Why did I not think of addressing cognate suffixes (-ly and -mente) before?
Technically, the two suffixes, -ly and -ment(e), are not cognates. To be a cognate, two forms must be derived from the same common ancestor. If you like, Gothic leiks, Old English līk 'body, corpse' (which survives in some placenames, e.g., Litchfield, VA, in the USA), and German Leiche 'corpse' are cognates. English sing and Latin canere are not, they they have the same meaning. The suffix -ly in English has two related uses: (1) to derive adjectives from nouns, e.g., fatherly, and (2) to derive adverbs from adjectives, e.g., perfectly. The suffix is inherited from the substantive līk in Old English and is cognate with German -lich. The Romance suffix -ment(e) is usual thought to be inherited from the Latin noun mens, mentis, 'mind'. They are a similar kind of historical development, but not cognates. Some actual affixes that are cognates would be Latin in-, Greek ou- (as in Utopia 'no-place'), Sanskrit a- (as in ahimsa 'non-violence'), and English un- as negative prefixes.
Likewise, I would use the term etymological meaning, rather than literal meaning. For me, words can have literal meanings and metaphorical ones. That's more about usage. Words may have multiple meanings, e.g., set.
