English has/had so many of these trios: where/there/here, whereby/thereby/ hereby, whither/thither/hither, whence/thence/hence ... and when/then/NOW? Whatever happened to "hen?" Did that form exist in an earlier stage or dialect?
1Now does stand out like a sore thumb in the paradigm, doesn't it?– jlawlerJul 8, 2017 at 20:57
Yes, it did.
From Middle English henne, heonne, hinne, from earlier henene, heonenen, henen, from Old English heonan, hionan, heonane, heonone (“hence, from here, away, from how”), from Proto-Germanic *hina, *hinanō (“from here”), from Proto-Indo-European *ḱe-, *ḱey- (“this, here”). Cognate with Dutch heen (“away”), German hin (“hence, from here”), Danish hen (“away, further, on”). See also hence.
EDITED TO ADD: To be sure, this answer addresses the "Whatever happened to "hen?" Did that form exist in an earlier stage or dialect?" part of your question. hen did not really mean 'now'.
6This word doesn't mean "now", though. Jul 7, 2017 at 2:24
True, I guess I was answering just the " Whatever happened to "hen?" Did that form exist in an earlier stage or dialect?" part. I'll clarify. Jul 7, 2017 at 13:43
1Another cognate: Modern West Frisian "hinne". Jul 23, 2017 at 22:58
The answer to your question is no, and, what's more, the deictic (IE *ke > hen)-related forms that Mark Beadles refers to cannot be directly connected with the bound morph -en of Modern English then because -en (just as -ere, -us - as in thus, etc.) is not deictic in then, there, thus, etc. What is deictic in all such complex words is the bound morph th-, not the -en (-ere, -us, etc.) that follow it. And, of course, the reason why now does not fit in the pseudoparadigm attempted above and why -en can never have meant now in English, is that now is deictic, whereas -en cannot be.
This impossibility can be explained in a principled way: a word (or phrase) containing two deictic elements, be they free or bound forms, would violate a universal principle, i.e., that variables (say a time variable, as in th-en) cannot be doubly bound. Hence, no deictic can possibly 'bind' (read 'apply to', if you prefer) a constituent that already contains an internal deictic: once the (time/place, etc.) variable is bound, a second deictic would have no variable to bind, biuniqueness would be violated, and the resulting word/phrase would be necessarily ill-formed.
That is, in short, why then can not be analysed as [t(h) + -(h)en = now] and why the answer to your question is No, English could never have used -hen, (-en, rather), to mean now, even if deictic hen-related forms did exist in, presumably, all primitive IE languages, as Mark Beadles pointed out.
1Great answer as for the "No" in "Was there a word 'hen' to mean 'now'". But how about some thoughts on "Why is there an apparent gap in the regularity of the paradigm", or, alternatively, "Why is it the above merely an 'attempted pseudoparadigm', not a valid pattern?" Jul 16, 2017 at 20:03
1a word (or phrase) containing two deictic elements, be they free or bound forms, would violate a universal principle: maybe I'm misunderstanding you, but how about e.g. this book here? Jul 17, 2017 at 0:56
2I don't think the asker meant to call -en deictic. Rather, -en seems like it could be a time-related ending: wh-ither = interrogative + allative, h-ither = demonstrative + locative, th-ither = demonstrative + locative; so wh-en = interrogative + time, *h-en = (potentially) demonstrative + time, th-en = demonstrative + time.– Draconis ♦Jul 17, 2017 at 4:11
@TKR I should have explained the phrasal cases too: I was referring to two th- DETERMINERS, as in '*the this new book', '*that that book over there', etc. Such Det's bind the variable associated with the NP's 'new book', 'book over there', because they c-command them. In 'This book here', 'That book over there', 'here' or 'there' do NOT bind the NP variable because they do NOT c-command the WHOLE NP, they BELONG to it. In particular, in your example 'this book here', 'here' modifies 'book', but is NOT a Det, cannot bind the variable of 'book here', and no double-binding problem arises.– user6814Jul 17, 2017 at 7:51
1Isn't that just part of the general rule that English disallows multiple determiners (with an exception for quantifiers)? In any case if it's a syntactic rule about determiners, I don't see how it would apply to single words. Jul 17, 2017 at 22:33