ᴹ√(M)BAD “duress, prison, doom, hell; *judge”
The most notable derivative of this root is Q. Mandos, which in Tolkien’s earlier writings was the name of the Vala, but in later writings technically only the name of his halls. The first iteration of this root was unglossed {ᴱ√MANÐAN >>} ᴱ√Mᵇ(A)NÐ(A)N from the Qenya Lexicon of the 1910s, with a more “normal” Quenya form of MAND; its derivative ᴱQ. Mandos was glossed “hell” at this early stage (QL/58, 60). In a list of roots at the end of the M-section Tolkien gave it as MṆÐṆ and its Gnomish variant band- was glossed “bind”, which was the most likely meaning of the root at this early stage (QL/63).
In The Etymologies of the 1930s the root appeared as ᴹ√BAD along with a strengthened form ᴹ√MBAD glossed “duress, prison, doom, hell” (Ety/BAD, MBAD). Earlier entries for the strengthened root ᴹ√MBAD had glosses “hell” >> “doom” (EtyAC/MBAD). The strengthened root had derivatives like N. bann “duress, prison” and cognate ᴹQ. manda as in ᴹQ. Angamanda “Hell, Iron Prison” (Ety/MBAD). Tolkien said the unstrengthened root ᴹ√BAD was “not in Q” and it had derivatives like primitive verb ᴹ✶bad- “judge”, N. bauð “judgement”, and N. badhron “judge” (Ety/BAD), the Noldorin words having been transferred from ᴹ√MBAD (EtyAC/MBAD).
Thus it seems the unstrengthened root meant ᴹ√BAD “judge” and the strengthened root ᴹ√MBAD meant “duress, prison”. Tolkien’s continued use of names like Q. Mandos “Castle of Custody” and S. Angband “Iron Prison” indicate the ongoing validity of strengthened √MBAD, but there are other later root for “judge” such as *√NAM or √NDAB (PE22/154; VT42/34). These roots served as the basis for the true name of Mandos: Q. Námo or Návo (S/28; PE21/85; PE22/154). For purposes of Neo-Eldarin, however, I think it best to assume √BAD survived as a Sindarin-only root for “judge”, since the Sindar had no direct encounter with the Vala and would be less likely to use this true name for Mandos.
ᴹ√MBAKH “exchange”
This root was the basis for words having to do with trade. It first appeared as unglossed ᴱ√VAKA in the Qenya Lexicon of the 1910s, with derivatives like ᴱQ. vakse “sale” and ᴱQ. vaktele “trade” (QL/99). It also had derivatives in the contemporaneous Gnomish Lexicon like G. bag- “sell, trade” (GL/21); a similar set of derivatives appeared in Early Noldorin word lists of the 1920s (PE13/138). The root appeared as ᴹ√MBAKH “exchange” in The Etymologies of the 1930s with derivatives like ᴹQ. manka-/N. banga- “trade”, N. bachor “pedlar”, and ᴹQ. makar “tradesman” (Ety/MBAKH). An earlier version of this entry instead had ᴹ√MBAƷ, but this was deleted (EtyAC/MBAƷ, MBAKH). The word ᴹQ. makar as well as a deleted variants māka- of ᴹQ. manka- and bagor of N. bachor (EtyAC/MBAKH) imply Tolkien also considered a variant form ᴹ√MBAK.
ᴹ√MBAL “[ᴱ√] crush, *pound; hurt, pain, damage, give maim to”
An unglossed root in The Etymologies of the 1930s serving as the basis for the word ᴹQ. malle “street” and ᴹQ. ambal “shaped stone, flag”, the second gloss probably meaning “flag stone” (Ety/MBAL). In the the Qenya Lexicon of the 1910s, the strengthened root ᴱ√MBALA was glossed “crush, hurt, pain, damage, give maim to” with Gnomish variant bal- and derivatives like ᴱQ. maldor “agony” and G. bal “anguish, pain” (QL/58; GL/21); the root entry is given under ᴱ√MALA³, but a list of roots at the end of the M-section make it clear the strengthened root is in fact ᴱ√MBALA (QL/63).
The Qenya Lexicon also has an unstrengthened root ᴱ√MALA¹ “crush, squeeze, pulp” to which ᴱ√MBALA was related, whose derivatives include ᴱQ. mal- “to crush”, ᴱQ. malle “street” and G. mal “paved way, road” (QL/58; GL/56). There was another variant root ᴱ√MLŘL [MḶÐḶ] “crush, pound”, also connected to ᴱ√MALA¹ and with Gnomish variant blid[h]-, though it had no obvious derivatives in the contemporaneous Gnomish Lexicon (QL/62-63). In the The Etymologies of the 1930s, it seems the “torment” words were transferred to ᴹ√(Ñ)GWAL: in the 1910s it seems in initial element of G. Balrog was from ᴱ√MBALA (GL/21; LT1A/Balrog) but in the 1930s it was from N. baul “torment” < ᴹ√(Ñ)GWAL (Ety/ÑGWAL). Thus 1930s ᴹ√MBAL was probably limited to meanings like “*crush, pound”.
√MBAR “settle; establish, fix, decide, determine, make a decision; dwell, [ᴹ√] inhabit, [ᴱ√] live”
This root was connected to the sense “dwell” for all of Tolkien’s life. In the Qenya Lexicon of the 1910s it appeared as ᴱ√MBARA “dwell, live”, though in Qenya it was blended with ᴱ√ŊWAÐA “tarry, linger” since initial [ŋʷ] became [m] (QL/60). In Gnomish the two roots remained distinct: G. bar “home” (GL/21) vs. G. gwadh- “dwell” (GL/46). However, other Gnomish derivatives hint at an unstrengthened variant ᴱ√MARA: G. mar “Earth, ground, soil” and G. Môr “The Earth”, of which Tolkien said “mar and môr probably conceal two roots and a confusion; mbara” (GL/56). The source of this confusion might be ᴱ√MARA¹ “ripe” (QL/59, 63).
