This word analysis focuses on two opposing concepts: “before” and “after”. These saw numerous vacillations over Tolkien’s life, making it difficult to incorporate them into a coherent paradigm.
In English and many other languages, these relationships can be used both temporally and spatially. Temporally, they refer to a sequence of events, “before” = “preceding” and “after” = “following”. Spatially, these terms generally refer to the orientation of a body: “before” = “in front of” and “after” = “behind, in back of”, though this spatial definition is strongest in the case of “before”. This makes sense if you view temporal events in terms of what is “visible”. The past is visible and knowable, and is therefore “before you”, but the future is unknown and invisible, and this therefore “after you”. Similarly, someone moving in front of you is coming before, whereas someone moving behind you is coming after.
One tantalizing idea that Tolkien hints at is that in Elvish, these temporal and spatial relations maybe be reversed from English, as first pointed out by Patrick Wynne (VT49/32, note #12). In one note from the 1950s Tolkien wrote: “The Eldar regarded all that was past as behind them, their faces being towards the future. With reference to Time therefore words with a basic sense ‘behind, at the back’ = before, and those originally meaning ‘in front, ahead’ = after” (VT49/12). This idea is very appealing, and is consistent with the idea that the Elves have great powers of foresight.
There are a number of attested words and roots that mean “before (of time); after (of space)” or vice versa, notably:
- Q. epe “before (in all relations but time)” but “after (of time)” (PE17/56, VT49/12)
- Q. opo “before, in front; after of time” (VT49/12)
- Q. nó “at back (of place), before (of time)” (VT49/32)
- ᴱ√NOWO “ahead, in front; after of time, tomorrow” (QL/66)
- ᴱ√DYĒ root. “behind, back (before of time)” (QL/105)
There are no clear examples of this in Sindarin/Noldorin/Gnomish, however. If this was an aspect of Elvish thinking, it may have been limited to Quenya concept. We instead see:
- G. ob “after, close behind (time and place)” (GL/61)
In Qenya and Gnomish Lexicons, the two basic roots for “before” and “after” seems to be those given above, ᴱ√NOWO and ᴱ√DYĒ. But their derivatives seem to be limited to the temporal dimension, for example:
- ᴱ√DYĒ: ᴱQ. die, G. gîr “yesterday” (QL/105, GL/38)
- ᴱ√NOWO: ᴱQ. nuo, G. nôn “tomorrow” and ᴱQ. nó “after (only of time)”, G. no “then, next, again; then, after(wards)”, nui “afterwards, again, later” (QL/66, GL/61)
Spatially there is also the root:
- ᴱ√POT-I: ᴱQ. pote “after, behind (of place)” (QL/75), G. ob “after, close behind (time and place)” (GL/61)
There is no clear preposition for “before, preceding, in front of” in QL or GL, either temporally or spatially. But not long after Tolkien used the ᴱQ. la or ala for concept. ᴱQ. la “before” vs. no “after” appears in the Early Qenya Grammar of the 1920s, probably spatial (PE14/48); la “before” also appearing in the Early Qenya Dictionary, vs. no “behind” (PE15/70). ᴱQ. ala “before” (PE16/62) appears in the drafts of the Oilima Markiya poem from the 1920s, used temporally.
But even in these early stages we see the vacillations that plagued these terms emerge. In particular, in the final version of the Oilima Markiya poem, Tolkien used ᴱQ. ala as “after” (MC/214, 221). The phrase in which it appeared remained essentially the same: ala fuin oilimaite, but its translation was changed from “before the last night” to “after the last night” (PE16/62, MC/221).
The Etymologies from the 1930s is strangely silent on these concepts. In the drafts of the Lord of the Rings appendices we see hints of these ideas in some alternate names of the months (PM/135-136):
- ᴹQ. Erlaire, N. Ebloer “June”
- ᴹQ. Nólaire, N. Cadloer “July”
Though untranslated, these seem to be “*First or Before Summer” and “*Second or After Summer”. ᴹQ. nó “after” seems to retain its Early Qenya sense, but N. cad “after” is nowhere else attested. However, there is support for this interpretation in the word ᴹQ. kata “after” seen in phrase kanya ére kata tulma alwara “to be wise after the event is useless”, also written in the 1940s (PE22/124). Whether or not N. eb means “before” or “first” is unclear, but since it distinct for ᴹQ. er “first”, the sense “before” seems probable.
