Why "admit" with T but "admissible" with SS? [duplicate]

I have noticed that when the suffix -ible is added to "admit", it becomes "admissible" rather than "admittible".

There are few other examples:

• "omit" = "omissible" not "omittible".
• "permit" = "permissible"
• "transmit" = "transmissible"

These are words that end with it and have ssi in the suffixed forms. I didn't find other words that show the same pattern. Most it words don't have ssi forms in suffixed words.

Also:

• commit = commission
• emit = emission
• admit = admission

I know that S and T are both alveolar sounds (produced at alveolar ridge) but this relationship seems to be confusing. I googled a lot but didn't find anything helpful. I asked some people and they told me "because Latin did it", but no one seemed to know the reason.

Does anyone know why the t changes to ss in these words?

Solution 1:

The etymological reason is because in Latin, words like omissible, permissible, transmissible were built on perfect passive participle stems, which were marked with a -t- suffix. Because of a sound change between Proto-Indo-European and Latin, the -t- suffix is not directly visible, but instead results in -ss- in the adjectives you mention, which are all derivatives of one Latin verb, mittō. There are other adjectives ending in -ible that instead show addition of unchanged -t-, as in perceptible, or a change to single -s-, as in divisible; whether the perfect passive participle stem of a verb has -t-, -ss- or -s- is usually regularly based on the phonetic context in Latin, although there are some irregularly formed perfect passive participles.

The relevant sound change is that when a stem ending in a dental obstruent in Proto-Indo-European (*t, *d, or *dʰ) is combined with a suffix starting with the dental *t, the two dental consonants are replaced in Latin with -ss-, which was further simplified to single -s- everywhere except for after a short vowel.

Sound changes are frequently not simple to explain. As you said, t and s are both produced with similar parts of the tongue.

The "sigmatic aorist" is a construction with a distinct origin that corresponded to a distinct Latin stem (the verbal perfect stem) that appeared in finite past tense forms of verbs. Its relevance to your question is indirect. Stems ending in a dental obstruent in Proto-Indo-European also developed forms with -ss- or -s- in Latin when combined with the *s suffix of the sigmatic aorist. Perhaps in part because of this, in Latin, as in English, there was sometimes analogical influence between the perfect stem used in finite verb forms and the non-finite perfect passive participle stem. However, in the case of mittō, the stems are in fact different (perfect mīsī vs. participle missus), and you can see that that the words you mention are formed on the participle stem, not on the aorist stem.

The Latin suffix -ibilis did not only appear after passive participle stems; it could also be added to other word forms. According to "The morphome vs similarity-based syncretism: Latin t-stem derivatives" by Donca Steriade,

adjectives like duct-i-bilis and duct-ilis ‘which can be led’ are based on a t-stem (cf. duct-us ‘led’) and denote only passive ability; only root-based duc-ibilis can mean ‘who can lead’.

(page 117)