(original) (raw)
{\rtf1\ansi\ansicpg1252\uc1 \deff0\deflang1033\deflangfe1044{\fonttbl{\f0\froman\fcharset0\fprq2{\*\panose 02020603050405020304}Times New Roman;}{\f1\fswiss\fcharset0\fprq2{\*\panose 020b0604020202020204}Arial;} {\f2\fmodern\fcharset0\fprq1{\*\panose 02070309020205020404}Courier New;}{\f3\froman\fcharset2\fprq2{\*\panose 05050102010706020507}Symbol;}{\f4\froman\fcharset0\fprq2{\*\panose 00000000000000000000}Times;} {\f5\fswiss\fcharset0\fprq2{\*\panose 00000000000000000000}Helvetica;}{\f6\fmodern\fcharset0\fprq1{\*\panose 00000000000000000000}Courier;}{\f7\fswiss\fcharset0\fprq2{\*\panose 00000000000000000000}Geneva;} {\f8\froman\fcharset0\fprq2{\*\panose 00000000000000000000}Tms Rmn;}{\f9\fswiss\fcharset0\fprq2{\*\panose 00000000000000000000}Helv;}{\f10\froman\fcharset0\fprq2{\*\panose 00000000000000000000}MS Serif;} {\f11\fswiss\fcharset0\fprq2{\*\panose 00000000000000000000}MS Sans Serif;}{\f12\froman\fcharset0\fprq2{\*\panose 00000000000000000000}New York;}{\f13\fswiss\fcharset0\fprq2{\*\panose 00000000000000000000}System;} {\f14\fnil\fcharset2\fprq2{\*\panose 05000000000000000000}Wingdings;}{\f15\fswiss\fcharset0\fprq2{\*\panose 020b0604030504040204}Tahoma;}{\f16\fnil\fcharset2\fprq2{\*\panose 00000000000000000000}Marlett;} {\f17\fnil\fcharset2\fprq2{\*\panose 00000000000000000000}MS Reference 1;}{\f18\fnil\fcharset2\fprq2{\*\panose 00000000000000000000}MS Reference 2;}{\f19\froman\fcharset2\fprq2{\*\panose 05020603050405020304}Times New Roman Special G1;} {\f20\froman\fcharset2\fprq2{\*\panose 05020603050405020304}Times New Roman Special G2;}{\f21\fswiss\fcharset0\fprq2{\*\panose 020b0506020202030204}Arial Narrow;}{\f22\fswiss\fcharset0\fprq2{\*\panose 020b0a04020102020204}Arial Black;} {\f23\froman\fcharset0\fprq2{\*\panose 02020404030301010803}Garamond;}{\f24\froman\fcharset0\fprq2{\*\panose 02050604050505020204}Bookman Old Style;}{\f25\fswiss\fcharset0\fprq2{\*\panose 020b0806030902050204}Impact;} {\f26\fswiss\fcharset2\fprq2{\*\panose 050b0604020202020204}Arial Special G1;}{\f27\fswiss\fcharset2\fprq2{\*\panose 050b0604020202020204}Arial Special G2;}{\f28\fswiss\fcharset2\fprq2{\*\panose 050b0506020202030204}Arial Narrow Special G1;} {\f29\fswiss\fcharset2\fprq2{\*\panose 050b0506020202030204}Arial Narrow Special G2;}{\f30\fscript\fcharset0\fprq2{\*\panose 030f0702030302020204}Comic Sans MS;}{\f31\fswiss\fcharset0\fprq2{\*\panose 020b0604030504040204}Verdana;} {\f32\fnil\fcharset0\fprq2{\*\panose 00000000000000000000}Angerthas;}{\f33\fnil\fcharset0\fprq2{\*\panose 00000000000000000000}Angerthas Moria;}{\f34\fswiss\fcharset0\fprq2{\*\panose 020b0500000000000000}Tengwar Quenya;} {\f35\fswiss\fcharset0\fprq2{\*\panose 020b0500000000000000}Tengwar Sindarin;}{\f36\fswiss\fcharset0\fprq2{\*\panose 020b0500000000000000}AngloSaxon Runes;}{\f37\fnil\fcharset0\fprq2{\*\panose 00000400000000000000}Tengwar Parmaite;} {\f38\froman\fcharset0\fprq0{\*\panose 00000000000000000000}.;}{\f129\fswiss\fcharset238\fprq2 Tahoma CE;}{\f130\fswiss\fcharset204\fprq2 Tahoma Cyr;}{\f132\fswiss\fcharset161\fprq2 Tahoma Greek;}{\f133\fswiss\fcharset162\fprq2 Tahoma Tur;} {\f134\fswiss\fcharset186\fprq2 Tahoma Baltic;}{\f165\fswiss\fcharset238\fprq2 Arial Narrow CE;}{\f166\fswiss\fcharset204\fprq2 Arial Narrow Cyr;}{\f168\fswiss\fcharset161\fprq2 Arial Narrow Greek;}{\f169\fswiss\fcharset162\fprq2 Arial Narrow Tur;} {\f170\fswiss\fcharset186\fprq2 Arial Narrow Baltic;}{\f171\fswiss\fcharset238\fprq2 Arial Black CE;}{\f172\fswiss\fcharset204\fprq2 Arial Black Cyr;}{\f174\fswiss\fcharset161\fprq2 Arial Black Greek;}{\f175\fswiss\fcharset162\fprq2 Arial Black Tur;} {\f176\fswiss\fcharset186\fprq2 Arial Black Baltic;}{\f177\froman\fcharset238\fprq2 Garamond CE;}{\f178\froman\fcharset204\fprq2 Garamond Cyr;}{\f180\froman\fcharset161\fprq2 Garamond Greek;}{\f181\froman\fcharset162\fprq2 Garamond Tur;} {\f182\froman\fcharset186\fprq2 Garamond Baltic;}{\f183\froman\fcharset238\fprq2 Bookman Old Style CE;}{\f184\froman\fcharset204\fprq2 Bookman Old Style Cyr;}{\f186\froman\fcharset161\fprq2 Bookman Old Style Greek;} {\f187\froman\fcharset162\fprq2 Bookman Old Style Tur;}{\f188\froman\fcharset186\fprq2 Bookman Old Style Baltic;}{\f189\fswiss\fcharset238\fprq2 Impact CE;}{\f190\fswiss\fcharset204\fprq2 Impact Cyr;}{\f192\fswiss\fcharset161\fprq2 Impact Greek;} {\f193\fswiss\fcharset162\fprq2 Impact Tur;}{\f194\fswiss\fcharset186\fprq2 Impact Baltic;}}{\colortbl;\red0\green0\blue0;\red0\green0\blue255;\red0\green255\blue255;\red0\green255\blue0;\red255\green0\blue255;\red255\green0\blue0;\red255\green255\blue0; \red255\green255\blue255;\red0\green0\blue128;\red0\green128\blue128;\red0\green128\blue0;\red128\green0\blue128;\red128\green0\blue0;\red128\green128\blue0;\red128\green128\blue128;\red192\green192\blue192;}{\stylesheet{ \nowidctlpar\widctlpar\adjustright \fs28\lang1044\cgrid \snext0 Normal;}{\*\cs10 \additive Default Paragraph Font;}}{\*\revtbl {Unknown;}}{\info{\title THE EVOLUTION FROM PRIMITIVE ELVISH TO QUENYA}{\author Helge K\'e5re Fauskanger}{\operator Helge K\'e5re Fauskanger} {\creatim\yr2003\mo3\dy22\hr22\min37}{\revtim\yr2003\mo3\dy22\hr22\min38}{\version1}{\edmins1}{\nofpages60}{\nofwords23124}{\nofchars131807}{\*\company }{\nofcharsws161868}{\vern73}}\paperw11906\paperh16838\margl1417\margr1417\margt1417\margb1417 \deftab708\widowctrl\ftnbj\aenddoc\hyphhotz425\formshade\viewkind4\viewscale100\pgbrdrhead\pgbrdrfoot \fet0\sectd \linex0\headery709\footery709\colsx709\endnhere\sectdefaultcl {\*\pnseclvl1\pnucrm\pnstart1\pnindent720\pnhang{\pntxta .}}{\*\pnseclvl2 \pnucltr\pnstart1\pnindent720\pnhang{\pntxta .}}{\*\pnseclvl3\pndec\pnstart1\pnindent720\pnhang{\pntxta .}}{\*\pnseclvl4\pnlcltr\pnstart1\pnindent720\pnhang{\pntxta )}}{\*\pnseclvl5\pndec\pnstart1\pnindent720\pnhang{\pntxtb (}{\pntxta )}}{\*\pnseclvl6 \pnlcltr\pnstart1\pnindent720\pnhang{\pntxtb (}{\pntxta )}}{\*\pnseclvl7\pnlcrm\pnstart1\pnindent720\pnhang{\pntxtb (}{\pntxta )}}{\*\pnseclvl8\pnlcltr\pnstart1\pnindent720\pnhang{\pntxtb (}{\pntxta )}}{\*\pnseclvl9\pnlcrm\pnstart1\pnindent720\pnhang {\pntxtb (}{\pntxta )}}\pard\plain \nowidctlpar\widctlpar\tx567\adjustright \fs28\lang1044\cgrid {\b\fs32\lang1033 THE EVOLUTION FROM PRIMITIVE ELVISH TO QUENYA \par A Comprehensive Survey}{\fs32\lang1033 \par }{\lang1033 \par Whereas many students of J.R.R. Tolkien's languages would be inclined to focus on their "classical" form \endash that is, Quenya and Sindarin as they are supposed to have existed at the time when the narratives of }{\i\lang1033 LotR}{\lang1033 and the } {\i\lang1033 Silmarillion}{\lang1033 took place \endash the author himself is often seen t o focus rather on the long evolution of these languages from the earliest forms of Elvish. Observes Christopher Tolkien, "My father was perhaps more interested in the processes of change than he was in displaying the structure and use of the languages at any given time" (LR:342). Elsewhere, he comments on how his father worked out a "minutely refined historical development of Quenya and Sindarin" (PM:367). \par \tab To the academic, the fact that the historical development was so important to Tolkien should be suffi cient to warrant a careful study of it. A deeper understanding of how the languages were supposed to have evolved during the long ages of Arda will also lead to a fuller apprehension of the method, spirit and focus of Tolkien's work. Yet this study does n o t have to be wholly "academic": People who want to develop a Tolkien-language like Quenya into a useable system, even coining new words from Tolkien's roots to supplement the existing vocabulary, must know what rules Tolkien himself applied when developin g Quenya vocabulary. It is a matter of debate whether a post-Tolkien Quenya word can be counted as a "genuine" word, but everybody would agree that a word that was not derived in accordance with Tolkien's rules and does not fit into his system cannot be co unted as genuine in any sense. \par \tab What follows is a comprehensive (not "exhaustive"!) list of phonological changes that occurred during the evolution of Quenya. Though I briefly describe the nature of the changes, I have focused more on providing a survey of the observed changes than on formulating abstract "rules" that may seem to govern the sound-shifts. The changes have been listed in what may be assumed to be something }{\i\lang1033 roughly}{\lang1033 similar to their chronological order. The material has been grouped under the following headings: }{\i\lang1033 Primitive Quendian}{\lang1033 , }{\i\lang1033 Common Eldarin}{\lang1033 , }{\i\lang1033 Pre-Record Quenya}{\lang1033 , }{\i\lang1033 Changes Producing Valinorean Quenya}{\lang1033 , and }{\i\lang1033 Changes Producing Exilic Quenya}{\lang1033 . In some cases, it is obvious where a certain phonological change belong (Tolkien may even have made an explic it statement about it). In many cases, it is less than obvious. It should be understood that this is in no way a "canonical" of "definite" listing as regards the chronology of Quenya sound-shifts. This said, the material here presented is quite sufficient to explain how the vast majority of Quenya words reached their "classical" or "contemporary" form. \par \tab I mention many of the clues and deductions that may throw some light on the chronological order of the changes, but when one studies these things in detail, it soon becomes obvious that the material is not entirely consistent. For instance, Tolkien seems to have experienced considerable trouble trying to figure out precisely }{\i\lang1033 when}{\lang1033 the shortening of the long final vowels of Primitive Elvish took place. Some of his "reconstructed" forms may seem to imply that this happened already during the Common Eldarin stage. On the other hand, the form }{\b\lang1033 Valin\'f4r\'ea}{\lang1033 "Valinor" in WJ:413 implies that the long final vowels were still in place as late as after initial }{\b\lang1033 b}{\lang1033 - had become }{\b\lang1033 v}{\lang1033 -, a change entirely characteristic of Quenya. (See also Appendix: Some Special Cases.) \par \par Some practical matters: In his notes, Tolkien mentions many primitive, "reconstructed" forms, and usually he even places an asterisk before such ancestral forms to mar k them as "unattested". Since Tolkien's own supposed "reconstructions" must be assumed to be }{\i\lang1033 rather more authoritative}{\lang1033 than my own, very much genuine reconstructions, I have distinguished between them here. Tolkien's alleged reconstructions are marked with the symbol \'a4 instead of *, the asterisk being reserved for my own reconstructions \endash forms that are really unattested in Tolkien's published material. (A double asterisk ** marks a wrong form.) Most of the forms here marked \'a4 are found in the }{\i\lang1033 Etymologies}{\lang1033 . To the extent primitive forms here mentioned cannot be readily located, the wordlist appended to my article about Primitive Elvish will provide references to the primary sources (see }{\cgrid0 http://www.uib.no/People/hnohf/primelv.htm}{\\lang1033 ). \par \tab In the sources, Tolkien usually marks long vowels by means of a }{\i\lang1033 macron}{\lang1033 , a line above the vowel. We will here use circumflexes instead: }{\b\lang1033 \'e2}{\lang1033 , }{\b\lang1033 \'ea}{\lang1033 , }{\b\lang1033 \'ee}{ \lang1033 , }{\b\lang1033 \'f4}{\lang1033 , }{\b\lang1033 \'fb}{\lang1033 . However, in the Quenya forms we will eventually reach, long vowels are marked by means of an accent instead: }{\b\lang1033 \'e1}{\lang1033 , }{\b\lang1033 \'e9}{\lang1033 , }{ \b\lang1033 \'ed}{\lang1033 , }{\b\lang1033 \'f3}{\lang1033 , }{\b\lang1033 \'fa}{\lang1033 . \par \tab The nasal sound }{\i\lang1033 ng}{\lang1033 as in English }{\i\lang1033 king}{\lang1033 is here represented by }{\b\lang1033 \'f1}{\lang1033 , a spelling often used by Tolkien himself. However, in accordance with Tolkien's own spelling the clusters }{\b\lang1033 ng}{\lang1033 and }{\b\lang1033 nk}{\lang1033 (}{\b\lang1033 nc}{\lang1033 ) in the middle of words are so represented, though this is technically }{\b\lang1033 \'f1g}{\lang1033 , }{\b\lang1033 \'f1k}{\lang1033 . \par \tab }{\i\lang1033 Syllabic}{\lang1033 consonants are here separated from the preceding part of the word by means of a dot, e.g. *}{\b\lang1033 hek\'b7l}{\lang1033 (the final }{\b\lang1033 l}{\lang1033 being syllabic, somewhat like English "little" is pronounced }{\i\lang1033 lit\'b7l}{\lang1033 .) An initial syllabic consonant is followed by a dot: *}{\b\lang1033 n\'b7d\'fbn\'ea}{\lang1033 . A syllabic consonant in the middle of a word is both preceded and followed by a dot: \'a4}{\b\lang1033 k}{\b\lang1033\super w}{\b\lang1033 ent\'b7r\'b7o}{\lang1033 . \par \tab In Tolkien's spelling of primitive forms, the semi-vowel }{\i\lang1033 y}{\lang1033 as in English }{\i\lang1033 you}{\lang1033 is inconsistently represented as either }{\b\lang1033 y}{\lang1033 or }{\b\lang1033 j}{\lang1033 (and in the }{ \i\lang1033 Etymologies}{\lang1033 as published in LR, Christopher Tolkien in most cases altered his father's original spelling, changing }{\b\lang1033 j}{\lang1033 to }{\b\lang1033 y}{\lang1033 ). In this study, we have consistently used the grapheme } {\b\lang1033 j}{\lang1033 when actual, primitive words are quoted, switching to }{\b\lang1033 y}{\lang1033 when we reach Quenya forms. (However, for easy reference, we maintain }{\i\lang1033 Y}{\lang1033 when referring to entry-heads in the }{ \i\lang1033 Etymologies}{\lang1033 , e.g. }{\i\lang1033 YUR}{\lang1033 rather than }{\i\lang1033 JUR}{\lang1033 . Inevitably there are here some orthographic inconsistencies, as when we quote the primitive form \'a4}{\b\lang1033 w\'e2j\'e2}{\lang1033 while referring to the entry }{\i\lang1033 WAY}{\lang1033 in Etym, but the graphemes }{\b\lang1033 j}{\lang1033 and }{\b\lang1033 y}{\lang1033 in any case refer to the same sound in all Elvish forms.) \par \tab The }{\i\lang1033 aspirated}{\lang1033 sounds of primitive Elvish are here marked by means of a raised }{\b\lang1033 h}{\lang1033 , as in }{\b\lang1033 p}{\b\lang1033\super h}{\lang1033 , }{\b\lang1033 t}{\b\lang1033\super h}{\lang1033 , }{ \b\lang1033 k}{\b\lang1033\super h}{\lang1033 . (In published sources, Tolkien simply wrote }{\b\lang1033 ph}{\lang1033 , }{\b\lang1033 th}{\lang1033 , }{\b\lang1033 kh}{\lang1033 .) \par \tab Elvish has certain }{\i\lang1033 labialized}{\lang1033 sounds, consonants pronounced with poised lips. The labialized quality of the consonant is here indicated by means of a raised }{\b\lang1033 w}{\lang1033 , e.g. }{\b\lang1033 k}{ \b\lang1033\super w}{\lang1033 , }{\b\lang1033 g}{\b\lang1033\super w}{\lang1033 , }{\b\lang1033 n}{\b\lang1033\super w}{\lang1033 , }{\b\lang1033 \'f1}{\b\lang1033\super w}{\lang1033 . (However, in normal Quenya orthography }{\b\lang1033 k}{ \b\lang1033\super w}{\lang1033 is written }{\b\lang1033 qu}{\lang1033 , and the other groups are represented as simple digraphs in -}{\b\lang1033 w}{\lang1033 , e.g. }{\b\lang1033 nw}{\lang1033 . The distinction between }{\b\lang1033 nw}{\lang1033 as a unitary, labialized consonant }{\b\lang1033 n}{\b\lang1033\super w}{\lang1033 and }{\b\lang1033 nw}{\lang1033 as a genuine cluster }{\b\lang1033 n}{\lang1033 + }{\b\lang1033 w}{\lang1033 is not upheld in the normal orthogr aphy. Indeed Tolkien does not uphold this distinction in any of his published writings.) \par \tab }{\i\lang1033 Palatalized}{\lang1033 consonants are here indicated by means of a raised }{\b\lang1033 j}{\lang1033 , e.g. }{\b\lang1033 n}{\b\lang1033\super j}{\b\lang1033 }{\lang1033 (= palatalized }{\b\lang1033 n}{\lang1033 , like Spanish }{ \b\lang1033 \'f1}{\lang1033 ) and }{\b\lang1033 t}{\b\lang1033\super j}{\b\lang1033 }{\lang1033 (more or less like the initial }{\i\lang1033 t}{\lang1033 of English "tune"). In Quenya orthography, the same sounds are represented as }{\b\lang1033 ny}{ \lang1033 and }{\b\lang1033 ty}{\lang1033 , though these combinations may also represent genuine clusters }{\b\lang1033 n}{\lang1033 + }{\b\lang1033 y}{\lang1033 , }{\b\lang1033 t}{\lang1033 + }{\b\lang1033 y}{\lang1033 . \par \tab The English alphabet has separate letters for only two }{\i\lang1033 spirant}{\lang1033 sounds, }{\b\lang1033 f}{\lang1033 and }{\b\lang1033 v}{\lang1033 . We will also use the special letters }{\b\lang1033 \'fe}{\lang1033 (more or less = English }{\i\lang1033 th}{\lang1033 as in }{\i\lang1033 think}{\lang1033 ) and }{\b\lang1033 \'f0}{\lang1033 (= soft English }{\i\lang1033 th}{\lang1033 , as in }{\i\lang1033 these}{\lang1033 ). The "back spirant", spirant }{\b\lang1033 G}{\lang1033 , Tolkien sometimes represented as a symbol resembling the number 3, and we will use }{\b\lang1033 3}{\lang1033 here as well. For }{\i\lang1033 ach-Laut}{\lang1033 we will resort to the digraph }{\b\lang1033 ch}{\lang1033 (as in Scottish }{\i\lang1033 loch}{\lang1033 \endash NOT as in English }{\i\lang1033 church}{\lang1033 , a sound that does not occur in any known Eldarin language). \par \tab }{\i\lang1033 Unvoiced}{\lang1033 }{\b\lang1033 w}{\lang1033 , }{\b\lang1033 n}{\lang1033 , }{\b\lang1033 m}{\lang1033 are here represented as digraphs }{\b\lang1033 hw}{\lang1033 , }{\b\lang1033 hn}{\lang1033 , }{\b\lang1033 hm}{\lang1033 . The digraph }{\b\lang1033 hy}{\lang1033 likewise represents voiceless }{\b\lang1033 y}{\lang1033 (= German }{\i\lang1033 ich-Laut}{\lang1033 ), as in the normal orthography of Quenya. \par \tab It will be noted that by introducing these special spellings, the exact pronunciation of primitive words "reconstructed" by Tolkien is made more specific than his original spellings are. While I do believe the interpretat ions here set forth are generally sound, they should not be taken as "Tolkien fact". For instance, the word here cited as \'a4}{\b\lang1033 k}{\b\lang1033\super w}{\b\lang1033 ent\'b7r\'b7o}{\lang1033 (with the symbol \'a4 to indicate a primitive form actually given in Tolkien's material) simply appears as "kwentro" in the source. While there are good reasons to assume that the initial "kw" is meant to be a unitary, labialized consonant }{\b\lang1033 k}{\b\lang1033\super w }{\lang1033 , and that the }{\b\lang1033 r}{\lang1033 at some point became syllabic, these phonological details are not directly confirmed by Tolkien's spelling and do not have absolute authority. \par \tab As this discussion reaches a recognizable form of Quenya, we introduce LotR-style Quenya orthography. (The relevant spelling conventions are set out in the Introduction to my Quenya course.) We have already commented on the use of }{\b\lang1033 y}{ \lang1033 instead of }{\b\lang1033 j}{\lang1033 ; also notice the use of }{\b\lang1033 c}{\lang1033 rather than }{\b\lang1033 k}{\lang1033 as well as }{\b\lang1033 x}{\lang1033 rather than }{\b\lang1033 cs}{\lang1033 /}{\b\lang1033 ks}{\lang1033 , plus the marking of certain vowels with diaereses (}{\b\lang1033 \'eb}{\lang1033 is so marked when occurring finally and in the combinations }{\b\lang1033 \'eba}{\lang1033 , }{\b\lang1033 \'ebo}{\lang1033 ). These divergences from the spelling of primitive forms do not suggest any phonological distinctions. \par \par We will now attempt to survey the changes that occurred during the evolution of Quenya from the earliest forms of Elvish, as far as Tolkien's intentions can be reconstructed. It must be understood that this discussion focuses on }{\i\lang1033 phonological }{\lang1033 developments: Of course there were also changes that had nothing to do with the regular sound-shifts (as when the old plural ending -}{\b\lang1033 i}{\lang1033 was in many cases replaced by the ending -}{\b\lang1033 r}{\lang1033 \endash according to PM:402 a "new device"). \par }{\fs36\lang1033 \par }{\i\fs32\lang1033 PRIMITIVE QUENDIAN}{\fs32\lang1033 \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par The "primeval" stage of Elvish speech, the language originally invented by the Elves at Cuivi\'e9nen, cannot be seen as a static and unchanging entity even within the primitive period. Tolkien explicitly refers to the "common }{\i\lang1033 development}{ \lang1033 of primitive Quendian" (VT39:8, emphasis added). There would be divergent developments even within the most primitive period, but th ere was apparently one unique point in pre-history when the various ultra-primitive language-constructions were unified and coordinated so that all Elves achieved one single language: "The hundred and forty-four Quendi [that originally awoke] dwelt long t ogether by the lake, until they all became of one mind and speech, and were glad" (WJ:423). \par \par Wrote Tolkien in an early source: "Roots...are n ot words in use at all, but serve as an elucidation of the words grouped together and a connection between them" (LT1:246). In all later forms of Elvish, the roots (stems, bases) can certainly be treated as abstract raw-material for actual words, mere ske letons to be fleshed out with various derivational endings, producing the kind of words that can actually be used in speech. Yet if there was ever a stage where the roots themselves existed as }{\i\lang1033 actual words}{\lang1033 , it must be the most primitive form of Elvish. At t he very beginning, we must probably imagine a fleeting era when as yet undifferentiated morphemes exist in a linguistic limbo, predating all derivation, all inflection, all distinguishing between different parts of speech. The first word ever spoken by th e Elves was supposedly }{\b\lang1033 ele}{\lang1033 , "a primitive exclamation, 'lo!', 'behold!' made by the Elves when they first saw the stars" (WJ:360). \par \par Later proto-words, if that is what they are, seem in most cases to have consisted of two (usually identical) short vowels separated by a medial consonant. Usually there was also an initial consonant, though it could be missing (as in }{\b\lang1033 ele}{ \lang1033 ). Forms later considered "roots", such as }{\i\lang1033 DELE}{\lang1033 "walk", }{\i\lang1033 HEKE}{\lang1033 "aside" or }{\i\lang1033 ABA}{\lang1033 "refuse" (WJ:360, 361), may be seen as representing actual, ultra-pr imitive words. This pattern came to be especially preferred in the case of stems with a }{\i\lang1033 verbal}{\lang1033 significance. At some point there occurred a "reorganisation of [...the] basic structure" of the language (WJ:392), some originally monosyllabic stems being expanded to conform with it. Hence the primeval root having to do with vocal speech, }{ \i\lang1033 K}{\i\lang1033\super W}{\i\lang1033 E}{\lang1033 , came to appear in elaborated forms }{\i\lang1033 K}{\i\lang1033\super W}{\i\lang1033 ENE}{\lang1033 and }{\i\lang1033 K}{\i\lang1033\super W}{\i\lang1033 ETE}{\lang1033 (the former underlying the very word }{\b\lang1033 Quenya}{\lang1033 , the latter producing the Q verb }{\b\lang1033 quet}{\lang1033 - "say, speak"). \par \par The "reorganisation" Tolkien hints at seems to suggest that the first Elves eventually developed a feeling for distinct parts of speech, their language gradually assuming a more well-defined structure. There appear derivational elements assisting the transfer of morphemes fr om one part of speech to another, or modifying the meaning of the proto-words, which words eventually begin to assume the function of "roots" from which actual words are }{\i\lang1033 derived}{\lang1033 . \par \par Later, in Valinor, the Eldar would analyze a Quenya word like }{\b\lang1033 tuil\'eb}{\lang1033 "spring" as representing }{\i\lang1033 tu-yu-le}{\lang1033 , becoming }{\i\lang1033 tuyle}{\lang1033 > }{\i\lang1033 tuile}{\lang1033 by the loss of the second vowel of the root }{\i\lang1033 TUJU}{\lang1033 (}{\i\lang1033 TUYU}{\lang1033 ) "sprout, bud" (VT39:7). Tolkien seems to hint that this analysis was essentially correct, or at least "later Loremasters" are said to have held similar views \endash "when Quendian origins had been further investigated" (VT39:11 No. 6). Given the form of the root }{\i\lang1033 TUJU}{\lang1033 , it might seem that the word for "spring" was originally simply this verbal root with the abstract ending -}{\b\lang1033 l \'ea}{\lang1033 attached: hence *}{\b\lang1033 tujul\'ea}{\lang1033 "sprouting, budding". \par \par However, Tolkien in the }{\i\lang1033 Etymologies}{\lang1033 listed \'a4}{\b\lang1033 tuil\'ea}{\lang1033 as the primitive form producing Quenya }{\b\lang1033 tuil\'eb}{\lang1033 . One of the very first phonological changes in Quendian may have been a primitive }{\i\lang1033 syncope}{\lang1033 whereby the second of two identical short vowels in many cases disappeared before a derivational ending: \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 *}{\b\lang1033 tujul\'ea}{\lang1033 > (*}{\b\lang1033 tujl\'ea}{\lang1033 >) \'a4}{\b\lang1033 tuil\'ea}{\lang1033 "spring" (or "budding") \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par Many of the words that seem to have been syncopated very early involve two-syllable bases (sometimes "original", sometimes extended from simpler roots, e.g. }{\i\lang1033 \'c1NAK}{\lang1033 from }{\i\lang1033 NAK}{\lang1033 or }{\i\lang1033 WENED}{ \lang1033 from }{\i\lang1033 WEN[E]}{\lang1033 ). In most cases only the fact that Tolkien listed the base itself (with the second vowel intact) allows us to deduce or theorize that some kind of syncope is operating. Some examples: \par \par \tab From the base }{\i\lang1033 \'c1SAT}{\lang1033 : *}{\b\lang1033 asat\'f4}{\lang1033 > \'a4}{\b\lang1033 ast\'f4}{\lang1033 "dust" (Quenya }{\i\lang1033 asto}{\lang1033 , Sindarin }{\i\lang1033 ast}{\lang1033 ) \par \tab From }{\i\lang1033 \'c1NAK}{\lang1033 : *}{\b\lang1033 anak\'e2}{\lang1033 > \'a4}{\b\lang1033 ank\'e2}{\lang1033 "jaw" (Q }{\i\lang1033 anca}{\lang1033 , S }{\i\lang1033 anc}{\lang1033 ) \par \tab From }{\i\lang1033 PARAK}{\lang1033 : *}{\b\lang1033 parak\'e2}{\lang1033 > *}{\b\lang1033 park\'e2}{\lang1033 "dry" (Q }{\i\lang1033 parca}{\lang1033 , S }{\i\lang1033 parch}{\lang1033 ) \par \tab From }{\i\lang1033 T\'c1WAR}{\lang1033 : *}{\b\lang1033 tawar\'ea}{\lang1033 > \'a4}{\b\lang1033 taur\'ea}{\lang1033 "forest" (Q }{\i\lang1033 taur\'eb}{\lang1033 , S }{\i\lang1033 taur}{\lang1033 ) \par \tab From }{\i\lang1033 T\'c9LEK}{\lang1033 : *}{\b\lang1033 teleku}{\lang1033 > *}{\b\lang1033 telku}{\lang1033 "leg" (Q }{\i\lang1033 telco}{\lang1033 , S }{\i\lang1033 telch}{\lang1033 ) \par \tab From }{\i\lang1033 \'daLUG}{\lang1033 : *}{\b\lang1033 ulugund\'f4}{\lang1033 > \'a4}{\b\lang1033 ulgund\'f4}{\lang1033 "monster" (Q }{\i\lang1033 ulundo}{\lang1033 , S }{\i\lang1033 ulun(d)}{\lang1033 ) \par \tab From }{\i\lang1033 WENED}{\lang1033 : *}{\b\lang1033 wened\'ea}{\lang1033 > *}{\b\lang1033 wend\'ea}{\lang1033 "maiden" (Q }{\i\lang1033 vend\'eb}{\lang1033 , S }{\i\lang1033 gwend}{\lang1033 ) \par \par }{\fs20\lang1033 NOTE: It will be noticed that Tolkien in many cases indicated (by means of an accent) that it is the }{\i\fs20\lang1033 first}{\fs20\lang1033 vowel of the relevant two-syllable bases that receives the stress. Where the stress was on the second vowel, no syncope occurred at this stage. Thus \'a4}{\b\fs20\lang1033 mor\'f3k\'f4}{\fs20\lang1033 "bear" from }{\i\fs20\lang1033 MOR\'d3K}{ \fs20\lang1033 did not become **}{\b\fs20\lang1033 mork\'f4}{\fs20\lang1033 (though the word did eventually end up as }{\b\fs20\lang1033 morco}{\fs20\lang1033 in Quenya, because of a later syncope). \par }{\lang1033 \par Here we list the eventual Sindarin forms as well as the Quenya forms, to demonstrate that these vowels must have been syncopated quite early. In Quenya, the second of two identical short vowels would have a "second chance" to become syncopated, but this happened after the branch of Elvish leading to Sindarin had separated from the branch leading to Quenya: When Quenya and Sindarin "agree" in having syncopated a certai n vowel, we can often assume that the syncope occurred already in their common ancestral language. In Etym, the entry }{\i\lang1033 USUK}{\lang1033 provides a good example: Tolkien listed a primitive word }{\b\lang1033 us}{\lang1033 (}{\b\lang1033 u}{ \lang1033 )}{\b\lang1033 k-w\'ea}{\lang1033 "reek", producing Q }{\b\lang1033 usqu\'eb}{\lang1033 and S }{\b\lang1033 osp}{\lang1033 . It seems that ultra-primitive \'a4}{\b\lang1033 usuk-w\'ea}{\lang1033 turned into \'a4}{\b\lang1033 usk-w\'ea}{ \lang1033 (or *}{\b\lang1033 usk}{\b\lang1033\super w}{\b\lang1033 \'ea}{\lang1033 , the labio-velar }{\b\lang1033 k}{\b\lang1033\super w}{\lang1033 becoming }{\b\lang1033 p}{\lang1033 in the Lindarin branch that Sindarin is a part of). *}{\b\lang1033 Usuk}{\b\lang1033\super w}{\b\lang1033 \'ea}{\lang1033 with no syncope might (via Old Sindarin **}{\b\lang1033 usupe}{\lang1033 , **}{\b\lang1033 uhupe}{\lang1033 ) have produced Sindarin **}{\b\lang1033 uhub}{\lang1033 instead of }{\b\lang1033 osp}{ \lang1033 . \par \par In the branch of Elvish that leads to Sindarin, an }{\i\lang1033 unaccented vowel}{\lang1033 in the first syllable of a word is very often syncopated, resulting in an initial consonant cluster: for instance, the base }{\i\lang1033 BOR\'d3N}{\lang1033 - produces Sindarin words in }{\b\lang1033 br}{\lang1033 - (e.g. }{\b\lang1033 br\'fbn}{\lang1033 "old" from }{\i\lang1033 b'r\'f4n\'e2}{\lang1033 , the first vowel of the original base dropping out). However, in the branch of Elvish that leads to Quenya, the first vowel of a base is nearly always preserved even where it was originally unaccented. We seem to have only one divergent example: the Quenya word }{ \b\lang1033 r\'e1ca}{\lang1033 "wolf" Tolkien referred to \'a4}{\i\lang1033 d'r\'e2k}{\lang1033 , a syncopated form of the stem }{\i\lang1033 DAR\'c1K}{\lang1033 . Actually the complete primitive word must have been *}{\i\lang1033 d'r\'e2k\'e2}{ \lang1033 , reduced from eaven earlier *}{\i\lang1033 dar\'e1k\'e2}{\lang1033 . \par \par As is evident from the example *}{\b\lang1033 tujul\'ea}{\lang1033 > (*}{\b\lang1033 tujl\'ea}{\lang1033 >) \'a4}{\b\lang1033 tuil\'ea}{\lang1033 "spring", a semi-vowel }{\b\lang1033 j}{\lang1033 occurring at the end of a root might turn into a full vowel }{\b\lang1033 i}{\lang1033 when it occurred in front of a consonant (or was brought into contact with one after the syncope). In a similar environment, }{\b\lang1033 w}{\lang1033 became }{ \b\lang1033 u}{\lang1033 . In the }{\i\lang1033 Etymologies}{\lang1033 , the bases }{\i\lang1033 T\'c1WAR}{\lang1033 (extended from }{\i\lang1033 TAWA}{\lang1033 , VT39:7) and }{\i\lang1033 TUY}{\lang1033 (= }{\i\lang1033 TUJU}{\lang1033 , VT39:7) produce primitive forms like \'a4}{\b\lang1033 taur\'ea}{\lang1033 "forest" and \'a4}{\b\lang1033 tuil\'ea}{\lang1033 "spring". Such words may be reworked from *}{\b\lang1033 tawr\'ea}{\lang1033 and *}{\b\lang1033 tujl\'ea}{\lang1033 , the syncopated forms of *}{\b\lang1033 tawar\'ea}{\lang1033 and *}{\b\lang1033 tujul\'ea}{\lang1033 . Cf. also one word derived from the base }{\i\lang1033 LAWAR}{\lang1033 : \'a4}{\b\lang1033 laur\'ea}{\lang1033 "golden light" (for *}{\b\lang1033 law'r\'ea}{\lang1033 ?) \par \par }\pard \nowidctlpar\widctlpar\tx567\adjustright {\lang1033 The syncope of the second of two identical short vowels in concomitant syllables is not entirely regular in the primitive language. There are quite a few examples of words that escaped this syncope. Examples of words in which the second vowel was }{ \i\lang1033 not}{\lang1033 lost include \'a4}{\b\lang1033 galad\'e2}{\lang1033 "tree", }{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 jeles\'ea}{\lang1033 "glass" and \'a4}{\b\lang1033 kjelep\'ea}{\lang1033 "silver" (cf. Sindarin }{\b\lang1033 gal} {\b\ul\lang1033 a}{\b\lang1033 dh}{\lang1033 , }{\b\lang1033 hel}{\b\ul\lang1033 e}{\lang1033 , }{\b\lang1033 cel}{\b\ul\lang1033 e}{\b\lang1033 b}{\lang1033 with the second vowel still intact: though it was eventually syncopated in Quenya }{\b\lang1033 al}{\b\ul\lang1033 '}{\b\lang1033 da}{\lang1033 , }{\b\lang1033 hyel}{\b\ul\lang1033 '}{\b\lang1033 l\'eb}{\lang1033 ,}{\b\lang1033 tyel}{\b\ul\lang1033 '}{\b\lang1033 p\'eb}{\lang1033 , this is a later deve lopment). It may indeed be doubted whether there was a primitive syncope at all, since some words mysteriously seem to have escaped it. When Tolkien lists a word like \'a4}{\b\lang1033 hekla}{\lang1033 "outcast" as a derivative of a root he quotes as }{ \i\lang1033 HEKE}{\lang1033 (WJ:361) are we to assume that \'a4}{\b\lang1033 hekla}{\lang1033 must necessarily be shortened from even earlier *}{\b\lang1033 hekela}{\lang1033 by a primitive syncope, or did the derivational patterns of Primitive Quendian allow the suppression of the second vowel of the root from the start? Regarding the situation in Common Eldarin , Tolkien did note that both leaving out and retaining the second root-vowel in derivatives were "legitimate and regular" procedures (VT41:9) \endash apparently simply a matter of choice. Anyhow: whether in PQ we have a somewhat sporadic and unpredictible syncop e, or a somewhat unpredictible suppression of reduplicated root-vowels in derivatives, is mainly an academic question. Of course, Tolkien may have imagined it differently at different times \endash and in any case even he himself preferred to deal with Primitive Quendian as a somewhat vague and hypothetical entity: The precise details were not necessarily to be pinned down in all cases, but were allowed to remain obscured by the passage of long ages. \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par Besides this early syncope, if it ever occurred, some }{\i\lang1033 final}{\lang1033 vowels may also have disappeared in the early period \endash so early that their former presence seems to have no influence on any of the later stages. In VT39:6, Tolkien states that in Quenya, "all final consonants had probably lost a vowel, if remote Quendian origins were considered". Earlier, in the }{ \i\lang1033 Etymologies}{\lang1033 , he had nonetheless cited a few forms that he himself identified as Primitive Quendian even though they ended in a consonant: \'a4}{\b\lang1033 atar}{\lang1033 "father", \'a4}{\b\lang1033 d\'ear}{\lang1033 "man", \'a4 }{\b\lang1033 m\'e23}{\lang1033 "hand" (but in both earlier and later sources, the primitive word for "hand" is said to have been \'a4}{\b\lang1033 maha}{\lang1033 ). Perhaps we are to understand that there was an even more primitive stage of PQ where these words did have a final vowel, but what its quality might have been, we can only speculate about. \par \par The normal plural marker -}{\b\lang1033 \'ee}{\lang1033 evidently had to be }{\i\lang1033 shortened}{\lang1033 following a long vowel (or any vowel?); the Primitive Quendian plural "Lindar" (Elves of the Third Clan) is given as \'a4}{\b\lang1033 Lind \'e2i}{\lang1033 , not **}{\b\lang1033 Lind\'e2\'ee}{\lang1033 (WJ:378). If the latter form ever existed, it must have been altered very early. Perhaps the plural of a word with a short final vowel, such as }{\b\lang1033 swanda}{\lang1033 "sponge", likewise appeared as *}{\b\lang1033 swandai}{\lang1033 rather than *}{\b\lang1033 swanda\'ee}{\lang1033 (the ultra-primitive form?) But a short final -}{\b\lang1033 e}{\lang1033 seems to have been displaced before }{\b\lang1033 \'ee}{ \lang1033 , as when the plural of \'a4}{\b\lang1033 k}{\b\lang1033\super w}{\b\lang1033 ende}{\lang1033 "elf" was \'a4}{\b\lang1033 k}{\b\lang1033\super w}{\b\lang1033 end\'ee}{\lang1033 (for ultra-primitive *}{\b\lang1033 k}{\b\lang1033\super w}{ \b\lang1033 ende\'ee}{\lang1033 ?) \par \par We will list some more very early changes, but whether they occurred in late Primitive Quendian or early Common Eldarin is difficult to say. Tolkien stated that the metathesis }{\b\lang1033 bm}{\lang1033 > }{\b\lang1033 mb}{\lang1033 occurre d in Common Eldarin at the latest, but added that it was "possibly earlier" (WJ:416), which necessarily implies that it may have occurred already in Primitive Quendian: \par \par \tab \'a4}{\b\lang1033 labm\'e2}{\lang1033 > \'a4}{\b\lang1033 lamb\'e2}{\lang1033 "[physical] tongue" (Q }{\i\lang1033 lamba}{\lang1033 ) \par \tab \'a4}{\b\lang1033 labm\'ea}{\lang1033 > \'a4}{\b\lang1033 lamb\'ea}{\lang1033 "tongue, language" (Q }{\i\lang1033 lamb\'eb}{\lang1033 ) \par \par (Given the form of the root }{\i\lang1033 LABA}{\lang1033 whence these words are derived, we may perhaps assume that their ultra-primitive, pre-syncope forms were *}{\b\lang1033 labam\'e2}{\lang1033 , *}{\b\lang1033 labam\'ea}{\lang1033 .) The change }{ \b\lang1033 bm}{\lang1033 > }{\b\lang1033 mb}{\lang1033 is an example of metathesis of a stop and a nasal sharing the same place of articulation; this is also the case with }{\b\lang1033 dn}{\lang1033 > }{\b\lang1033 nd}{\lang1033 , which seems to be another example of very early metathesis: \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \'a4}{\b\lang1033 adn\'f4}{\lang1033 > *}{\b\lang1033 and\'f4}{\lang1033 "gate" (Q }{\i\lang1033 ando}{\lang1033 ) \par \'a4}{\b\lang1033 sjadn\'f4}{\lang1033 > \'a4}{\b\lang1033 sjand\'f4}{\lang1033 "cleaver" (Q }{\i\lang1033 hyando}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par A long vowel -}{\b\lang1033 \'e2}{\lang1033 , -}{\b\lang1033 \'ea}{\lang1033 , -}{\b\lang1033 \'f4}{\lang1033 , or -}{\b\lang1033 \'fb}{\lang1033 + -}{\b\lang1033 i}{\lang1033 becomes a diphthong -}{\b\lang1033 ai}{\lang1033 , -}{\b\lang1033 ei}{ \lang1033 , -}{\b\lang1033 oi}{\lang1033 , or -}{\b\lang1033 ui}{\lang1033 : \par \par \tab \'a4}{\b\lang1033 Lind\'e2i}{\lang1033 > \'a4}{\b\lang1033 Lindai}{\lang1033 "Lindar" (WJ:378 vs. WJ:385), sg. }{\b\lang1033 Lind\'e2 \par }{\lang1033 \tab *}{\b\lang1033 orn\'eai}{\lang1033 > \'a4}{\b\lang1033 ornei}{\lang1033 pl. "trees" (SD:302), sg. }{\b\lang1033 orn\'ea}{\lang1033 \par \par (If the combination -}{\b\lang1033 \'eei}{\lang1033 ever occurred [e.g. *}{\b\lang1033 t\'e2r\'eei}{\lang1033 "queens"?], it was presumably contracted to -}{\b\lang1033 \'ee}{\lang1033 .) \par \par The combination Tolkien spells "KW" is probably to be taken as a labialized sound }{\b\lang1033 k}{\b\lang1033\super w}{\lang1033 ; thus, this is a single consonant rather than a cluster }{\b\lang1033 k}{\lang1033 + }{\b\lang1033 w}{\lang1033 . However, it seems that even genuine clusters }{\b\lang1033 k}{\lang1033 + }{\b\lang1033 w}{\lang1033 very early were assimilated to unitary consonants }{\b\lang1033 k}{\b\lang1033\super w}{\lang1033 : \par \par \tab \'a4}{\b\lang1033 hek-w\'e2}{\lang1033 (root }{\i\lang1033 HEK(E)}{\lang1033 + ending -}{\b\lang1033 w\'e2}{\lang1033 ) > *}{\b\lang1033 hek}{\b\lang1033\super w}{\b\lang1033 \'e2}{\lang1033 (Q }{\i\lang1033 hequa}{\lang1033 "except") \par \tab \'a4}{\b\lang1033 sal\'e1k-w\'ea}{\lang1033 (root }{\i\lang1033 SAL\'c1K}{\lang1033 + ending -}{\b\lang1033 w\'ea}{\lang1033 ) > *}{\b\lang1033 sal\'e1k}{\b\lang1033\super w}{\b\lang1033 \'ea}{\lang1033 (Q }{\i\lang1033 salqu\'eb}{\lang1033 "grass") \par \par Notice that the latter example will later undergo syncope in Quenya: the second }{\b\lang1033 a}{\lang1033 has disappeared in the Quenya form. However, as will be discussed below, such syncope did }{\i\lang1033 not}{\lang1033 occur in front of a consonant cluster. We are therefore able to tell that by the time the syncope occurred, the original cluster }{\b\lang1033 kw}{\lang1033 had merged into a single labialized consonant }{\b\lang1033 k}{\b\lang1033\super w}{\lang1033 . This }{\b\lang1033 k}{\b\lang1033\super w}{\lang1033\super }{\lang1033 survived into Quenya (in regularlized spelling written }{\b\lang1033 qu}{\lang1033 , but the fact that Tolkien in early sources represents it by means of the single grapheme }{ \b\lang1033 q}{\lang1033 also suggests that it is to be taken as a single consonant \endash }{\i\lang1033 salqu\'eb}{\lang1033 is indeed spelt "salqe" in the primary source). \par \par The aspirates are reduced to unaspirated stops before certain consonants. The aspirates }{\b\lang1033 k}{\b\lang1033\super h}{\lang1033 and }{\b\lang1033 t}{\b\lang1033\super h}{\lang1033 (we have no example for }{\b\lang1033 p}{\b\lang1033\super h}{ \lang1033 ) evidently become simple }{\b\lang1033 k}{\lang1033 , }{\b\lang1033 t}{\lang1033 before }{\b\lang1033 s}{\lang1033 . \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 From root }{\i\lang1033 LOK}{\i\lang1033\super H}{\lang1033 : *}{\b\lang1033 lok}{\b\lang1033\super h}{\b\lang1033 s\'ea}{\lang1033 > *}{\b\lang1033 loks\'ea}{\lang1033 (Q }{\i\lang1033 lox\'eb} {\lang1033 = }{\i\lang1033 loks\'eb}{\lang1033 ) \par From root }{\i\lang1033 K}{\i\lang1033\super H}{\i\lang1033 OT}{\i\lang1033\super H}{\lang1033 : *}{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 ot}{\b\lang1033\super h}{\b\lang1033 s\'ea}{\lang1033 > \'a4}{\b\lang1033 k}{\b\lang1033\super h}{ \b\lang1033 ots\'ea}{\lang1033 "assembly" (no known Q descendant) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par (The example \'a4}{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 ots\'ea}{\lang1033 , "reconstructed" by Tolkien himself, could also be taken to mean that aspirates could not occur before }{\b\lang1033 s}{\lang1033 even in Primitive Quendian, some phonological restraint prohibiting such combinations. The question has only academic interest.) \par \par Possibly, }{\b\lang1033 k}{\b\lang1033\super h}{\lang1033 becomes simple }{\b\lang1033 k}{\lang1033 before }{\b\lang1033 t}{\lang1033 : The Quenya verb }{\b\lang1033 rihta}{\lang1033 - "to jerk" may descend from either *}{\b\lang1033 rikt\'e2}{ \lang1033 or }{\b\lang1033 rik}{\b\lang1033\super h}{\b\lang1033 t\'e2}{\lang1033 ; Tolkien gave the root as }{\i\lang1033 RIK(H)}{\lang1033 , either }{\i\lang1033 RIK}{\i\lang1033\super H}{\lang1033 or }{\i\lang1033 RIK}{\lang1033 . But the Quenya form would be }{\b\lang1033 rihta}{\lang1033 - in any case, so we cannot be sure.}{\b\lang1033 \par }{\lang1033 \par }{\i\fs32\lang1033 COMMON ELDARIN \par }{\lang1033 \par Common Eldarin was the language spoken by the Eldar (as opposed to the Avari) during the Great March from Cuivi\'e9nen. This is the last common ancestor of all the later Eldarin tongues \endash Quenya, Telerin, Sindarin and Nandorin. Common Eldarin should not be thought of as an entirely homogeneous structure; apart from the linguistic changes that occurred during the March, there were a lready some minor dialectal differences. For instance, it seems that the Teleri had already turned the original labialized sound }{\b\lang1033 k}{\b\lang1033\super w}{\lang1033 into }{\b\lang1033 p}{\lang1033 (whereas it remained }{\b\lang1033 k}{ \b\lang1033\super w}{\lang1033 = }{\b\lang1033 qu}{\lang1033 among the Noldor and the Vanyar). \par \par The aspirate }{\b\lang1033 t}{\b\lang1033\super h}{\lang1033 and the voiced stop }{\b\lang1033 d}{\lang1033 both became }{\b\lang1033 s}{\lang1033 before }{\b\lang1033 t}{\lang1033 : \par \par \tab *}{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 ot}{\b\lang1033\super h}{\b\lang1033 t\'e2}{\lang1033 > *}{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 ost\'e2}{\lang1033 "to collect" (Q *}{\i\lang1033 chosta}{\lang1033 - > }{\i\lang1033 hosta}{\lang1033 -; cf. root }{\i\lang1033 K}{\i\lang1033\super H}{\i\lang1033 OT}{\i\lang1033\super H}{\lang1033 ) \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 *}{\b\lang1033 kot}{\b\lang1033\super h}{\b\lang1033 t\'e2}{\lang1033 > *}{\b\lang1033 kost\'e2}{\lang1033 "to quarrel" (Q }{\i\lang1033 costa}{\lang1033 -; Tolkien changed the root from }{ \i\lang1033 KOT}{\lang1033 to }{\i\lang1033 KOT}{\i\lang1033\super H}{\lang1033 ) \par *}{\b\lang1033 ledt\'e2}{\lang1033 > *}{\b\lang1033 lest\'e2}{\lang1033 "to leave" (Q }{\i\lang1033 lesta}{\lang1033 -, cf. the entries }{\i\lang1033 \'c9LED}{\lang1033 , }{\i\lang1033 LED}{\lang1033 in Etym) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par Another example of }{\b\lang1033 dt}{\lang1033 > }{\b\lang1033 st}{\lang1033 was \'a4}{\b\lang1033 wed-t\'e2}{\lang1033 > *}{\b\lang1033 west\'e2}{\lang1033 "to swear" (Q *}{\i\lang1033 westa}{\lang1033 > }{\i\lang1033 vesta}{\lang1033 ), but Tolkien dropped this word because it clashed with [*}{\b\lang1033 best\'e2}{\lang1033 >] }{\i\lang1033 vesta}{\lang1033 "to marry". See entries }{\i\lang1033 WED}{\lang1033 and }{\i\lang1033 BES}{\lang1033 in Etym, and cf. }{\b\lang1033 lesta}{ \lang1033 above. (It is not entirely clear whether the word }{\b\lang1033 lesta}{\lang1033 was rejected or not; the entry }{\i\lang1033 \'c9LED}{\lang1033 where it occurred was "replaced" by another version \endash but the phonological rule }{ \b\lang1033 dt}{\lang1033 > }{\b\lang1033 st}{\lang1033 may be "valid" all the same.) \par \par A }{\b\lang1033 t}{\lang1033 may also become }{\b\lang1033 s}{\lang1033 before another }{\b\lang1033 t}{\lang1033 }{\i\lang1033 when they are perceived to belong to different morphemes:}{\lang1033 \par \par \tab *}{\b\lang1033 put-t\'e2}{\lang1033 (root }{\i\lang1033 PUT}{\lang1033 + verbal ending -}{\b\lang1033 t\'e2}{\lang1033 ) > *}{\b\lang1033 pust\'e2}{\lang1033 "to stop" (Q }{\i\lang1033 pusta}{\lang1033 ) \par }\pard \qj\fi708\nowidctlpar\widctlpar\adjustright {\lang1033 *}{\b\lang1033 lot-t\'e2}{\lang1033 (root }{\i\lang1033 LOT}{\lang1033 + the same ending) > *}{\b\lang1033 lost\'e2}{\lang1033 "to bloom" (Q }{\i\lang1033 losta}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par The latter example comes from VT42:18, Tolkien explicitly commenting on the change "}{\i\lang1033 t}{\lang1033 -}{\i\lang1033 t}{\lang1033 in inflexion > }{\i\lang1033 st}{\lang1033 ". This change does not occur when }{\b\lang1033 tt}{\lang1033 occurs within a unitary word. Indeed the same root }{\i\lang1033 PUT}{\lang1033 (see LR:382 s.v. }{\i\lang1033 PUS}{\lang1033 ) that is the source of }{\b\lang1033 pusta}{\lang1033 - also yields Quenya }{\b\lang1033 putta}{\lang1033 "stop (in punctuation)", which is clearly meant to represent primitive *}{\b\lang1033 putt\'e2}{\lang1033 . It may be similar in sound to the primitive word that produced Q }{\b\lang1033 pusta}{\lang1033 , but here the double }{\b\lang1033 tt}{\lang1033 occurs as a medial fortification of the final consonant of the stem }{\i\lang1033 PUT}{\lang1033 . This *}{\b\lang1033 putt\'e2}{\lang1033 is therefore perceived as a unitary word, and the change }{\b\lang1033 tt}{\lang1033 > }{\b\lang1033 st}{ \lang1033 does not occur. \par \par The above-mentioned Quenya word }{\b\lang1033 costa}{\lang1033 "to quarrel" was originally an example of }{\b\lang1033 tt}{\lang1033 > }{\b\lang1033 st}{\lang1033 , Tolkien referring it to \'a4}{\b\lang1033 kot-t[\'e2]}{\lang1033 , but then he changed the root from }{\i\lang1033 KOT}{\lang1033 to }{\i\lang1033 KOT}{\i\lang1033\super H}{\lang1033 . \par \par The nasals }{\b\lang1033 m}{\lang1033 and }{\b\lang1033 \'f1}{\lang1033 both became }{\b\lang1033 n}{\lang1033 when preceding a }{\b\lang1033 t}{\lang1033 or }{\b\lang1033 d}{\lang1033 : \par \par \tab *}{\b\lang1033 kemtan\'f4}{\lang1033 > *}{\b\lang1033 kentan\'f4}{\lang1033 "potter" (Q }{\i\lang1033 centano}{\lang1033 \endash the first element seems to represent the root }{\i\lang1033 KEM}{\lang1033 "earth") \par \tab *}{\b\lang1033 pilimda}{\lang1033 (final vowel uncertain, may also be -}{\b\lang1033 o}{\lang1033 or -}{\b\lang1033 e}{\lang1033 ) > *}{\b\lang1033 pilinda}{\lang1033 "arrow" (Q }{\i\lang1033 pilin}{\lang1033 with stem }{\i\lang1033 pilind}{ \lang1033 -; cf. the base }{\i\lang1033 P\'cdLIM}{\lang1033 , LR:382) \par \tab *}{\b\lang1033 wi\'f1t\'e2}{\lang1033 > *}{\b\lang1033 wint\'e2}{\lang1033 "to scatter" (Q }{\i\lang1033 winta}{\lang1033 > *}{\i\lang1033 vinta}{\lang1033 ; if this word is to be related to the noun }{\i\lang1033 wing\'eb}{\lang1033 as suggested in PM:376, it seems to require a root *}{\i\lang1033 WI\'d1}{\lang1033 ) \par \par Final -}{\b\lang1033 w\'f4}{\lang1033 changes to -}{\b\lang1033 w\'e2}{\lang1033 : \par \par \tab \'a4}{\b\lang1033 nidw\'f4}{\lang1033 > *}{\b\lang1033 nidw\'e2}{\lang1033 "cushion" (Q }{\i\lang1033 nirwa}{\lang1033 ) \par \par Final -}{\b\lang1033 m\'e2}{\lang1033 becomes -}{\b\lang1033 m\'ea}{\lang1033 after a dental consonant: \par \par \tab \'a4}{\b\lang1033 jatm\'e2}{\lang1033 > *}{\b\lang1033 jatm\'ea}{\lang1033 "bridge" (Q }{\i\lang1033 yanw\'eb}{\lang1033 ) \par \par (Cf. also "Old Noldorin"/Old Sindarin }{\b\lang1033 katwe}{\lang1033 from earlier \'a4}{\b\lang1033 katw\'e2}{\lang1033 , demonstrating that -}{\b\lang1033 w\'e2}{\lang1033 likewise becomes -}{\b\lang1033 w\'ea}{\lang1033 > -}{\b\lang1033 we}{ \lang1033 following a dental consonant \endash see the entry }{\i\lang1033 KAT}{\lang1033 in Etym.) \par \par Medial }{\b\lang1033 \'f1}{\b\lang1033\super w}{\lang1033 turns into }{\b\lang1033 \'f1g}{\b\lang1033\super w}{\lang1033 : \par \par \tab \'a4}{\b\lang1033 li\'f1}{\b\lang1033\super w}{\b\lang1033 i}{\lang1033 > *}{\b\lang1033 li\'f1g}{\b\lang1033\super w}{\b\lang1033 i}{\lang1033 "fish" (Q }{\i\lang1033 lingw\'eb}{\lang1033 ) \par \tab \'a4}{\b\lang1033 ne\'f1}{\b\lang1033\super w}{\b\lang1033 i}{\lang1033 > *}{\b\lang1033 ne\'f1g}{\b\lang1033\super w}{\b\lang1033 i}{\lang1033 "nose" (Q }{\i\lang1033 nengw\'eb}{\lang1033 ) \par \par Evidently quite early (or even from the "start"), a }{\i\lang1033 long final}{\lang1033 }{\i\lang1033 vowel}{\lang1033 is shortened if the word in which it occurs appears as the first element of a compound: \par \par \tab \'a4}{\b\lang1033 and\'e2}{\lang1033 "long" + \'a4}{\b\lang1033 -mbund\'e2}{\lang1033 "snouted" = \'a4}{\b\lang1033 andamb\'fbnd\'e2}{\lang1033 (not *}{\i\lang1033 and}{\b\i\lang1033 \'e2}{\i\lang1033 mbund\'e2}{\lang1033 ) "long-snouted" \par \tab \'a4}{\b\lang1033 lass\'ea}{\lang1033 "leaf" + \'a4-}{\b\lang1033 k}{\b\lang1033\super w}{\b\lang1033 el\'eane}{\lang1033 "fading" = \'a4}{\b\lang1033 lassek}{\b\lang1033\super w}{\b\lang1033 el\'eane}{\lang1033 (not *}{\i\lang1033 lass}{ \b\i\lang1033 \'ea}{\i\lang1033 k}{\i\lang1033\super w}{\i\lang1033 el\'eane}{\lang1033 ) "autumn" \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \'a4}{\b\lang1033 tuil\'ea}{\lang1033 "spring" + \'a4}{\b\lang1033 lind\'f4}{\lang1033 "singer" = \'a4}{\b\lang1033 tuilelind\'f4}{\lang1033 (not *}{\i\lang1033 tuil}{\b\i\lang1033 \'ea}{ \i\lang1033 lind\'f4}{\lang1033 ) "spring-singer, swallow" \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par In some cases, but not regularly, the nasal element of the initial nasalized stops }{\b\lang1033 nd}{\lang1033 , }{\b\lang1033 \'f1g}{\lang1033 , }{\b\lang1033 mb}{\lang1033 develops into an independent syllabic consonant: \par \par \tab \'a4}{\b\lang1033 nd\'fbn\'ea}{\lang1033 > *}{\b\lang1033 n\'b7d\'fbn\'ea}{\lang1033 "sunset" (Q }{\i\lang1033 and\'fan\'eb}{\lang1033 ) \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 *}{\b\lang1033 \'f1gj\'f4 > }{\lang1033 *}{\b\lang1033 \'f1\'b7gj\'f4}{\lang1033 "grandchild, descendant" (Q }{\i\lang1033 indyo}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \tab *}{\b\lang1033 \'f1g}{\b\lang1033\super w}{\b\lang1033 al\'ea > }{\lang1033 *}{\b\lang1033 \'f1\'b7g}{\b\lang1033\super w}{\b\lang1033 al\'ea}{\lang1033 (Q }{\i\lang1033 ungwal\'eb}{\lang1033 ) \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 *}{\b\lang1033 mbarta}{\lang1033 > *}{\b\lang1033 m\'b7barta}{\lang1033 "fate" (Q }{\i\lang1033 umbar}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par In origin, this development seems to have functioned as a kind of "strengthening" that may also modify the meaning of the word: From \'a4}{\b\lang1033 \'f1g\'f4l\'ea}{\lang1033 "philosophy" is derived *}{\b\lang1033 \'f1\'b7g\'f4l\'ea}{\lang1033 "deep lore, magic", and these words continue to coexist (eventually producing Quenya }{\b\lang1033 \'f1\'f3l\'eb}{\lang1033 and }{\b\lang1033 ingol\'eb}{\lang1033 , respectively). But in most cases the "original" form with no syllabic nasal seems to have been lost completely, replaced by the "strengthened" form. \par \par Later Quenya developments suggest that the initial nasal of the group }{\b\lang1033 \'f1g}{\b\lang1033\super w}{\lang1033 - became labialized by assimilation to the following }{\b\lang1033 g}{\b\lang1033\super w}{\lang1033 , so that the whole group is technically }{\b\lang1033 \'f1}{\b\lang1033\super w}{\b\lang1033 g}{\b\lang1033\super w}{\lang1033 (or }{\b\lang1033 \'f1}{\b\lang1033\super w}{\b\lang1033 \'b7g}{\b\lang1033\super w}{\lang1033 where the nasal came to be syllabic); perhaps *}{\b\lang1033 \'f1\'b7g}{\b\lang1033\super w}{\b\lang1033 al\'ea}{\lang1033 above should more strictly be represented as *}{\b\lang1033 \'f1}{\b\lang1033\super w}{\b\lang1033 \'b7g}{\b\lang1033\super w}{ \b\lang1033 al\'ea}{\lang1033 . \par \par Some cases of haplology: \par \par \tab \'a4}{\b\lang1033 kuk\'fbw\'e2}{\lang1033 > *}{\b\lang1033 k\'fbw\'e2}{\lang1033 or *}{\b\lang1033 kuw\'e2}{\lang1033 "dove" (Q }{\i\lang1033 cua}{\lang1033 ) \par \tab \'a4}{\b\lang1033 tuilelind\'f4}{\lang1033 > *}{\b\lang1033 tuilind\'f4}{\lang1033 "spring-singer" = "swallow" (Q }{\i\lang1033 tuilindo}{\lang1033 ) \par \par The changes above may have occurred in the earliest stages of Common Eldarin. Some of the changes listed below may be imagined to have occurred later \endash say, after the Eld ar had crossed the Hithaeglir. (By this time, Common Eldarin would not be entirely "common" anymore, since the Nandor had already left the march.) \par \par It may seem that relatively early, the aspirates }{\b\lang1033 t}{\b\lang1033\super h}{\lang1033 , }{\b\lang1033 p}{\b\lang1033\super h}{\lang1033 , }{\b\lang1033 k}{\b\lang1033\super h}{\lang1033 became normal }{\b\lang1033 t}{\lang1033 , }{\b\lang1033 p}{\lang1033 , }{\b\lang1033 k}{\lang1033 immediately following another consonant. The word Quenya word }{\b\lang1033 ilfirin}{\lang1033 "immortal" is evidently meant to be derived from *}{\b\lang1033 l\'b7p}{\b\lang1033\super h}{\b\lang1033 irin}{ \lang1033 - (root }{\i\lang1033 P}{\i\lang1033\super H}{\i\lang1033 IR}{\lang1033 ), but in the relevant entry in Etym, Tolkien seems to indicate that this should regularly have produced Quenya *}{\b\lang1033 ilpirin}{\lang1033 instead. The actual f orm Quenya }{\b\lang1033 ilfirin}{\lang1033 was apparently reformed after }{\b\lang1033 firin}{\lang1033 "mortal" (< *}{\b\lang1033 p}{\b\lang1033\super h}{\b\lang1033 irin}{\lang1033 -). \par \par During the Common Eldarin stage, }{\b\lang1033 tk}{\b\lang1033\super h}{\lang1033 was apparently assimilated to }{\b\lang1033 kk}{\lang1033 . Perhaps }{\b\lang1033 tk}{\b\lang1033\super h}{\lang1033 first became }{\b\lang1033 tk}{\lang1033 in accordance with the rule set out above, and this in turn becomes }{\b\lang1033 kk}{\lang1033 (evidence from Sindarin indicates that }{\b\lang1033 tk}{\lang1033 of whatever origin normally became }{\b\lang1033 kk}{\lang1033 , whence Grey-elven }{ \b\lang1033 ch}{\lang1033 ). The later Quenya word }{\b\lang1033 eccaira}{\lang1033 "remote, far" must come from *}{\b\lang1033 ekkair\'e2}{\lang1033 , but since the root is given as }{\i\lang1033 K}{\i\lang1033\super H}{\i\lang1033 AYA}{\lang1033 , we must probably assume that at an even older stage, the word appeared as *}{\b\lang1033 etk}{\b\lang1033\super h}{\b\lang1033 air\'e2}{\lang1033 with the prefix }{\b\lang1033 et}{\lang1033 - "out-" prefixed. \par \par In one attested case, }{\b\lang1033 tk}{\lang1033 does not become }{\b\lang1033 kk}{\lang1033 , but undergoes metathesis to }{\b\lang1033 kt}{\lang1033 instead: \'a4}{\b\lang1033 et}{\i\lang1033 -}{\b\lang1033 kel\'ea}{\lang1033 *"out-flow" early becomes *}{\b\lang1033 ektel\'ea}{\lang1033 (later \'a4}{\b\lang1033 ektele}{\lang1033 , whence}{\b\lang1033 }{\lang1033 Quenya }{\b\lang1033 ehtel\'eb}{\lang1033 "spring, issue of water"). \par }{\fs36\lang1033 \par }{\b\lang1033 S}{\lang1033 is voiced to }{\b\lang1033 z}{\lang1033 before a voiced stop: \par \par \tab \'a4}{\b\lang1033 esd\'ea}{\lang1033 > \'a4}{\b\lang1033 ezd\'ea}{\lang1033 "rest" (WJ:403 seems to place this change in the CE era) \par \par Another example of the same would seem to be \'a4}{\b\lang1033 mazg\'e2}{\lang1033 "pliant"; given the base }{\i\lang1033 MASAG}{\lang1033 , this must come from earlier *}{\b\lang1033 masg\'e2}{\lang1033 . \par \par }{\b\lang1033 D}{\lang1033 becomes }{\b\lang1033 t}{\lang1033 before }{\b\lang1033 s}{\lang1033 : \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \'a4}{\b\lang1033 sjads\'ea}{\lang1033 > \'a4}{\b\lang1033 sjats\'ea}{\lang1033 "cleft, gash" (Q }{\i\lang1033 hyats\'eb}{\lang1033 )}{\fs36\lang1033 \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par Before an unvoiced consonant, the back spirant }{\b\lang1033 3}{\lang1033 lost its voicing by assimilation, producing }{\b\lang1033 ch}{\lang1033 . The first of the examples below is explicitly identified as an "Eld" (Common Eldarin) form in Etym. \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \'a4}{\b\lang1033 ma3t\'e2}{\lang1033 > \'a4}{\b\lang1033 macht\'e2}{\lang1033 - "to handle" (Q }{\i\lang1033 mahta}{\lang1033 -) }{\i\fs20\lang1033 (In Etym, entry MA3, the CE word is spelt }{ \b\i\fs20\lang1033 maht\'e2}{\i\fs20\lang1033 -, but it is clear that Tolkien is here using the letter }{\b\i\fs20\lang1033 h}{\i\fs20\lang1033 for [x] rather than a simple breath-}{\b\i\fs20\lang1033 h}{\i\fs20\lang1033 ; cf. Quenya }{ \b\i\fs20\lang1033 mahta}{\i\fs20\lang1033 -, where it is clear from the descriptions in LotR Appendix E that }{\b\i\fs20\lang1033 ht}{\i\fs20\lang1033 represents [xt].)