We have one attested Sindarin interrogative pronoun man, from the untranslated phrase:
- ar·phent Rían Tuor·na: man agorech? “*and said Rían to Tuor: what have we done?” (VT50/5).
It is almost certainly derived from the interrogative root √MA that is the basis for question words in Quenya as well. In this phrase, man is the object of the sentence. If it follows the usual pattern for pronouns, it seems likely that man would be its adjectival form also, but *ma would be its subject form: ma egin i Edhel “who saw the Elf”? Most Neo-Sindarin writers (myself included) assume ma/man can mean both “what” (inanimate) and “who” (animate).
In Quenya, an indicative statement can be turned into a question simply by adding the interrogative particle ma, as in Q. ma Varda enquantuva i yulma nin sí “will Varda now refill the cup for me?” (PE22/161). It’s possible the same is true in Sindarin, as in: ma Elbereth adbannath i iulf annin hí (adbanna- and iulf are neologisms for “refill” and “cup”). However, most Neo-Sindarin writers assume that simple questions are indicated by intonation alone: Elbereth adbannath i iulf annin hí? (rising voice indicating a question). This is supported by the only other attested questions in Sindarin, linnathol? “will you sing (please)?” (PE22/167).
Question words beyond “who, what” are more difficult to deduce. Most Quenya question words are based on the Quenya case system, such as Q. masse “where”, which is simply ma plus the locative suffix -sse. We know that Sindarin does not make general use of such suffixes, since “the inclusion of these elements in organized declensions with plural and dual forms is a special development of Quenya” (PE21/79). It is conceivable that Sindarin could make use of these ancient suffixes in isolated words (as seems to be the case with ennas “there”) but the ancient Quenya case system is unlikely to be the basis of interrogative words in Sindarin more generally.
In her book A Fan’s Guide to Neo-Sindarin, Fiona Jallings suggested that question words can be formed by attaching prepositions to man “who/what” (FGNS/220-221). I mostly recommend her paradigm for Neo-Sindarin question words:
- an + man = *amman “for what, why”.
- mi + man = *mivan “in/at what, where”.
- na + man = *navan “at/to what, where, whither”.
- o + man = *oman [stop mutation] “from what, whence”.
- na(n) + man = *naman [nasal mutation] “with what, how”.
- man + lû = *mallú “what time, when”.
Fiona suggested neman for “when”, based on the theory that ned means “during” (so “when” = “during what”), but recent publications strongly indicate ned probably means “and one more”. A conversation on the Vinyë Lambengolmor Discord from September 2018 came up with mallú = man + lû for “when”, and that’s my current recommendation for “when” as a question word.
Conceptual Development: Gnomish in the 1910s had a relatively comprehensive paradigm of questions words based on the root ᴱ√DO: don “who”, dir “who (f.)”, G. deg “what”, dui “where”, dos “when” (GL/30); in the 1910s the root ᴱ√MA was the root of indefinite words like G. madheg “something” (GL/55). There is no indication that interrogative √DO survived past the 1910s.