No one has addressed my question: why was so much of memini discarded? Remembering is a common concept. The only other defective I know is odi, another popular pastime.
If you look at the cognates in related languages, I think it is pretty clear that any hypothetical present tense of meminī would not have meant 'I remember'. For example, Skt.
मन्यते [manyate] means 'think'. The meaning 'remember' is thus in line with the normal semantics of the Indo-European perfect stem, which often denotes a state that is characterised by regular performance of the action denoted by the present stem. From regularly thinking of something to remembering it is an easy semantic development. (Resultative perfects seem to be a newer phenomenon; the resultative use of the perfect is rare in early Greek). What was discarded was then a present stem meaning 'think' or similar, though the causative formation
moneō 'cause to think of', 'remind' has been retained. Perhaps this would have given something like *
menior had it survived, but probably it was lost at a far-off prehistoric stage when Latin looked somewhat different anyway.
Is it related to memor? Compare tremo/tremor, caleo/calor and many other such - do they imply memo/memor or memeo/memor?
The older view was that
memor is from a different root, *
(s)mer- 'remind'. More recently scholars (e.g. Sihler) have reckoned with a fossilised perfect participle of
meminī, following the perfect participle in -
os- which might be familiar from Greek -ως, *
me-mn-os- > *
memnor- >
memor. This works if one assumes simplification
mn >
m (and then some sort of analogical lengthening of the vowel to ō). If this is the case, it also answers your question about the perfect participle.
