√PĀ/APA “touch; after, behind of place”
This root and ones like it were used for various spacial and temporal relationships during Tolkien’s life. Perhaps the earliest iteration in this chain of developments was the root ᴱ√PE “at, by” from the Qenya Lexicon of the 1910s, with derivatives like ᴱQ. penasta “being near at hand, support, backing” and ᴱQ. peanta- “give into one’s hands, give instructions to, enjoin” (QL/72). In Noldorin Word-lists from the 1920s Tolkien gave primitive ᴱ✶apa > ᴱN. ó “to, on” and ᴱT. pa, presumably of the same meaning (PE13/151). In a late note of unclear date Tolkien wrote:
√apa “touch” (not with fingers but of the contact of surfaces, esp. vertical surface). So prep. apa, pa “touching, against”. In Q. this, espec. in form pā̆ is used as prep. = touching, as regards, concerning. Also the verb ape is used fig. as to touch one, concern, affect. In literal sense [touch] the strengthened weak verb appa is mostly used (VT44/26).
In another note from this period Tolkien glossed apa as “above but touching” (VT44/26).
In other words from the 1950s and 60s, apa- is given the temporal sense “after”, most notably in Q. Apanónar/S. Abonnen “After-born” (WJ/386-387), the former appearing in the published version of The Silmarillion (S/103). It was also used temporally in Q. apacenyë “of foresight” (MR/216). In a table of spatial and temporal roots from 1969, Tolkien gave √PĀ/PATA “after, behind of place” (PE22/147) and in another 1969 note Tolkien glossed apa, pā as both “before of time” and “after (later than)”, with both being deleted (VT44/36). Other similar roots in the semantic space of “after” include √EPE and √OPO.
Neo-Eldarin: As indicated by the discussion above, Tolkien’s treatment spatial and temporal prepositional elements was wildly inconsistent. √APA was one of the elements Tolkien used for “before” and “after”, but also as “on” or “touching”. For purposes of Neo-Eldarin, the most useful definition of √PĀ/APA is “touch”, for which we have no other good late roots, so I would assume this was its primary meaning. I generally prefer ✶epe for “after (of time); before (of space)” and ✶nō for its opposite. However, I think apa- might be a euphonic variant of epe-, perhaps limited to “after” of time, to justify Q. Apanónar/S. Abonnen “After-born”.
√PAK “close, shut”
This root appears in a list of roots from 1959-60 as the opposite of √LAT “open, unenclosed, free to entry”, with a single derivative Q. pahta “closed, shut, private” (VT41/6). It seems to refer to the state of being closed, as opposed to √SKOL which refers to the act of closing. In the much earlier Gnomish Lexicon of the 1910s, primitive ᴱ√PAKA had derivatives like G. paga “court, paved floor” and G. pagra- “pave” (GL/63), but this seems to be unrelated semantically to later √PAK “close, shut”.
√PAKAT “talk, speech”
This root appeared in notes probably dating to the early 1960s as the basis for words for “speech”, probably replacing another root √KARAP of similar meaning, and with a deleted variant √PATAK (PE17/126). The root √PAKAT itself was probably a restoration of a (hypothetical) early root *ᴱ√PAKATA, the likely basis for “speech” words in the Gnomish Lexicon of the 1910s such as G. pactha- “utter, speak, talk” and G. paithron “orator” (GL/63).
√PAL “wide, broad, extended; [ᴹ√] wide (open); [ᴱ√] flatness”
This root was well established in Tolkien’s mind, but its meaning shifted gradually over time. It first appeared as ᴱ√PALA “flatness” in the Qenya Lexicon of the 1910s with derivatives like ᴱQ. palume “plain”, ᴱQ. palo “plane surface, plain, the flat”, and ᴱQ. Paluren “the Wide World” (elsewhere ᴱQ. Palúrien), but also ᴱQ. palwa- “make wander” and ᴱQ. palāva “wandering” (QL/71). The Gnomish name G. Bladorwen “the Wide Earth” points to an actual root of *ᴱ√BALA, as do words like G. bladwen “plain” and G. blant “flat” (GL/23). However, the words G. paltha “blade (of swords, knives, oars, etc.); wide flat leaf; page of book” and G. plados “oar” seem to be derived directly from ᴱ√PALA, though the verb G. palta- “beat” [sic., rather than expected paltha-] hints that this variant of the root may mean “beat” rather than “flat” (GL/63-64); see the entry for ᴹ√PALAP “*beat” for further discussion.