}{\fs20\lang1033 \par }{\lang1033 *}{\b\lang1033 wa3s\'ea}{\lang1033 > \'a4}{\b\lang1033 wachs\'ea}{\lang1033 "stain" (Q *}{\i\lang1033 wax\'eb}{\lang1033 > }{\i\lang1033 vax\'eb}{\lang1033 ) }{\i\fs20\lang1033 (again Tolkien's actual spelling of the primitive word is "wahs \'ea")}{\lang1033 \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par }{\i\lang1033 THE LOSS OF MEDIAL }{\b\i\lang1033 3}{\i\lang1033 (}{\b\i\lang1033 H}{\i\lang1033 ) AND }{\b\i\lang1033 \'d1}{\i\lang1033 : }{\lang1033 Except initially, these sounds normally disappeared, though }{\b\lang1033 3}{\lang1033 became }{ \b\lang1033 ch}{\lang1033 before unvoiced consonants (see above) and }{\b\lang1033 \'f1}{\lang1033 survived before }{\b\lang1033 g}{\lang1033 and }{\b\lang1033 k}{\lang1033 (including labialized }{\b\lang1033 g}{\b\lang1033\super w}{\lang1033 , }{ \b\lang1033 k}{\b\lang1033\super w}{\lang1033 \endash the labialized sound }{\b\lang1033 \'f1}{\b\lang1033\super w}{\lang1033 arising from }{\b\lang1033 \'f1}{\lang1033 + }{\b\lang1033 w}{\lang1033 had already become }{\b\lang1033 \'f1g}{ \b\lang1033\super w}{\lang1033 and thus preserved }{\b\lang1033 \'f1}{\lang1033 ). Much later, the language would re-acquire }{\b\lang1033 3}{\lang1033 , developed from earlier }{\b\lang1033 g}{\lang1033 , only to lose it once again (even initially). The loss of "original" }{\b\lang1033 3}{\lang1033 and this "later" }{\b\lang1033 3}{\lang1033 has certain parallels, but whereas the "later" }{\b\lang1033 3}{\lang1033 survived into historical times (it was recorded in R\'familian orthography, at least in initial position), the "original" }{\b\lang1033 3}{\lang1033 or }{\b\lang1033 h}{\lang1033 was lost very early: Tolkien also referred to the loss of intervocalic }{ \b\lang1033 h}{\lang1033 as "prehistoric" (VT39:11). Even more definite is the statement made in WJ:368: "Medial }{\i\lang1033 h}{\lang1033 was very early lost without trace in CE" \endash tying this change to the Common Eldarin stage. \par \par A final }{\b\lang1033 3}{\lang1033 following }{\b\lang1033 u}{\lang1033 might develop into }{\b\lang1033 w}{\lang1033 (or perhaps rather: the loss of }{\b\lang1033 3}{\lang1033 triggered the development of a semi-vowel -}{\b\lang1033 w}{\lang1033 from the corresponding full vowel -}{\b\lang1033 u}{\lang1033 , as if to compensate for the lost -}{\b\lang1033 3}{\lang1033 ): \par \par \tab \tab \'a4}{\b\lang1033 ku3}{\lang1033 > \'a4}{\b\lang1033 kuw}{\lang1033 "bow" (Q }{\i\lang1033 c\'fa}{\lang1033 ) \par \par Sometimes, the loss of intervocalic }{\b\lang1033 \'f1}{\lang1033 and }{\b\lang1033 3}{\lang1033 /}{\b\lang1033 h}{\lang1033 would not cause any further changes, but simply leave two vowels in hiatus; occasionally a new diphthong might arise after the loss, like }{\b\lang1033 ai}{\lang1033 in the second example below: \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 *}{\b\lang1033 te\'f1\'e2}{\lang1033 > *}{\b\lang1033 te\'e2}{\lang1033 > "indicates" (Q }{\i\lang1033 t\'eba}{\lang1033 ; cf. VT39:6) \par \'a4}{\b\lang1033 ma3iti}{\lang1033 > *}{\b\lang1033 maiti}{\lang1033 "handy" (Q }{\i\lang1033 mait\'eb}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par The primitive dual \'a4}{\b\lang1033 pe\'f1\'fb}{\lang1033 producing Quenya }{\b\lang1033 peu}{\lang1033 "pair of lips" would also be an example of a new diphthong arising after the loss of }{\b\lang1033 \'f1}{\lang1033 . Whether }{\b\lang1033 e\'fb}{ \lang1033 became }{\b\lang1033 eu}{\lang1033 "directly", or remained as two vowels in hiatus until the much later shortening of the final vowels, has only academic interest. \par \par The Quenya genitive ending -}{\b\lang1033 o}{\lang1033 is derived from the root }{\i\lang1033 HO}{\lang1033 or }{\i\lang1033 3O}{\lang1033 after this loss of medial }{\b\lang1033 h}{\lang1033 /}{\b\lang1033 3}{\lang1033 (a primitive postposition }{ \b\lang1033 3\'f4}{\lang1033 "from" being directly suffixed and turning into *-}{\b\lang1033 \'f4}{\lang1033 , later becoming -}{\b\lang1033 o}{\lang1033 ). \par \par Two identical short vowels brought into contact by the loss of }{\b\lang1033 \'f1}{\lang1033 or }{\b\lang1033 3}{\lang1033 /}{\b\lang1033 h}{\lang1033 would merge into one (long) vowel: \par }{\b\lang1033 \par }{\lang1033 \tab \'a4}{\b\lang1033 pe\'f1e}{\lang1033 > }{\b\lang1033 p\'ea}{\lang1033 "lip" (VT39:11; still }{\i\lang1033 p\'e9}{\lang1033 in Quenya) \par \tab \'a4}{\b\lang1033 maha}{\lang1033 > }{\b\lang1033 m\'e2}{\lang1033 "hand" (VT39:11; still }{\i\lang1033 m\'e1}{\lang1033 in Quenya) \par \par }{\fs20\lang1033 NOTE: These examples are interesting, indicating that the loss of medial }{\b\fs20\lang1033 \'f1}{\fs20\lang1033 and }{\b\fs20\lang1033 h}{\fs20\lang1033 occurred very early, since short final -}{\b\fs20\lang1033 a}{\fs20\lang1033 and -}{\b\fs20\lang1033 e}{\fs20\lang1033 were lost during the Common Eldarin stage; see below. \'a4}{\b\fs20\lang1033 Pe\'f1e}{\fs20\lang1033 was evidently contracted to }{\b\fs20\lang1033 p\'ea}{\fs20\lang1033 }{\i\fs20\lang1033 before}{ \fs20\lang1033 \'a4}{\b\fs20\lang1033 pe\'f1e}{\fs20\lang1033 would have been reduced to **}{\b\fs20\lang1033 p\'ea\'f1}{\fs20\lang1033 (with the same lengthening of the vowel as in PQ \'a4}{\b\fs20\lang1033 k}{\b\fs20\lang1033\super w}{ \b\fs20\lang1033 ene}{\fs20\lang1033 > CE \'a4}{\b\fs20\lang1033 k}{\b\fs20\lang1033\super w}{\b\fs20\lang1033 \'ean}{\fs20\lang1033 ).While *}{\b\fs20\lang1033 p\'ea\'f1}{\fs20\lang1033 would probably also have produced Quenya }{\b\fs20\lang1033 p\'e9} {\fs20\lang1033 , this does not seem to have been the way Tolkien imagined it, since he referred to }{\b\fs20\lang1033 p\'ea}{\fs20\lang1033 as a "contraction" of \'a4}{\b\fs20\lang1033 pe\'f1e}{\fs20\lang1033 (VT39:11). \par }{\lang1033 \par However, if the last vowel was }{\i\lang1033 long}{\lang1033 , and the loss of }{\b\lang1033 \'f1}{\lang1033 or }{\b\lang1033 3}{\lang1033 brought it into contact with another vowel of similar quality (long or short), the first vowel might be dissimilated, like }{\b\lang1033 e\'ea}{\lang1033 > }{\b\lang1033 i\'ea}{\lang1033 : \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \'a4}{\b\lang1033 te3\'ea}{\lang1033 or *}{\b\lang1033 te\'f1\'ea}{\lang1033 > (*}{\b\lang1033 te\'ea}{\lang1033 >) *}{\b\lang1033 ti\'ea}{\lang1033 "path" (Q }{\i\lang1033 ti\'eb}{\lang1033 ; see the entries }{\i\lang1033 TE3}{\lang1033 , }{\i\lang1033 TE\'d1}{\lang1033 in Etym, the latter replacing the former) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par Possibly the form *}{\b\lang1033 te\'ea}{\lang1033 survived for a while, the actual dissimilation to *}{\b\lang1033 ti\'ea}{\lang1033 taking place much closer to historical times: Similar dissimilation can be observed later as well, as when \'a4}{ \b\lang1033 w\'ea3\'ea}{\lang1033 produces Quenya }{\b\lang1033 vi\'eb}{\lang1033 "manhood" (via *}{\b\lang1033 w\'ea\'ea}{\lang1033 > *}{\b\lang1033 w\'ee\'ea}{\lang1033 ). But in this case }{\b\lang1033 3}{\lang1033 is not original; it is altered from }{\b\lang1033 g}{\lang1033 , since the root not *}{\i\lang1033 WE3}{\lang1033 , but }{\i\lang1033 WEG}{\lang1033 (LR:398). The change }{\b\lang1033 g}{\lang1033 > }{\b\lang1033 3}{\lang1033 (and its subsequent loss) occurred much later. \par \par Where }{\b\lang1033 3}{\lang1033 or }{\b\lang1033 \'f1}{\lang1033 disappeared before a consonant, the preceding vowel was lengthened in compensation: \par \par \tab \'a4}{\b\lang1033 do3m\'ea}{\lang1033 > *}{\b\lang1033 d\'f4m\'ea}{\lang1033 "night" (Q }{\i\lang1033 l\'f3m\'eb}{\lang1033 ) \par \tab \'a4}{\b\lang1033 te3r\'e2}{\lang1033 or \'a4}{\b\lang1033 te\'f1r\'e2}{\lang1033 > *}{\b\lang1033 t\'ear\'e2}{\lang1033 "straight, right" (Q }{\i\lang1033 t\'e9ra}{\lang1033 ; again see }{\i\lang1033 TE3}{\lang1033 , }{\i\lang1033 TE\'d1}{ \lang1033 in Etym) \par \tab \'a4}{\b\lang1033 wa3r\'e2}{\lang1033 > *}{\b\lang1033 w\'e2r\'e2}{\lang1033 "dirty" (Q *}{\i\lang1033 w\'e1ra}{\lang1033 > }{\i\lang1033 v\'e1ra}{\lang1033 ) \par \par It should be noted that while }{\i\lang1033 medial}{\lang1033 }{\b\lang1033 h}{\lang1033 /}{\b\lang1033 3}{\lang1033 disappeared, it surprisingly seems to have survived }{\i\lang1033 finally}{\lang1033 , being lost only much later in that position. This phenomenon is not explicitly referred to in Tolkien's published writings, but we shall probably need final }{\b\lang1033 3}{\lang1033 later, to explain the Book Quenya accusative \endash formed by lengthening the final vowel of a noun. This probably indicates the former presence of a final }{\b\lang1033 3}{\lang1033 as a primitive accusative ending. But if we let final }{\b\lang1033 3}{\lang1033 disappear already in Common Eldarin, it would be lost too early to exert any influence over the final vowels of later Quenya. So a word like *}{\b\lang1033 lass\'ea3}{\lang1033 or *}{\b\lang1033 lass\'eah}{\lang1033 , the probable primitive accusative of \'a4}{\b\lang1033 lass\'ea}{\lang1033 "leaf", must remain unchanged for now. \par \par During the Common Eldarin stage, the short final vowels -}{\b\lang1033 a}{\lang1033 , -}{\b\lang1033 e}{\lang1033 and -}{\b\lang1033 o}{\lang1033 were lost. Examples: \par \par Loss of final short -}{\b\lang1033 a}{\lang1033 ; where the ending -}{\b\lang1033 la}{\lang1033 is so reduced, the remaining }{\b\lang1033 l}{\lang1033 came to be syllabic: \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \'a4}{\b\lang1033 swanda}{\lang1033 > *}{\b\lang1033 swand}{\lang1033 "sponge" (Q }{\i\lang1033 hwan}{\lang1033 with stem }{\i\lang1033 hwand}{\lang1033 -) \par \'a4}{\b\lang1033 jakta}{\lang1033 > *}{\b\lang1033 jakt}{\lang1033 "neck" (Q }{\i\lang1033 yat}{\lang1033 with stem }{\i\lang1033 yaht}{\lang1033 -) \par \'a4}{\b\lang1033 hekla}{\lang1033 > *}{\b\lang1033 hek\'b7l}{\lang1033 "outcast" (Q }{\i\lang1033 hecil}{\lang1033 ) \par \'a4}{\b\lang1033 makla}{\lang1033 > *}{\b\lang1033 mak\'b7l}{\lang1033 "sword" (Q }{\i\lang1033 macil}{\lang1033 ) \par \'a4}{\b\lang1033 tankla}{\lang1033 > *}{\b\lang1033 tank\'b7l}{\lang1033 "brooch" (Q }{\i\lang1033 tancil}{\lang1033 ) \par \'a4}{\b\lang1033 tekla}{\lang1033 > *}{\b\lang1033 tek\'b7l}{\lang1033 "pen" (Q }{\i\lang1033 tecil}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par Loss of final short -}{\b\lang1033 e}{\lang1033 : \par \tab \'a4}{\b\lang1033 i-ndise}{\lang1033 > }{\b\lang1033 indis}{\lang1033 "bride" \par \tab PQ \'a4}{\b\lang1033 k}{\b\lang1033\super w}{\b\lang1033 ene}{\lang1033 "person" > CE \'a4}{\b\lang1033 k}{\b\lang1033\super w}{\b\lang1033 \'ean}{\lang1033 (WJ:360) \par \tab \'a4}{\b\lang1033 nere}{\lang1033 "man" > }{\b\lang1033 n\'ear}{\lang1033 (WJ:393) \par \par Loss of final short -}{\b\lang1033 o}{\lang1033 : \par \tab \'a4}{\b\lang1033 abaro}{\lang1033 "refuser" > \'a4}{\b\lang1033 abar}{\lang1033 (WJ:371 identifies the latter as a CE form) \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \'a4}{\b\lang1033 nd\'f4ro}{\lang1033 > *}{\b\lang1033 nd\'f4r}{\lang1033 "land" (Q }{\i\lang1033 n\'f3r}{\lang1033 , WJ:413) \par \'a4}{\b\lang1033 Spanturo}{\lang1033 > *}{\b\lang1033 Spantur}{\lang1033 "Cloud-lord" (Q }{\i\lang1033 Fantur}{\lang1033 ) \par \'a4}{\b\lang1033 tollo}{\lang1033 > *}{\b\lang1033 toll}{\lang1033 "island" (Q }{\i\lang1033 tol}{\lang1033 with stem }{\i\lang1033 toll}{\lang1033 -) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \tab \'a4}{\b\lang1033 Denwego}{\lang1033 > *}{\b\lang1033 Denweg}{\lang1033 (personal name, Q }{\i\lang1033 Lenw\'eb}{\lang1033 ) \par \par As is evident from the examples \'a4}{\b\lang1033 k}{\b\lang1033\super w}{\b\lang1033 ene}{\lang1033 "person" > CE \'a4}{\b\lang1033 k}{\b\lang1033\super w}{\b\lang1033 \'ean}{\lang1033 (WJ:360) and \'a4}{\b\lang1033 nere}{\lang1033 "man" > }{ \b\lang1033 n\'ear}{\lang1033 (WJ:393) already quoted, a word that becomes monosyllabic by the loss of a final vowel has its stem-vowel lengthened in compensation. But where the word occurs compounded or with endings and thus is not a monosyllable, the vowel remains short. Therefore the CE plural of \'a4}{\b\lang1033 k}{\b\lang1033\super w}{\b\lang1033 \'ean}{\lang1033 is given as \'a4}{\b\lang1033 k}{\b\lang1033\super w}{\b\lang1033 en\'ee}{\lang1033 (WJ:360), not **}{ \b\lang1033 k}{\b\lang1033\super w}{\b\lang1033 \'ean\'ee}{\lang1033 . This variation is still reflected in Quenya }{\b\lang1033 qu\'e9n}{\lang1033 pl. }{\b\lang1033 queni}{\lang1033 (WJ:361; likewise }{\b\lang1033 n\'e9r}{\lang1033 "man" pl. }{ \b\lang1033 neri}{\lang1033 , MR:213). \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 Where the loss of a final short vowel made a semi-vowel -}{\b\lang1033 w}{\lang1033 final, it turned into a full vowel }{\b\lang1033 u}{\lang1033 : \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \par (\'a4}{\b\lang1033 angwa}{\lang1033 >) CE *}{\b\lang1033 angw}{\lang1033 > *}{\b\lang1033 angu}{\lang1033 "snake" (Q }{\i\lang1033 ango}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par }{\fs20\lang1033 [Compare the entry-head }{\i\fs20\lang1033 ANGWA}{\fs20\lang1033 in Etym with the stem-form }{\b\fs20\lang1033 angu}{\fs20\lang1033 - of the Quenya noun }{\b\fs20\lang1033 ango}{\fs20\lang1033 , seen in the compound }{\b\fs20\lang1033 ang}{\b\fs20\ul\lang1033 u}{\b\fs20\lang1033 l\'f3ce}{\fs20\lang1033 , entry }{\i\fs20\lang1033 LOK}{\fs20\lang1033 . Like -}{\b\fs20\lang1033 w}{\fs20\lang1033 became -}{\b\fs20\lang1033 u}{\fs20\lang1033 , a -}{\b\fs20\lang1033 j}{\fs20\lang1033 becoming final would likewise turn into -}{\b\fs20\lang1033 i}{\fs20\lang1033 ; cf. the example \'a4}{\b\fs20\lang1033 talrunja}{\fs20\lang1033 "sole of foot" producing Quenya }{\b\fs20\lang1033 tallun\'eb}{\fs20\lang1033 , evidently via *}{ \b\fs20\lang1033 talrunj}{\fs20\lang1033 and then *}{\b\fs20\lang1033 talruni}{\fs20\lang1033 ; see below. In this case the loss of a final -}{\b\fs20\lang1033 a}{\fs20\lang1033 is later than CE, indicating that the rule for how final -}{ \b\fs20\lang1033 w}{\fs20\lang1033 and -}{\b\fs20\lang1033 j}{\fs20\lang1033 behaved (turning into -}{\b\fs20\lang1033 u}{\fs20\lang1033 and -}{\b\fs20\lang1033 i}{\fs20\lang1033 ) continued to operate in the earliest stages of Quenya.]}{ \i\fs20\lang1033 \par }{\lang1033 \par A final semi-vowel following a long vowel could likewise develop into a separate vowel: \par \par \tab (PQ ?*}{\i\lang1033 r\'e2wa}{\lang1033 >) CE *}{\b\lang1033 r\'e2w}{\lang1033 > \'a4}{\b\lang1033 r\'e2u}{\lang1033 "lion" \par \par This is our sole certain example of this phenomenon (Tolkien only provided the root }{\i\lang1033 RAW}{\lang1033 and the form \'a4}{\b\lang1033 r\'e2u}{\lang1033 ). This -}{\b\lang1033 u}{\lang1033 was later lost, evidently also during the Common Eldarin stage. It is unclear whether we should generalize a rule "a short final vowel immediately following a long vowel is lost": \par \par \tab \'a4}{\b\lang1033 r\'e2u}{\lang1033 > }{\b\lang1033 r\'e2}{\lang1033 "lion" (still }{\i\lang1033 r\'e1}{\lang1033 in Q) \par \par }{\fs20\lang1033 NOTE: If the entry-head }{\i\fs20\lang1033 NOWO}{\fs20\lang1033 in Etym is taken as representing a PQ word *}{\b\fs20\lang1033 nowo}{\fs20\lang1033 , Quenya }{\b\fs20\lang1033 n\'f3}{\fs20\lang1033 "conception, idea" }{\i\fs20\lang1033 might}{\fs20\lang1033 arise in this way: PQ *}{\b\fs20\lang1033 nowo}{\fs20\lang1033 first becomes CE *}{\b\fs20\lang1033 n\'f4w}{\fs20\lang1033 (just like }{\b\fs20\lang1033 k}{\b\fs20\lang1033\super w}{\b\fs20\lang1033 ene}{\fs20\lang1033 becomes } {\b\fs20\lang1033 k}{\b\fs20\lang1033\super w}{\b\fs20\lang1033 \'ean}{\fs20\lang1033 ), this *}{\b\fs20\lang1033 n\'f4w}{\fs20\lang1033 then turns into *}{\b\fs20\lang1033 n\'f4u}{\fs20\lang1033 , and the final -}{\b\fs20\lang1033 u}{\fs20\lang1033 is subsequently lost leaving only *}{\b\fs20\lang1033 n\'f4}{\fs20\lang1033 = Q }{\b\fs20\lang1033 n\'f3}{\fs20\lang1033 . \par }{\lang1033 \par After the loss of a final vowel by any of the processes described above, the plural forms were evidently reformed by analogy with the new simplex form. In Primitive Quendian, the plural of \'a4}{\b\lang1033 swanda}{\lang1033 "sponge" must have been *}{ \b\lang1033 swanda\'ee}{\lang1033 or (later?) *}{\b\lang1033 swandai}{\lang1033 , but when \'a4}{\b\lang1033 swanda}{\lang1033 was reduced to *}{\b\lang1033 swand}{\lang1033 in CE, its plural was altered to *}{\b\lang1033 swand\'ee}{\lang1033 (still reflected in Q }{\i\lang1033 hwandi}{\lang1033 ) \endash as if this were a normal consonant stem. The original plural of PQ *}{\b\lang1033 r\'e2wa}{\lang1033 (or *}{\b\lang1033 r\'e2we}{\lang1033 , *}{\b\lang1033 r\'e2wo}{\lang1033 ) "lion" must likewise have been altered from *}{\b\lang1033 r\'e2wai}{\lang1033 (or *}{\b\lang1033 r\'e2wei}{\lang1033 , *}{\b\lang1033 r\'e2woi}{\lang1033 ) to *}{\b\lang1033 r\'e2w\'ee}{\lang1033 . Here, }{\b\lang1033 w}{\lang1033 was not final and therefore did not become }{\b\lang1033 u}{\lang1033 , later lost as described above. So in Quenya, the plural of }{\b\lang1033 r\'e1}{\lang1033 "lion" is still }{\b\lang1033 r\'e1vi}{\lang1033 , reflecting CE *}{\b\lang1033 r\'e2w\'ee} {\lang1033 and the original root }{\i\lang1033 RAW}{\lang1033 . \par \par Evidently during the latter part of the Common Eldarin stage, short final -}{\b\lang1033 i}{\lang1033 turned into -}{\b\lang1033 e}{\lang1033 : \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \'a4}{\b\lang1033 d\'f4mi}{\lang1033 > *}{\b\lang1033 d\'f4me}{\lang1033 "twilight" (Q }{\i\lang1033 l\'f3m\'eb}{\lang1033 ) \par \'a4}{\b\lang1033 pori}{\lang1033 > *}{\b\lang1033 pore}{\lang1033 "flour, meal" (Q }{\i\lang1033 por\'eb}{\lang1033 )}{\b\lang1033 \par }{\lang1033 *}{\b\lang1033 kari}{\lang1033 > *}{\b\lang1033 kare}{\lang1033 "does" }{\fs20\lang1033 (aorist form of the verb }{\b\fs20\lang1033 kar}{\fs20\lang1033 -) }{\lang1033 (Q }{\i\lang1033 car\'eb}{\lang1033 ) \par *}{\b\lang1033 karini}{\lang1033 > *}{\b\lang1033 karine}{\lang1033 "I do" }{\fs20\lang1033 (aorist with pronominal ending; cf. Old "Noldorin"/Sindarin }{\b\fs20\lang1033 yurine}{\fs20\lang1033 "I run", see note below) }{\lang1033 (Q }{\i\lang1033 carin}{\lang1033 ) \par *}{\b\lang1033 talruni}{\lang1033 > *}{\b\lang1033 talrune}{\lang1033 "sole of foot" (Q }{\i\lang1033 tallun\'eb}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par The vowels did }{\i\lang1033 not}{\lang1033 change their quality when they were not final, and so in Quenya we still have the variation seen for instance in the aorist, as in }{\b\lang1033 care}{\lang1033 "does" besides }{\b\lang1033 carin}{\lang1033 "I do" (the first evolving from *}{\b\lang1033 kari}{\lang1033 with the change }{\b\lang1033 i}{\lang1033 > }{\b\lang1033 e}{\lang1033 , the second from *}{\b\lang1033 karini}{\lang1033 with a 1st person ending, the pronominal suffix *-}{\b\lang1033 ni}{\lang1033 precluding the change }{\b\lang1033 i}{\lang1033 > }{\b\lang1033 e}{\lang1033 because the vowel was not final). \par \par }{\fs20\lang1033 NOTE: We date this change to the Common Eldarin stage because it is also seen in Old "Noldorin"/Sindarin: The form }{\b\fs20\lang1033 yurine}{\fs20\lang1033 "I run" (}{\i\fs20\lang1033 YUR}{\fs20\lang1033 ) exemplifies an I-stem *}{ \b\fs20\lang1033 yuri}{\fs20\lang1033 - with an ending -}{\b\fs20\lang1033 ne}{\fs20\lang1033 "I", whereas the form }{\b\fs20\lang1033 trenare}{\fs20\lang1033 "he recounts" (}{\i\fs20\lang1033 NAR}{\i\fs20\lang1033\super 2}{\fs20\lang1033 ) shows that the change from older *}{\b\fs20\lang1033 trenari}{\fs20\lang1033 has already occurred: Add the ending -}{\b\fs20\lang1033 ne}{\fs20\lang1033 "I" to }{\b\fs20\lang1033 trenare}{\fs20\lang1033 so that the final -}{\b\fs20\lang1033 e}{ \fs20\lang1033 is not final, and we would evidently see *}{\b\fs20\lang1033 trenari-ne}{\fs20\lang1033 with the original quality preserved. The ending -}{\b\fs20\lang1033 ne}{\fs20\lang1033 "I" would itself exemplify the change from -}{ \b\fs20\lang1033 i}{\fs20\lang1033 to -}{\b\fs20\lang1033 e}{\fs20\lang1033 , since it must come from earlier }{\b\fs20\lang1033 ni}{\fs20\lang1033 ; compare the stem }{\i\fs20\lang1033 NI}{\i\fs20\lang1033\super 2}{\fs20\lang1033 , simply defined "I". Another relevant "ON"/OS form is }{\b\fs20\lang1033 barane}{\fs20\lang1033 "brown", which is surely meant to represent older *}{\b\fs20\lang1033 barani}{\fs20\lang1033 (cf. Quenya }{\b\fs20\lang1033 varn\'eb}{\fs20\lang1033 with stem }{\b\fs20\lang1033 varni}{\fs20\lang1033 -). However, the change -}{\b\fs20\lang1033 i}{\fs20\lang1033 > -}{\b\fs20\lang1033 e}{\fs20\lang1033 did not happen at the }{\i\fs20\lang1033 beginning}{\fs20\lang1033 of the Common Eldarin stage; this is evident from the fact that Tolkien refers to }{\b\fs20\lang1033 phini}{\fs20\lang1033 "a hair" [> Q }{\b\fs20\lang1033 fin\'eb}{\fs20\lang1033 ] as a Common El darin form in PM:362 (cf. PM:340 for Q form). It may also be noted that }{\i\fs20\lang1033 Nandorin}{\fs20\lang1033 , Green-Elven, seems to have branched off from Common Eldarin before this change occurred; see the entry }{\b\fs20\lang1033 lygn}{ \fs20\lang1033 in the wordlist appended to my article about Nandorin. \par \par }{\lang1033 As a parallel to the change from short -}{\b\lang1033 i}{\lang1033 to -}{\b\lang1033 e}{\lang1033 , short final -}{\b\lang1033 u}{\lang1033 becomes -}{\b\lang1033 o}{\lang1033 : \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 *}{\b\lang1033 kuru}{\lang1033 > }{\b\lang1033 kuro}{\lang1033 "skillful device" (in Q still }{\i\lang1033 curo}{\lang1033 with stem }{\i\lang1033 curu}{\lang1033 -; the examples below must also have stems in -}{\i\lang1033 u}{\lang1033 when endings are added to them) \par \'a4}{\b\lang1033 smalu}{\lang1033 > *}{\b\lang1033 smalo}{\lang1033 "pollen" (Q }{\i\lang1033 malo}{\lang1033 ) \par \'a4}{\b\lang1033 tundu}{\lang1033 > *}{\b\lang1033 tundo}{\lang1033 "hill, mound" (Q }{\i\lang1033 tundo}{\lang1033 ) \par \'a4}{\b\lang1033 ranku}{\lang1033 > *}{\b\lang1033 ranko}{\lang1033 "arm" (Q }{\i\lang1033 ranco}{\lang1033 ) \par *}{\b\lang1033 angu}{\lang1033 > *}{\b\lang1033 ango}{\lang1033 "snake" (Q }{\i\lang1033 ango}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\fs20\lang1033 \par }{\i\fs32\lang1033 PRE-RECORD QUENYA \par }{\lang1033 As far as the imaginary Arda timeline is concerned, the March is now past, and the Lindarin branch of Elvish (that leads to Sindarin and Amanya Telerin) has now definitely separated from the "Vanya-\'d1 oldorin branch" (VT39:10) that we will continue to pursue here. The Vanyar and the Noldor have settled in Aman, but the linguistic development of this period still has to be "reconstructed"; Quenya sti ll has not been reduced to writing. An actual word of what seems to be pre-record Quenya is found in WJ:402: }{\b\lang1033 Mbelek\'f4ro}{\lang1033 is there said to be the "oldest Q[uenya] form" of the name }{\b\lang1033 Melkor}{\lang1033 . This word indicates that the primitive nasalized stops }{\b\lang1033 mb}{\lang1033 , }{\b\lang1033 nd}{\lang1033 , }{\b\lang1033 \'f1g}{\lang1033 were still in place initially, and that the syncope still had not taken place: }{\b\lang1033 Mbelek}{ \lang1033 - would later become }{\b\lang1033 Melk}{\lang1033 -. Tolkien actually asterisked the form }{\b\lang1033 Mbelek\'f4ro}{\lang1033 , so it is clear that R\'famil still had not invented writing. Hence this can properly be called }{\i\lang1033 pre-record Quenya}{\lang1033 . \par \tab It was in the period that now followed that the language underwent the greatest changes. }{"Its altering ...[came]...in the softening and harmonizing of the sounds and patterns of the Quendian tongue to forms that seemed to the Noldor more beaut iful" (WJ:20).}{\lang1033 \par \par A long final vowel occurring at the end of a compound is normally }{\i\lang1033 shortened}{\lang1033 ; Tolkien explicitly dated this change to "the earlier stages of Quenya" (VT39:6): \par \par \tab *}{\b\lang1033 Eruk}{\b\lang1033\super h}{\b\lang1033 \'een\'e2}{\lang1033 > *}{\b\lang1033 Eruk}{\b\lang1033\super h}{\b\lang1033 \'eena}{\lang1033 "Child of Eru" (*}{\b\lang1033 Er\'fb}{\lang1033 , *}{\b\lang1033 Eru}{\lang1033 - "God" + }{ \b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 \'een\'e2}{\lang1033 "child"; cf. WJ:403 s.v. }{\i\lang1033 Atan}{\lang1033 ) \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 *}{\b\lang1033 tal-runj\'e2}{\lang1033 > \'a4}{\b\lang1033 talrunja}{\lang1033 "sole of foot" (*}{\b\lang1033 runj\'e2}{\lang1033 "sole" combined with }{\b\lang1033 tal}{\lang1033 - "foot"; Q } {\i\lang1033 tallun\'eb}{\lang1033 ) \par *}{\b\lang1033 pel-taks\'ea}{\lang1033 > \'a4}{\b\lang1033 peltakse}{\lang1033 "pivot" (\'a4}{\b\lang1033 taks\'ea}{\lang1033 "nail" combined with }{\b\lang1033 pel}{\lang1033 - "revolve on fixed point"; Q }{\i\lang1033 peltas}{\lang1033 with stem }{ \i\lang1033 peltax}{\lang1033 -) \par \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 Cf. also -}{\b\lang1033 wego}{\lang1033 as the "compound form" of }{\b\lang1033 weg\'f4}{\lang1033 "man" (LR:398 s.v. }{\i\lang1033 WEG}{\lang1033 ). \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 However, the plural ending -}{\b\lang1033 \'ee}{\lang1033 is not subject to this change (\'a4}{\b\lang1033 kala-k}{\b\lang1033\super w}{\b\lang1033 end\'ee}{\lang1033 "light-elves" was not reduced to **}{\b\lang1033 kala-k}{\b\lang1033\super w}{\b\lang1033 endi}{\lang1033 ). It is also evident that many Quenya compounds were formed }{\i\lang1033 after}{\lang1033 this rule ceased to be effective \endash for instance, }{\b\lang1033 Elent\'e1ri}{\lang1033 "Star-Queen" would rather have been **}{\b\lang1033 Elentar}{\lang1033 if this descended from *}{\b\lang1033 Elent\'e2r\'ee}{\lang1033 at the oldest stage. One example mentioned above, *}{\b\lang1033 kemtan\'f4}{\lang1033 > *}{\b\lang1033 kentan\'f4}{\lang1033 "potter", may likewise actually be a later formation (suggesting that the rule }{\b\lang1033 mt}{\lang1033 > }{\b\lang1033 nt}{\lang1033 continued to operate in the language): If this had been reduced to *}{\b\lang1033 kentano}{\lang1033 in early Quenya, the Quenya form might have been **}{\b\lang1033 centan}{\lang1033 instead of }{\b\lang1033 centano}{\lang1033 because of the loss of certain final vowels that we will now discuss. \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par At some later stage, in words of three or more syllables, a short final vowel may occasionally be lost (but not wholly regularly so): \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \'a4}{\b\lang1033 Awadelo}{\lang1033 > *}{\b\lang1033 Awadel}{\lang1033 "Away-goer" (Q }{\i\lang1033 Auzel}{\lang1033 > }{\i\lang1033 Aurel}{\lang1033 ; this word refers to the Elves that departed from Beleriand and went to Valinor, WJ:360, 363; the form below is a variant) \par \'a4}{\b\lang1033 Aw\'e2delo}{\lang1033 > *}{\b\lang1033 Aw\'e2del}{\lang1033 "Away-goer" (Q }{\i\lang1033 Oazel}{\lang1033 > }{\i\lang1033 Oarel}{\lang1033 ) \par \'a4}{\b\lang1033 Mbelek\'f4ro}{\lang1033 > *}{\b\lang1033 Mbelek\'f4r}{\lang1033 "Melkor" (Q }{\i\lang1033 Melkor}{\lang1033 ) \par (\'a4}{\i\lang1033 etsiri}{\lang1033 >) late CE *}{\b\lang1033 etsire}{\lang1033 > Q }{\b\lang1033 etsir}{\lang1033 "mouth of a river" \par (*}{\i\lang1033 karini}{\lang1033 >) late CE *}{\b\lang1033 karine}{\lang1033 > Q }{\b\lang1033 carin}{\lang1033 "I do" \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par However, this loss seems to occur quite regularly in the case of the final vowels that had been shortened in compounds (another change dated by Tolkien to "the earlier stages of Quenya", VT39:6): \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (*}{\i\lang1033 Eru-k}{\i\lang1033\super h}{\i\lang1033 \'een\'e2}{\lang1033 >) *}{\b\lang1033 Eruk}{\b\lang1033\super h}{\b\lang1033 \'eena}{\lang1033 > *}{\b\lang1033 Eruk}{\b\lang1033\super h }{\b\lang1033 \'een}{\lang1033 "Child of Eru" (later Q }{\i\lang1033 Eruhin}{\lang1033 with stem }{\i\lang1033 Eruh\'edn}{\lang1033 -) \par (*}{\i\lang1033 pel-taks\'ea}{\lang1033 >) \'a4}{\b\lang1033 peltakse}{\lang1033 > *}{\b\lang1033 peltaks}{\lang1033 "pivot" (later Q }{\i\lang1033 peltas}{\lang1033 with stem }{\i\lang1033 peltax}{\lang1033 -) \par (*}{\i\lang1033 k}{\i\lang1033\super h}{\i\lang1033 ala-tirn\'f4}{\lang1033 >) \'a4}{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 alatirno >}{\lang1033 *}{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 alatirn}{\lang1033 "fish-watcher" (name of a bird, English kingfisher; in later Q }{\i\lang1033 halatir}{\lang1033 with stem }{\i\lang1033 halatirn}{\lang1033 -) \par (*}{\i\lang1033 Mand-ost\'f4}{\lang1033 >) \'a4}{\b\lang1033 Mandosto}{\lang1033 > *}{\b\lang1033 Mandost}{\lang1033 (later Q }{\i\lang1033 Mandos}{\lang1033 with stem }{\i\lang1033 Mandost}{\lang1033 -, MR:350 cf. VT39:6)}{\b\lang1033 \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par }{\fs20\lang1033 [Perhaps \'a4}{\b\fs20\lang1033 Mbelek\'f4ro}{\fs20\lang1033 > *}{\b\fs20\lang1033 Mbelek\'f4r}{\fs20\lang1033 also belongs in this list, if the "oldest Q form" represents a compound *}{\b\fs20\lang1033 Mbelek}{\fs20\lang1033 - "mighty" + -}{\b\fs20\lang1033 \'f4ro}{\fs20\lang1033 , shortening of *}{\b\fs20\lang1033 \'f4r\'f4}{\fs20\lang1033 ?"rising" or ?"one who rises". The name }{\b\fs20\lang1033 Melkor}{\fs20\lang1033 \endash as it later appears \endash supposedly means "mighty-rising", also translated "He that arises in might".]