掲示板 Forums - A question about a (な) adjective

The difference becomes clear if you write it with kanji.

If I write ふるい (old) with kanji it is 古い

古 is ふる and we have an い that comes after the kanji. That's what い adjectives look like i.e. some kanji followed by hiragana ending in い

If I write きれい with kanji it is 奇麗.

奇 is き and 麗 is れい. We don't have anything after the kanji. So it doesn't look like an い adjective.

Re-reading my post, I never addressed the original concern: Why is 嫌い a na-adjective?

Originally, the verb 嫌う was nominalized using the conjunctive form to create the noun 嫌い = hatred. Over time it came to be used as an adjective and, stemming from a noun, was treated as a na-adjective (which conjugate like nouns anyway).

嫌い is by no means the only such adjective, but it is by far the most common.

A few other examples:

場違いな発言 (originates from 違う) = out-of-place remark

不似合いな夫婦 (originates from 合う) = ill-matched couple

大食いな人 (originates from 食う) = glutton

幸いな結末 (originates from 幸ふ) = happy ending

It might be easier to just memorize the handful of "exceptions" when you come across them, rather than think through their etymology, but there is some historical reasoning behind it all.