In The Etymologies of the 1930s the root appeared as ᴹ√MBAR “dwell, inhabit” (Ety/MBAR); an earlier version of this entry had the extra gloss “build” (EtyAC/MBAR). In The Etymologies, it seems this root was a strengthed form of ᴹ√BAR “raise” (Ety/BAR), but the root ᴹ√MBARAT “fate” seems to be unrelated at this point (Ety/MBARAT). The root mbar- was mentioned in the Outline of Phonetic Development (OP1) from the 1930s (PE19/36) and again as √MBAR “{house >>} dwell” in the second version of Tengwesta Qenderinwa (TQ2) from around 1950 (PE18/92).
The root √MBAR still had the basic sense “dwell” in Notes on Galadriel’s Song (NGS) from the late 1950s or early 1960s (PE17/64); in those notes it seems it was distinct from both √BAR “raise” and √MAR(AT) or √MBART “doom, fate” (PE17/66). But in notes on Ambar, umbar, Turambar from the late 1960s, Tolkien shifted the meaning of the root:
English “settle” in its various branches of meaning closely resembles the development of the meanings of √MBAR: thus settlement can mean the act of colonizing or taking up an abode, or the area or place so occupied (by a family or community); or (the terms of) an agreement fixed after debate. The development was not, however, the same: the senses of settle proceed from a sense “place in or take up” a firm position, especially in a place that seems suitable; from which the sense of settling affairs that were in confusion or doubt arose. √MBAR meant basically to make a decision, and the meanings relating to dwelling or occupying land proceeded from that (PE17/105).
This new definition for √MBAR was to support a more direct connection between this root and words like Q. umbar “fate”, hence establishing an etymological relationship to extended √MBARAT for the first time, for example with √MBAR > ✶mbartā- “to define, decree, destine” (PE17/104). For purposes of Neo-Eldarin, I think this late 1960s meaning “settle, decide, establish; (by extension) dwell” is probably the best.
√M(B)ARAT “doom, fate”
This root and ones like it were connected to “fate” throughout Tolkien’s life. It first appeared as ᴱ√MṚTṚ or ᴱ√M(B)ṚTṚ “chance” in the Qenya Lexicon of the 1910s with derivatives like ᴱQ. mart “a piece of luck”, ᴱQ. mart- “it happens” and ᴱQ. marto “fortune, fate, lot” (QL/63). Thus in QL it seems to have more to do with happenstance than destiny, but in the contemporaneous Gnomish Lexicon its derivatives were G. Bridwen “Fate personified”, G. mart(os) “portion, fate, lot”, G. martion “fated, doomed, fey”, and G. umbart or umrod “fate” (GL/24, 56, 75), so perhaps the root meaning had already begun to shift.
In The Etymologies of the 1930s the root appeared as ᴹ√MBARAT “fate” (Ety/MBARAT); this replaced an earlier unstrengthened version ᴹ√MAR-TA, as reflected in the change of the Noldorin form from N. marth >> N. barth (EtyAC/MBARAT), later still N. ammarth >> S. amarth (Ety/MBARAT; SA/amarth). In Notes on Galadriel’s Song (NGS) from the late 1950s or early 1960s Tolkien explicitly said that the root √MAR(AT) or √MBART “doom, fate” was not connected to √MBAR “dwell” (PE17/66), but in notes from the late 1960s he changed his mind and made an etymological connection between the two roots: see √MBAR for details.
√MBAS “bake, [ᴱ√] cook; [ᴹ√] knead”
This root was connected to bread and baking for all of Tolkien’s life. It first appeared as ᴱ√M(B)ASA “cook, bake” in the Qenya Lexicon of the 1910s with derivatives like ᴱQ. maksa- “cook” and ᴱQ. masta “bread” (QL/59). In this period, both unstrengthend and strengthened forms were used as indicated by the Gnomish roots bas or mas (QL/59), both reflected by words in the contemporaneous Gnomish Lexicon like G. bast “bread”, G. brath- “to cook”, and G. mast “feed, food, nourishment; fodder” (GL/22, 24, 56), though the mas- words seem to be influenced by ᴱ√MATA “eat”.
The root appeared as ᴹ√MBAS in The Etymologies of the 1930s with the derivatives like ᴹQ. masta/N. bast “bread” and ᴹQ. masta- “bake” (Ety/MBAS). When this root was first written it had the gloss “knead” and an unstrengthened form ᴹ√MAS, but both of these were crossed through (EtyAC/MBAS). The sense “knead” was transferred to the root ᴹ√MASAG, which still shows signs of unstrengthened ᴹ√MAS (Ety/MASAG). The root √MBAS “bake” appeared again in notes from the 1960s discussing the origin of S. lembas (PE17/51).
ᴹ√MBAW “compel, force, subject, oppress”
A root in The Etymologies of the 1930s glossed “compel, force, subject, oppress”, with derivatives ᴹQ. maure/N. baur “need” and N. baugla- “to oppress” (Ety/MBAW), the latter serving as the basic for one of the names of Morgoth: N. Bauglir “Constrainer” (LR/206). Tolkien’s continued use of S. Bauglir in later writings (S/104) indicate the ongoing validity of this root.