}{\lang1033 \par \par Where a semi-vowel }{\b\lang1033 j}{\lang1033 became final because of the loss of a final vowel, it turned into a full vowel -}{\b\lang1033 i}{\lang1033 : \par }{\b\lang1033 \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \'a4}{\b\lang1033 talrunja}{\lang1033 > CE }{\b\lang1033 (}{\lang1033 *}{\b\lang1033 talrunj}{\lang1033 >) *}{\b\lang1033 talruni}{\lang1033 "sole of foot" (Q }{\i\lang1033 tallun\'eb}{ \lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par As is evident from the Quenya form, short final -}{\b\lang1033 i}{\lang1033 at some point became -}{\b\lang1033 e}{\lang1033 just as in Common Eldarin, so it would seem that this rule continued to operate in the language: *}{\b\lang1033 talruni}{ \lang1033 became *}{\b\lang1033 talrune}{\lang1033 (later Q }{\i\lang1033 tallun\'eb}{\lang1033 ). This parallels the change of -}{\b\lang1033 w}{\lang1033 > -}{\b\lang1033 u}{\lang1033 > -}{\b\lang1033 o}{\lang1033 in Common Eldarin (as when \'a4}{ \b\lang1033 angwa}{\lang1033 "snake" becomes *}{\b\lang1033 angw}{\lang1033 > *}{\b\lang1033 angu}{\lang1033 > Q }{\b\lang1033 ango}{\lang1033 ), and it is essentially the same rule continuing to operate in the language (but after this stage, it apparently ceases to function). \par \par Very early, there were some instances of }{\i\lang1033 metathesis}{\lang1033 , certain sounds often switching place. One example is DL > LD, as in \'a4}{\b\lang1033 Edl\'f4}{\lang1033 > *}{\b\lang1033 Eld\'f4}{\lang1033 (Q }{\b\lang1033 Eldo}{\lang1033 , a variant of }{\b\lang1033 Elda}{\lang1033 "Elf", though ultimately with another etymology \endash see WJ:363). Tolkien wrote that "transposition of }{\i\lang1033 dl}{\lang1033 in primary contact to the favoured sequence }{\i\lang1033 ld}{\lang1033 not infrequently [but not universally either?] occurred in the pre-record period of Quenya" (WJ:363). Later, more systematic metathesis would occur. \par \par This may be the stage where Quenya loses many of the initial consonant clusters of the primitive language. The initial combinations }{\b\lang1033 sk}{\lang1033 -, }{\b\lang1033 sp}{\lang1033 -, }{\b\lang1033 st}{\lang1033 - evidently turned into aspirates }{\b\lang1033 k}{\b\lang1033\super h}{\lang1033 -, }{\b\lang1033 p}{\b\lang1033\super h}{\lang1033 -, }{\b\lang1033 t}{\b\lang1033\super h}{\lang1033 -, merging with the original aspirates (but medially these groups were apparently unchanged, cf. }{\b\lang1033 sk}{\lang1033 surviving in a Quenya word like }{\b\lang1033 rusco}{\lang1033 "fox"). This change is not directly "attested" in Tolkien's published papers, but we know he imagined such a change in the branch of Eldarin that lead to Sindarin. (In both Quenya and Sindarin the original aspirates }{\i\lang1033 and}{\lang1033 the initial combinations of }{\b\lang1033 s}{\lang1033 + }{\b\lang1033 k}{\lang1033 , }{\b\lang1033 p}{\lang1033 or }{\b\lang1033 t}{\lang1033 will later become spirants }{\b\lang1033 ch}{\lang1033 [> }{\b\lang1033 h}{\lang1033 ], }{\b\lang1033 f}{ \lang1033 , }{\b\lang1033 \'fe}{\lang1033 , respectively. However, Amanya Telerin retains the combinations in }{\b\lang1033 s}{\lang1033 - unchanged, so this change must have happened independently in Quenya and Sindarin: If it occurred already in Common Eldarin, we would have seen it in Telerin as well.) \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \'a4}{\b\lang1033 skarw\'ea}{\lang1033 > *}{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 arw\'ea}{\lang1033 "wound" (Q *}{\i\lang1033 charw\'eb}{\lang1033 > }{\i\lang1033 harw\'eb}{ \lang1033 ) \par \'a4}{\b\lang1033 skjapat}{\lang1033 - > *}{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 japat}{\lang1033 "shore" (Q }{\i\lang1033 hyapat}{\lang1033 ) \par *}{\b\lang1033 sk}{\b\lang1033\super w}{\b\lang1033 arm\'e2}{\lang1033 > *}{\b\lang1033 k}{\b\lang1033\super hw}{\b\lang1033 arm\'e2}{\lang1033 "crossbar" (Q }{\i\lang1033 hwarma}{\lang1033 )}{\b\lang1033 \par }{\lang1033 \'a4}{\b\lang1033 spang\'e2}{\lang1033 > *}{\b\lang1033 p}{\b\lang1033\super h}{\b\lang1033 ang\'e2}{\lang1033 "beard" (Q }{\i\lang1033 fanga}{\lang1033 ) \par \'a4}{\b\lang1033 stank\'e2}{\lang1033 > *}{\b\lang1033 t}{\b\lang1033\super h}{\b\lang1033 ank\'e2}{\lang1033 "split" (Q *}{\i\lang1033 \'feanca}{\lang1033 > }{\i\lang1033 sanca}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par The initial groups }{\b\lang1033 sl}{\lang1033 - and }{\b\lang1033 sr}{\lang1033 - become }{\b\lang1033 hl}{\lang1033 - and }{\b\lang1033 hr}{\lang1033 -, sc. unvoiced }{\b\lang1033 l}{\lang1033 and }{\b\lang1033 r}{\lang1033 : \par \par \tab \'a4}{\b\lang1033 sr\'e2w\'ea}{\lang1033 > *}{\b\lang1033 hr\'e2w\'ea}{\lang1033 "flesh" (Q }{\i\lang1033 hr\'e1v\'eb}{\lang1033 ) \par \tab *}{\b\lang1033 sl\'f4k\'ea}{\lang1033 > *}{\b\lang1033 hl\'f4k\'ea}{\lang1033 "dragon" (Q }{\i\lang1033 hl\'f3c\'eb}{\lang1033 ) \par \par }{\fs20\lang1033 NOTE: In the scenario of the }{\i\fs20\lang1033 Etymologies}{\fs20\lang1033 , original initial }{\b\fs20\lang1033 sl}{\fs20\lang1033 - seems to be simplified to }{\b\fs20\lang1033 l}{\fs20\lang1033 - at a relat ively early stage; as one of the primitive derivatives of the stem }{\i\fs20\lang1033 SLIG}{\fs20\lang1033 we have \'a4}{\b\fs20\lang1033 lig\'e2}{\fs20\lang1033 (presumably representing even earlier *}{\b\fs20\lang1033 slig\'e2}{\fs20\lang1033 , cf. other primitive derivatives from the same root, \'a4}{\b\fs20\lang1033 slign\'ea}{\fs20\lang1033 or \'a4}{\b\fs20\lang1033 sling\'ea}{\fs20\lang1033 ). In Etym, all Quenya words derived from stems in }{\b\fs20\lang1033 sl}{\fs20\lang1033 - show simple }{\b\fs20\lang1033 l}{\fs20\lang1033 -, which is of course logical enough if }{\b\fs20\lang1033 s}{\fs20\lang1033 early disappeared with no trace. (Etym provides no example of how initial }{\b\fs20\lang1033 sr}{\fs20\lang1033 - came out in Quenya.) But later, Tolkien decided that primitive initial }{\b\fs20\lang1033 sr}{\fs20\lang1033 - and }{\b\fs20\lang1033 sl}{\fs20\lang1033 - produced Quenya }{\b\fs20\lang1033 hr}{\fs20\lang1033 - and }{\b\fs20\lang1033 hl}{\fs20\lang1033 -, though this had come to be pronounced as normal }{\b\fs20\lang1033 r}{\fs20\lang1033 - and }{\b\fs20\lang1033 l}{\fs20\lang1033 - by the Third Age. So when Etym has a primitive word like (say) \'a4}{\b\fs20\lang1033 sl\'eew\'ea}{\fs20\lang1033 yielding Quenya }{\b\fs20\lang1033 l\'edv\'eb}{\fs20\lang1033 instead of *}{\b\fs20\lang1033 hl\'edv\'eb}{\fs20\lang1033 , we may see }{\b\fs20\lang1033 l\'edv\'eb}{\fs20\lang1033 as representing the Third Age }{\i\fs20\lang1033 pronunciation}{ \fs20\lang1033 . But in reality, Tolkien probably did not imagine any intermediate stage with }{\b\fs20\lang1033 hl}{\fs20\lang1033 - when he wrote Etym in the thirties; the introduction of }{\b\fs20\lang1033 hr}{\fs20\lang1033 - and }{\b\fs20\lang1033 hl }{\fs20\lang1033 - as Quenya sounds was a later idea. In the pre-LotR period it would seem that he meant }{\b\fs20\lang1033 sl}{\fs20\lang1033 - to have been simplified to }{\b\fs20\lang1033 l}{\fs20\lang1033 - already in the primitive days. Cf. }{ \b\fs20\lang1033 sm}{\fs20\lang1033 - and }{\b\fs20\lang1033 sn}{\fs20\lang1033 - below.}{\lang1033 \par \par The initial groups }{\b\lang1033 sm}{\lang1033 - and }{\b\lang1033 sn}{\lang1033 - are simplified to }{\b\lang1033 m}{\lang1033 - and }{\b\lang1033 n}{\lang1033 - (conceivably there may have been a intermediate stage with unvoiced }{\b\lang1033 hm}{ \lang1033 - and }{\b\lang1033 hn}{\lang1033 -, paralleling }{\b\lang1033 hl}{\lang1033 - and }{\b\lang1033 hr}{\lang1033 - above, but there is no direct evidence that Tolkien imagined this to have been the case): \par \par \tab \'a4}{\b\lang1033 smalin\'e2}{\lang1033 > *}{\b\lang1033 malin\'e2}{\lang1033 "yellow" (Q }{\i\lang1033 malina}{\lang1033 ) \par \tab *}{\b\lang1033 snard\'e2}{\lang1033 > *}{\b\lang1033 nard\'e2}{\lang1033 "knot" (Q }{\i\lang1033 narda}{\lang1033 ) \par \par Perhaps it is at this stage certain primitive initial clusters in -}{\b\lang1033 j}{\lang1033 are altered. Primitive }{\b\lang1033 kj}{\lang1033 becomes }{\b\lang1033 t}{\b\lang1033\super j}{\lang1033 (via }{\b\lang1033 tj}{\lang1033 ?), primitive }{ \b\lang1033 sj}{\lang1033 becomes }{\b\lang1033 hy}{\lang1033 (= }{\i\lang1033 ich-Laut}{\lang1033 ), whereas primitive }{\b\lang1033 gj}{\lang1033 , }{\b\lang1033 dj}{\lang1033 and }{\b\lang1033 \'f1gj}{\lang1033 are all reduced to a simple semivowel }{\b\lang1033 j}{\lang1033 -: \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \'a4}{\b\lang1033 kjelep\'ea}{\lang1033 > *}{\b\lang1033 t}{\b\lang1033\super j}{\b\lang1033 elep\'ea}{\lang1033 "silver" (Q }{\i\lang1033 tyelp\'eb}{\lang1033 ) \par \'a4}{\b\lang1033 sjalm\'e2}{\lang1033 > *}{\b\lang1033 hyalm\'e2}{\lang1033 "shell, conch" (Q }{\i\lang1033 hyalma}{\lang1033 ) \par \'a4}{\b\lang1033 gjern\'e2}{\lang1033 > *}{\b\lang1033 jern\'e2}{\lang1033 "old, worn" (Q }{\i\lang1033 yerna}{\lang1033 ) \par *}{\b\lang1033 djelt\'e2}{\lang1033 > *}{\b\lang1033 jelt\'e2}{\lang1033 "to loathe" (Q }{\i\lang1033 yelta}{\lang1033 ) \par *}{\b\lang1033 \'f1gjalm\'ea}{\lang1033 > *}{\b\lang1033 jalm\'ea}{\lang1033 "clamour" (Q }{\i\lang1033 yalm\'eb}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par The sound of }{\b\lang1033 hw}{\lang1033 (unvoiced }{\b\lang1033 w}{\lang1033 ) enters the language as a simplification of initial }{\b\lang1033 sw}{\lang1033 -: \par \tab \par \tab (\'a4}{\i\lang1033 swanda}{\lang1033 >) *}{\b\lang1033 swand}{\lang1033 > *}{\b\lang1033 hwand}{\lang1033 "sponge" (Q }{\i\lang1033 hwan}{\lang1033 with stem }{\i\lang1033 hwand}{\lang1033 -) \par }\pard \qj\nowidctlpar\widctlpar\adjustright {\lang1033 \tab \'a4}{\b\lang1033 swesta}{\lang1033 - (read *}{\b\lang1033 swest\'e2}{\lang1033 -) > *}{\b\lang1033 hwest\'e2}{\lang1033 - "to puff" (Q }{\i\lang1033 hwesta}{\lang1033 -) \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 *}{\b\lang1033 swinj\'e2}{\lang1033 - > *}{\b\lang1033 hwinj\'e2}{\lang1033 "to swirl" (Q }{\i\lang1033 hwinya}{\lang1033 , cf. the entry }{\i\lang1033 SWIN}{\lang1033 in Etym) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par Another source of }{\b\lang1033 hw}{\lang1033 is the aspirate }{\b\lang1033 p}{\b\lang1033\super h}{\lang1033 , where it occurred before the vowel }{\b\lang1033 u}{\lang1033 (cf. VT41:8): \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 *}{\b\lang1033 p}{\b\lang1033\super h}{\b\lang1033 uin\'ea}{\lang1033 > *}{\b\lang1033 hwuin\'ea}{\lang1033 "gloom, deep shadow" (Q }{\i\lang1033 huin\'eb}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par }{\i\lang1033 BEGINNING OF THE RETRACTION PERIOD:}{\lang1033 In primitive Elvish, it was not predictable (from the shape of a word) which syllable received the stress. In a little less than 30 cases, Tolkien in the }{\i\lang1033 Etymologies}{\lang1033 indicated by means of an accent which syllable is stressed (plus some additional cases where he indicated which syllable in the }{\i\lang1033 root}{\lang1033 is stressed). For instance, it would seem that the primitive words }{\b\lang1033 bar\'e1d\'e2}{ \lang1033 "lofty", }{\b\lang1033 ber\'e9k\'e2}{\lang1033 "wild" and }{\b\lang1033 mor\'f3k\'f4}{\lang1033 "bear" were stressed on the second-to-last syllable. But at some early point in the evolution of Quenya, the stress was moved to the }{ \i\lang1033 first syllable}{\lang1033 in nearly all cases. (The only known exception is the augment, or prefixed reduplication, in verbal stems: according to WJ:366 this remained unstressed.) So now the three words above became *}{\b\lang1033 b\'e1rad \'e2}{\lang1033 , }{\b\lang1033 b\'e9rek\'e2}{\lang1033 , }{\b\lang1033 m\'f3rok\'f4}{\lang1033 \endash but since practically all words are henceforth stressed on the first syllable, there is no point in marking the stress at all. The period from now and until the emergence of the later Latin-style stress-patterns we know and love from L otR-style Quenya is called the }{\i\lang1033 retraction period}{\lang1033 (WJ:366). It seems to cover the entire formative phase of the Quenya language, so the original stress-patterns of Primitive Elvish actually have minimal impact on the evolution of High-Elven. \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 Exactly when Tolkien imagined the retraction period to have begun we cannot know, but it should be noted that the vowels that were originally accented in our three examples (}{\b\lang1033 bar\'e1d \'e2}{\lang1033 , }{\b\lang1033 ber\'e9k\'e2}{\lang1033 , }{\b\lang1033 mor\'f3k\'f4}{\lang1033 ), have all been }{\i\lang1033 lost}{\lang1033 in their Quenya descendants: }{\i\lang1033 Varda}{\lang1033 , }{\i\lang1033 verca}{\lang1033 , }{\i\lang1033 morco}{\lang1033 . They disappeared in the Quenya Syncope (see below). It is not probable that they would have been syncopated if they had still been accented, so the retraction period must at any rate have begun before the syncope set in. *}{\b\lang1033 Sal\'e1k}{\b\lang1033\super w}{\b\lang1033 \'ea}{\lang1033 "grass" must also have transferred its accent to the first syllable (*}{\b\lang1033 s\'e1lak}{\b\lang1033\super w}{\b\lang1033 \'ea}{\lang1033 ) before it underwent syncope to produce the immediate ancestor of Quenya }{\i\lang1033 salqu\'eb}{\lang1033 .}{\b\lang1033 \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \tab Shorter words that Tolkien indicated were accented on the final syllable, like the verb }{\b\lang1033 barj\'e2}{\lang1033 - "protect" or the related adjective (or participle?) }{\b\lang1033 barn \'e2}{\lang1033 "safe, protected" were also stressed on the first syllable from now on. The Quenya descendants }{\i\lang1033 varya}{\lang1033 - and }{\i\lang1033 varna}{\lang1033 are still so accented. \par \par }{\i\lang1033 THE QUENYA SYNCOPE:}{\lang1033 Where a word of three or more syllables commences in two concomitant syllables that contain }{\i\lang1033 identical short vowels}{\lang1033 , the second of these vowels is normally syncopated (cf. Letters:426): \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 *}{\b\lang1033 doron\'f4}{\lang1033 > *}{\b\lang1033 dorn\'f4}{\lang1033 "oak" (Q }{\i\lang1033 norno}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \tab \'a4}{\b\lang1033 galad\'e2}{\lang1033 > *}{\b\lang1033 gald\'e2}{\lang1033 "tree" (Q }{\i\lang1033 alda}{\lang1033 ) \par \tab (\'a4}{\i\lang1033 karani}{\lang1033 >) late CE *}{\b\lang1033 karane}{\lang1033 > *}{\b\lang1033 karne}{\lang1033 "red" (Q }{\i\lang1033 carn\'eb}{\lang1033 ) \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 kjelep\'ea}{\lang1033 >) *}{\b\lang1033 t}{\b\lang1033\super j}{\b\lang1033 elep\'ea}{\lang1033 > *}{\b\lang1033 t}{\b\lang1033\super j}{\b\lang1033 elp\'ea}{\lang1033 "silver" (Q }{\i\lang1033 tyelpe}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \tab *}{\b\lang1033 morok\'f4}{\lang1033 (no longer stressed on the second }{\b\lang1033 o}{\lang1033 !) > }{\b\lang1033 mork\'f4}{\lang1033 "bear" (Q }{\i\lang1033 morco}{\lang1033 ) \par \tab \'a4}{\b\lang1033 neres\'e2}{\lang1033 > *}{\b\lang1033 ners\'e2}{\lang1033 (Q }{\i\lang1033 Nessa}{\lang1033 , name of a Vali\'eb: WJ:416) \par \tab (\'a4}{\i\lang1033 p}{\i\lang1033\super h}{\i\lang1033 oroti}{\b\lang1033 >) }{\lang1033 late CE *}{\b\lang1033 p}{\b\lang1033\super h}{\b\lang1033 orote}{\lang1033 > *}{\b\lang1033 p}{\b\lang1033\super h}{\b\lang1033 orte}{\lang1033 "northern" (Q }{\i\lang1033 fort\'eb}{\lang1033 ) \par \tab \'a4}{\b\lang1033 silim\'e2}{\lang1033 > *}{\b\lang1033 silm\'e2}{\lang1033 "shining white" (adj.) (Q }{\i\lang1033 silma}{\lang1033 ) \par \tab \'a4}{\b\lang1033 turum\'e2}{\lang1033 > *}{\b\lang1033 turm\'e2}{\lang1033 "shield" (Q }{\i\lang1033 turma}{\lang1033 ) \par \tab \'a4}{\b\lang1033 Ulum\'f4}{\lang1033 > *}{\b\lang1033 Ulm\'f4}{\lang1033 , name of the Vala of all waters (Q }{\i\lang1033 Ulmo}{\lang1033 ) \par \tab (\'a4}{\i\lang1033 salakw\'ea}{\lang1033 >) *}{\b\lang1033 salak}{\b\lang1033\super w}{\b\lang1033 \'ea}{\lang1033 > *}{\b\lang1033 salk}{\b\lang1033\super w}{\b\lang1033 \'ea}{\lang1033 "grass" (Q }{\i\lang1033 salqu\'eb}{\lang1033 ) \par \par However, the syncope did NOT take place if there was a consonant cluster }{\i\lang1033 before}{\lang1033 or }{\i\lang1033 following}{\lang1033 the second vowel. The word *}{\b\lang1033 aklar\'ea}{\lang1033 "glory" (Q }{\i\lang1033 alcar\'eb}{\lang1033 ), where the second vowel follows the group }{\b\lang1033 kl}{\lang1033 , cannot be syncopated to **}{\b\lang1033 akl'r\'ea}{\lang1033 . A word like *}{\b\lang1033 kjelepn\'e2}{\lang1033 "of silver" could not become *}{\b\lang1033 kjelpn\'e2}{\lang1033 , and therefore the second }{\b\lang1033 e}{\lang1033 still persists in Quenya }{\i\lang1033 telemna}{\lang1033 (altered from *}{\i\lang1033 tyelemna}{\lang1033 under the influence of Telerin, that had }{\i\lang1033 t}{\lang1033 for }{\i\lang1033 ty}{ \lang1033 ). A }{\i\lang1033 double}{\lang1033 consonant likewise hinders syncope in the previous syllable: P }{\b\lang1033 t}{\b\lang1033\super j}{\b\lang1033 uluss\'ea}{\lang1033 "poplar-tree" is not reduced to **}{\b\lang1033 t}{\b\lang1033\super j}{ \b\lang1033 ulss\'ea}{\lang1033 (the second }{\i\lang1033 u}{\lang1033 is still present in Quenya }{\i\lang1033 tyuluss\'eb}{\lang1033 ). \par \par }{\fs20\lang1033 NOTE: The fact that double consonants would prevent the syncope from taking place, confirms one sound-shift set out above: }{\b\fs20\lang1033 k}{\fs20\lang1033 + }{\b\fs20\lang1033 w}{\fs20\lang1033 had already merged into a single labial consonant }{\b\fs20\lang1033 k}{\b\fs20\lang1033\super w}{\fs20\lang1033 . If \'a4}{\b\fs20\lang1033 sal\'e1k-w\'ea}{\fs20\lang1033 (LR:385) had not become *}{\b\fs20\lang1033 s\'e1lak}{\b\fs20\lang1033\super w}{ \b\fs20\lang1033 \'ea}{\fs20\lang1033 at this stage, there would still have been a consonant cluster following the second }{\b\fs20\lang1033 a}{\fs20\lang1033 , and the syncope producing the form *}{\b\fs20\lang1033 salk}{\b\fs20\lang1033\super w}{ \b\fs20\lang1033 \'ea}{\fs20\lang1033 (directly underlying Q }{\b\fs20\lang1033 salqu\'eb}{\fs20\lang1033 ) could not have taken place. I cannot readily explain why \'a4}{\b\fs20\lang1033 atakw\'ea}{\fs20\lang1033 "building" (LR:390, later presumably *} {\b\fs20\lang1033 atak}{\b\fs20\lang1033\super w}{\b\fs20\lang1033 \'ea}{\fs20\lang1033 ) is not syncopated to *}{\b\fs20\lang1033 atk}{\b\fs20\lang1033\super w}{\b\fs20\lang1033 \'ea}{\fs20\lang1033 ; the second }{\b\fs20\lang1033 a}{\fs20\lang1033 is still present in Quenya }{\b\fs20\lang1033 ataqu\'eb}{\fs20\lang1033 . But a form *}{\b\fs20\lang1033 atqu\'eb}{\fs20\lang1033 would probably be impossible in Quenya, and a combination like *}{\b\fs20\lang1033 tqu}{\fs20\lang1033 = *}{ \b\fs20\lang1033 tk}{\b\fs20\lang1033\super w}{\fs20\lang1033 would maybe be disliked also at earlier stages of Vanya-Noldorin; perhaps the syncope was sometimes avoided where certain "undesirable" consonant clusters would otherwise arise. Cf. also a word like \'a4}{\b\fs20\lang1033 tat}{\b\fs20\lang1033\super h}{ \b\fs20\lang1033 ar\'ea}{\fs20\lang1033 yielding Quenya }{\b\fs20\lang1033 tasar\'eb}{\fs20\lang1033 ("willow") rather than syncopated (**}{\b\fs20\lang1033 tat}{\b\fs20\lang1033\super h}{\b\fs20\lang1033 ar\'ea}{\fs20\lang1033 >**}{\b\fs20\lang1033 ta \'fere}{\fs20\lang1033 >) **}{\b\fs20\lang1033 tass\'eb}{\fs20\lang1033 . \endash The syncope did }{\i\fs20\lang1033 not}{\fs20\lang1033 regularly take place in the branch of Elvish leading to Sindarin; hence primitive \'a4}{\b\fs20\lang1033 karani}{ \fs20\lang1033 (late CE *}{\b\fs20\lang1033 karane}{\fs20\lang1033 , which would also be the Old Sindarin form) comes out as }{\b\fs20\lang1033 caran}{\fs20\lang1033 in Grey-elven. Where S indarin words are syncopated, it may at least sometimes be the result of the more sporadic syncope taking place already in Primitive Quendian. Therefore we postulated above that (\'a4}{\i\fs20\lang1033 usuk-w\'ea}{\fs20\lang1033 >) *}{\b\fs20\lang1033 usuk}{\b\fs20\lang1033\super w}{\b\fs20\lang1033 \'ea}{\fs20\lang1033 "reek" became *}{\b\fs20\lang1033 usk}{\b\fs20\lang1033\super w}{\b\fs20\lang1033 \'ea}{\fs20\lang1033 very early and did not have to wait for the Quenya syncope: Though the Quenya form would be }{\b\fs20\lang1033 usqu\'eb}{\fs20\lang1033 in any case, "Noldorin"/Sindarin }{\b\fs20\lang1033 osp}{\fs20\lang1033 demands a syncopated form *}{\b\fs20\lang1033 usk}{\b\fs20\lang1033\super w}{ \b\fs20\lang1033 \'ea}{\fs20\lang1033 existing already in Common Eldarin (in the Lindarin dialect of CE actually *}{\b\fs20\lang1033 usp\'ea}{\fs20\lang1033 ). \par }{\lang1033 \par A }{\i\lang1033 long}{\lang1033 vowel could not be syncopated, so a word like }{\b\lang1033 ter\'eaw\'e2}{\lang1033 was not reduced to **}{\b\lang1033 terw\'e2}{\lang1033 (Quenya }{\i\lang1033 tereva}{\lang1033 "fine, acute" still preserves the second }{\i\lang1033 e}{\lang1033 ; as we see it did become short later, but by then the syncopation had ceased to operate). Another example of the same is \'a4}{\b\lang1033 nar\'e2ka}{\lang1033 (read *}{\b\lang1033 nar \'e2k\'e2}{\lang1033 ), that was }{\i\lang1033 not}{\lang1033 reduced to **}{\b\lang1033 nark\'e2}{\lang1033 (Quenya }{\i\lang1033 naraca}{\lang1033 "rushing, violent"). Also contrast the development of two variant forms of the same word (WJ:360, 363): CE \'a4}{\b\lang1033 awa-delo}{\lang1033 "away-goer" becomes syncopated to *}{\b\lang1033 aw'del}{\lang1033 > *}{\b\lang1033 audel}{\lang1033 (Quenya }{\i\lang1033 Auzel}{\lang1033 > }{\i\lang1033 Aurel}{\lang1033 ), whereas the variant }{\b\lang1033 aw\'e2-delo}{\lang1033 shows no syncope because of the long }{\b\lang1033 \'e2}{\lang1033 : This produces Quenya }{\i\lang1033 Oazel}{ \lang1033 > }{\i\lang1033 Oarel}{\lang1033 , the group }{\i\lang1033 oa}{\lang1033 later arising from earlier }{\b\lang1033 awa}{\lang1033 (after it had been shortened from original }{\b\lang1033 aw\'e2}{\lang1033 , but too late for the syncope to take effect). \par \par }{\fs20\lang1033 NOTE: The fact that the syncope only occurred in words of }{\i\fs20\lang1033 more than two syllables}{\fs20\lang1033 is responsible for the fact that a Quenya word like }{\b\fs20\lang1033 toron}{\fs20\lang1033 "brother" is reduced to }{ \b\fs20\lang1033 torn}{\fs20\lang1033 - before an ending (e.g. pl. }{\b\fs20\lang1033 torni}{\fs20\lang1033 ). The singular }{\b\fs20\lang1033 toron}{\fs20\lang1033 had only two syllables and was t herefore unaffected by the syncope, whereas the three-syllable plural form *}{\b\fs20\lang1033 toron\'ee}{\fs20\lang1033 was reduced to *}{\b\fs20\lang1033 torn\'ee}{\fs20\lang1033 (> Quenya }{\b\fs20\lang1033 torni}{\fs20\lang1033 ). But sometimes inflected forms were later re-formed on analogy with the singular form, the syncopated vowel being reintroduced. The noun }{\b\fs20\lang1033 elen}{\fs20\lang1033 "star" is an example of this: "The pl. form }{\i\fs20\lang1033 eleni}{ \fs20\lang1033 , without syncope, is re-formed after the singular" (WJ:262). But a syncopated form *}{\b\fs20\lang1033 eln\'ee}{\fs20\lang1033 had actually existed at an earlier stage, yielding an alternative Quenya plural }{\b\fs20\lang1033 eldi}{ \fs20\lang1033 that might turn up in verse (for the change }{\b\fs20\lang1033 ln}{\fs20\lang1033 > }{\b\fs20\lang1033 ld}{\fs20\lang1033 , see below). Other Quenya examples of what seems to be analogical plurals are }{\b\fs20\lang1033 talami}{ \fs20\lang1033 "floors" and }{\b\fs20\lang1033 filici}{\fs20\lang1033 "birds" (instead of syncopated **}{\b\fs20\lang1033 talmi}{\fs20\lang1033 , **}{\b\fs20\lang1033 filci}{\fs20\lang1033 ; cf. syncopated }{\b\fs20\lang1033 nelci}{\fs20\lang1033 rather than **}{\b\fs20\lang1033 neleci}{\fs20\lang1033 for "teeth"). \par }{\lang1033 \par Where the syncope brought }{\b\lang1033 j}{\lang1033 or }{\b\lang1033 w}{\lang1033 into direct contact with a consonant, they became vowels }{\b\lang1033 i}{\lang1033 , }{\b\lang1033 u}{\lang1033 and could form a diphthong with the previous vowel: \par \par \tab \'a4}{\b\lang1033 gajak\'e2}{\lang1033 > (*}{\b\lang1033 gajk\'e2}{\lang1033 >) *}{\b\lang1033 gaik\'e2 }{\lang1033 \~"fell, terrible, dire" (Quenya }{\i\lang1033 aica}{\lang1033 , PM:363) \par \tab (\'a4}{\i\lang1033 awa}{\b\lang1033 -}{\i\lang1033 delo}{\lang1033 >) *}{\b\lang1033 awa-del}{\lang1033 > (*}{\b\lang1033 awdel}{\lang1033 >) *}{\b\lang1033 audel}{\lang1033 "Away-goer" (Quenya }{\i\lang1033 Auzel}{\lang1033 > }{\i\lang1033 Aurel}{\lang1033 ) \par \tab *}{\b\lang1033 tawarin\'e2}{\lang1033 > (*}{\b\lang1033 tawrin\'e2}{\lang1033 >) *}{\b\lang1033 taurin\'e2}{\lang1033 (Q }{\i\lang1033 taurina}{\lang1033 "of wood") \par \par At some early point (but }{\i\lang1033 after}{\lang1033 the final separation of the branches of Eldarin that lead to Quenya on the one hand and Amanya Telerin on the other, and evidently also after the syncope), the consonant }{\b\lang1033 d}{\lang1033 is changed in many positions. Initial }{\b\lang1033 d}{\lang1033 is consistently altered in the proto-Quenya branch. It normally becomes }{\b\lang1033 l}{\lang1033 : \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 d\'f4mi}{\lang1033 >) late CE *}{\b\lang1033 d\'f4me}{\lang1033 > *}{\b\lang1033 l\'f4me}{\lang1033 "twilight" (still }{\i\lang1033 l\'f3m\'eb}{\lang1033 in later Quenya) \par \'a4}{\b\lang1033 datt\'e2}{\lang1033 > *}{\b\lang1033 latt\'e2}{\lang1033 "hole, pit" (Q }{\i\lang1033 latta}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par If the word \'a4}{\b\lang1033 d'r\'e2k[\'e2]}{\lang1033 "wolf" still survived with its initial cluster }{\b\lang1033 dr}{\lang1033 - (produced by syncope of the original base }{\i\lang1033 DAR\'c1K}{\lang1033 -) intact, it may have been simplified to *}{ \b\lang1033 r\'e2k\'e2}{\lang1033 at this point. If the form *}{\b\lang1033 lr\'e2k\'e2}{\lang1033 ever existed, it hardly lasted long. The original initial cluster }{\b\lang1033 dr}{\lang1033 - has in any case been simplified in the later Quenya form } {\b\lang1033 r\'e1ca}{\lang1033 , whenever the simplification may have occurred. \par \par Occasionally, initial }{\b\lang1033 d}{\lang1033 becomes }{\b\lang1033 n}{\lang1033 instead of }{\b\lang1033 l}{\lang1033 : \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \'a4}{\b\lang1033 d\'ear}{\lang1033 > }{\b\lang1033 n\'ear}{\lang1033 "man" (still }{\i\lang1033 n\'e9r}{\lang1033 in Quenya) \par (*}{\i\lang1033 doron\'f4}{\lang1033 >) *}{\b\lang1033 dorn\'f4}{\lang1033 > }{\b\lang1033 norn\'f4}{\lang1033 "oak" (Q }{\i\lang1033 