative, it is called spirantization. Most specifically, when a stop or fricative preceding a high front vowel/semivowel undergoes change, it is called palatalization.

Some other alternations involving Latinate words which presently show a palatalization alternation:

  1. submit/submission (t-->sh)
  2. electric/electricity (k-->s)
  3. affect/affection (t-->ch)
  4. confuse/confusion (z-->zh)

Note that in some cases the alternation is reflected in the spelling (as in submit/submission), but in others the same consonant letter is used with different sound values (as in confuse/confusion). As a commenter points out, submissio was already spelled with an s in Latin.

  • @StoneyB indeed it is (just checked). let me throw in a qualifier. – jlovegren Dec 22 '12 at 16:47
  • Yeah, Grimm's Law, for example, consists of a dozen individual lenition changes resulting in a regular pattern. – John Lawler Dec 22 '12 at 19:13
1

As StoneyB says, this is a common consonant shift between different Latin stems from the same root.

Its origins are quite old: it comes from what seems to have been a Proto-Indo-European phenomenon whereby a dental consonant, like /d/ or /t/, was converted to a sibilant affricate (like [ts]) before a suffix starting with a dental consonant. In Latin, these affricate-dental consonant clusters were further simplified to /ss/ or /s/ (generally /ss/ after a short vowel, and /s/ after a long vowel).

The Latin verb mittō of course has a root ending in /t/, and when you add a suffix starting in -t- to this it results in miss-, as in mission (the Latin source of -ion, the noun-forming suffix -iō, was typically preceded by -t-). There are a number of similar examples in English such as secede, secession; compatible, compassion; erode, erosion; invade, invasion.

Interestingly enough, this sound change was also inherited by the Germanic branch of Indo-European, and there was even a similar development to -ss-/-s- based on the length of the preceding vowel. We don't see these alternations in the inflectional paradigms of Germanic verbs, but we do see it in some other words such as the adjective wise (from Proto-Indo-European *weyd- + the participial suffix *-tos, the same source as the Latin word vīsus, the past participle of the word videō "to see", related to the Latin noun vīsiō which is the source of the English noun vision).

In general, these kinds of changes fall under the rubric of "morphology" and they are certainly dependent on the language of origin of the word. Latin or Greek-derived words in English show a number of morphological processes that don't generally apply in native vocabulary.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.