Keiyōdōshi (形容動詞 – ‘adjectival verbs’) are a class of declinable words such 静か and 堂々 that take the auxiliary verbs だ, なり, and たり. The classification is widely accepted in Japan and is recognised by highly reputable dictionaries such as Daijirin and the current Kōjien.
It was not, however, accepted by Shinmura Izuru (新村出), the original editor of the Kōjien (広辞苑). The fourth edition of the Kōjien was the last to include his dissenting opinion in a meaningful manner:
また、別説では、文語の第一類は形容詞連用形と「あり」とに還元し、第二類・第三類は名詞と指定の助動詞との結合とみなして形容動詞を認めない。本辞典はこの最後の説に従う。
[…] In another theory, keiyōdōshi are not acknowledged [to exist]: the first form 1 is reduced to あり and the continuative form of the adjective, and the second2 and third3 forms are considered to be combinations of nouns and the specified auxiliary verbs. This dictionary follows this last theory.
The fifth and sixth editions seem to accept the existence of keiyōdōshi, and only mention the controversy in passing:
形容動詞を品詞の一としない説もある。
There is also a theory in which keiyōdōshi are not recognised as a part of speech.
Shinmura’s theory interests me because it seems to consider なり and たり more analytically—after all, the auxiliary verb なり is formed by combining に and あり, and たり by combining と and あり.
To say 静かなる山 (‘a still/quiet mountain’) is to say ‘mountain existing in [a state of] stillness/quietness’. This is a little different grammatically from an expression such as 静かだ (which can be translated as ‘it is still/quiet’) in which 静か is used somewhat like an adjective.
Very strictly speaking, it is hard to justify putting a d– sound after a なり word such as 静か. If we look at the Daijirin and Daijisen, we can see that even though they say that 静かd– is permissible in spoken language, all the usage examples they provide for 静か are of 静かn–.