norno}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par In the latter case, }{\b\lang1033 d}{\lang1033 evidently becomes }{\b\lang1033 n}{\lang1033 because of assimilation with the other }{\b\lang1033 n}{\lang1033 in the word. In WJ:413-414, Tolkien comments on the derivation of the Q adjective }{ \b\lang1033 norna}{\lang1033 "tough, stiff", which is implied to come from *}{\b\lang1033 dorn\'e2}{\lang1033 ; he notes that "this is probably one of the cases in which Q initial }{\i\lang1033 d}{\lang1033 became }{\i\lang1033 n}{\lang1033 -, not }{ \i\lang1033 l}{\lang1033 -, by assimilation to an }{\i\lang1033 n}{\lang1033 occurring later in the word". (The Quenya words }{\b\lang1033 nenda}{\lang1033 "slope" and }{\b\lang1033 nend\'eb}{\lang1033 "sloping", derived from a root }{\i\lang1033 DEN}{ \lang1033 , seem to display the same assimilation: primitive *}{\b\lang1033 dend\'e2}{\lang1033 , *}{\b\lang1033 dend\'ea}{\lang1033 . However, Tolkien struck out the entry }{\i\lang1033 DEN}{\lang1033 [LR:354].) As for }{\b\lang1033 d\'ear}{\lang1033 > }{\b\lang1033 n\'ear}{\lang1033 "man", the }{\i\lang1033 Etymologies}{\lang1033 implies that this unexpected development is "partly due to [influence from] }{\i\lang1033 N\'ce}{\lang1033 , }{\i\lang1033 NIS}{\lang1033 woman, partly to strengthened stem }{\i\lang1033 ndere}{\lang1033 bridegroom" (LR:354 s.v. }{\i\lang1033 DER}{\lang1033 ); therefore this word did not end up as **}{\b\lang1033 l\'e9r}{\lang1033 in Quenya (LR:377 s.v. }{\i\lang1033 N\'ce}{ \i\lang1033\super 1}{\lang1033 ). \par \par }{\fs20\lang1033 NOTE: But in a post-LotR source, Tolkien derived Quenya }{\b\fs20\lang1033 n\'e9r}{\fs20\lang1033 from a root }{\i\fs20\lang1033 NERE}{\fs20\lang1033 instead (WJ:393), eliminating the complications above: In the new scenario, the initial }{\b\fs20\lang1033 n}{\fs20\lang1033 of }{\b\fs20\lang1033 n\'e9r}{\fs20\lang1033 had been }{\b\fs20\lang1033 n}{\fs20\lang1033 all along. \par }{\lang1033 \par Medial, intervocalic }{\b\lang1033 d}{\lang1033 may also become }{\b\lang1033 l}{\lang1033 by assimilation with another }{\b\lang1033 l}{\lang1033 in the same word, but this only happens sporadically: \par \par \tab \'a4}{\b\lang1033 g-lada}{\lang1033 - > *}{\b\lang1033 glala}{\lang1033 - verbal stem "to laugh" (Q }{\i\lang1033 lala}{\lang1033 -, PM:359) \par \par Contrast \'a4}{\b\lang1033 l\'e2da}{\lang1033 (read *}{\b\lang1033 l\'e2d\'e2}{\lang1033 ) as a primitive word for "flat"; this did }{\i\lang1033 not}{\lang1033 become *}{\b\lang1033 l\'e2l\'e2}{\lang1033 (which would have produced Quenya **}{ \i\lang1033 l\'e1la}{\lang1033 , but we actually see }{\i\lang1033 l\'e1ra}{\lang1033 , the consonant }{\i\lang1033 r}{\lang1033 regularly evolving from post-vocalic }{\i\lang1033 d}{\lang1033 \endash see below). \par \par The primitive word *}{\b\lang1033 l\'e2d\'e2}{\lang1033 derived from a stem }{\i\lang1033 DAL}{\lang1033 seems to exemplify another phenomenon: Occasionally, the initial change }{\b\lang1033 d}{\lang1033 > }{\b\lang1033 l}{\lang1033 would cause an }{ \b\lang1033 l}{\lang1033 occurring later in the word to be }{\i\lang1033 dissimilated}{\lang1033 to }{\b\lang1033 d}{\lang1033 . In effect, }{\b\lang1033 d}{\lang1033 and }{\b\lang1033 l}{\lang1033 switch places, as when the root }{\i\lang1033 DAL}{ \lang1033 seems to have become *}{\i\lang1033 LAD}{\lang1033 in the derivative *}{\b\lang1033 l\'e2d\'e2}{\lang1033 (whence Q }{\b\lang1033 l\'e1ra}{\lang1033 ). Tolkien explicitly deals with another example of this phenomenon in WJ:363: "by dissimilation", the root }{\i\lang1033 DEL}{\lang1033 "go, proceed" turned into }{\i\lang1033 led}{\lang1033 - in Quenya. (The }{\i\lang1033 Etymologies}{\lang1033 only lists }{\i\lang1033 LED}{\lang1033 - "go, fare, travel" [LR:368]; when writing Etym, Tolkien had apparently not yet "discovered" that this is actually a dissimilated form of }{\i\lang1033 DEL}{\lang1033 -!) \par \par According to WJ:363, there is yet another way }{\b\lang1033 d}{\lang1033 can become }{\b\lang1033 l}{\lang1033 in Quenya: "}{\i\lang1033 dj}{\lang1033 became }{\i\lang1033 ly}{\lang1033 medially". This is from the same paragraph that discusses }{ \i\lang1033 DEL}{\lang1033 vs. }{\i\lang1033 LED}{\lang1033 , so we are obviously to assume that Quenya }{\b\lang1033 lelya}{\lang1033 - "go" descends from older *}{\b\lang1033 ledj\'e2}{\lang1033 -. However, this do es not agree very well with one example from the }{\i\lang1033 Etymologies}{\lang1033 : The root }{\i\lang1033 MAD}{\lang1033 yields a Quenya word }{\b\lang1033 marya}{\lang1033 "pale", clearly meant to come from *}{\b\lang1033 madj\'e2}{\lang1033 (post-vocalic }{\b\lang1033 d}{\lang1033 regularly evolving into }{\b\lang1033 r}{\lang1033 via }{\b\lang1033 \'f0}{\lang1033 and }{\b\lang1033 z}{\lang1033 ; see below). According to the rule Tolkien later set out, *}{\b\lang1033 madj\'e2}{\lang1033 should have yielded Quenya *}{\b\lang1033 malya}{\lang1033 instead, the cluster }{\b\lang1033 dj}{\lang1033 becoming }{\b\lang1033 ly}{\lang1033 before }{\b\lang1033 d}{\lang1033 had the chance to become }{\b\lang1033 r}{\lang1033 . But instead of altering the word, we may simply see }{\b\lang1033 marya}{\lang1033 as a lone sample of an irregular development. \par \par As we have seen, Tolkien could not quite make up his mind whether a specific primitive sound was to be reconstructed as }{\b\lang1033 h}{\lang1033 (presumably like English }{\b\lang1033 h}{\lang1033 ), or as a back-spirant }{\b\lang1033 3}{\lang1033 . But if this sound was not }{\b\lang1033 h}{\lang1033 all along, the back-spirant }{\b\lang1033 3}{\lang1033 was changed to }{\b\lang1033 h}{\lang1033 at this stage at the very latest. We know this because Quenya was about to acquire }{\b\lang1033 3}{ \lang1033 from another source (changed from initial and post-vocalic }{\b\lang1033 g}{\lang1033 ), and this sound would }{\i\lang1033 not}{\lang1033 behave like the older (?}{\b\lang1033 3}{\lang1033 >) }{\b\lang1033 h}{\lang1033 . The latter survived into Quenya as }{\b\lang1033 h}{\lang1033 whereas the "new" }{\b\lang1033 3}{\lang1033 would eventually be lost completely. Therefore, they cannot have been identical during the next stages of the linguistic evolution. \par \par }{\i\lang1033 THE SPIRANTS ARISE: }{\lang1033 So far proto-Quenya did not have any }{\i\lang1033 spirant}{\lang1033 sounds (except possibly }{\b\lang1033 3}{\lang1033 , which had by this stage at the latest finally become }{\b\lang1033 h}{\lang1033 , as noted abo ve). But now a wide range of spirants would develop from two sources: the voiced stops and the unvoiced aspirates. Turning these sounds into spirants may be one of the changes Tolkien referred to as a "softening" of the older form of Elvish (WJ:20). \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 Initial and post-vocalic }{\b\lang1033 g}{\lang1033 and }{\b\lang1033 b}{\lang1033 became }{\b\lang1033 3}{\lang1033 and }{\b\lang1033 v}{\lang1033 , respectively: \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par \tab (\'a4}{\i\lang1033 galad\'e2}{\lang1033 >) *}{\b\lang1033 gald\'e2}{\lang1033 > *}{\b\lang1033 3ald\'e2}{\lang1033 "tree" (Q }{\i\lang1033 alda}{\lang1033 , but }{\i\lang1033 3alda}{\lang1033 in the archaic R\'familian spelling \endash see below) \par \tab *}{\b\lang1033 weg\'ea}{\lang1033 > \'a4}{\b\lang1033 we3\'ea}{\lang1033 "manhood" (Q *}{\i\lang1033 wi\'eb}{\lang1033 > }{\i\lang1033 vi\'eb}{\lang1033 ) \par \tab (\'a4}{\i\lang1033 Denwego}{\lang1033 > *}{\i\lang1033 Lenweg}{\lang1033 >) *}{\b\lang1033 Lenweg}{\lang1033 > *}{\b\lang1033 Lenwe3}{\lang1033 (personal name, Q }{\i\lang1033 Lenw\'eb}{\lang1033 , WJ:412) \par \tab (\'a4}{\i\lang1033 lugni}{\lang1033 >) *}{\b\lang1033 lugne}{\lang1033 > *}{\b\lang1033 lu3ne}{\lang1033 "blue" (Q }{\i\lang1033 l\'fan\'eb}{\lang1033 , possibly *}{\i\lang1033 lu3ne}{\lang1033 in R\'familian spelling)}{\b\i\lang1033 \par }{\lang1033 \tab \'a4}{\b\lang1033 bes\'fb}{\lang1033 > *}{\b\lang1033 ves\'fb}{\lang1033 "married pair" (Q *}{\i\lang1033 vezu }{\lang1033 > }{\i\lang1033 veru}{\lang1033 ) \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \'a4}{\b\lang1033 laib\'ea}{\lang1033 > }{\b\lang1033 laiv\'ea}{\lang1033 "ointment" (Q }{\i\lang1033 laiv\'eb}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par It seems that }{\b\lang1033 g}{\lang1033 became }{\b\lang1033 3}{\lang1033 following }{\b\lang1033 l}{\lang1033 as well: \par \par \tab \'a4}{\b\lang1033 ulgund\'f4}{\lang1033 > *}{\b\lang1033 ul3und\'f4}{\lang1033 "monster" (Q }{\i\lang1033 ulundo}{\lang1033 ) \par \par It may be that not just }{\b\lang1033 g}{\lang1033 , but also }{\b\lang1033 b}{\lang1033 turned into a spirant following }{\b\lang1033 l}{\lang1033 (hence }{\b\lang1033 lb}{\lang1033 > }{\b\lang1033 lv}{\lang1033 ), but the evidence is somewhat divergent. The }{\i\lang1033 Etymologies}{\lang1033 has it that the cluster }{\b\lang1033 lb}{\lang1033 produces Quenya }{\b\lang1033 lw }{\lang1033 (as in \'a4}{\b\lang1033 golb\'e2}{\lang1033 > Q }{\b\lang1033 olwa}{ \lang1033 "branch", entry }{\i\lang1033 G\'d3LOB}{\lang1033 ). But Tolkien seems to have abandoned this idea; in a post-LotR source the word for "branch" is quoted as }{\b\lang1033 olba}{\lang1033 instead (PM:340) \endash as if the original cluster }{ \b\lang1033 lb}{\lang1033 was unchanged in Quenya. However, LotR Appendix E informs us that "for }{\i\lang1033 lv}{\lang1033 , not for }{\i\lang1033 lw}{\lang1033 , many speakers, especially Elves, used }{\i\lang1033 lb}{\lang1033 ". So perhaps \'a4}{ \b\lang1033 golb\'e2}{\lang1033 does properly yield Quenya *}{\b\lang1033 olva}{\lang1033 , but many speakers would alter it back to }{\b\lang1033 olba}{\lang1033 , and it is this "alternative" form that is quoted in PM:340. The cluster }{\b\lang1033 lv} {\lang1033 is not very common in Quenya, but in some of the cases where it does occur (e.g. in the word }{\b\lang1033 elv\'eba}{\lang1033 "starlike"), it almost certainly comes from earlier }{\b\lang1033 lb}{\lang1033 . We can probably feel free to postulate a regular development }{\b\lang1033 lb}{\lang1033 > }{\b\lang1033 lv}{\lang1033 , though "many speakers, especially Elves" would later undo this change \endash apparently to distinguish this cluster more clearly from }{\b\lang1033 lw}{\lang1033 . \par \par Paralleling the changes }{\b\lang1033 b}{\lang1033 > }{\b\lang1033 v}{\lang1033 and }{\b\lang1033 g}{\lang1033 > }{\b\lang1033 3}{\lang1033 , post-vocalic }{\b\lang1033 d}{\lang1033 became }{\b\lang1033 \'f0}{\lang1033 . The only direct evidence for this is Tolkien's statement (in WJ:363) that a primitive word \'a4}{\b\lang1033 edelo}{\lang1033 would have produced Quenya \'a4}{\b\lang1033 ello}{\lang1033 via \'a4}{\b\lang1033 e\'f0lo}{\lang1033 ; the word \'a4}{ \b\lang1033 ello}{\lang1033 was not actually found, but the development }{\b\lang1033 d}{\lang1033 > }{\b\lang1033 \'f0}{\lang1033 must be "valid" all the same. Hence: \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (*}{\i\lang1033 tada}{\lang1033 >) CE \'a4}{\b\lang1033 tad}{\lang1033 > *}{\b\lang1033 ta\'f0}{\lang1033 "thither" (Q *}{\i\lang1033 taz}{\lang1033 , in Exilic }{\i\lang1033 tar}{\lang1033 ) \par (\'a4}{\i\lang1033 awadelo}{\lang1033 >) *}{\b\lang1033 Audel}{\lang1033 > *}{\b\lang1033 au\'f0el}{\lang1033 > }{\b\lang1033 auzel}{\lang1033 (Q }{\i\lang1033 Auzel}{\lang1033 > }{\i\lang1033 Aurel}{\lang1033 ) \par (\'a4}{\i\lang1033 nidw\'f4}{\lang1033 > *}{\i\lang1033 nidw\'e2}{\lang1033 >) *}{\b\lang1033 nidwa}{\lang1033 > *}{\b\lang1033 ni\'f0wa}{\lang1033 "cushion" (Q *}{\i\lang1033 nizwa}{\lang1033 > }{\i\lang1033 nirwa}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par As for the first example, compare the primitive allative ending -}{\b\lang1033 da}{\lang1033 (WJ:366), hence *}{\b\lang1033 ta-da}{\lang1033 "that-to" = to that, thither. (Quenya }{\b\lang1033 mir}{\lang1033 "to the inside; into" must likewise desc end from *}{\b\lang1033 mida}{\lang1033 via }{\b\lang1033 mid}{\lang1033 > }{\b\lang1033 mi\'f0}{\lang1033 > }{\b\lang1033 miz}{\lang1033 .) \par \par }{\fs20\lang1033 NOTE: No examples of }{\i\fs20\lang1033 initial}{\fs20\lang1033 }{\b\fs20\lang1033 d}{\fs20\lang1033 becoming }{\b\fs20\lang1033 \'f0}{\fs20\lang1033 can be quoted, since in this position }{\b\fs20\lang1033 d}{\fs20\lang1033 had already become }{\b\fs20\lang1033 l}{\fs20\lang1033 (or exceptionally }{\b\fs20\lang1033 n}{\fs20\lang1033 ). Hence a primitive word like \'a4}{\b\fs20\lang1033 do3m\'ea}{\fs20\lang1033 produces Quenya }{\b\fs20\lang1033 l\'f3m\'eb}{\fs20\lang1033 rather than }{\i\fs20\lang1033 (\'a4}{\b\i\fs20\lang1033 do3m\'ea}{\i\fs20\lang1033 > *}{\b\i\fs20\lang1033 d\'f4m\'ea}{\i\fs20\lang1033 > **}{\b\i\fs20\lang1033 \'f0\'f4me}{\i\fs20\lang1033 > **}{\b\i\fs20\lang1033 z\'f4me}{\i\fs20\lang1033 >) }{ \fs20\lang1033 **}{\b\fs20\lang1033 r\'f3m\'eb}{\fs20\lang1033 . \par }{\lang1033 \par Possibly about the same point in time, the unvoiced }{\i\lang1033 aspirates}{\lang1033 also turn into spirant sounds. This seems to have occurred in all positions, though most of our examples involve initial aspirates. The change from aspirates to spirants se ems to have happened more or less independently in all Eldarin tongues; for instance, Tolkien noted that }{\b\lang1033 k}{\b\lang1033\super h}{\lang1033 "became in all Eldarin tongues the spirant" \endash sc. the spirant }{\b\lang1033 ch}{\lang1033 , }{ \i\lang1033 ach}{\lang1033 -Laut (VT41:9). Likewise, the aspirate }{\b\lang1033 p}{\b\lang1033\super h}{\lang1033 became the spirant }{\b\lang1033 f}{\lang1033 , whereas }{\b\lang1033 t}{\b\lang1033\super h}{\lang1033 became the spirant }{\b\lang1033 \'fe}{\lang1033 : \par \par \tab (\'a4}{\i\lang1033 k}{\i\lang1033\super h}{\i\lang1033 \'eet}{\i\lang1033\super h}{\i\lang1033 i}{\lang1033 >) *}{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 \'eet}{\b\lang1033\super h}{\b\lang1033 e}{\lang1033 > *}{\b\lang1033 ch\'ee\'fee}{ \lang1033 "mist" (Q *}{\i\lang1033 ch\'ed\'fe\'eb}{\lang1033 > }{\i\lang1033 h\'eds\'eb}{\lang1033 ) \par \tab \'a4}{\b\lang1033 tat}{\b\lang1033\super h}{\b\lang1033 ar}{\lang1033 > *}{\b\lang1033 ta\'fear}{\lang1033 "willow-tree" (Q *}{\i\lang1033 ta\'fear}{\lang1033 > }{\i\lang1033 tasar}{\lang1033 ) \par \tab \'a4}{\b\lang1033 neth-r\'e2}{\lang1033 > *}{\b\lang1033 ne\'fera}{\lang1033 "young" (Q }{\i\lang1033 nessa}{\lang1033 ) \par \tab \'a4}{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 jeles\'ea}{\lang1033 > *}{\b\lang1033 chjeles\'ea}{\lang1033 "glass" (Q }{\i\lang1033 hyell\'eb}{\lang1033 ) \par \tab \'a4}{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 it}{\b\lang1033\super h}{\b\lang1033 me}{\lang1033 (read *}{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 it}{\b\lang1033\super h}{\b\lang1033 m\'ea}{\lang1033 ) > *}{\b\lang1033 chi\'fem\'ea}{ \lang1033 "fog" (Q *}{\i\lang1033 chi\'few\'eb}{\lang1033 > }{\i\lang1033 hisw\'eb}{\lang1033 ) \par \tab (\'a4}{\i\lang1033 t}{\i\lang1033\super h}{\i\lang1033 indi}{\lang1033 >) *}{\b\lang1033 t}{\b\lang1033\super h}{\b\lang1033 inde}{\lang1033 > }{\b\lang1033 \'feinde}{\lang1033 "grey" (Q }{\i\lang1033 \'feind\'eb}{\lang1033 > }{\i\lang1033 sind \'eb}{\lang1033 , WJ:384) \par \tab (\'a4}{\i\lang1033 phini}{\lang1033 >) *}{\b\lang1033 p}{\b\lang1033\super h}{\b\lang1033 ine}{\lang1033 > }{\b\lang1033 fine}{\lang1033 "a single hair" (still }{\i\lang1033 fin\'eb}{\lang1033 in Q, PM:362/340) \par \par Since it seems that Quenya had already turned original initial }{\b\lang1033 sk}{\lang1033 -, }{\b\lang1033 sp}{\lang1033 -, }{\b\lang1033 st}{\lang1033 - into aspirates }{\b\lang1033 k}{\b\lang1033\super h}{\lang1033 -, }{\b\lang1033 p}{ \b\lang1033\super h}{\lang1033 -, }{\b\lang1033 t}{\b\lang1033\super h}{\lang1033 -, they now became spirants just like the original aspirates (hence Tolkien was free to change his mind about whether }{\b\lang1033 find\'eb}{\lang1033 "tress, hair" comes from \'a4}{\b\lang1033 spind\'ea}{\lang1033 or \'a4}{\b\lang1033 p}{\b\lang1033\super h}{\b\lang1033 ind\'ea}{\lang1033 : LR:387 vs. PM:362!) \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 skarw\'ea}{\lang1033 >) *}{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 arw\'ea}{\lang1033 > *}{\b\lang1033 charw\'ea}{\lang1033 "wound" (Q *}{\i\lang1033 charw\'eb}{ \lang1033 > }{\i\lang1033 harw\'eb}{\lang1033 ) \par (\'a4}{\i\lang1033 skjapat-}{\lang1033 >) *}{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 japat}{\lang1033 > *}{\b\lang1033 chjapat}{\lang1033 "shore" (Q }{\i\lang1033 hyapat}{\lang1033 ) \par (\'a4}{\i\lang1033 sk}{\i\lang1033\super w}{\i\lang1033 arm\'e2}{\lang1033 >) *}{\b\lang1033 k}{\b\lang1033\super hw}{\b\lang1033 arm\'e2}{\lang1033 > *}{\b\lang1033 ch}{\b\lang1033\super w}{\b\lang1033 arm\'e2}{\lang1033 "crossbar" (Q }{\i\lang1033 hwarma}{\lang1033 ) \par (\'a4}{\i\lang1033 spang\'e2}{\lang1033 >) *}{\b\lang1033 p}{\b\lang1033\super h}{\b\lang1033 ang\'e2}{\lang1033 > *}{\b\lang1033 fang\'e2}{\lang1033 "beard" (Q }{\i\lang1033 fanga}{\lang1033 ) \par (\'a4}{\i\lang1033 stank\'e2}{\lang1033 >) *}{\b\lang1033 t}{\b\lang1033\super h}{\b\lang1033 ank\'e2}{\lang1033 > *}{\b\lang1033 \'feanc\'e2}{\lang1033 "split" (Q *}{\i\lang1033 \'feanca}{\lang1033 > }{\i\lang1033 sanca}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par It should be noted that the }{\b\lang1033 \'fe}{\lang1033 that now arose in proto-Quenya was not quite like the sound of English }{\i\lang1033 th}{\lang1033 as in }{\i\lang1033 think}{\lang1033 . This English sound is interdental, sc. pronounced with the tip of the tongue between the teeth. The Elvish }{\b\lang1033 \'fe}{\lang1033 was strictly dental, made with the tongue-tip behind the back of the upper front teeth, and therefore closer to }{ \b\lang1033 s}{\lang1033 (and in the Noldorin dialect, it would indeed merge with }{\b\lang1033 s}{\lang1033 later). Likewise, }{\b\lang1033 f}{\lang1033 was not originally quite like English }{\b\lang1033 f}{\lang1033 ; the }{\b\lang1033 f}{\lang1033 of Amanya Elvish was originally a }{\i\lang1033 bilabial}{\lang1033 sound, pronounced using the lips only, without any friction between the teeth and the lower lip. It was quite close to }{\b\lang1033 hw}{\lang1033 (unvoiced }{\b\lang1033 w}{\lang1033 ), and later \endash in the Vanyarin dialect \endash the sounds would indeed threaten to fall together. See VT41:7-8 concerning all this. \par \par The combination }{\b\lang1033 chj}{\lang1033 , whether descended from }{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 j }{\lang1033 or }{\b\lang1033 skj}{\lang1033 , becomes }{\b\lang1033 hy}{\lang1033 (= }{\i\lang1033 ich-Laut}{\lang1033 ): \par \par \tab (\'a4}{\i\lang1033 k}{\i\lang1033\super h}{\i\lang1033 jeles\'ea}{\lang1033 >) *}{\b\lang1033 chjeles\'ea}{\lang1033 > *}{\b\lang1033 hyeles\'ea}{\lang1033 "glass" (Q }{\i\lang1033 hyell\'eb}{\lang1033 ) \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 skjapat-}{\lang1033 > *}{\i\lang1033 k}{\i\lang1033\super h}{\i\lang1033 japat}{\lang1033 >) *}{\b\lang1033 chjapat}{\lang1033 > }{\b\lang1033 hyapat}{\lang1033 "shore" (still so in Q) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par The labialized combination }{\b\lang1033 ch}{\b\lang1033\super w}{\lang1033 becomes }{\b\lang1033 hw}{\lang1033 (= unvoiced }{\i\lang1033 w}{\lang1033 ): \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \par (*}{\i\lang1033 sk}{\i\lang1033\super w}{\i\lang1033 arm\'e2}{\lang1033 > }{\i\lang1033 k}{\i\lang1033\super hw}{\i\lang1033 arm\'e2}{\lang1033 >) *}{\b\lang1033 ch}{\b\lang1033\super w}{\b\lang1033 arm\'e2}{\lang1033 > }{\b\lang1033 hwarm\'e2 }{ \lang1033 "crossbar" (Q }{\i\lang1033 hwarma}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par Because of this early merger of }{\b\lang1033 ch}{\b\lang1033\super w}{\lang1033 with }{\b\lang1033 hw}{\lang1033 , "distinct signs for }{\i\lang1033 chw}{\lang1033 and }{\i\lang1033 hw}{\lang1033 were not required" in F\'eb anorian writing (LotR Appendix E). There really was a merger, for it should be noted that the sounds }{\b\lang1033 hw}{\lang1033 and }{\b\lang1033 hy}{\lang1033 were not quite new to the language: It seems that at this stage, proto-Quenya already had }{ \b\lang1033 hw}{\lang1033 derived from }{\b\lang1033 sw}{\lang1033 and from the aspirate }{\b\lang1033 p}{\b\lang1033\super h}{\lang1033 in front of }{\b\lang1033 u}{\lang1033 , and possibly also }{\b\lang1033 hy}{\lang1033 derived from earlier }{ \b\lang1033 sj}{\lang1033 - (see above). \par \par }{\i\lang1033 THE SHORTENING OF FINAL VOWELS IN POLYSYLLABIC WORDS:}{\lang1033 At some point perhaps not very far from the beginning of the "historical" period, the }{\i\lang1033 long final vowels}{\lang1033 so characteristic of Primitive Elvish are finally shortened. (The late Common Eldarin and early proto-Quenya rule that final short -}{\b\lang1033 i}{\lang1033 and -}{\b\lang1033 u}{\lang1033 turn into -}{\b\lang1033 e}{\lang1033 and -}{\b\lang1033 o}{ \lang1033 is no longer operating, so the final -}{\b\lang1033 u}{\lang1033 and -}{\b\lang1033 i}{\lang1033 that now develop from earlier -}{\b\lang1033 \'fb}{\lang1033 and -}{\b\lang1033 \'ee}{\lang1033 are }{\i\lang1033 not}{\lang1033 so changed.) \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \'a4}{\b\lang1033 korn\'e2}{\lang1033 > Q }{\b\lang1033 corna}{\lang1033 "round" \par \'a4}{\b\lang1033 Valin\'f4r\'ea}{\lang1033 (WJ:413) > Q }{\b\lang1033 Valin\'f3re}{\lang1033 "Valinor" \par \'a4}{\b\lang1033 t\'e2r\'ee}{\lang1033 > Q }{\b\lang1033 t\'e1ri}{\lang1033 "queen" \par \'a4}{\b\lang1033 kund\'fb}{\lang1033 > Q }{\b\lang1033 cundu}{\lang1033 "prince" \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par Evidently by analogy, the shortened vowels were also introduced in positions where the vowels were not absolutely final, as in the inflected forms of nouns. For instance, the dative of }{\b\lang1033 lass\'eb}{\lang1033 "leaf" (earlier \'a4}{\b\lang1033 lass\'ea}{\lang1033 ) is }{\b\lang1033 lassen}{\lang1033 , not **}{\b\lang1033 lass\'e9n}{\lang1033 . \par }{\b\lang1033 \par }{\lang1033 The example }{\b\lang1033 Valin\'f4r\'ea}{\lang1033 from WJ:413 is of interest regarding the sequence of the changes: It demonstrates that the change from voiced plosives to spirants, like }{\b\lang1033 b}{\lang1033 > }{\b\lang1033 v,}{ \lang1033 occurred }{\i\lang1033 before}{\lang1033 the shortening of the long final vowels. \par \par }{\fs20\lang1033 NOTE: Tolkien may have imagined a different sequence when he wrote the }{\i\fs20\lang1033 Etymologies}{\fs20\lang1033 , where we find some primitive words that already seem to have shortened the long final vowels, but still have not turned the voiced stops and the unvoiced aspirates into spirants. For instance, we have \'a4}{\b\fs20\lang1033 r\'e2ba}{\fs20\lang1033 (> Q }{\i\fs20\lang1033 r\'e1va}{\fs20\lang1033 "wild") and \'a4}{\b\fs20\lang1033 k}{\b\fs20\lang1033\super h}{\b\fs20\lang1033 agda}{\fs20\lang1033 (> Q *}{\i\fs20\lang1033 chahta}{\fs20\lang1033 > }{\i\fs20\lang1033 hahta}{\fs20\lang1033 "mound") with }{\b\fs20\lang1033 b}{\fs20\lang1033 and }{\b\fs20\lang1033 k}{\b\fs20\lang1033\super h}{\fs20\lang1033 still intact, though the final -}{\b\fs20\lang1033 a}{\fs20\lang1033 has already been shortened. It }{\i\fs20\lang1033 must}{ \fs20\lang1033 have been long at the oldest stage, or it would not have survived into Quenya (cf. \'a4}{\b\fs20\lang1033 swanda}{\fs20\lang1033 > Q }{\i\fs20\lang1033 hwan}{\fs20\lang1033 "sponge", not **}{\i\fs20\lang1033 hwanda}{\fs20\lang1033 ). But then there are also words in the }{\i\fs20\lang1033 Etymologies}{\fs20\lang1033 that would seem to confirm that the spirants entered the language }{\i\fs20\lang1033 before}{\fs20\lang1033 the final vowels were shortened. Cf. \'a4}{\b\fs20\lang1033 we3 \'ea}{\fs20\lang1033 (> Q }{\i\fs20\lang1033 vi\'eb}{\fs20\lang1033 "manhood"), that comes from a root }{\i\fs20\lang1033 WEG}{\fs20\lang1033 and must therefore have been *}{\b\fs20\lang1033 weg\'ea}{\fs20\lang1033 at an older stage; the form \'a4}{ \b\fs20\lang1033 we3\'ea}{\fs20\lang1033 cannot be Primitive Quendian or Common Eldarin, but is rather proto-Quenya, after the voiced stops had become spirants in many positions. Yet in \'a4}{\b\fs20\lang1033 we3\'ea}{\fs20\lang1033 the long final vowel is still in place, confirming the order of changes suggested by }{\b\fs20\lang1033 Valin\'f4r\'ea}{\fs20\lang1033 in WJ:413. The }{\i\fs20\lang1033 Etymologies}{\fs20\lang1033 material is thus not wholly consistent; we may simply assume that the macron denoting a long final vowel has dropped out in a number of cases. This is also seen in later sources; in WJ:361 a certain Primitive Quendian verb is quoted as \'a4}{ \b\fs20\lang1033 hek-t\'e2}{\fs20\lang1033 , but in WJ:363 we have \'a4}{\b\fs20\lang1033 hek-ta}{\fs20\lang1033 instead \endash plainly an error, whether the author or the transcriber is to be blamed. \par }{\lang1033 \par However, Quenya still had some long final vowels at a later stage, in the }{\i\lang1033 accusative}{\lang1033 form of nouns. In the Plotz Letter, Tolkien indicated that in the conservative "Book Quenya", a noun like }{\b\lang1033 cirya}{\lang1033 "ship" had the accusative form }{\b\lang1033 ciry\'e1}{\lang1033 , whereas }{\b\lang1033 lass\'eb}{\lang1033 "leaf" had the accusative form }{\b\lang1033 lass\'e9}{\lang1033 . How are we to account for these forms? It has been su ggested that at an older stage, there was an accusative marker *-}{\b\lang1033 h}{\lang1033 . Whereas }{\i\lang1033 medial}{\lang1033 }{\b\lang1033 h}{\lang1033 was lost already in Common Eldarin (WJ:368), it may be that }{\i\lang1033 final}{\lang1033 -}{\b\lang1033 h}{\lang1033 survived much longer. If older \'a4}{\b\lang1033 lass\'ea}{\lang1033 "leaf" had an accusative form *}{\b\lang1033 lass\'eah}{\lang1033 , the long vowel of the latter word would initially be shortened by analogy with the development of the simplex form: the new accusative would be *}{\b\lang1033 lasseh}{\lang1033 . However, final *-}{\b\lang1033 h}{\lang1033 was apparently lost before the beginning of the historical period, and the preceding vowel was lengthened in compensation. Hence presumably *}{\b\lang1033 lasseh}{\lang1033 > acc. }{\b\lang1033 lass\'e9}{\lang1033 , *}{\b\lang1033 ciryah}{\lang1033 > acc. }{\b\lang1033 ciry\'e1}{\lang1033 , long final vowels arising anew. (Presumably we would even see long final vowels in the accusatives of nouns that originally had }{\i\lang1033 short}{\lang1033 final vowels.) \par \par }{\i\lang1033 Devoicing of medial combinations: }{\lang1033 Perhaps at this point, a number of medial clusters of two voiced sounds are seen to become }{\i\lang1033 unvoiced}{\lang1033 . Examples include }{\b\lang1033 zd}{\lang1033 > }{\b\lang1033 st}{ \lang1033 , }{\b\lang1033 zg}{\lang1033 > }{\b\lang1033 sk}{\lang1033 , and }{\b\lang1033 gd}{\lang1033 > }{\b\lang1033 kt}{\lang1033 : \par \par \tab (\'a4}{\i\lang1033 esd\'ea}{\lang1033 > \'a4}{\i\lang1033 ezd\'ea}{\lang1033 >) }{\b\lang1033 ezde }{\lang1033 > }{\b\lang1033 est\'eb}{\lang1033 "rest" (Q }{\i\lang1033 Est\'eb}{\lang1033 as name of a Vali\'eb, WJ:403) \par \tab (\'a4}{\i\lang1033 mizd\'ea}{\lang1033 >) *}{\b\lang1033 mizde}{\lang1033 > Q }{\b\lang1033 mist\'eb}{\lang1033 "fine rain" \par \tab (*}{\i\lang1033 masg\'e2}{\lang1033 > \'a4}{\i\lang1033 mazg\'e2}{\lang1033 >) *}{\b\lang1033 mazga}{\lang1033 > *}{\b\lang1033 maska}{\lang1033 "soft" (Q }{\i\lang1033 maxa}{\lang1033 ) \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\b\lang1033 (}{\lang1033 \'a4}{\i\lang1033 k}{\i\lang1033\super h}{\i\lang1033 agda}{\lang1033 [read *}{\i\lang1033 k}{\i\lang1033\super h}{\i\lang1033 agd\'e2}{\lang1033 ] >) *}{\b\lang1033 chagda}{ \lang1033 > *}{\b\lang1033 chakta}{\lang1033 "mound" (Q *}{\i\lang1033 chahta}{\lang1033 > }{\i\lang1033 hahta}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par Moreover, we can deduce that }{\b\lang1033 bd}{\lang1033 at one stage turned into }{\b\lang1033 pt}{\lang1033 : \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (*}{\i\lang1033 libd\'e2}{\lang1033 >) \'a4}{\b\lang1033 libda}{\lang1033 > *}{\b\lang1033 lipta}{\lang1033 "soap" (Q }{\i\lang1033 lipsa}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par We rely on the example \'a4}{\b\lang1033 libda}{\lang1033 to place the devoicing of