In The Etymologies of the 1930s this root appeared as ᴹ√PAL¹ “wide (open)”, still serving as the basis for the name ᴹQ. Palúrien along with other derivatives like ᴹQ. palar “flat field, ‘wang’, plain” and ᴹQ. palme/N. palath “surface” (Ety/PAL). A later addition to the entry was ᴹQ. palan “far” and ᴹQ. palantir “far-seeing stone”. The root √PAL remained the basis for Q. palan “far” in Tolkien’s later writings, and appeared with the glosses “broad, wide” (PE17/65) and “wide, extended” (VT47/8). The latter gloss appeared in notes from the late 1960s having to do with Eldarin Hands, Fingers and Numerals, where Tolkien indicated the root had two extended forms palat- and palan, and that its meaning was “originally also with the implication that the area was more or less flat and even, without hindrance to movement, or view”, and so serving as the basis for the words Q. palta/S. plad “palm, flat of the hand” (VT47/8-9). Thus even in Tolkien’s later writings, it seems the root retained the sense “flatness”.
A root appearing in both The Etymologies of the 1930s as well as the contemporaneous Tengwesta Qenderinwa (first version, TQ1) as both ᴹ√PAL “beat” and extended form ᴹ√PALAP of similar meaning (Ety/PALAP; PE18/33). In The Etymologies, the entry for short ᴹ√PAL was erased (EtyAC/PAL), and the extended root ᴹ√PALAP had a couple of derivatives: ᴹQ. palpa- “to beat, batter” and N. blab- “flap, beat (wings etc.)” (Ety/PALAP). The short form ᴹ√PAL may also have been the basis for some words in the earlier Gnomish Lexicon of the 1910s: G. palta- “beat” [sic., rather than expected paltha-], G. paltha “blade (of swords, knives, oars, etc.); wide flat leaf; page of book” and G. plados “oar” (GL/63-64), though these words might be blended with root ᴱ√PALA “flatness” from the contemporaneous Qenya Lexicon (QL/71); see that entry for discussion.
√PAN “arrange, set in order; [ᴹ√] place, set, fix in place (especially of wood)”
Tolkien introduced this root early and it retained more or less the same meaning throughout his life. It first appeared as ᴱ√PANA “arrange” in the Qenya Lexicon of the 1910s with derivatives like ᴱQ. pano “series, course, plan, arrangement” and ᴱQ. panya- “plan, arrange, intend, mean” (QL/72). It also had derivatives in the contemporaneous Gnomish Lexicon such as G. pan “arrangement, settlement; place, spot” and G. panta- “set, put, place, arrange, settle” (GL/63).
In The Etymologies of the 1930s it appeared as ᴹ√PAN “place, set, fix in place (especially of wood)”, with derivatives like ᴹQ. panya-/N. penia- “fix, set”, ᴹQ. pano “piece of shaped wood”, and N. pân “plank, fixed board (especially in a floor)” (Ety/PAN). √PAN appeared unglossed in the second version of the Tengwesta Qenderinwa from around 1950 (TQ2: PE18/89) and the root √PAN “arrange, set in order” was mentioned in passing in etymological notes from the late 1960s (PE17/108).
√PAR¹ “learn; arrange, [ᴹ√] compose, put together”; √PAR² “peel”
This root was the basis for Q. parma “book”, but Tolkien vacillated on the exact sense for the root. It first appeared as unglossed ᴱ√PARA in the Qenya Lexicon of the 1910s with a single derivative: ᴱQ. parma “skin, bark; parchment; †book, writings” (QL/72). In the contemporaneous Gnomish Lexicon, words like G. padhwen “bark” and G. paglos “parchment” are probably related, along with deleted word G. pand “bark; book” (GL/63). These Gnomish forms hint that the root might actually be PAŘA [PAÐA]. Another set of words appearing in the same part of the Gnomish Lexicon and thus probably also related are: G. past “skin”, G. pasta- “skin, peel, flay” and G. path “peel, skin of fruit, fine bark (paper)” (GL/63).
The root ᴹ√PAR reappeared in The Etymologies of the 1930s with a completely different meaning “compose, put together”, though still with the derivative ᴹQ. parma/N. parf “book” along with ON. partha- “arrange, compose” (Ety/PAR). The root ᴹ√PAR “compose, arrange” also appeared in the first version of the Tengwesta Qenderinwa (TQ1) also from the 1930s, again as the basis for ᴹQ. parma “book” (PE18/51). The root √PAR¹ “arrange” > Q. parma “writing, composition, book” appeared again in the second version of the Tengwesta Qenderinwa (TQ2) from around 1950 (PE18/101).
In notes from around 1959 Tolkien reverted to the earlier meaning of the root. In etymological notes from 1959 Tolkien wrote “√PAR-, peel (hence bark, book). [S.] paran, Q. parne, bald, bare” (PE17/171). In a list of Sindarin words from the same period he wrote:
S paran, naked, bare. Cf. Dol Baran. √PAR “peel”. Cf. Q parna, bare. (Q parma, peel, applied to bark or skin, hence “book”). Q. parca, naked, of persons. S parch (PE17/86).
