In this question, asking why vowels get changed when prefixes/suffixes get added to a word, Aricauria states (in their very popular answer) that this is a side-effect of the way English organises stressed syllables in connected speech.
Aricauria does a very good job of describing what a stress timed language is. But the phenomenon of vowels changing when word endings change is not limited to stress-timed languages. This can be demonstrated by some simple examples of similar changes occuring in, for example, Spanish (a strongly syllable timed language.) This phenomenon is by no means exclusive to stress timed languages.
That said, everyone is having such a party over there talking about stress timing (Araucaria has over 250 upvotes, and the top comment "I'd never heard of stress timing before" has over 60.)
Under the circumstances, it seemed absurd to leave a comment suggesting that Araucaria edit his/her answer so I wrote a short, polite answer with some examples of similar vowel changes from Spanish (see below) to demonstrate this.
Despite gaining a modest number of upvotes, my answer has been deleted by a moderator. I fail to see how this improves the site. I think pointing out that such vowel changes also exist in non stress-timed languages was a useful addition.
Here is my deleted answer for reference.
Araucaria's answer is almost perfect. I just have one thing to add: this phenomenon is not limited to English, nor to stress-timed languages. In Spanish (a syllable-timed language) we find similar behaviour. Students of Spanish as a foreign language are exposed from a very early stage to "radical changing verbs" where the last vowel sound of the verb stem is stressed strongly in certain persons (and is a dipthong - usually) and weakly in other persons (and is a different, simple vowel sound.) Here are some examples, with the correct stress indicated by capitalization. (English meanings should be obvious.) The full present-tense conjugations are available here and in many other places.
Infinitive (to do) 1st person singular (I do) 1st person plural (we do) resolvER resUELvo resolvEMos preferIR prefIERo preferIMos repetIR repITo repetIMos
This also occurs (more rarely) where adjectives are derived from nouns, e.g. Venezuela -> Venezuelan:
VenezUELa -> venezolANo
While that is interesting it is not an answer to the question, nor related to English
To this I would like to respond: Actually it is about English, because it is pointing out that the observed phenomenon is not necessarily caused by stress-timing. If I had been more blunt, my answer might not have been deleted.
I just have one correction to makeinstead of
just one thing to add(though it remains deleted so you probably won't be able to see it.) I don't know if that'll be enough to get it undeleted. When I see people enjoying the discovery of stress-timed / syllable-timed for the first time, I don't want to rain on their parade. In his defence, the moderator went in to send the discussion about timing to Chat and clean things up a bit. I think he was wrong to delete my answer but indeed I may not have been clear enough about my reason for posting.
Venezuela -> venezolanothe fundamental reason (easier pronunciation) is the same, but the rules are different. In Spanish the affected vowel goes from stressed to unstressed, whereas in most of your English examples it remains stressed but gets shorter.