these combinations in the period postdating the shortening of the long final vowels. The example \'a4}{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 agda}{\lang1033 points in the same direction, but other evidence suggests that the initial aspirate }{\b\lang1033 k}{\b\lang1033\super h}{\lang1033 - should already have turned into a spirant }{\b\lang1033 ch}{\lang1033 - when the final vowel had become short. \par \par According to the }{\i\lang1033 Etymologies}{\lang1033 , the combination }{\b\lang1033 pt}{\lang1033 (of whatever origin) later turned into }{\b\lang1033 ps}{\lang1033 : \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 libda}{\lang1033 >) *}{\b\lang1033 lipta}{\lang1033 > Q }{\b\lang1033 lipsa}{\lang1033 "soap" \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \tab (*}{\i\lang1033 lepet\'ea}{\lang1033 >) *}{\b\lang1033 lepte}{\lang1033 > Q }{\b\lang1033 leps\'eb}{\lang1033 "finger" \par \tab }{\b\lang1033 \par }{\lang1033 For the primitive form *}{\b\lang1033 lepet\'ea}{\lang1033 , cf. the root }{\i\lang1033 LEPET}{\lang1033 and the "Noldorin" cognate }{\b\lang1033 lhebed}{\lang1033 (Sindarin *}{\b\lang1033 lebed}{\lang1033 ), LR:368. Conceivably the most primitive form should rather be *}{\b\lang1033 lepeti}{\lang1033 , but the immediate ancestor of Quenya }{\b\lang1033 leps\'eb}{\lang1033 would still be *}{\b\lang1033 lepte}{\lang1033 . \par \par It may be that Tolkien dropped the "}{\b\lang1033 pt}{\lang1033 > }{\b\lang1033 ps}{\lang1033 " rule; SD:68, 72 gives }{\b\lang1033 raccalepta}{\lang1033 for "claw-fingered". This adjective }{\b\lang1033 lepta}{\lang1033 "fingered" must probably be derived from *}{\b\lang1033 lepet\'e2}{\lang1033 via *}{\b\lang1033 lept\'e2}{\lang1033 ; in any case, the word must be closely related to }{\b\lang1033 leps\'eb}{\lang1033 "finger". A late source also has }{\b\lang1033 tapta}{\lang1033 rather than *}{\b\lang1033 tapsa}{\lang1033 as a word for "impeded" (VT39:17). Some would emend }{\b\lang1033 leps\'eb}{\lang1033 "finger" and }{\b\lang1033 lipsa}{\lang1033 "soap" to *}{\b\lang1033 lept\'eb}{\lang1033 , *}{ \b\lang1033 lipta}{\lang1033 \endash to go with Tolkien's later insights on Quenya phonology. Perhaps we may rather assume that there were minor dialectal variants. \par \par "Before the separation of Vanyar and \'d1oldor," both clans alter }{\b\lang1033 hw}{\lang1033 - to }{\b\lang1033 h}{\lang1033 - before the vowel }{\b\lang1033 u}{\lang1033 - (VT41:8): \par \par \tab (*}{\i\lang1033 p}{\i\lang1033\super h}{\i\lang1033 uin\'ea}{\lang1033 > *}{\i\lang1033 hwuin\'ea}{\lang1033 >) *}{\b\lang1033 hwuin\'eb}{\lang1033 > }{\b\lang1033 huin\'eb}{\lang1033 "gloom, deep shadow" \par \par In the }{\i\lang1033 Etymologies}{\lang1033 , this word has a Quenya side-form }{\b\lang1033 fuin\'eb}{\lang1033 (LR:382), but according to VT41:8, }{\b\lang1033 fuin\'eb}{\lang1033 is actually a Telerin form. \par \par Initially, the primitive nasalized explosives }{\b\lang1033 mb}{\lang1033 -, }{\b\lang1033 nd}{\lang1033 -, }{\b\lang1033 \'f1g}{\lang1033 - are normally reduced to }{\b\lang1033 m}{\lang1033 -, }{\b\lang1033 n}{\lang1033 -, and }{\b\lang1033 \'f1}{ \lang1033 -. (As for certain alternative developments, as when \'a4}{\b\lang1033 nd\'fbn\'ea}{\lang1033 produces Quenya }{\b\lang1033 and\'fan\'eb}{\lang1033 instead of **}{\b\lang1033 n\'fan\'eb}{\lang1033 , see below.) If we can trust the examples \'a4}{\b\lang1033 ndeuna}{\lang1033 and \'a4}{\b\lang1033 ndeuro}{\lang1033 found in the entry }{\i\lang1033 NDEW}{\lang1033 in the }{\i\lang1033 Etymologies}{\lang1033 , this simplification took place }{\i\lang1033 after}{\lang1033 the shortening of the final long vowels (or we would see *}{\b\lang1033 ndeun\'e2}{\lang1033 and *}{\b\lang1033 ndeur\'f4}{\lang1033 with the long vowels intact): \par \tab \par }\pard \qj\nowidctlpar\widctlpar\adjustright {\lang1033 \tab (*}{\i\lang1033 ndeun\'e2}{\lang1033 >) \'a4}{\b\lang1033 ndeuna}{\lang1033 > Q }{\b\lang1033 neuna}{\lang1033 "second" \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \tab (*}{\i\lang1033 ndeur\'f4}{\lang1033 >) \'a4}{\b\lang1033 ndeuro}{\lang1033 > Q }{\b\lang1033 neuro}{\lang1033 "successor" \par }\pard \qj\nowidctlpar\widctlpar\adjustright {\lang1033 \tab (\'a4}{\i\lang1033 \'f1golod\'f4}{\lang1033 > *}{\i\lang1033 \'f1gold\'f4}{\lang1033 >) *}{\b\lang1033 \'f1goldo}{\lang1033 > Q }{\b\lang1033 \'d1oldo}{\lang1033 "Noldo" \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \tab (*}{\i\lang1033 \'f1g}{\i\lang1033\super w}{\i\lang1033 alm\'ea}{\lang1033 >) *}{\b\lang1033 \'f1g}{\b\lang1033\super w}{\b\lang1033 alme}{\lang1033 > Q }{\b\lang1033 \'f1walm\'eb}{\lang1033 "torment" }{\fs20\lang1033 (spelt }{\b\fs20\lang1033 ngwalme}{\fs20\lang1033 in LotR, Appendix E)}{\lang1033 \par \tab (\'a4}{\i\lang1033 mband\'f4}{\lang1033 >) *}{\b\lang1033 mbando}{\lang1033 > Q }{\b\lang1033 mando}{\lang1033 "custody" (MR:350) \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (*}{\i\lang1033 mbarat\'e2}{\lang1033 > *}{\i\lang1033 mbart\'e2}{\lang1033 >) *}{\b\lang1033 mbarta}{\lang1033 > Q }{\b\lang1033 marta}{\lang1033 "fey, fated" }{\fs20\lang1033 (for the middle vowel of *}{\i\fs20\lang1033 mbarat\'e2}{\fs20\lang1033 , later syncopated, cf. the base }{\i\fs20\lang1033 MBAR}{\i\fs20\ul\lang1033 A}{\i\fs20\lang1033 T}{\fs20\lang1033 and the "Noldorin"/Sindarin cognate }{\i\fs20\lang1033 bar}{ \i\fs20\ul\lang1033 a}{\i\fs20\lang1033 d}{\fs20\lang1033 )}{\lang1033 \par }\pard \qj\nowidctlpar\widctlpar\adjustright {\lang1033 \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 Properly, this reduction only occurs initially, but in some cases, it can be observed medially as well. The primitive word \'a4}{\b\lang1033 andambund\'e2}{\lang1033 "long-snouted" produces Quenya }{ \b\lang1033 andamunda}{\lang1033 "elephant" (not **}{\b\lang1033 andambunda}{\lang1033 ). The latter element of a compound might be "reformed after the simplex" (LR:350); in this way Tolkien in the }{\i\lang1033 Etymologies}{\lang1033 explained how \'a4 }{\b\lang1033 bal\'ee-nd\'f4re}{\lang1033 "Valar-land" could end up as }{\b\lang1033 Valinor}{\lang1033 instead of *}{\b\lang1033 Valindor}{\lang1033 : Occurring as a "simplex", by itself, *}{\b\lang1033 nd\'f4r\'ea}{\lang1033 yielded Quenya }{ \b\lang1033 n\'f3r\'eb}{\lang1033 "land" \endash and what would be *}{\b\lang1033 Valindor}{\lang1033 was altered to }{\b\lang1033 Valinor}{\lang1033 by analogy (but see WJ:413, note 25, for a somewhat different explanation which Tolkien came up with decades later). So *}{\b\lang1033 andambunda}{\lang1033 might become }{\b\lang1033 andamunda}{\lang1033 by analogy with a related word like }{ \b\lang1033 mundo}{\lang1033 "snout" (itself < \'a4}{\b\lang1033 mbundu}{\lang1033 ). Where the second element of a compound normally does not occur as a separate word, }{\b\lang1033 nd}{\lang1033 /}{\b\lang1033 mb}{\lang1033 /}{\b\lang1033 \'f1g}{ \lang1033 is normally preserved unchanged also medially. Prominent examples are the endings -}{\b\lang1033 ndil}{\lang1033 "friend" and -}{\b\lang1033 ndur}{\lang1033 "servant", as in }{\b\lang1033 Eldandil}{\lang1033 "Elf-friend" (WJ:410). There are no separate "simplex" words **}{\b\lang1033 nil}{\lang1033 and **}{\b\lang1033 nur}{\lang1033 , though longer derivatives from the same stems are found in Quenya (e.g. }{\b\lang1033 nildo}{\lang1033 "friend" < *}{ \b\lang1033 ndild\'f4}{\lang1033 ). Indeed we may also include the element -}{\b\lang1033 ndor}{\lang1033 "land", as in }{\b\lang1033 Valandor}{\lang1033 "Vala-land" (WJ:413, after Tolkien had rejected or forgotten his original explanation of }{ \b\lang1033 Valinor}{\lang1033 \endash in }{\i\lang1033 Quendi and Eldar}{\lang1033 he invoked, not analogy with a simplex form, but rather confusion with a distinct root that never had initial }{\b\lang1033 nd}{\lang1033 - to explain why we do not see *}{\b\lang1033 Valindor}{\lang1033 ). \par \par }{\i\lang1033 NEW VOWELS:}{\lang1033 Around this stage, a number of vowels appeared "out of nothing" in Quenya. This usually occurred to turn }{\i\lang1033 syllabic consonants}{\lang1033 into normal syllables including a vowel. Henceforth, syllabic consonants were not tolerated in the language. \par \par This explains why in some cases the pre-nasalized initial stops }{\b\lang1033 mb}{\lang1033 , }{\b\lang1033 nd}{\lang1033 , }{\b\lang1033 \'f1g}{\lang1033 were }{\i\lang1033 not}{\lang1033 simplified to }{\b\lang1033 m}{\lang1033 , }{\b\lang1033 n}{ \lang1033 , }{\b\lang1033 \'f1}{\lang1033 as set out above. As described above, during the Common Eldarin stage the nasal element had occasionally been "strengthened" so that it had come to constitute a syllable by itself. Tolkien explicitly deals with one example in PM:360: The primitive noun \'a4}{\b\lang1033 \'f1golod\'f4}{\lang1033 could appear both with and without syllabic }{\b\lang1033 \'f1}{\lang1033 . With no syllabic }{\b\lang1033 \'f1}{\lang1033 it produced Quenya }{ \b\lang1033 \'d1oldo}{\lang1033 as described above. But where a syllabic }{\b\lang1033 \'f1}{\lang1033 occurred (*}{\b\lang1033 \'f1\'b7golod\'f4}{\lang1033 ), it came out as }{\b\lang1033 Ingoldo}{\lang1033 in Quenya, an }{\b\lang1033 i}{\lang1033 developing before the syllabic consonant. }{\b\lang1033 Ingoldo}{\lang1033 was a "strengthened" form, "being in full and more dignified form is more or less equivalent to '}{\i\lang1033 the}{\lang1033 \'d1 oldo, one eminent in the kindred' " (PM:360). As is evident from this example, an }{\b\lang1033 i}{\lang1033 developed before syllabic }{\b\lang1033 \'f1}{\lang1033 (except in the case of }{\b\lang1033 \'f1\'b7g}{\b\lang1033\super w}{\lang1033 , see below) but before a syllabic }{\b\lang1033 n}{\lang1033 an }{\b\lang1033 a}{\lang1033 appears, and before a syllabic }{\b\lang1033 m}{\lang1033 there develops a }{\b\lang1033 u}{\lang1033 : \par \tab \par \tab (\'a4}{\i\lang1033 n\'b7d\'fbn\'ea}{\lang1033 >) *}{\b\lang1033 n\'b7d\'fbne}{\lang1033 > }{\b\lang1033 and\'fbne }{\lang1033 "sunset" (still }{\i\lang1033 and\'fan\'eb}{\lang1033 in Q) \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (*}{\i\lang1033 \'f1\'b7gj\'f4}{\b\lang1033 >) }{\lang1033 *}{\b\lang1033 \'f1\'b7gjo}{\lang1033 > *}{\b\lang1033 i\'f1gjo}{\lang1033 "grandchild, descendant" (Q }{\i\lang1033 indyo}{\lang1033 ) \par (*}{\i\lang1033 m\'b7barta}{\lang1033 [final vowel uncertain] >) *}{\b\lang1033 m\'b7bart}{\lang1033 > *}{\b\lang1033 umbart}{\lang1033 "fate" (Q }{\i\lang1033 umbar}{\lang1033 with stem }{\i\lang1033 umbart}{\lang1033 -) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par Syllabic }{\b\lang1033 \'f1}{\lang1033 before }{\b\lang1033 g}{\b\lang1033\super w}{\lang1033 however prefers the vowel }{\b\lang1033 u}{\lang1033 (as we noted above, it seems that this }{\b\lang1033 \'f1}{\lang1033 had itself become labialized to }{ \b\lang1033 \'f1}{\b\lang1033\super w}{\lang1033 already in Common Eldarin, by assimilation to the following }{\b\lang1033 g}{\b\lang1033\super w}{\lang1033 \endash and its labialized quality colours the quality of the vowel that develops before it): \par \par \tab (*}{\i\lang1033 \'f1\'b7g}{\i\lang1033\super w}{\i\lang1033 al\'ea}{\b\lang1033 >) }{\lang1033 *}{\b\lang1033 \'f1}{\b\lang1033\super w}{\b\lang1033 \'b7g}{\b\lang1033\super w}{\b\lang1033 ale}{\lang1033 >}{\b\lang1033 u\'f1}{\b\lang1033\super w}{ \b\lang1033 g}{\b\lang1033\super w}{\b\lang1033 ale}{\lang1033 "torture" (spelt "ungwal\'eb" in Q) \par \par Syllabic }{\b\lang1033 l}{\lang1033 triggered the development of the vowel }{\b\lang1033 i}{\lang1033 . As recorded in the entry }{\i\lang1033 LA}{\lang1033 in the }{\i\lang1033 Etymologies}{\lang1033 , there was a negation prefix \'a4}{\b\lang1033 la}{ \lang1033 - "not-, un-" that had lost its vowel (not a regular development), so that only a syllabic (or "vocalic") }{\b\lang1033 l}{\lang1033 was left: This prefix became }{\b\lang1033 il}{\lang1033 - in Quenya, as in }{\b\lang1033 ilfirin}{\lang1033 "immortal" (for *}{\b\lang1033 l\'b7p}{\b\lang1033\super h}{\b\lang1033 irin}{\lang1033 -). It has been suggested that the Quenya verb }{\b\lang1033 ilca}{\lang1033 - "to gleam" comes from the same root that produces Sindarin }{\b\lang1033 lach}{ \lang1033 "flame", evidently *}{\i\lang1033 LAK}{\lang1033 -; if so }{\b\lang1033 ilca}{\lang1033 - could come from an original *}{\b\lang1033 l\'b7k\'e2}{\lang1033 , an unusual primitive form suppressing the stem-vowel. An }{\b\lang1033 i}{\lang1033 likewise developed before the }{\i\lang1033 final}{\lang1033 syllabic }{\b\lang1033 l}{\lang1033 's that had become final after the loss of final short -}{\b\lang1033 a}{\lang1033 in Common Eldarin. This change gives these words their "final" Quenya shape (the difference }{\b\lang1033 k}{\lang1033 /}{\b\lang1033 c}{\lang1033 is of course purely orthographic): \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 hekla }{\lang1033 >) *}{\b\lang1033 hek\'b7l}{\lang1033 > Q }{\b\lang1033 hecil}{\lang1033 "outcast" \par (\'a4}{\i\lang1033 makla}{\lang1033 >) *}{\b\lang1033 mak\'b7l}{\lang1033 > Q }{\b\lang1033 macil}{\lang1033 "sword" \par (\'a4}{\i\lang1033 tankla}{\lang1033 >) *}{\b\lang1033 tank\'b7l}{\lang1033 > Q }{\b\lang1033 tancil}{\lang1033 "brooch" \par (\'a4}{\i\lang1033 tekla}{\lang1033 >) *}{\b\lang1033 tek\'b7l}{\lang1033 > Q }{\b\lang1033 tecil}{\lang1033 "pen" \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par It may be that the Q word }{\b\lang1033 siril}{\lang1033 "rivulet" comes from *}{\b\lang1033 sirla}{\lang1033 >}{\b\lang1033 }{\lang1033 *}{\b\lang1033 sir\'b7l}{\lang1033 ; if so, this would demonstrate that the vowel }{\b\lang1033 i}{\lang1033 developed before a final syllabic }{\b\lang1033 l}{\lang1033 }{\i\lang1033 before}{\lang1033 the assimilation }{\b\lang1033 rl}{\lang1033 > }{\b\lang1033 ll}{\lang1033 occurred (see below) \endash the intruding vowel breaking up the cluster before it could be assimilated. \par \par In one case, an }{\b\lang1033 a}{\lang1033 rather than an }{\b\lang1033 i}{\lang1033 develops before }{\b\lang1033 l}{\lang1033 . In the }{\i\lang1033 Etymologies}{\lang1033 , Tolkien derived the noun }{\b\lang1033 cala}{\lang1033 "light" from a primitive form "k'l\'e2", perhaps = }{\b\lang1033 k\'b7l\'b7\'e2}{\lang1033 . But since the stem is given as }{\i\lang1033 KAL}{\lang1033 , it would seem very much easier to derive }{\b\lang1033 cala}{\lang1033 from *}{\b\lang1033 kal\'e2 }{\lang1033 , instead of supposing a weird primitive form that suppresses the stem-vowel only to have an identical vowel arising later. \par \par In some cases an }{\b\lang1033 a}{\lang1033 also developed before }{\b\lang1033 r}{\lang1033 , because it had become syllabic in certain positions: \par \par \tab (*}{\i\lang1033 k}{\i\lang1033\super w}{\i\lang1033 entr\'f4}{\lang1033 >) \'a4}{\b\lang1033 k}{\b\lang1033\super w}{\b\lang1033 ent\'b7r\'b7o}{\lang1033 > Q }{\b\lang1033 quentaro}{\lang1033 "narrator" }{\fs20\lang1033 (the distinction }{ \b\fs20\lang1033 k}{\b\fs20\lang1033\super w}{\fs20\lang1033 /}{\b\fs20\lang1033 qu}{\fs20\lang1033 is of course orthographic only) \par }{\lang1033 \par Here we have a syllabic }{\b\lang1033 r}{\lang1033 developing after a consonant cluster. In the case of Q }{\b\lang1033 tambaro}{\lang1033 "knocker" (used = "woodpecker"), which Tolkien indicates was ultimately derived from \'a4}{\b\lang1033 tamr\'f4}{ \lang1033 , we must probably assume an intermediate form *}{\b\lang1033 tamb\'b7r\'b7o}{\lang1033 . There are no other examples of the strange development }{\b\lang1033 mr}{\lang1033 > }{\b\lang1033 mbr}{\lang1033 , but if a group }{\b\lang1033 mbr}{ \lang1033 arose, the further development would confirm that }{\b\lang1033 r}{\lang1033 became syllabic following a consonant cluster. The word }{\b\lang1033 Istar}{\lang1033 "wizard" is perhaps to be derived from *}{\b\lang1033 istro}{\lang1033 > early Q *}{\b\lang1033 ist\'b7r}{\lang1033 , given the Sindarin cognate }{\i\lang1033 Ithron}{\lang1033 (apparently representing an extended form *}{\b\lang1033 istro-n-}{\lang1033 ; the point is that for }{\b\lang1033 st}{\lang1033 to have become }{ \i\lang1033 th}{\lang1033 = }{\i\lang1033 \'fe}{\lang1033 , it must at an earlier stage have been in direct contact with the following }{\b\lang1033 r}{\lang1033 ). \par }\pard \qj\nowidctlpar\widctlpar\adjustright {\lang1033 \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 Around this stage, various medial consonant clusters underwent changes. Some were complete }{\i\lang1033 assimilations}{\lang1033 resulting in double consonants: \par \par LR > LL: \tab (\'a4}{\i\lang1033 talrunja}{\lang1033 >) *}{\b\lang1033 talrune}{\lang1033 > Q }{\b\lang1033 tallun\'eb}{\lang1033 "sole of foot" \par }\pard \fi708\li708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 kalr\'f4}{\lang1033 >) *}{\b\lang1033 kalro}{\lang1033 > Q }{\b\lang1033 callo}{\lang1033 "hero" \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 NL > LL: \tab (*}{\i\lang1033 nenl\'ea}{\lang1033 >) \'a4}{\b\lang1033 nenle}{\lang1033 > Q }{\b\lang1033 nell\'eb}{\lang1033 "brook" \par LS > LL: \tab (\'a4}{\i\lang1033 t\'e9les\'e2}{\lang1033 > *}{\i\lang1033 tels\'e2}{\lang1033 >) *}{\b\lang1033 telsa}{\lang1033 > Q }{\b\lang1033 tella}{\lang1033 "hindmost" \tab \tab \tab \tab \tab (\'a4}{\i\lang1033 k}{\i\lang1033\super hj}{ \i\lang1033 eles\'ea}{\lang1033 > *}{\i\lang1033 ch}{\i\lang1033\super j}{\i\lang1033 eles\'ea}{\lang1033 > *}{\i\lang1033 hyeles\'ea}{\lang1033 > *}{\i\lang1033 hyels\'ea}{\lang1033 >) *}{\b\lang1033 hyelse}{\lang1033 > Q }{\b\lang1033 hyell\'eb}{ \lang1033 "glass" \par RS > SS:\tab (\'a4}{\i\lang1033 neres\'e2}{\lang1033 >) *}{\b\lang1033 Nersa}{\lang1033 > Q }{\b\lang1033 Nessa}{\lang1033 (name of a Vali\'eb) (WJ:416) \par SM > MM: \tab (\'a4}{\i\lang1033 asmal\'ea}{\lang1033 >) *}{\b\lang1033 asmale}{\lang1033 > Q }{\b\lang1033 ammal\'eb}{\lang1033 "yellow bird" \par SR > RR: \tab \'a4}{\b\lang1033 mi-srawanwe}{\lang1033 > *}{\b\lang1033 mirrawanwe}{\lang1033 "incarnate" (Q }{\i\lang1033 mirroanw\'eb}{\lang1033 , MR:350) \par \'feR > \'fe\'fe: \tab (\'a4}{\i\lang1033 net}{\i\lang1033\super h}{\i\lang1033 -r\'e2}{\lang1033 >) *}{\b\lang1033 ne\'fer\'e2}{\lang1033 > *}{\b\lang1033 ne\'fe\'fea}{\lang1033 (Q }{\i\lang1033 nessa}{\lang1033 "young") \par \par Regarding medial }{\b\lang1033 sm}{\lang1033 , the examples are somewhat divergent. In \'a4}{\b\lang1033 asmal\'ea}{\lang1033 > Q *}{\b\lang1033 ammal\'eb}{\lang1033 we see a straightforward assimilation. Yet Tolkien (in LR:362 s.v. }{\i\lang1033 KAS}{ \lang1033 ) has primitive \'a4}{\b\lang1033 kas-sa}{\lang1033 /\'a4}{\b\lang1033 kas-ma}{\lang1033 yield Quenya }{\b\lang1033 cassa}{\lang1033 "helmet"; if }{\b\lang1033 cassa}{\lang1033 is to be referred to }{\b\lang1033 kas-ma}{\lang1033 , we must postulate a progressive assimilation }{\b\lang1033 sm}{\lang1033 > }{\b\lang1033 ss}{\lang1033 (and then }{\b\lang1033 ammal\'eb}{\lang1033 should have been **}{\b\lang1033 assal\'eb}{\lang1033 instead!) However, }{\b\lang1033 cassa}{ \lang1033 is perhaps rather to be derived from \'a4}{\b\lang1033 kas-sa}{\lang1033 . In a later source, the Quenya word for "helm(et)" appears as }{\b\lang1033 carma}{\lang1033 instead (PM:260). If this is to be derived from \'a4}{\b\lang1033 kas-ma}{ \lang1033 , the group }{\b\lang1033 sm}{\lang1033 is }{\i\lang1033 not}{\lang1033 assimilated to }{\b\lang1033 mm}{\lang1033 (or }{\b\lang1033 ss}{\lang1033 ), but simply remains unchanged. (Later, }{\b\lang1033 s}{\lang1033 occurring between two voiced sounds is regularly voiced to }{\b\lang1033 z}{\lang1033 , in turn producing Exilic Quenya }{\b\lang1033 r}{\lang1033 \endash hence the form }{\b\lang1033 carma}{\lang1033 . But then Q }{\b\lang1033 ammal\'eb}{\lang1033 from \'a4}{\b\lang1033 asmal\'ea}{\lang1033 should have become **}{\b\lang1033 armal\'eb}{\lang1033 instead!) It seems that we shall simply have to accept that medial }{\b\lang1033 sm}{\lang1033 could develop in various ways. \par \par In WJ:363, Tolkien also indicates that }{\b\lang1033 \'f0l}{\lang1033 would be assimilated to }{\b\lang1033 ll}{\lang1033 , though no good example can be quoted (Tolkien mentioned that a form \'a4}{\b\lang1033 e\'f0lo}{\lang1033 }{\i\lang1033 would have }{\lang1033 produced Quenya \'a4}{\b\lang1033 ello}{\lang1033 , but this word he asterisked, for it was "not found"). The group }{\b\lang1033 \'f0r}{\lang1033 would likewise have been assimilated to }{\b\lang1033 rr}{\lang1033 , and here we may have a direct example, though a garbled one: From a primitive form }{\b\lang1033 n}{\b\lang1033\super j}{\b\lang1033 adr\'f4}{\lang1033 , literally *"gnawer", Tolkien derived a Quenya word for "rat" that his son read as "nyano" (LR:379). However, this is certainly a misreading for }{\b\lang1033 nyarro}{\lang1033 , representing older *}{\b\lang1033 n}{\b\lang1033\super j}{\b\lang1033 a\'f0ro}{\lang1033 . (In the normal Roman spelling of Quenya, }{\b\lang1033 n}{\b\lang1033\super j}{\lang1033 is represented as }{\b\lang1033 ny}{\lang1033 , but this is only a matter of orthography, like }{\b\lang1033 k}{\b\lang1033\super w}{\lang1033 = }{\b\lang1033 qu} {\lang1033 .) Since post-vocalic }{\b\lang1033 \'f0}{\lang1033 would later become }{\b\lang1033 r}{\lang1033 (via }{\b\lang1033 z}{\lang1033 ) anyway, we cannot be certain that this assimilation really occurred. If not, the word for "rat" would appear as *}{\b\lang1033 nyazro}{\lang1033 in the Vanyarin dialect.}{\i\lang1033 \par }{\lang1033 \par The example \'a4}{\b\lang1033 nenle}{\lang1033 (not older *}{\b\lang1033 nenl\'ea}{\lang1033 ) given in the }{\i\lang1033 Etymologies}{\lang1033 would seem to indicate that the assimilation }{\b\lang1033 nl}{\lang1033 > }{\b\lang1033 ll}{\lang1033 occurred }{\i\lang1033 after}{\lang1033 the shortening of the long final vowels, so perhaps all of these assimilations can be placed at this stage. \par \par There are also some cases of }{\i\lang1033 partial}{\lang1033 assimilation/dissimilation (not producing double consonants): \par \par LN > LD: (\'a4}{\i\lang1033 skaln\'e2}{\lang1033 > *}{\i\lang1033 k}{\i\lang1033\super h}{\i\lang1033 aln\'e2}{\lang1033 >) *}{\b\lang1033 chalna}{\lang1033 > *}{\b\lang1033 chalda}{\lang1033 "hidden" (in later Q }{\i\lang1033 halda}{\lang1033 ) \par \par Cf. also WJ:362 about the plural form (\'a4}{\i\lang1033 elen\'ee}{\lang1033 > *}{\i\lang1033 eln\'ee }{\lang1033 >) *}{\b\lang1033 elni}{\lang1033 "stars" becoming }{\b\lang1033 eldi}{\lang1033 (though it was usually replaced by the analogical form }{ \b\lang1033 eleni}{\lang1033 ).}{\b\lang1033 \par }{\lang1033 \par }{\fs20\lang1033 NOTE: It seems that occasionally, *}{\b\fs20\lang1033 ln}{\fs20\lang1033 may yield Quenya }{\b\fs20\lang1033 ll}{\fs20\lang1033 instead of (or as well as) }{\b\fs20\lang1033 ld}{\fs20\lang1033 : cf. \'a4}{\b\fs20\lang1033 ndulna}{ \fs20\lang1033 > Q }{\b\fs20\lang1033 nulla }{\fs20\lang1033 or }{\b\fs20\lang1033 nulda}{\fs20\lang1033 "secret" (LR:355 s.v. }{\i\fs20\lang1033 DUL}{\fs20\lang1033 ). Such attested past tense-forms as }{\b\fs20\lang1033 will\'eb}{\fs20\lang1033 "f lew" and }{\b\fs20\lang1033 ull\'eb}{\fs20\lang1033 "poured" may then represent older *}{\b\fs20\lang1033 wilne}{\fs20\lang1033 , *}{\b\fs20\lang1033 ulne}{\fs20\lang1033 ; cf. -}{\b\fs20\lang1033 n\'eb}{\fs20\lang1033 as a common past tense ending in Quenya. But it may also be that these past tense forms represent older nasal-infixed stems, so that the double }{\b\fs20\lang1033 ll}{\fs20\lang1033 seen in Quenya actually comes from }{\b\fs20\lang1033 nl}{ \fs20\lang1033 rather than }{\b\fs20\lang1033 ln}{\fs20\lang1033 . \par }{\lang1033 \par PN and BN both > MN: \par (\'a4}{\i\lang1033 Utupnu}{\lang1033 >) *}{\b\lang1033 Utupno}{\lang1033 > Q }{\b\lang1033 Utumno}{\lang1033 (Melkor's first stronghold) \par (\'a4}{\i\lang1033 stabn\'f4}{\lang1033 >) *}{\b\lang1033 \'feabno}{\lang1033 > Q *}{\b\lang1033 \'feamno}{\lang1033 "carpenter" (in later Q }{\i\lang1033 samno}{\lang1033 ) \par (\'a4}{\i\lang1033 tubn\'e2}{\lang1033 >) *}{\b\lang1033 tubna}{\lang1033 > Q }{\b\lang1033 tumna}{\lang1033 "low" \par \par (The reconstruction \'a4}{\b\lang1033 Utupnu}{\lang1033 is found in MR:69; Etym gives \'a4}{\b\lang1033 Utubnu}{\lang1033 instead [LR:394 s.v. }{\i\lang1033 TUB}{\lang1033 ], but the Quenya form would be }{\b\lang1033 Utumno}{\lang1033 in either case.) }{\i\lang1033 \par }{\lang1033 \par }{\fs20\lang1033 NOTE: In one exceptional case, }{\b\fs20\lang1033 bn}{\fs20\lang1033 becomes }{\b\fs20\lang1033 mb}{\fs20\lang1033 instead: The Quenya adjective }{\b\fs20\lang1033 lemba}{\fs20\lang1033 "left behind" is explicitly referred to \'a4}{ \b\fs20\lang1033 lebn\'e2}{\fs20\lang1033 (LR:368 s.v. }{\i\fs20\lang1033 LEB}{\fs20\lang1033 ). This must be due to an early, irregular metathesis \'a4}{\b\fs20\lang1033 lebn\'e2}{\fs20\lang1033 > *}{\b\fs20\lang1033 lenb\'e2}{\fs20\lang1033 > *}{ \b\fs20\lang1033 lemb\'e2}{\fs20\lang1033 , perhaps helped by the alternative root }{\i\fs20\lang1033 LEM}{\fs20\lang1033 listed in the same entry in Etym (}{\i\fs20\lang1033 LEM}{\fs20\lang1033 could have produced a form *}{\b\fs20\lang1033 lemb\'e2}{ \fs20\lang1033 already in the primitive language). In the case of }{\b\fs20\lang1033 samb\'eb}{\fs20\lang1033 "room" (LR:387 s.v. }{\i\fs20\lang1033 STAB}{\fs20\lang1033 ), Tolkien listed both \'a4}{\b\fs20\lang1033 stabn\'ea}{\fs20\lang1033 and \'a4}{ \b\fs20\lang1033 stamb\'ea}{\fs20\lang1033 as primitive forms, but these are evidently meant as variants co-existing already in the primitive language: While Quenya }{\b\fs20\lang1033 samb\'eb}{\fs20\lang1033 clearly comes from \'a4}{\b\fs20\lang1033 stamb\'ea}{\fs20\lang1033 , the "Noldorin"/Sindarin word }{\b\fs20\lang1033 thafn}{\fs20\lang1033 is to be derived from \'a4}{\b\fs20\lang1033 stabn\'ea}{\fs20\lang1033 ; the true "N"/S cognate of Q }{\b\fs20\lang1033 samb\'eb}{\fs20\lang1033 is actually }{\b\fs20\lang1033 tham}{\fs20\lang1033 , listed in the same entry. Thus the meaning cannot be that \'a4}{\b\fs20\lang1033 stabn\'ea}{\fs20\lang1033 became or was replaced by \'a4}{\b\fs20\lang1033 stamb\'ea}{\fs20\lang1033 , so we have no second attestation of the development }{\b\fs20\lang1033 bn}{\fs20\lang1033 > Q }{\b\fs20\lang1033 mb}{\fs20\lang1033 after all. On the contrary, the entry }{\i\fs20\lang1033 STAB}{\fs20\lang1033 in Etym confirms the regular development }{\b\fs20\lang1033 bn}{\fs20\lang1033 > Q }{\b\fs20\lang1033 mn}{\fs20\lang1033 , since the Q word }{\b\fs20\lang1033 samna}{\fs20\lang1033 "wooden post" there listed is certainly meant to come from *}{ \b\fs20\lang1033 stabn\'e2}{\fs20\lang1033 . Cf. also the adjective }{\b\fs20\lang1033 lumna}{\fs20\lang1033 "burdensome", derived from a root }{\i\fs20\lang1033 DUB}{\fs20\lang1033 ; the primitive form is clearly meant to be *}{\b\fs20\lang1033 dubn\'e2 }{\fs20\lang1033 . \par }{\lang1033 \par The nasal }{\b\lang1033 m}{\lang1033 becomes }{\b\lang1033 w}{\lang1033 following }{\b\lang1033 n}{\lang1033 or }{\b\lang1033 \'fe}{\lang1033 : \par \par (\'a4}{\i\lang1033 tinm\'ea}{\lang1033 >) *}{\b\lang1033 tinme}{\lang1033 > Q }{\b\lang1033 tinw\'eb}{\lang1033 "sparkle, glint; star" \par (\'a4}{\i\lang1033 k}{\i\lang1033\super h}{\i\lang1033 it}{\i\lang1033\super h}{\i\lang1033 me}{\lang1033 [read *}{\i\lang1033 k}{\i\lang1033\super h}{\i\lang1033 it}{\i\lang1033\super h}{\i\lang1033 m\'ea}{\lang1033 ] > *}{\i\lang1033 chi\'fem\'ea}{ \lang1033 >) *}{\b\lang1033 chi\'feme}{\lang1033 > Q *}{\b\lang1033 chi\'few\'eb}{\lang1033 "fog" (in later Q }{\b\lang1033 hisw\'eb}{\lang1033 ) \par \par Certain other changes involved }{\i\lang1033 metathesis}{\lang1033 , two sounds switching place: \par \par KL > LK (LC): (*}{\i\lang1033 aklar\'ea}{\lang1033 >) *}{\b\lang1033 aklare}{\lang1033 > Q }{\b\lang1033 alcar\'eb}{\lang1033 "glory" (also short }{\i\lang1033 alcar}{\lang1033 ) \par KR > RK (RC): (*}{\i\lang1033 lakr\'e2}{\lang1033 >) \'a4}{\b\lang1033 lakra}{\lang1033 }{\b\lang1033 > }{\lang1033 Q}{\b\lang1033 larca }{\lang1033 "swift" \par TR > RT: (\'a4}{\i\lang1033 netere}{\lang1033 >) *}{\b\lang1033 netre}{\lang1033 > Q }{\b\lang1033 nert\'eb}{\lang1033 "nine" (VT41:26; in this source Tolkien explicitly refers to changes TL, TR "etc." > LT, RT) \par }\pard \qj\nowidctlpar\widctlpar\adjustright {\b\lang1033 \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 Again, we have an example (\'a4}{\b\lang1033 lakra}{\lang1033 instead of older *}{\b\lang1033 lakr\'e2}{\lang1033 ) that would seem to place these changes in the period post-dating the shortening of the final vowels. (The example \'a4}{\b\lang1033 akr\'e2}{\lang1033 "narrow" > Q }{\b\lang1033 arca}{\lang1033 [LR:348] would then skip over the intermediate form *}{\b\lang1033 akra}{\lang1033 .) \par \par Tolkien is seen to have changed his mind repeatedly about the behaviour of the medial cluster }{\b\lang1033 sk}{\lang1033 : Did it survive unchanged to become Quenya }{\b\lang1033 sc}{\lang1033 , or did it undergo metathesis to produce Quenya }{ \b\lang1033 ks}{\lang1033 = }{\b\lang1033 x}{\lang1033 ? In the entry }{\i\lang1033 EZGE}{\lang1033 in Etym (suggesting a primitive word *}{\b\lang1033 ezg\'ea}{\lang1033 ) a Quenya word }{\b\lang1033 esc\'eb}{\lang1033 is listed; this }{\b\lang1033 sc }{\lang1033 represents the }{\b\lang1033 sk}{\lang1033 that arose after the regular devoicing of older }{\b\lang1033 zg}{\lang1033 . However, the entry }{\i\lang1033 EZGE}{\lang1033 was struck out. In other entries, Tolkien let }{\b\lang1033 sk}{ \lang1033 (whether original or devoiced from }{\b\lang1033 zg}{\lang1033 ) undergo metathesis to become Quenya }{\b\lang1033 x}{\lang1033 (actually spelt }{\b\lang1033 ks}{\lang1033 in Etym). Thus the word \'a4}{\b\lang1033 mazg\'e2}{\lang1033 (entry }{\i\lang1033 MASAG}{\lang1033 ) yields Quenya }{\b\lang1033 maxa}{\lang1033 (}{\b\lang1033 maksa}{\lang1033 ), and given the primitive form, this must be metathesized from an intermediate form *}{\b\lang1033 maska}{\lang1033 . The entry }{\i\lang1033 MISK}{\lang1033 indicates that original }{\b\lang1033 sk}{\lang1033 (not devoiced from }{\b\lang1033 zg}{\lang1033 ) behaves in the same way; here we find a Q word }{\b\lang1033 mixa}{\lang1033 (}{\b\lang1033 miksa}{\lang1033 ) "wet", which given the entry-head must descend from *}{\b\lang1033 misk\'e2}{\lang1033 at the oldest stage. \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 However, Tolkien seems to have changed his mind back again later: In a post-LotR source we find a word like }{\b\lang1033 rusca}{\lang1033 "red-brown" (VT41:10), which must be connected to the entry }{\i\lang1033 RUSK\'c2}{\lang1033 in Etym (in the later source, Tolkien actually struck out }{\b\lang1033 rusca}{\lang1033 , but retained related words like }{\b\lang1033 rusco}{\lang1033 "fox"). So now there is }{\i\lang1033 no}{\lang1033 metathesis }{\b\lang1033 sk}{\lang1033 > }{\b\lang1033 ks}{\lang1033 , just as suggested by the rejected entry }{\i\lang1033 EZGE}{\lang1033 in Etym. If we are to go with Tolkien's final (?) insights on this matter, the Quenya words }{\b\lang1033 mixa}{\lang1033 "wet", }{\b\lang1033 maxa}{\lang1033 "pliant, soft" and }{\b\lang1033 max\'eb}{\lang1033 "dough" (< *}{\b\lang1033 misk\'e2}{ \lang1033 , \'a4}{\b\lang1033 mazg\'e2}{\lang1033 , \'a4}{\b\lang1033 mazg\'ea}{\lang1033 ) would have to be altered to *}{\b\lang1033 misca}{\lang1033 , *}{\b\lang1033 masca}{\lang1033 and *}{\b\lang1033 masc\'eb}{\lang1033 . Perhaps we may simply assume that there were some minor dialectal differences and divergent developments within Quenya, and accept all words as they stand.