In another note from this period he gave a very similar derivation with √PAR > Q. parma = [originally] “bark” [later] “parchment, book” noting that the first Elvish writing materials were bark, but he then rejected this etymology (PE17/171).
In notes from the 1960s Tolkien gave:
√PAR- “learn, to acquire information, not by experience or observation, but by communication”, by the instruction, or accounts of others in words or writing: parma, a book (or written document of some size). To read a book in Elvish was often expressed so: paranye (apārien) parmanen, I am learning (have learnt) by means of a book (PE17/180).
This last meaning of the root is further supported by the phrase Q. cuita’r parë “live and learn” from Late Notes on Verbs composed in 1969 (PE22/154).
Thus the semantic evolution of root seems to be 1910s “*peel” >> 1930s-1950 “arrange, compose” >> 1959 “peel” >> 1960s “learn”.
Neo-Eldarin: Tolkien’s shifting definitions of the root √PAR make it tricky to use in the context of Neo-Eldarin. While it was the source of Q. parma “book” for all of Tolkien’s life, the exact mechanism of how √PAR was connected to “book” underwent a number of changes. Of these, I think the use of √PAR = “peel” is the one that can be most easily discarded. While this does leave S. paran “bare” from the name S. Dol Baran with no etymology, that word might be salvaged by assuming it was derived instead of an unrelated (hypothetical) root *√PARAN.
The other two meanings of the root, “compose” (1930s-1950) and “learn” (1960s) are both popular parts of Neo-Eldarin. The verb Q. par- has become the basis verb for “learn” verbs in Neo-Eldarin since its publication in PE17 in 2007. However, the sense “compose, arrange” is also well established. I think it best to assume this root means both “compose, put together” in general as well as “compose (information)” = “learn”, to retain both these senses. As for “arrange”, that seems to be better covered by √PAN.
ᴹ√PAT “[ᴱ√] open, wide, spreading”
An unglossed root in The Etymologies of the 1930s with derivatives like ᴹQ. panta- “to unfurl, spread out, open” and N. panna- “to open, enlarge”, along with ᴹQ. panta “open” and N. pann “wide” (Ety/PAT). It seems to be a later iteration of ᴱ√PATA² or ᴱ√PṆTṆ from the Qenya Lexicon of the 1910s glossed “open, spread out, show” with very similar derivatives like ᴱQ. panta “open, wide, spreading” and ᴱQ. panta- “open, unfold, spread” (QL/72). In Tolkien’s later writings, *√PAT seems to mean “step” or “walk”; see that entry for details.
Neo-Eldarin: Despite the later shift in meaning for √PAT, I think this earlier sense “open (wide)” can be salvaged by assuming that form of the root is actually *√PANAT or *√PANTA, which is consistent with most of its derivatives. For the non-verbal senses of “open”, I think √LAT¹ is generally better.
*√PAT “step, walk”
Based on words like Q. pata-/S. pad- “walk” and S. pâd “step; track, road” (PE17/34), it seems Tolkien imagined had a root *√PAT = “walk” or “step” in his later conception of the languages. This was probably a later iteration of unglossed ᴱ√PATA¹ from the Qenya Lexicon of the 1910s with derivatives like ᴱQ. pata- “rap, tap (of feet)”, ᴱQ. patake “clatter”, and ᴱQ. patinka “shoe, slipper” (QL/72). Tolkien compared this early root to ᴱ√PETE which had derivatives like ᴱQ. pete- “knock, strike” (QL/73), so likely ᴱ√PATA¹ = “tap” (light) vs. ᴱ√PETE = “knock” (heavy). In the contemporaneous Gnomish Lexicon this root had derivatives like G. padra- “walk”, with a much clear connection Tolkien’s later ideas from the 1950s and 1960s. In The Etymologies of the 1930s, Tolkien gave N. pâd as an element Tharbad, whose gloss is unclear but seems to be “?Crossway” (Ety/THAR); this might indicate some continuity of thought between 1910s ᴱ√PATA¹ “*tap” and 1950s/60s *√PAT “walk”.
The root ᴹ√PAT did appear in its own entry in the The Etymologies of the 1930s, but it had derivatives like ᴹQ. panta-/N. panna- “to open” (Ety/PAT), making it more likely a continuation of ᴱ√PATA² or ᴱ√PṆTṆ “open, spread out, show” from the the Qenya Lexicon of the 1910s, which coexisted with ᴱ√PATA¹ “*tap” (QL/72). See the entry on ᴹ√PAT for further discussion.
√PATH “smooth (to feel), silky”
This root first appeared as unglossed ᴹ√PATH in The Etymologies of the 1930s with derivatives like ᴹQ. pasta/N. path “smooth” and N. pathw “level place, sward” (Ety/PATH). It reappeared in the Outline of Phonology from the early 1950s as √PATH “to smooth” (PE19/88), and again in etymological notes from 1959 as √PĂTH/PAS “smooth (to feel), silky” (PE17/171).