By the 1950s, the sense of Q. nó seems to have flipped. In the so-called “Merin Sentence” Tolkien used nó as “before”: merin sa haryalye alasse nó vanyalye Ambarello “I hope that you have happiness before you pass from the world”. In a note from the 1950s, we have nō “at back (of place), before (of time)” vs. opo “in front (of place)” (VT49/32).
Around this time Tolkien seems to use three similar terms for “after (of time)”: epe, apa, opo. The form apa is only used temporally, most notably in Apanónar “After-born” as a name for Men from the Silmarillion, which is as close as we get to a canonical definition of any of these terms. The other two terms could also be used with a reversed spatial sense: opo “before, in front; after of time” (VT49/12) and epe “before (in all relations but time)” but “after (of time)” (PE17/56, VT49/12). The form Q. epe seems to be the best attested, several several sentences (PE17/56-57), as well as epesse “after name” and epetai “consequently, (lit.) following which (fact)” (PM/339; UT/266; VT49/8, 12). Of these only ab “before” appears in Sindarin: in S. Abonnen “After-born”, cognate of Q. Apanónar.
The vacillations continued in Tolkien’s later writings. In the 1960s he used Q. noa (likely related to Q. nó) as both “tomorrow” and “yesterday”, the second case a reduction of noa ré “former day” (VT49/34, note #23).
In a note from the late 1960s, Tolkien seems to have begun work on a new paradigm (PE22/147), citing the roots:
- √PHĀ/APHA: “before of time”
- √KHYA: “before of place”
- √KĀ/KATA: “after of time”
- √PĀ/PATA: “after of place”
The third of these seems to be a restoration of N. cad, ᴹQ. kata “after”, the latter of which also appears as Q. cata “behind, at back of place” beside variants ca and cana in the 1950s (VT43/29), though this is spatial, not temporal. However, the only Quenya or Sindarin word Tolkien derived directly from this paradigm was Q. fai, in an untranslated phrase Q. Eru fai, sî, euva, most likely “*Eru (was) before, (is) now, will be (after)”. It’s unclear whether this new paradigm would have taken hold or was a transient idea.
Assembling this inconsistency into a coherent paradigm requires making some arbitrary choices. In Quenya, I think the best option is nó “before (of time); at back (of place), *behind” and epë “after (of time); before (of space), *in front of”. Among other things this lets us use noa for “former, yesterday”, for which we have no other late words; for “tomorrow” we still have ᴹQ. enar (probably “(lit.) yonder day”). Quenya also seems have variant apa of epe, used less often.
In Sindarin, we can use the cognate ab for “after (of time)”, though reusing N. cad for this purpose is probably also fine. Unfortunately, this leaves us no good Sindarin word for “before (of time)”, since eb (cognate of Q. epe) would be unsuitable. The cognate of Q. nó “before” would be ᴺS. naw, but this word already has several meanings. The best idea I have is restoring G. nui “after(wards)” with a flipped meaning as ᴺS. nui “before”, derived from a primitive adjective *nōya.
For spatial relationships in Sindarin, I would not retain the Quenya paradigm of reversing before and after, since there is no evidence of it. For “after (of space)” I would use N. adel “behind, in rear (of)”. For “before (of space)” I would coin an adjective from the noun N. nîf “front“, perhaps ᴺS. niven, and perhaps create a preposition ᴺS. anif or anef (the latter assuming the root changed from ᴹ√NIB to √NEB) by analogy with adel.
I admit that I don’t find these Sindarin neologisms completely satisfactory, but the only other option I see is creating a bunch of neologisms from the late PHĀ/KHYA/KĀ/PĀ paradigm given above, and I find this even worse.
Addendum: Since I originally wrote the article and posted it on G+, I’ve given some thought to restoring the eb “before”, cad “after” paradigm in Sindarin, as it appeared around 1950. This isn’t consistent with Q. epe “before (of space), after (of time)”, but perhaps there was some semantic drift in Sindarin, with eb picking up its spatial meaning. If so, S. ab “after” and S. eb “before” would have opposite meanings.