}{\i\lang1033 \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par The group }{\b\lang1033 rg}{\lang1033 becomes }{\b\lang1033 ry}{\lang1033 : (\'a4}{\i\lang1033 targ\'e2}{\lang1033 >) *}{\b\lang1033 targa}{\lang1033 > }{\b\lang1033 tarya}{\lang1033 "stiff". \par \par }{\fs20\lang1033 NOTE: In one case, Tolkien in the }{\i\fs20\lang1033 Etymologies}{\fs20\lang1033 likewise lets }{\b\fs20\lang1033 lg}{\fs20\lang1033 become }{\b\fs20\lang1033 ly}{\fs20\lang1033 , Telerin }{\b\fs20\lang1033 felga}{\fs20\lang1033 "cave" having the Quenya cognate }{\b\fs20\lang1033 felya}{\fs20\lang1033 (}{\i\fs20\lang1033 P}{\i\fs20\lang1033\super H}{\i\fs20\lang1033 \'c9LEG}{\fs20\lang1033 ). The primitive form is clearly meant to be *}{\b\fs20\lang1033 p}{ \b\fs20\lang1033\super h}{\b\fs20\lang1033 elg\'e2}{\fs20\lang1033 . However, this does not agree with another example, quoted above: \'a4}{\b\fs20\lang1033 ulgund\'f4}{\fs20\lang1033 "monster" does not yield Quenya **}{\b\fs20\lang1033 ulyundo}{ \fs20\lang1033 , but }{\b\fs20\lang1033 ulundo}{\fs20\lang1033 (evidently via *}{\b\fs20\lang1033 ul3und\'f4}{\fs20\lang1033 ). We can probably afford to ignore the entire entry }{\i\fs20\lang1033 P}{\i\fs20\lang1033\super H}{\i\fs20\lang1033 \'c9LEG}{ \fs20\lang1033 in Etym, for it reflects an idea Tolkien later rejected: He is here explaining the origin of the first element of the name }{\i\fs20\lang1033 Felagund}{\fs20\lang1033 , which in the }{\i\fs20\lang1033 Etymologies}{\fs20\lang1033 is "Noldorin"/Sindarin for Cave-Prince (cf. the entry }{\i\fs20\lang1033 KUND\'db}{\fs20\lang1033 ). However, Tolkien later decided that Felagund was not an Elvish name at all. In a post-LotR source it was re-explained as an adaptation of a Dwarvish (Khuzdul) term for Cave-Hewer: }{\i\fs20\lang1033 Felakgundu}{\fs20\lang1033 or }{\i\fs20\lang1033 Felaggundu}{\fs20\lang1033 \endash and it is now actually the second element }{\i\fs20\lang1033 gundu}{\fs20\lang1033 rather than the initial element }{\i\fs20\lang1033 felag}{\fs20\lang1033 - that means "cave" or "underground dwelling" (PM:352). So from a certain point of view, Quenya }{\b\fs20\lang1033 felya}{\fs20\lang1033 "cave" and all its cognates must be rejected a s Elvish words. It would seem, then, that Tolkien considerately eliminated a troublesome example for us. I would not let }{\b\fs20\lang1033 lg}{\fs20\lang1033 produce Quenya }{\b\fs20\lang1033 ly}{\fs20\lang1033 , but let }{\b\fs20\lang1033 lg}{ \fs20\lang1033 behave as in \'a4}{\b\fs20\lang1033 ulgund\'f4}{\fs20\lang1033 > (*}{\b\fs20\lang1033 ul3ond\'f4}{\fs20\lang1033 >) }{\b\fs20\lang1033 ulundo}{\fs20\lang1033 instead.}{\lang1033 \par \par At some point after the shortening of the long final vowels (as is evident from the example \'a4}{\b\lang1033 ektele}{\lang1033 ), the group }{\b\lang1033 kt}{\lang1033 becomes }{\b\lang1033 cht}{\lang1033 , but we are now so close to classical Quenya that we will represent }{\b\lang1033 cht}{\lang1033 as }{\b\lang1033 ht}{\lang1033 in accordance with normal Quenya orthography: \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 et-kel\'ea}{\lang1033 >) \'a4}{\b\lang1033 ektele > }{\lang1033 Q }{\b\lang1033 ehtel\'eb}{\lang1033 "spring, issue of water" \par (\'a4}{\i\lang1033 hek-t\'e2}{\lang1033 >) \'a4}{\b\lang1033 hek-ta}{\lang1033 > Q }{\b\lang1033 hehta}{\lang1033 - "forsake" (WJ:361, 363) \par (*}{\i\lang1033 k}{\i\lang1033\super h}{\i\lang1033 agd\'e2}{\lang1033 > *}{\i\lang1033 k}{\i\lang1033\super h}{\i\lang1033 akt\'e2}{\lang1033 >) *}{\b\lang1033 chakta}{\lang1033 > *}{\b\lang1033 chahta}{\lang1033 "mound" (in later Q }{\i\lang1033 hahta}{\lang1033 ) \par (\'a4}{\i\lang1033 jakta}{\lang1033 >) *}{\b\lang1033 jakt}{\lang1033 > *}{\b\lang1033 jaht}{\lang1033 "neck" (Q }{\i\lang1033 yat}{\lang1033 with stem }{\i\lang1033 yaht}{\lang1033 -) \par (\'a4}{\i\lang1033 nukt\'e2}{\lang1033 >) *}{\b\lang1033 nukta}{\lang1033 - > Q}{\b\lang1033 nuhta}{\lang1033 - "to stunt" (WJ:413) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par In a few cases, this change caused confusion. The verb (\'a4}{\i\lang1033 makt\'e2}{\lang1033 - >) *}{\b\lang1033 makta}{\lang1033 - "wield a weapon" had until now been distinct from (PQ *}{\i\lang1033 ma3t\'e2}{\lang1033 > CE *}{\i\lang1033 macht\'e2}{ \lang1033 - >) *}{\b\lang1033 machta}{\lang1033 - "to handle". But now *}{\b\lang1033 makta}{\lang1033 - came to be pronounced just like *}{\b\lang1033 machta}{\lang1033 -, and the verbs fell together in the form we would normally spell }{\b\lang1033 mahta}{\lang1033 - in Quenya. The }{\i\lang1033 Etymologies}{\lang1033 records how the verbs became "blended" (LR:371 s.v. }{\i\lang1033 MAK}{\lang1033 ). \par \par The verb }{\b\lang1033 rihta}{\lang1033 "to jerk" may descend from either *}{\b\lang1033 rikt\'e2}{\lang1033 or }{\b\lang1033 rik}{\b\lang1033\super h}{\b\lang1033 t\'e2}{\lang1033 ; Tolkien somewhat indecisively listed the root as "}{\i\lang1033 RIK(H) }{\lang1033 " (LR:383) \endash either }{\i\lang1033 RIK}{\lang1033 or }{\i\lang1033 RIK}{\i\lang1033\super H}{\lang1033 , or perhaps rather both. It may be that }{\b\lang1033 k}{\b\lang1033\super h}{\lang1033 before }{\b\lang1033 t}{\lang1033 was reduced to }{\b\lang1033 k}{\lang1033 already in Common Eldarin; if so Q }{\b\lang1033 rihta}{\lang1033 would come from *}{\b\lang1033 rikt\'e2}{\lang1033 anyhow. But if *}{\b\lang1033 rik}{\b\lang1033\super h}{\b\lang1033 t\'e2}{\lang1033 survived in this form, it became Q }{\b\lang1033 rihta}{\lang1033 (= }{\b\lang1033 richta}{\lang1033 ) already in the general change }{\b\lang1033 k}{\b\lang1033\super h}{\lang1033 > }{\b\lang1033 ch}{\lang1033 and did not have to wait for the change }{\b\lang1033 kt}{\lang1033 > }{\b\lang1033 ht}{\lang1033 (= }{\b\lang1033 cht}{\lang1033 ).}{\i\lang1033 \par }{\lang1033 \par Eventually }{\b\lang1033 ch}{\lang1033 (of whatever origin) before }{\b\lang1033 t}{\lang1033 would turn into }{\i\lang1033 ich-Laut}{\lang1033 following the vowels }{\b\lang1033 e}{\lang1033 and }{\b\lang1033 i}{\lang1033 ; cf. the later Quenya pronunciation of words like }{\b\lang1033 nehta}{\lang1033 "spear-head" or }{\b\lang1033 tihta}{\lang1033 - "to blink". We cannot know at what stage this change happened (since it is not reflected in normal Quenya orthography anyway \endash it may well be late). In any case, }{\b\lang1033 ch}{\lang1033 remained unchanged following the vowels }{\b\lang1033 a }{\lang1033 , }{\b\lang1033 o}{\lang1033 , and }{\b\lang1033 u}{\lang1033 ; cf. the later Quenya pronunciation of words like }{\b\lang1033 mahta}{\lang1033 - "to handle/wield", }{\b\lang1033 ohta}{\lang1033 "war", }{\b\lang1033 luhta}{\lang1033 - "to enchant". \par \par Various combinations of }{\i\lang1033 voiced stops}{\lang1033 + }{\b\lang1033 m}{\lang1033 also underwent changes. The }{\i\lang1033 Etymologies}{\lang1033 quotes \'a4}{\b\lang1033 nakma}{\lang1033 (rather than older *}{\b\lang1033 nakm\'e2}{ \lang1033 ) as the origin of the Quenya word }{\i\lang1033 nangwa}{\lang1033 "jaw"; if we can trust this piece of information, this change also occurred after the shortening of the final long vowels. (In the case of a word like }{\i\lang1033 sungwa}{ \lang1033 "drinking-vessel", said to be derived from \'a4}{\b\lang1033 sukm\'e2}{\lang1033 , we may assume that Tolkien simply did not mention the intermediate stage *}{\b\lang1033 sukma}{\lang1033 with the final vowel shortened.) \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 KM and GM both > NGW (technically \'d1GW): \par (*}{\i\lang1033 nakm\'e2}{\lang1033 >) \'a4}{\b\lang1033 nakma}{\lang1033 > Q }{\b\lang1033 nangwa }{\lang1033 "jaw" \par (\'a4}{\i\lang1033 sagm\'e2}{\lang1033 >) *}{\b\lang1033 sagma}{\lang1033 > Q }{\b\lang1033 sangwa}{\lang1033 "poison"}{\i\lang1033 \par }{\lang1033 \tab \tab \tab \tab \par TM > NW: \par (*}{\i\lang1033 mbaratm\'ea}{\lang1033 >) *}{\b\lang1033 maratme}{\lang1033 > Q }{\b\lang1033 maranw\'eb}{\lang1033 "destiny" \par (\'a4}{\i\lang1033 jatm\'e2}{\lang1033 > *}{\i\lang1033 jatm\'ea}{\lang1033 >) *}{\b\lang1033 jatme}{\lang1033 > Q }{\b\lang1033 yanw\'eb }{\lang1033 "bridge" }{\fs20\lang1033 (the difference }{\b\fs20\lang1033 j}{\fs20\lang1033 /}{\b\fs20\lang1033 y}{ \fs20\lang1033 is of course orthographic only) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par The group *}{\b\lang1033 dm}{\lang1033 would probably also become }{\b\lang1033 nw}{\lang1033 , but we have no examples. \par \par Probably already in the pre-record period, initial }{\b\lang1033 w}{\lang1033 is lost before long }{\b\lang1033 \'f4}{\lang1033 (WJ:367). Hence the prefix \'a4}{\b\lang1033 w\'f4}{\lang1033 -, meaning "together" referring to the union of two things, was reduced to }{\b\lang1033 \'f3}{\lang1033 - in Quenya (when unstressed shortened to }{\b\lang1033 o}{\lang1033 -), and we have for instance *}{\b\lang1033 w\'f4-mentie}{\lang1033 > Q }{ \b\lang1033 omenti\'eb}{\lang1033 "meeting (of two)", *}{\b\lang1033 w\'f4-n\'f4na}{\lang1033 > Q adj. }{\b\lang1033 on\'f3na}{\lang1033 "twin-born" (WJ:367). \par \par Post-vocalic }{\b\lang1033 s}{\lang1033 followed by a voiced sound (any vowel or a voiced consonant) becomes }{\b\lang1033 z}{\lang1033 . (This }{\b\lang1033 z}{\lang1033 will eventually become }{\b\lang1033 r}{\lang1033 in Exilic Quenya.) \par \par \tab (\'a4}{\i\lang1033 t}{\i\lang1033\super h}{\i\lang1033 aus\'e2}{\lang1033 > *}{\i\lang1033 \'feaus\'e2}{\lang1033 >) *}{\b\lang1033 \'feausa}{\lang1033 > *}{\b\lang1033 \'feauza}{\lang1033 "foul" (Q *}{\i\lang1033 \'feauza}{\lang1033 > }{ \i\lang1033 saura}{\lang1033 ) \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 bes\'fb}{\lang1033 > *}{\i\lang1033 ves\'fb}{\lang1033 >) *}{\b\lang1033 vesu}{\lang1033 > *}{\b\lang1033 vezu}{\lang1033 "married pair" (Q *}{\i\lang1033 vezu }{\lang1033 > }{\i\lang1033 veru}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \tab (\'a4}{\i\lang1033 besn\'f4}{\lang1033 > *}{\i\lang1033 vesn\'f4}{\lang1033 >) *}{\b\lang1033 vesno}{\lang1033 > *}{\b\lang1033 vezno}{\lang1033 "husband" (Q *}{\i\lang1033 vezno}{\lang1033 > }{\i\lang1033 verno}{\lang1033 ) \par \tab (*}{\i\lang1033 pisj\'e2}{\lang1033 >) *}{\b\lang1033 pisya}{\lang1033 > *}{\b\lang1033 pizya}{\lang1033 "syrup" (Q *}{\i\lang1033 pizya}{\lang1033 > }{\i\lang1033 pirya}{\lang1033 ) \par \par The voicing }{\b\lang1033 s}{\lang1033 > }{\b\lang1033 r}{\lang1033 does }{\i\lang1033 not}{\lang1033 take place if }{\b\lang1033 s}{\lang1033 is followed by an unvoiced consonant (including another }{\b\lang1033 s}{\lang1033 , in the case of double } {\b\lang1033 ss}{\lang1033 ): \par \par \tab (\'a4}{\i\lang1033 usuk-w\'ea}{\lang1033 > *}{\i\lang1033 usk}{\i\lang1033\super w}{\i\lang1033 \'ea}{\lang1033 >) }{\b\lang1033 usk}{\b\lang1033\super w}{\b\lang1033 e}{\lang1033 = Q }{\b\lang1033 usqu\'eb}{\lang1033 "reek" (not **}{\b\lang1033 uzqu\'eb}{\lang1033 > **}{\b\lang1033 urqu\'eb}{\lang1033 ) \par \tab (\'a4}{\i\lang1033 bess\'ea}{\lang1033 > *}{\i\lang1033 vess\'ea}{\lang1033 >) }{\b\lang1033 vesse}{\lang1033 = Q }{\b\lang1033 vess\'eb}{\lang1033 "wife" (not **}{\b\lang1033 vezs\'eb}{\lang1033 > **}{\b\lang1033 vers\'eb}{\lang1033 ) \par \par The evidence is somewhat divergent regarding the behaviour of a (single) post-vocalic }{\b\lang1033 s}{\lang1033 occurring finally. In the }{\i\lang1033 Etymologies}{\lang1033 , Tolkien let }{\b\lang1033 s}{\lang1033 become }{\b\lang1033 z}{\lang1033 (> later }{\b\lang1033 r}{\lang1033 ) in this position as well, as when he derived the Quenya noun }{\b\lang1033 solor}{\lang1033 "surf" from older \'a4}{\b\lang1033 solos}{\lang1033 (implying the existence of an intermediate form *}{\b\lang1033 soloz} {\lang1033 ). The first word listed in the entry }{\i\lang1033 \'d3LOS}{\lang1033 is likewise }{\b\lang1033 olor}{\lang1033 ("dream"); the imagined development would be *}{\b\lang1033 olos}{\lang1033 > *}{\b\lang1033 oloz}{\lang1033 > }{\b\lang1033 olor}{\lang1033 . However, this very word reappears in UT:396, but now in the form }{\b\lang1033 olos}{\lang1033 \endash though its }{\i\lang1033 plural}{\lang1033 is given as }{\b\lang1033 olozi}{\lang1033 /}{\b\lang1033 olori}{\lang1033 . Hence Tolkien has altered the rules for what precise environment the voicing }{\b\lang1033 s}{\lang1033 > }{\b\lang1033 z}{\lang1033 occurs in. In Etym, the rule is that a post-vocalic }{\b\lang1033 s}{\lang1033 becomes voiced unless it is followed by an unvoiced sound. The later rule seems to be that }{\b\lang1033 s}{\lang1033 is voiced if it occurs }{\i\lang1033 between}{\lang1033 two voiced sounds (either two vowels, or a vowel and a voiced consonant). Thus }{\b\lang1033 olos}{\lang1033 is still }{\b\lang1033 olozi}{\lang1033 in the plural, where the plural ending makes }{\b\lang1033 s}{\lang1033 intervocalic so that the voicing can occur. \par \par }{\fs20\lang1033 NOTE: If we were to impose full "consistency" on the material, we should have to alter some of the words in the }{\i\fs20\lang1033 Etymologies}{\fs20\lang1033 . }{\b\fs20\lang1033 Solor}{\fs20\lang1033 "surf" would become *}{ \b\fs20\lang1033 solos}{\fs20\lang1033 (to go with }{\b\fs20\lang1033 olos}{\fs20\lang1033 rather than }{\b\fs20\lang1033 olor}{\fs20\lang1033 ), and }{\b\fs20\lang1033 seler}{\fs20\lang1033 "sister" (derived from older *}{\b\fs20\lang1033 \'feelez}{ \fs20\lang1033 , cf. the base }{\i\fs20\lang1033 T}{\i\fs20\lang1033\super H}{\i\fs20\lang1033 ELES}{\fs20\lang1033 ) would have to become *}{\b\fs20\lang1033 seles}{\fs20\lang1033 instead. }{\b\fs20\lang1033 Teler}{\fs20\lang1033 "Telerin Elf", in Etym derived from }{\i\fs20\lang1033 TELES}{\fs20\lang1033 , would have to become *}{\b\fs20\lang1033 Teles}{\fs20\lang1033 in the singular, though the plural could still be }{\b\fs20\lang1033 Teleri}{\fs20\lang1033 for older *}{\b\fs20\lang1033 Telezi}{\fs20\lang1033 (cf. }{\b\fs20\lang1033 olos}{\fs20\lang1033 having the pl. }{\b\fs20\lang1033 olozi}{\fs20\lang1033 > }{\b\fs20\lang1033 olori}{\fs20\lang1033 in UT:396). However, Tolkien in a post-LotR source implied that }{\b\fs20\lang1033 Teler}{\fs20\lang1033 comes from *}{\b\fs20\lang1033 Telero}{\fs20\lang1033 instead (WJ:371), so this word can stay as it is (the }{\b\fs20\lang1033 r}{\fs20\lang1033 now being original, not altered from }{\b\fs20\lang1033 z}{\fs20\lang1033 < }{ \b\fs20\lang1033 s}{\fs20\lang1033 ). Perhaps Tolkien changed the etymology precisely because he did not want the singular to be *}{\b\fs20\lang1033 Teles}{\fs20\lang1033 (as it would have to be according to his Revised Voici ng Rules). We could follow Tolkien's example and keep }{\b\fs20\lang1033 seler}{\fs20\lang1033 "sister" as it is by assuming that it "actually" comes from primitive *}{\i\fs20\lang1033 t}{\i\fs20\lang1033\super h}{\i\fs20\lang1033 elere}{\fs20\lang1033 (with a feminine ending *-}{\i\fs20\lang1033 re}{\fs20\lang1033 corresponding to masc. \'a4-}{\i\fs20\lang1033 ro}{\fs20\lang1033 added to *}{\i\fs20\lang1033 T}{\i\fs20\lang1033\super H}{\i\fs20\lang1033 ELE}{\fs20\lang1033 , the extended form of the root }{\i\fs20\lang1033 T}{\i\fs20\lang1033\super H}{\i\fs20\lang1033 EL}{\fs20\lang1033 listed in LR:390 \endash of which }{\i\fs20\lang1033 T}{\i\fs20\lang1033\super H}{\i\fs20\lang1033 ELES}{\fs20\lang1033 listed in the same place is merely another possible extension). \endash Concerning a curious example from the entry }{\i\fs20\lang1033 KAS}{\fs20\lang1033 , }{\b\fs20\lang1033 c\'e1r}{\fs20\lang1033 "head" with stem }{\b\fs20\lang1033 cas}{ \fs20\lang1033 -, see Appendix. We should probably read *}{\b\fs20\lang1033 c\'e1s}{\fs20\lang1033 with stem *}{\b\fs20\lang1033 car}{\fs20\lang1033 -. \par }{\lang1033 \par The spirant }{\b\lang1033 \'f0}{\lang1033 merges with the newly-developed sound }{\b\lang1033 z}{\lang1033 (and will also end up as }{\b\lang1033 r}{\lang1033 in Exilic Quenya): \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 Awadelo}{\lang1033 > *}{\i\lang1033 Audel}{\lang1033 >) *}{\b\lang1033 au\'f0el}{\lang1033 > }{\b\lang1033 auzel}{\lang1033 (Q }{\i\lang1033 Auzel}{\lang1033 > }{ \i\lang1033 Aurel}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \tab (\'a4}{\i\lang1033 nidw\'f4}{\lang1033 > *}{\i\lang1033 nidw\'e2}{\lang1033 >) *}{\b\lang1033 ni\'f0wa}{\lang1033 > *}{\b\lang1033 nizwa}{\lang1033 "cushion" (Q }{\i\lang1033 nizwa}{\lang1033 > }{\i\lang1033 nirwa}{\lang1033 ) \par \par When writing the }{\i\lang1033 Etymologies}{\lang1033 , Tolkien imagined that non-initial }{\b\lang1033 t}{\lang1033 became }{\b\lang1033 s}{\lang1033 before the vowel }{\b\lang1033 i}{\lang1033 (at what point in the imagined timeline this change was supposed to have occurred is difficult to say; probably relatively late, since there is no trace of such a shift in "Noldorin"/Sindarin). Therefore, the adjective (\'a4}{\i\lang1033 ma3iti}{ \lang1033 >) }{\b\lang1033 mait\'eb}{\lang1033 "handy, skilled" had the plural form }{\b\lang1033 maisi}{\lang1033 , ultimately < *}{\i\lang1033 ma3it\'ee}{\lang1033 . (From this example it is clear that this shift must have occurred after final short - }{\b\lang1033 i}{\lang1033 became -}{\b\lang1033 e}{\lang1033 in late Common Eldarin, or }{\b\lang1033 t}{\lang1033 would have become }{\b\lang1033 s}{\lang1033 in the singular form as well.) In the case of a cluster }{\b\lang1033 tt}{\lang1033 , only the second }{\b\lang1033 t}{\lang1033 changes to }{\b\lang1033 s}{\lang1033 before }{\b\lang1033 i}{\lang1033 , as when }{\b\lang1033 quelet}{\lang1033 "corpse" (stem }{\b\lang1033 quelett}{\lang1033 - < \'a4}{\i\lang1033 k}{\i\lang1033\super w} {\i\lang1033 elett}{\lang1033 -) has the plural }{\b\lang1033 queletsi}{\lang1033 . \endash However, it may be that Tolkien dropped this sound-shift after completing the }{\i\lang1033 Etymologies}{\lang1033 . In a much later (post-LotR) source, Tolkien wrote }{\b\lang1033 hlon\'edti tengwi}{\lang1033 , not ?}{\b\lang1033 hlon\'edsi tengwi}{\lang1033 , for "phonetic signs" (WJ:395). So perhaps the plural form of }{\b\lang1033 mait\'eb}{\lang1033 , }{ \b\lang1033 quelett}{\lang1033 - could simply be *}{\b\lang1033 maiti}{\lang1033 , *}{\b\lang1033 queletti}{\lang1033 as well. \par \par Sporadic losses of final vowels still occur; some adjectives in -}{\b\lang1033 ina}{\lang1033 are shortened: \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \par (*}{\i\lang1033 melin\'e2}{\lang1033 >) *}{\b\lang1033 melina}{\lang1033 > }{\b\lang1033 melin}{\lang1033 "dear" \par (*}{\i\lang1033 sk}{\i\lang1033\super w}{\i\lang1033 arin\'e2}{\lang1033 >) *}{\b\lang1033 hwarina}{\lang1033 > }{\b\lang1033 hwarin}{\lang1033 "crooked" \par (*}{\i\lang1033 p}{\i\lang1033\super h}{\i\lang1033 irin\'e2}{\lang1033 >) *}{\b\lang1033 firina}{\lang1033 > }{\b\lang1033 firin}{\lang1033 "dead" \par (*}{\i\lang1033 k}{\i\lang1033\super w}{\i\lang1033 alin\'e2}{\lang1033 >) *}{\b\lang1033 qualina}{\lang1033 > }{\b\lang1033 qualin}{\lang1033 "dead" }{\fs20\lang1033 (by unnatural cause)}{\lang1033 \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par The Quenya descendant of *}{\b\lang1033 latin\'e2}{\lang1033 Tolkien listed as }{\b\lang1033 latin}{\lang1033 (}{\b\lang1033 a}{\lang1033 ), apparently meaning that }{\b\lang1033 latina}{\lang1033 and shortened }{\b\lang1033 latin}{\lang1033 coexisted. The unattested longer forms of the adjectives above may also still be "valid". Most adjectives in -}{\b\lang1033 ina}{\lang1033 seem to persist in their full forms (}{\b\lang1033 calina}{\lang1033 "light", }{\b\lang1033 culina}{\lang1033 "flame-coloured", }{\b\lang1033 enwina}{\lang1033 "old", }{\b\lang1033 malina}{\lang1033 "yellow", etc. etc.) \par \par Evidently before the beginning of the "historical" period, the diphthongs }{\b\lang1033 ae}{\lang1033 , }{\b\lang1033 ao}{\lang1033 and }{\b\lang1033 ei}{\lang1033 turn into long vowels }{\b\lang1033 \'e9}{\lang1033 , }{\b\lang1033 \'f3}{\lang1033 , and }{\b\lang1033 \'ed}{\lang1033 , respectively. (}{\b\lang1033 Ae}{\lang1033 and }{\b\lang1033 ao}{\lang1033 did not change "directly"; Tolkien implied that between }{\b\lang1033 ae}{\lang1033 and }{\b\lang1033 \'e9}{\lang1033 there was an intermediate stage which he represented as }{\b\lang1033 \'e6}{\lang1033 , and before }{\b\lang1033 ao}{\lang1033 turned into }{\b\lang1033 \'f3}{\lang1033 this diphthong first became a special A-like version of the vowel }{\b\lang1033 \'f4}{\lang1033 . See VT39:10. These intermediate stages may have been short-lived.) \par \par Where one of the older diphthongs }{\b\lang1033 ao}{\lang1033 or }{\b\lang1033 ei}{\lang1033 (no examples for }{\b\lang1033 ae}{\lang1033 ) had occurred at the end of a word, the language now regained some long final vowels in polysyllabic words. They survived throughout the "historical period" in Valinor and we re directly recorded in the classical "Book Quenya" orthography (these long final vowels became short only after the Noldor went to Middle-earth). \par \par }{\b\lang1033 Ae}{\lang1033 becoming long }{\b\lang1033 \'e9}{\lang1033 : \par \par \tab (*}{\i\lang1033 mael\'e2}{\lang1033 >) *}{\b\lang1033 maela}{\lang1033 > }{\b\lang1033 m\'e9la}{\lang1033 "loving" }{\i\lang1033 (Tolkien mentioned no primitive forms, but this Quenya adjective is implied to be derived from an A-infixed form of the stem }{\lang1033 MEL}{\i\lang1033 -; see VT39:10)}{\lang1033 \par \par }{\b\lang1033 Ao}{\lang1033 becoming long }{\b\lang1033 \'f3}{\lang1033 : \par \par \tab (*}{\i\lang1033 kaol\'f4}{\lang1033 >) *}{\b\lang1033 kaolo}{\lang1033 > }{\b\lang1033 c\'f3lo}{\lang1033 "burden" }{\i\lang1033 (again, Tolkien in VT39:10 implies that an A-infixed variant of the stem }{\lang1033 KOL}{\i\lang1033 underlies this word)}{\b\lang1033 \par }{\lang1033 \par Also notice that in "Book Quenya", nouns in -}{\b\lang1033 a}{\lang1033 are seen to have genitive forms in -}{\b\lang1033 \'f3}{\lang1033 , e.g. }{\b\lang1033 ciry\'f3}{\lang1033 as the genitive of }{\b\lang1033 cirya}{\lang1033 "ship". The genitive form }{\b\lang1033 ciry\'f3}{\lang1033 arose at this point in the linguistic evolution and represents *}{\b\lang1033 kirjao}{\lang1033 , in turn derived from older }{\b\lang1033 kirj\'e2-h\'f4}{\lang1033 or *}{\b\lang1033 kirja-h \'f4}{\lang1033 . The vowel of the suffixed element -}{\b\lang1033 h\'f4}{\lang1033 "from, of" came into direct contact with the preceding vowel }{\b\lang1033 a}{\lang1033 after the "very early" loss of medial }{\b\lang1033 h}{\lang1033 treated above (WJ:368). Thus a final diphthong -}{\b\lang1033 ao}{\lang1033 arose. \par \par }{\b\lang1033 Ei}{\lang1033 becoming long }{\b\lang1033 \'ed}{\lang1033 : \par \par \tab (\'a4}{\i\lang1033 neiti}{\lang1033 >) *}{\b\lang1033 neite}{\lang1033 > Q }{\b\lang1033 n\'edt\'eb}{\lang1033 "moist" \par \~\~\~\~\~\~\~\~\~ (\'a4}{\i\lang1033 mein\'e2}{\lang1033 >) *}{\b\lang1033 meina}{\lang1033 > Q }{\b\lang1033 m\'edna}{\lang1033 adj. "eager to go" (VT39:11) \par \par Final -}{\b\lang1033 ei}{\lang1033 also becomes -}{\b\lang1033 \'ed}{\lang1033 , as in the plural of words ending in -}{\b\lang1033 e}{\lang1033 (the diphthong arising when the plural ending -}{\b\lang1033 i}{\lang1033 is added). The plural of }{ \b\lang1033 lass\'eb}{\lang1033 "leaf", that had long been *}{\b\lang1033 lassei}{\lang1033 , now becomes }{\b\lang1033 lass\'ed}{\lang1033 \endash the form recorded in "Book Quenya" orthography. \par \tab In one exceptional case, Tolkien let }{\b\lang1033 ei}{\lang1033 become }{\b\lang1033 ai}{\lang1033 instead of }{\b\lang1033 \'ed}{\lang1033 : In the }{\i\lang1033 Etymologies}{\lang1033 (entry }{\i\lang1033 WEY}{\lang1033 ), the name of the Vali \'eb }{\b\lang1033 Vair\'eb}{\lang1033 is referred to primitive }{\b\lang1033 weir\'ea}{\lang1033 , and there is a reference to a change "}{\i\lang1033 wei}{\lang1033 > }{\i\lang1033 wai}{\lang1033 " as if the preceding }{\b\lang1033 w}{\lang1033 has something with this unusual development. However, perhaps this example should be ignored. In a later source Tolkien derived Vair\'eb's name from an A-infixed form of the stem }{\i\lang1033 WIR}{\lang1033 (evidently meant to be *}{\i\lang1033 Wair \'ea}{\lang1033 ), in which case the dipthong }{\b\lang1033 ai}{\lang1033 would simply be unchanged since the primitive period. \endash VT39:10. \par \par Virtually all final consonant clusters are simplified (the ending -}{\b\lang1033 nt}{\lang1033 for dual dative is the sole known final cluster in later Quenya); the clusters are however preserved before endings, since they are not then final: \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 jakta}{\lang1033 > *}{\i\lang1033 jakt}{\lang1033 >) *}{\b\lang1033 jaht}{\lang1033 > Q }{\b\lang1033 yat}{\lang1033 "neck" (but with stem }{\b\lang1033 yaht}{\lang1033 -) \par (\'a4}{\i\lang1033 peltakse}{\lang1033 >) *}{\b\lang1033 peltaks}{\lang1033 > Q }{\b\lang1033 peltas}{\lang1033 "pivot" (with stem }{\b\lang1033 peltax}{\lang1033 - = }{\b\lang1033 peltaks}{\lang1033 -) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par Final -}{\b\lang1033 rt}{\lang1033 and -}{\b\lang1033 st}{\lang1033 become simple -}{\b\lang1033 r}{\lang1033 and -}{\b\lang1033 