√RĀ/ARA “noble, high, royal”
An invertible root meaning “noble”, itself an extension of √AR (PE17/147). Given its widespread use, Tolkien did not define this root until quite late. Its first clear mention was in Notes on Names (NN) and Quenya Notes (QN), both from 1957. In NN, Tolkien introduced invertible √ARA/RĀ to give an explanation of the element rod in names like S. Rodon “Vala” and S. Finrod, where Tolkien said:
S raud “tall, high, lofty, eminent”. √ARA, RĀ. Cf. Q aran “king”. Q arta “high” < árată. S raud < form (a)rā́tă (PE17/118).
Earlier in The Etymologies of the 1930s, the second element of N. Finrod was N. rhaud “metal” (Ety/RAUTĀ), whereas ᴹQ. haran/N. aran “king” was derived from ᴹ√ƷAR “have, hold” (Ety/ƷAR). As for the later root √ARA/RĀ, Tolkien gave a lengthier explanation in QN:
√AR. ARI-, ARA/RĀ, ARAT/ARAN. Original sense probably seen in adverb/preposition *ara/ar/ra, beyond, further than. The stems ARI/ARAN/ARAT mean “good, excellent, noble” — differing from √MAN in stating that any one specimen is “good of its kind”, excels, without necessarily implying that others are bad or marred. Hence ari-, good as prefix, ar/ara/aran as prefixes of excellence especially in royal names, aran king. Q arta noble, arato a “noble”, ráta- excel, surpass. S aran, arod (arāta); or raud excellent, noble, eminent (PE17/147).
Thus all these roots, including the extensions √ARAN and √ARAT, were elaborations on the sense “beyond” from base root √AR. The simple root √AR itself had a lengthy conceptual history; see that entry for details. As for ARA/RĀ, the connection between these forms and nobility seems to be firmly established by the late 1950s and afterwards, with primitive element Ara- “noble” mentioned in The Shibboleth of Fëanor from 1968 (PM/344, 363 note #43), and numerous other mentions of derived forms in documents from this period.
√RAB “astray, wandering, unsettled”
This root first appeared as unglossed ᴹ√RAB in The Etymologies of the 1930s with derivatives like ᴹQ. ráva/N. rhaw “wild, untamed” and ᴹQ. ravanda/N. rhofan “wilderness” (Ety/RAB; EtyAC/RAB), the latter an element in the name N. Rhovanion from Lord of the Rings drafts (TI/296). The Quenya/Noldorin r-/rh- variation was a result of the fact that initial [r-], [l-] were unvoiced in Noldorin (PE22/32).
The root √RAB reappeared in Tolkien’s later writings with the gloss “astray, wandering, unsettled”; Tolkien contrasted it with √RAN by saying “it differed from √RAN in that it referred to absence of direction or purpose, whereas √RAN meant to “err”, go aside from a course (commanded or self-chosen)” (PE17/78). In Sindarin, however, initial r-, l- were no longer unvoiced, so Tolkien coined a new intensified variant √S-RAB “wild in senses ‘not tamed, domesticated’, and hence often ‘fierce, savage, hostile’ (to Elves and Men)”. This intensified variant served as a new basis for S. Rhovanion “Wilderland”, and also had other derivatives like Q. hráva “wild” and Q. hravan “wild beast”.
In another place Tolkien defined this root as √S)ROB with variant √D)ROB as the basis for Rhovanion (PE17/99), also explaining the element Drû in Drúadan, the Sindarin name for the Woses. These roots produced different Quenya derivatives like hróva.
Neo-Eldarin: Since the Quenya form (h)ráva “wild” is better known, I think it best to ignore the √S)ROB and √D)ROB variants of this root for purposes of Neo-Eldarin. The Sindarin name for the Woses can be easily explain as a loan word from that people’s name for themselves: Drughu (UT/385).
In The Etymologies as published by Christopher Tolkien in The Lost Road, this “root” is given as RAMBĀ (Ety/RAMBĀ). However, Carl Hostetter and Patrick Wynne clarified that the root form was actually given as RAB⁽²⁾, and ᴹ✶rambā is simply the primitive form of ᴹQ. ramba/N. rham “wall” (EtyAC/RAMBĀ). The root form ᴹ√RAB² is consistent with other derivatives of the root: ᴹQ. ráva/N. rhaw “bank (especially of a river)”. Tolkien’s continued use of S. ram for “wall” in later writings (S/122; RC/512) indicates the ongoing validity of at least the strengthened form of this root.
√RAD “[ᴹ√] back, return”
This root first appeared in The Etymologies of the 1930s as ᴹ√RAD “back, return” with derivatives like ᴹQ. randa/N. anrand “cycle, age (100 Valian Years)” as well as Dor. radhon “east” (Ety/RAD). It reappeared (unglossed) in Outline of Phonology (OP2) as an example of a root where medial d dissimilated to l in Quenya rather than becoming r as usual: √RAD > Q. ral- (PE19/99). Neither the root or the verb were glossed in OP2, so it is unclear whether or not it retained it 1930s meaning, though there is no reason to believe it didn’t.
*√RAK “*break; [ᴱ√] pile up”
This root appeared as ᴱ√RAKA “pile up” in the Qenya Lexicon of the 1910s, with derivatives like ᴱQ. rakta- “pile, hoard, amass, collect” and ᴱQ. rakte “pile, heap” (QL/78). In the contemporaneous Gnomish Lexicon, there are similar forms with different meanings: G. ractha “breach” and G. rag- “break asunder, burst” (GL/64). This conceptual shift from “pile up” to “*break” is supported by the word ᴱQ. ranka “broken” in drafts of the ᴱQ. Oilima Markirya from the end of the 1920s (PE16/77). Q. rak- “break” reappeared in the revised version of that poem from the late 1960s (MC/222-223), implying the root remained valid or (more likely) was abandoned but later restored. Q. rakine “stripped” in phonetic notes from 1959-60 might be related.
ᴹ√RAK “stretch out, reach”
The first iteration of this root was as ᴱ√RAHA “stretch forward” in the Qenya Lexicon of the 1910s with derivatives like ᴱQ. rá “arm” and ᴱQ. ráma “wing” (QL/78). It also had derivatives in the contemporaneous Gnomish Lexicon such as G. rath “the full arm” and G. ram “wing, pinion” (GL/64-65). In The Etymologies of the 1930s, “wing” words were based on ᴹ√RAM (Ety/RAM), but “arm” words were derived from ᴹ√RAK “stretch out, reach”: ᴹQ. ranko/N. rhanc “arm” along with ᴹQ. rangwe/N. rhaew “fathom” (Ety/RAK). The relationship to earlier and later ᴱ√RAKA/*√RAK “break”, if any, is unclear.
ᴹ√RAM “*wing, fly”
An unglossed root in The Etymologies of the 1930s glossed with derivatives like ᴹQ. ráma “wing”, N. rhofal “pinion”, and N. rhenia- “fly, sail” (Ety/RAM). In the Qenya Lexicon of the 1910s, very similar “wing” words were instead derived from ᴱ√RAHA: ᴱQ. ráma and G. ram (QL/78; GL/64). In Tolkien’s later writings, he continued to use Q. ráma and S. raw for “wing” (PE17/63), indicating the ongoing validity of ᴹ√RAM.