The rule for regular past tense endings in English is to add the phoneme /d/ to the base of the word:
- rule /ru:l/ --> ruled /ru:ld/
If the last sound in the base is unvoiced (doesn't involve vibration of the vocal cords), then we use an unvoiced version of the suffix, namely /t/ to match it:
- tip /tɪp/ --> tipped /tɪpt/
However, some bases already end in /d/ or /t/. In such cases if we added another /d/ or /t/, the suffix would be inaudible. In any case, English does not allow geminate (double length) consonants, so such a word would be ungrammatical. We couldn't have the following for example, as the past form of mend:
or the following as the past form of the verb fit:
When the base of a regular verb ends in /d/ or /t/, to make the past tense form we insert a vowel, /ɪ/ before the suffix. Because this vowel is voiced (all vowels are voiced in English), the consonant after the vowel will be a /d/:
- /mendɪd/ mended
- /fɪtɪd/ fitted
The insertion of this vowel now means that the past tense suffix is clearly audible and prevents the words from breaking the phonotactic rules of English. Of course, the insertion of the vowel also results in the creation of an additional syllable.
Notice that this does not affect how we spell the word. In the spelling we just add the normal -ed written suffix to the base of the word regardless of whether the base of the word ends in /t/ or /d/. So the -ed ending results in the addition of a single phoneme but no extra syllable in most cases, but results in an additional syllable if the base ends in /d/ or /t/.
In some varieties of English the vowel which is inserted may be a schwa, /ə/ instead of the KIT vowel /ɪ/.