s}{\lang1033 , respectively: \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (*}{\i\lang1033 m\'b7barta}{\lang1033 > *}{\i\lang1033 m\'b7bart}{\lang1033 >) *}{\b\lang1033 umbart}{\lang1033 > Q }{\b\lang1033 umbar}{\lang1033 "fate" (with stem }{\b\lang1033 umbart}{ \lang1033 -) \par (*}{\i\lang1033 Mand-ost\'f4}{\lang1033 > \'a4}{\i\lang1033 Mandosto}{\lang1033 >) *}{\b\lang1033 Mandost}{\lang1033 > Q }{\b\lang1033 Mandos}{\lang1033 (with stem }{\b\lang1033 Mandost}{\lang1033 -, MR:350 cf. VT39:6) \par \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 Final clusters in -}{\b\lang1033 d}{\lang1033 loose this consonant, e.g. -}{\b\lang1033 nd}{\lang1033 becoming -}{\b\lang1033 n}{\lang1033 : \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \tab (\'a4}{\i\lang1033 swanda}{\lang1033 > *}{\i\lang1033 swand}{\lang1033 >) *}{\b\lang1033 hwand}{\lang1033 > Q }{\b\lang1033 hwan}{\lang1033 "sponge" (with stem }{\b\lang1033 hwand}{\lang1033 -) \par \tab (*}{\i\lang1033 t}{\i\lang1033\super h}{\i\lang1033 aurondo}{\lang1033 >) \'a4}{\b\lang1033 \'feaurond}{\lang1033 > Q *}{\b\lang1033 \'feauron}{\lang1033 (Exilic Quenya }{\i\lang1033 Sauron}{\lang1033 , quite possibly with stem }{\i\lang1033 Saurond}{\lang1033 -) \par \par For the latter example, see Letters:380. Another example involves a simpification -}{\b\lang1033 ld}{\lang1033 > -}{\b\lang1033 l}{\lang1033 : In the }{\i\lang1033 Etymologies}{\lang1033 scenario, the name }{\b\lang1033 Taniquetil}{\lang1033 has the stem }{\b\lang1033 Taniquetild}{\lang1033 - because the latter element of this name is reduced from }{\b\lang1033 tild\'eb}{\lang1033 "horn". The final -}{\b\lang1033 e}{\lang1033 was lost early, but the name must have appeared as *}{ \b\lang1033 Taniquetild}{\lang1033 at one stage. \par \tab Final -}{\b\lang1033 ng}{\lang1033 (actually -}{\b\lang1033 \'f1g}{\lang1033 ) loses the -}{\b\lang1033 g}{\lang1033 and }{\b\lang1033 \'f1}{\lang1033 becomes a normal }{\b\lang1033 n}{\lang1033 . The name of the Golden Tree }{\b\lang1033 Laurelin }{\lang1033 was sometimes taken to mean "Hanging Gold" and assigned the stem }{\b\lang1033 Laureling}{\lang1033 -, the latter element being related to }{\b\lang1033 linga}{\lang1033 - "hang, d angle". To the extent this interpretation was correct, the name may indeed have appeared as *}{\b\lang1033 Laureling}{\lang1033 (i.e., *}{\b\lang1033 Laureli\'f1g}{\lang1033 ) at one earlier stage. \par \par Final -}{\b\lang1033 m}{\lang1033 and -}{\b\lang1033 k}{\lang1033 become -}{\b\lang1033 n}{\lang1033 and -}{\b\lang1033 t}{\lang1033 , respectively. Thus the older base }{\i\lang1033 TALAM}{\lang1033 is the source of the Quenya word }{\b\lang1033 tala n}{\lang1033 "floor", with stem }{\b\lang1033 talam}{\lang1033 -. A word for "bird" derived from the stem }{\i\lang1033 P}{\i\lang1033\super H}{\i\lang1033 ILIK}{\lang1033 had by now probably reached the form }{\b\lang1033 filik}{\lang1033 ; in classical Quenya we find }{\b\lang1033 filit}{\lang1033 , with stem }{\b\lang1033 filic}{\lang1033 -. Cf. also }{\b\lang1033 nelet}{\lang1033 "tooth" from an original base }{\i\lang1033 NEL-EK}{\lang1033 (in this case the stem-form has undergone syncope: }{\b\lang1033 nelc}{\lang1033 -). \par \par Final double consonants are simplified, but remain before endings: \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 tollo}{\lang1033 >) *}{\b\lang1033 toll}{\lang1033 > Q }{\b\lang1033 tol}{\lang1033 "island" (stem }{\b\lang1033 toll}{\lang1033 -) \par *}{\b\lang1033 k}{\b\lang1033\super w}{\b\lang1033 elett}{\lang1033 > Q }{\b\lang1033 quelet}{\lang1033 "corpse" (stem *}{\b\lang1033 quelett}{\lang1033 -) \par (*}{\i\lang1033 nisse}{\lang1033 >) *}{\b\lang1033 niss}{\lang1033 > }{\b\lang1033 nis}{\lang1033 "woman" (stem }{\b\lang1033 niss}{\lang1033 -) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par (The simplex }{\b\lang1033 nis}{\lang1033 "woman" also occurs in the form }{\b\lang1033 n\'eds}{\lang1033 ; it is difficult to account for the long vowel, unless this form is influenced by the archaic word }{\b\lang1033 n\'ed}{\lang1033 of similar meaning.) \par \par }{\i\fs32\lang1033 Changes producing}{\fs32\lang1033 VALINOREAN QUENYA \par }{\lang1033 \par Probably before the beginning of the "historical" period, a double }{\b\lang1033 ii}{\lang1033 (evidently two vowels in hiatus) was simplified to }{\b\lang1033 i}{\lang1033 . The old perfect form \'a4}{\b\lang1033 a-w\'e2wii\'ea}{\lang1033 "has passed away" (WJ:366) would have become *}{\b\lang1033 aw\'e2wiie}{\lang1033 with a short final vowel by now; perhaps around this stage, it was further simplified to *}{\b\lang1033 aw\'e2wie}{\lang1033 . (In later Quenya this yielded }{\b\lang1033 av\'e1ni\'eb}{\lang1033 , which is }{\i\lang1033 not}{\lang1033 a regular phonetic development; the }{\b\lang1033 n}{\lang1033 replacing the last }{\b\lang1033 w}{\lang1033 intruded by association with the past tense ending -}{\b\lang1033 n\'eb}{ \lang1033 . See WJ:366.) \par \par At some relatively late stage, }{\b\lang1033 ai}{\lang1033 /}{\b\lang1033 ay}{\lang1033 (but }{\i\lang1033 not}{\lang1033 }{\b\lang1033 \'e2y}{\lang1033 with a long }{\b\lang1033 \'e2}{\lang1033 ) becomes }{\b\lang1033 e}{\lang1033 before an unaccented vowel (or at least before }{\b\lang1033 o}{\lang1033 and }{\b\lang1033 a}{\lang1033 ; examples for other vowels seem to be missing). Probably }{\b\lang1033 ay}{\lang1033 first became }{\b\lang1033 ai}{\lang1033 ; then it was reduced to }{\b\lang1033 e}{\lang1033 (}{\b\lang1033 \'eb}{\lang1033 ) in this position. \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 daio}{\lang1033 [read *}{\i\lang1033 daj\'f4}{\lang1033 ] > *}{\i\lang1033 laj\'f4}{\lang1033 >) *}{\b\lang1033 layo}{\lang1033 /*}{\b\lang1033 laio}{\lang1033 > }{ \b\lang1033 l\'ebo}{\lang1033 "shade" \par }\pard \qj\fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 gaj\'e2r}{\lang1033 > *}{\i\lang1033 3ay\'e2r}{\lang1033 >) *}{\b\lang1033 ay\'e1r/}{\lang1033 *}{\b\lang1033 ai\'e1r}{\lang1033 > *}{\b\lang1033 \'eb\'e1r}{\lang1033 "sea" (cf. WJ:400; in Exilic Quenya }{\i\lang1033 \'ebar}{\lang1033 ) \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (*}{\i\lang1033 p}{\i\lang1033\super h}{\i\lang1033 aj\'e2}{\lang1033 >) }{\b\lang1033 faya}{\lang1033 /*}{\b\lang1033 faia}{\lang1033 > }{\b\lang1033 f\'eba}{\lang1033 "spirit" \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par The latter word is the first element in the name of F\'ebanor, Quenya }{\b\lang1033 F\'eban\'e1ro}{\lang1033 ; the fact that it is quoted in the form }{\b\lang1033 Fayan\'e1ro}{\lang1033 in PM:343 would seem to indicate that }{\b\lang1033 aya}{\lang1033 > }{\b\lang1033 \'eba}{\lang1033 is a late change (at least post-dating the change }{\b\lang1033 p}{\b\lang1033\super h}{\lang1033 > }{\b\lang1033 f}{\lang1033 ). We do not asterisk the form }{\b\lang1033 faya}{\lang1033 , since it is attested in this name. It may be that Tolkien meant this change to have taken place during F\'ebanor's lifetime, since his name could also be quoted as }{\b\lang1033 Fayan\'e1ro}{\lang1033 . \par \par Paralleling this change, the group }{\b\lang1033 aw}{\lang1033 (not }{\b\lang1033 \'e2w}{\lang1033 ) before an unaccented vowel is reduced to }{\b\lang1033 o}{\lang1033 (in WJ:367, Tolkien hints that }{\b\lang1033 aw}{\lang1033 first became }{ \b\lang1033 au}{\lang1033 and then }{\b\lang1033 o}{\lang1033 ): \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 aw\'e2}{\lang1033 >) *}{\b\lang1033 awa}{\lang1033 > }{\b\lang1033 \'e1ua}{\lang1033 (WJ:367, the accent probably indicating stress rather than length) > }{\b\lang1033 oa}{ \lang1033 "away" (WJ:366, VT39:6) \par (*}{\i\lang1033 law\'e2}{\lang1033 >) \'a4}{\b\lang1033 lawa}{\lang1033 > *}{\b\lang1033 laua}{\lang1033 > }{\b\lang1033 loa}{\lang1033 "year" (VT42:10) \par (\'a4}{\i\lang1033 sraw\'e2}{\lang1033 >) *}{\b\lang1033 hrawa}{\lang1033 > *}{\b\lang1033 hraua}{\lang1033 > }{\b\lang1033 hroa}{\lang1033 "body" (MR:349-350) \par (\'a4}{\i\lang1033 mi-srawanwe}{\lang1033 >) *}{\b\lang1033 mirrawanwe}{\lang1033 > *}{\b\lang1033 mirrauanwe}{\lang1033 >Q }{\b\lang1033 mirroanw\'eb}{\lang1033 "incarnate" (MR:350) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par Strangely, Tolkien has the primive word \'a4}{\b\lang1033 tawar}{\lang1033 "wood" (as material) yielding Quenya }{\b\lang1033 tavar}{\lang1033 instead of **}{\b\lang1033 toar}{\lang1033 (LR:391 s.v. }{\i\lang1033 TAWAR}{\lang1033 ; as for }{ \b\lang1033 w}{\lang1033 becoming }{\b\lang1033 v}{\lang1033 in this position, see below). In a later source, a word for "wood" is however given as }{\b\lang1033 toa}{\lang1033 , regularly derived from a root }{\i\lang1033 TAWA}{\lang1033 (VT39:6). The }{\i\lang1033 TAWAR}{\lang1033 of the }{\i\lang1033 Etymologies}{\lang1033 could be simply an extended form of it. \par \par At some later point, when the rule turning }{\b\lang1033 ay}{\lang1033 /}{\b\lang1033 ai}{\lang1033 and + a vowel into }{\b\lang1033 e}{\lang1033 had ceased to function, the combination }{\b\lang1033 \'e2j}{\lang1033 (or in Quenya orthography }{ \b\lang1033 \'e1y}{\lang1033 ) became }{\b\lang1033 ai}{\lang1033 . The older forms with }{\b\lang1033 \'e1y}{\lang1033 seem to have survived into the historical period, so the Quenya designation of the lesser spirits of the people of the Valar can be cited as }{\b\lang1033 M\'e1yar}{\lang1033 (PM:363, 364). Yet the word normally appears in its later form }{\b\lang1033 Maiar}{\lang1033 (which did not turn into **}{\b\lang1033 M\'ebar}{\lang1033 since the older rule had ceased to operate, as noted above). Another example of the same is }{\b\lang1033 waia}{\lang1033 "envelope", ultimately derived from \'a4}{\b\lang1033 w\'e2j\'e2}{\lang1033 (Etym, entry }{\i\lang1033 WAY}{\lang1033 ). \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 Perhaps some words escaped this change; in the }{\i\lang1033 Etymologies}{\lang1033 , Tolkien cited the Quenya adverb "far off" as }{\b\lang1033 h\'e1ya}{\lang1033 without mentioning any later form *}{\b\lang1033 haia}{\lang1033 . (The word must ultimately be derived from }{\i\lang1033 k}{\i\lang1033\super h}{\i\lang1033 \'e2j\'e2}{\lang1033 , closely reflecting the root }{\i\lang1033 K}{\i\lang1033\super H}{ \i\lang1033 AYA}{\lang1033 . In Valinorean Quenya it would actually appear as *}{\b\lang1033 ch\'e1ya}{\lang1033 ; see below concerning the shift }{\b\lang1033 ch}{\lang1033 > }{\b\lang1033 h}{\lang1033 in Exilic.) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par The retraction period ends. The classical stress-patterns are established: Two-syllable words are nearly always stressed on the first syllable; longer words are henceforth accented on the second-to-last syllable if that is }{\i\lang1033 long}{\lang1033 ; otherwise the stress falls on the third syllable from the end. \par \par At some point after the establishing of the new stress-patterns, a long vowel immediately preceding or following the accented vowel is usually shortened (but not universally so as far as a preceding vowel is concerned; for instance, in WJ:368 Tolkien notes that the Quenya prefix }{\b\lang1033 h\'f3}{\lang1033 - "from" usually retained its long vowel "even when it had become unstressed"). \par \par Shortening of vowels immediately preceding the accented vowel: \par \par \tab (*}{\i\lang1033 w\'f4-n\'f4n\'e2}{\lang1033 > *}{\i\lang1033 \'f4-n\'f4na}{\lang1033 >) *}{\b\lang1033 \'f4n\'f4na}{\lang1033 > Q }{\b\lang1033 on\'f3na}{\lang1033 "twin-born"}{\b\lang1033 \par }{\lang1033 \tab (*}{\i\lang1033 \'f4mand\'ee}{\lang1033 >) *}{\b\lang1033 \'f4mandi}{\lang1033 > Q }{\b\lang1033 omandi}{\lang1033 pl. "vowels" (but singular *}{\i\lang1033 \'f4mand}{\lang1033 - > Q }{\i\lang1033 \'f3man}{\lang1033 with the long vowel intact since it is stressed)}{\b\lang1033 \par }{\lang1033 \par }{\fs20\lang1033 NOTE: The plural }{\b\fs20\lang1033 omandi}{\fs20\lang1033 appears as "amandi" in LR:379 s.v. }{\i\fs20\lang1033 OM}{\fs20\lang1033 , transparently a misreading or typo.}{\i\lang1033 \par }{\lang1033 \tab \par Many three-syllable w ords do not behave quite like we might expect. Even where there is a long vowel in the second-to-last syllable that "ought" to attract the stress, the accent remains on the first syllable even after the end of the retraction period, and the long vowel in the following syllable is }{\i\lang1033 shortened}{\lang1033 : \par \par \tab (\'a4}{\i\lang1033 Aw\'e2delo}{\lang1033 > *}{\i\lang1033 Aw\'e2\'f0el}{\lang1033 >) *}{\b\lang1033 O\'e2zel}{\lang1033 > Q }{\b\lang1033 Oazel}{\lang1033 "away-goer" (later > }{\i\lang1033 Oarel}{\lang1033 ) \par \tab (\'a4}{\i\lang1033 Ar\'e2t\'e2}{\lang1033 >) \'a4}{\b\lang1033 ar\'e2ta}{\lang1033 > }{\b\lang1033 arata}{\lang1033 "exalted" (PM:363, VT41:9; Q pl. }{\i\lang1033 Aratar}{\lang1033 used of the chief Valar) \par \tab (*}{\i\lang1033 nar\'e2k\'e2}{\lang1033 >) \'a4}{\b\lang1033 nar\'e2ka}{\lang1033 > Q }{\b\lang1033 naraca}{\lang1033 "harsh" \par \tab (\'a4}{\i\lang1033 ter\'eaw\'e2}{\lang1033 >) *}{\b\lang1033 ter\'eawa}{\lang1033 > *}{\b\lang1033 terewa}{\lang1033 "fine, acute" (> Q }{\i\lang1033 tereva}{\lang1033 ) \par \tab (\'a4}{\i\lang1033 ter\'ean\'ea}{\lang1033 >) *}{\b\lang1033 ter\'eane}{\lang1033 > Q }{\b\lang1033 teren\'eb}{\lang1033 "slender" \par \par Perhaps we should also include }{\b\lang1033 ontaro}{\lang1033 "parent" here, if the accompanying form }{\b\lang1033 ont\'e2ro}{\lang1033 listed in Etym (entry }{\i\lang1033 ONO}{\lang1033 ) is to be taken as an older form rather than a mere alternative. (It is indicated that }{\b\lang1033 ont\'e2ro}{\lang1033 was at some point accented on the second-to-last syllable, but during the retraction period, the stress would have fallen on }{ \b\lang1033 ont}{\lang1033 -.)}{\i\lang1033 \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 This development seems to be relat ively "regular" when we are dealing with an originally long vowel in the penultimate syllable of three-syllable words: The vowel is shortened and the stress settles on first syllable. Perhaps *}{\i\lang1033 w\'f4-n\'f4n\'e2}{\lang1033 behaves differently (producing Q }{\b\lang1033 on\'f3na}{\lang1033 instead of **}{\b\lang1033 \'f3nona}{\lang1033 ) because the initial syllable is a prefix so that this did not count as a quite regular three-syllable word? However, there are certainly further inconsistencies in the material on this point: The primitive form \'a4}{\b\lang1033 \'f1g\'f4l\'ea}{ \lang1033 (> *}{\b\lang1033 \'f1\'b7g\'f4l\'ea}{\lang1033 > *}{\b\lang1033 ing\'f4le}{\lang1033 ) produces Quenya }{\b\lang1033 ingol\'eb}{\lang1033 "philosophy" (PM:360), with the stress moving to the first syllable and the long vowel in the penultimate syllable becoming short, just as in }{\b\lang1033 naraca}{\lang1033 , }{\b\lang1033 teren\'eb}{\lang1033 and the other examples just listed. Yet primitive \'a4}{\b\lang1033 nd\'fbn\'ea}{\lang1033 (>*}{\b\lang1033 n\'b7d\'fbn\'ea}{\lang1033 > *}{\b\lang1033 and\'fbne}{\lang1033 ) yields Quenya }{\b\lang1033 and\'fan\'eb}{\lang1033 "sunset" with the long vowel in the penultimate syllable }{\i\lang1033 intact}{\lang1033 ; this must also be the syllable that receives the stress. These two examples ought to be completely parallel (incidentally even in the fact that the initial vowel developed from a syllabic nasal), so why do they develop differently? If }{\b\lang1033 ingol\'eb}{\lang1033 , why not **}{\b\lang1033 andun\'eb}{\lang1033 ? And if }{\b\lang1033 and\'fan\'eb}{\lang1033 , why not **}{\b\lang1033 ing\'f3l\'eb}{\lang1033 (and **}{\b\lang1033 nar\'e1ca}{\lang1033 , **}{\b\lang1033 ter\'e9va}{\lang1033 , **}{ \b\lang1033 ter\'e9n\'eb}{\lang1033 etc.)? It would seem that the outcome of the phonological evolution is not entirely predicable in this regard . Exceptionally the long vowel in the penultimate syllable attracted the stress and remained long, but in most three-syllable words the stress moved to the first syllable and the long vowel of the penultimate syllable became short. \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par On the other hand, another development has the stress moving towards the }{\i\lang1033 end}{\lang1033 of the word, a vowel being lengthened to accomplish this. A word ending in three short syllables would normally be stressed on the third syllable from the end. However, it seems that this stress-pattern is only tolerated in words that only consist of three syllables in all, so that the accented syllable is also the first syllable of the word. If the stress would normally fall on the third syllable from the end in a word of }{\i\lang1033 more}{\lang1033 than three syllables, and this syllable is short, the vowel in the second-to-last syllable is lengthened so as to attract the stress. Thus if we add the possessive ending -}{\b\lang1033 va}{\lang1033 to }{\b\lang1033 Eldali\'eb}{\lang1033 "Elf-people", the result is not **}{\b\lang1033 Eldalieva}{\lang1033 (accented on the short syllable }{\b\lang1033 li}{\lang1033 ), but }{\b\lang1033 Eldali\'e9va}{\lang1033 (accented on }{\b\lang1033 \'e9v}{\lang1033 ). Cf. also }{\b\lang1033 Cuivi \'e9nen}{\lang1033 , which is }{\b\lang1033 cuivi\'eb}{\lang1033 "awakening" + }{\b\lang1033 n\'e9n}{\lang1033 , -}{\b\lang1033 nen}{\lang1033 "water". \par }{\fs20\lang1033 NOTE: Curiously, the combination }{\b\fs20\lang1033 ui}{\fs20\lang1033 seems to count as two short syllables for this purpose, as is this were a combination of two short vowels in hiatus rather than the diphthong Tolkien stated }{ \b\fs20\lang1033 ui}{\fs20\lang1033 to be. Thus the possessive form of }{\b\fs20\lang1033 huin\'eb}{\fs20\lang1033 "shadow" is attested as }{\b\fs20\lang1033 huin\'e9va}{\fs20\lang1033 (LR:382) rather than **}{\b\fs20\lang1033 huineva}{\fs20\lang1033 , and a compound of }{\b\fs20\lang1033 tuil\'eb}{\fs20\lang1033 "spring" + }{\b\fs20\lang1033 r\'e9}{\fs20\lang1033 , -}{\b\fs20\lang1033 r\'eb}{\fs20\lang1033 "day" must appear as }{\b\fs20\lang1033 tuil\'e9r\'eb}{\fs20\lang1033 (mentioned in the LotR Appendices) rather than **}{\b\fs20\lang1033 tuiler\'eb}{\fs20\lang1033 . \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 The adjective }{\b\lang1033 vanima}{\lang1033 "beautiful", if supplied with the partitive plural ending -}{\b\lang1033 li}{\lang1033 to express "beautiful ones", must of course become *}{ \b\lang1033 vanim\'e1li}{\lang1033 . This is only attested in the genitive form }{\b\lang1033 vanim\'e1lion}{\lang1033 (occurring in Treebeard's greeting to Celeborn and Galadriel). The first edition of LotR had }{\b\lang1033 vanimalion}{\lang1033 , but this rejected form would conflict with the apparent rule that a noun of more than three syllables cannot end in three }{\i\lang1033 short}{\lang1033 syllables. Some compounds (like }{\b\lang1033 Altariel}{\lang1033 /}{\b\lang1033 \'d1altariel}{ \lang1033 = }{\b\lang1033 \'f1alta}{\lang1033 + }{\b\lang1033 riel}{\lang1033 , PM:347) seem to be exempted from this rule; on the other hand, we observe the excepted lengthening in }{\b\lang1033 An}{\b\ul\lang1033 \'e1}{\b\lang1033 rion}{\lang1033 "sun-son" = }{\b\lang1033 Anar}{\lang1033 + the patronymic ending -}{\b\lang1033 ion}{\lang1033 . \par }\pard \nowidctlpar\widctlpar\adjustright {\fs20\lang1033 NOTE: Since }{\b\fs20\lang1033 Anar}{\fs20\lang1033 "Sun" descends from \'a4}{\b\fs20\lang1033 an\'e2r}{\fs20\lang1033 - (LR:348), it may be that in }{\b\fs20\lang1033 An\'e1rion}{\fs20\lang1033 an old long vowel is simply }{\i\fs20\lang1033 preserved}{\fs20\lang1033 long because it happened to be stressed. (The simplex "sun" may still have been *}{\b\fs20\lang1033 An\'e1r}{\fs20\lang1033 in earlier forms of Quenya: according to what Tolkien wrote in the Plotz Letter, the shortening of long vowels in the final syllable of polysyllabic words is a feature of Exilic Quenya. See below.) \par }{\lang1033 \par Classical Quenya does not seem to permit a }{\i\lang1033 long vowel}{\lang1033 immediately before a consonant cluster. }{\b\lang1033 M\'e1nwen}{\lang1033 or }{\b\lang1033 M\'e1nw\'eb}{\lang1033 is quoted as the oldest Quenya form of the name }{ \b\lang1033 Manw\'eb}{\lang1033 (WJ:399); notice how }{\b\lang1033 \'e1}{\lang1033 was eventually shortened before the cluster }{\b\lang1033 nw}{\lang1033 . Likewise, }{\i\lang1033 F\'edriel's Song}{\lang1033 indicates that the noun }{\b\lang1033 m\'e1 }{\lang1033 "hand" appears as }{\b\lang1033 mannar}{\lang1033 in the (plural) allative \endash perhaps for older *}{\b\lang1033 m\'e1nnar}{\lang1033 . However, long vowels seem to be allowed before clusters ending in the semi-vowel }{\b\lang1033 y}{ \lang1033 : }{\i\lang1033 Nam\'e1ri\'eb}{\lang1033 has }{\b\lang1033 m\'e1ryat}{\lang1033 (rather than *}{\b\lang1033 maryat}{\lang1033 ) for "her hands". \par \par }{\i\lang1033 Anticipatory vowels developing following initial }{\b\i\lang1033 3}{\i\lang1033 + consonant:}{\lang1033 It seems that in some cases, a vowel developed following an initial consonant cluster in }{\b\lang1033 3}{\lang1033 - (representing earlier }{\b\lang1033 g}{\lang1033 -). This vowel seems to echo or rather "anticipate" the first vowel occurring after it, even where that vowel was the first element of a diphthong. Tolkien indicated that original \'a4}{\b\lang1033 grauk} {\lang1033 - underlies Quenya }{\b\lang1033 rauco}{\lang1033 and }{\b\lang1033 arauco}{\lang1033 "terrible creature" (WJ:415). Evidently \'a4}{\b\lang1033 grauk}{\lang1033 - (the complete primitive word must have been *}{\b\lang1033 grauk\'f4}{ \lang1033 or *}{\b\lang1033 grauku}{\lang1033 ) first became *}{\b\lang1033 3rauk}{\lang1033 - with the general change }{\b\lang1033 g}{\lang1033 > }{\b\lang1033 3}{\lang1033 . This }{\b\lang1033 3}{\lang1033 would later disappear (see below). *}{ \b\lang1033 3rauk}{\lang1033 - might either persist in this form, in which case it would simply be reduced to }{\b\lang1033 rauc}{\lang1033 - in Quenya, or it might develop an extended parallel form *}{\b\lang1033 3arauk}{\lang1033 -, the vowel developing after }{\b\lang1033 3}{\lang1033 anticipating the next vowel in the word (in this case the first element of the diphthong }{\b\lang1033 au}{\lang1033 ). This *}{\b\lang1033 3arauk}{\lang1033 - would later become Quenya }{ \b\lang1033 arauc}{\lang1033 -. \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 The }{\i\lang1033 Etymologies}{\lang1033 seems to include one example of this phenomenon, though Tolkien later struck out the relevant entry: The stem }{\i\lang1033 GLINDI}{\lang1033 produced Quenya }{\b\lang1033 ilin}{\lang1033 "pale blue" (evidently shortened from *}{\b\lang1033 ilind\'eb}{\lang1033 ). The idea must be that }{\i\lang1033 GLI-}{\lang1033 first becomes *}{\b\lang1033 3li}{\lang1033 - and then *}{\b\lang1033 3ili}{ \lang1033 - with an anticipatory vowel, eventually reduced to }{\b\lang1033 ili}{\lang1033 - with the later loss of }{\b\lang1033 3}{\lang1033 . Another possible example was apparently also rejected by Tolkien: In MR:284, the place-name "Avathar" (evidently = }{\b\lang1033 Ava\'fear}{\lang1033 ) is said to be ancient Quenya for "the Shadows". This must somehow be related to Sindarin }{ \b\lang1033 gwa\'fe}{\lang1033 "shadow" (see }{\i\lang1033 gwath}{\lang1033 in the }{\i\lang1033 Silmarillion}{\lang1033 Appendix). In the }{\i\lang1033 Etymologies}{\lang1033 , this word is derived from a root }{\i\lang1033 WAT}{\i\lang1033\super H}{ \lang1033 , and the prefixing of }{\b\lang1033 g}{\lang1033 - in the "Noldorin"/Sindarin branch is secondary. However, Tolkien may at one point have considered deriving }{\b\lang1033 gwa\'fe}{\lang1033 from a stem *}{\i\lang1033 GWAT}{\i\lang1033\super H }{\lang1033 instead, the initial }{\i\lang1033 G}{\lang1033 being original and therefore also able to influence the branch leading to Quenya. \par In this alternative scenario, the word "shade, shadow" might be *}{\b\lang1033 gwat}{\b\lang1033\super h}{\b\lang1033 \'e2}{\lang1033 in the primitive language, later becoming *}{\b\lang1033 3wa\'fea}{\lang1033 > *}{\b\lang1033 3awa\'fea}{\lang1033 in the proto-Quenya branch, in turn producing archaic Quenya *}{\b\lang1033 ava\'fea}{\lang1033 (see below concerning the change of post-vocalic }{\b\lang1033 w}{\lang1033 > }{\b\lang1033 v}{\lang1033 ). }{\b\lang1033 Ava\'fear}{\lang1033 could obviously be the plural of this word, so that the word could refer to the "Shadows" as Tolkien stated. This derivation is not entirely unproblematic, though: Instead of *}{\i\lang1033 GWAT}{\i\lang1033\super H}{\lang1033 and *}{\b\lang1033 gwat}{ \b\lang1033\super h}{\b\lang1033 \'e2}{\lang1033 we would expect *}{\i\lang1033 G}{\i\lang1033\super W}{\i\lang1033 AT}{\i\lang1033\super H}{\lang1033 and *}{\b\lang1033 g}{\b\lang1033\super w}{\b\lang1033 at}{\b\lang1033\super h}{\b\lang1033 \'e2}{ \lang1033 at the primitive stage; the initial combination would not strictly be a cluster }{\b\lang1033 g}{\lang1033 + }{\b\lang1033 w}{\lang1033 , but simply }{\b\lang1033 g}{\b\lang1033\super w}{\lang1033 = labialized }{\b\lang1033 g}{\lang1033 , a unitary consonant. Later we would then have }{\b\lang1033 3}{\b\lang1033\super w}{\lang1033 (rather than }{\b\lang1033 3w}{\lang1033 ). Hence }{\b\lang1033 3}{\lang1033 would not really be part of an initial consonant cluster, the environment that triggers the development of an anticipatory vowel. Could the labialized quality of }{\b\lang1033 3}{\b\lang1033\super w}{\lang1033 really be separated from it to materialize as a separate consonant }{\b\lang1033 w}{\lang1033 > }{\b\lang1033 v}{\lang1033 , an intruding vowel splitting an originally }{\i\lang1033 unitary}{\lang1033 consonant in two (}{\b\lang1033 3}{ \b\lang1033\super w}{\lang1033 > }{\b\lang1033 3aw}{\lang1033 -)? In the }{\i\lang1033 Etymologies}{\lang1033 , the stem }{\i\lang1033 G}{\i\lang1033\super W}{\i\lang1033 EN}{\lang1033 produces Quenya }{\b\lang1033 wend\'eb}{\lang1033 > }{\b\lang1033 vend\'eb}{\lang1033 "maiden", not **}{\b\lang1033 evend\'eb}{\lang1033 for older *}{\b\lang1033 3}{\b\lang1033\super w}{\b\lang1033 ende}{\lang1033 . (Cf. also \'a4}{\b\lang1033 g-lada}{\lang1033 - > *}{\b\lang1033 glala}{\lang1033 - > *}{\b\lang1033 3lala}{\lang1033 - simply becoming }{\b\lang1033 lala}{\lang1033 - in Quenya, PM:359 \endash but the hyphen inserted by Tolkien in the form \'a4}{\b\lang1033 g-lada}{\lang1033 - may indicate that the prefixing of }{\b\lang1033 g}{\lang1033 - w as peculiar to the Lindarin branch, so that the branch leading to Quenya never had anything but simple initial }{\b\lang1033 l}{\lang1033 -.) It seems that Tolkien eventually gave up trying to find an Elvish etymology for the name Avathar, explicitly denying that it was Elvish and instead suggesting that it was a borrowing from Valarin (WJ:404). But since our source for this is the same essay that provides the example \'a4}{\b\lang1033 grauk}{\lang1033 - > Q }{\b\lang1033 arauc}{\lang1033 -, the idea of anticapatory vowels as such was evidently not rejected. \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par The semi-vowel }{\b\lang1033 j}{\lang1033 is lost when it occurs after the corresponding full vowel }{\b\lang1033 i}{\lang1033 and before another vowel. }{\b\lang1033 W}{\lang1033 is likewise lost between the corrresponding full vowel }{\b\lang1033 u}{ \lang1033 and a following vowel: \par \par \tab (\'a4}{\i\lang1033 oij\'e2}{\lang1033 >) *}{\b\lang1033 oija}{\lang1033 > Q }{\b\lang1033 oia}{\lang1033 "ever"}{\fs20\lang1033 (but in later Exilic Quenya, the form*}{\b\fs20\lang1033 oiya}{\fs20\lang1033 might arise once again, because a glide consonant }{\b\fs20\lang1033 y}{\fs20\lang1033 occasionally developed between }{\b\fs20\lang1033 i}{\fs20\lang1033 and another vowel \endash see below)}{\lang1033 \par \tab (\'a4}{\i\lang1033 kuk\'fbw\'e2}{\lang1033 > *}{\i\lang1033 k\'fbw\'e2}{\lang1033 >) *}{\b\lang1033 k\'fbwa}{\lang1033 > Q }{\b\lang1033 cua}{\lang1033 "dove" \par \par Tolkien mentions }{\b\lang1033 k\'fbua}{\lang1033 as the immediate ancestor of }{\b\lang1033 cua}{\lang1033 , so it would seem that the }{\b\lang1033 w}{\lang1033 of }{\b\lang1033 k\'fbwa}{\lang1033 became }{\b\lang1033 u}{\lang1033 before it simply merged into the preceding }{\b\lang1033 u}{\lang1033 . Also notice how the long }{\b\lang1033 \'fb}{\lang1033 becomes short when brought into contact with a following vowel. Such shortening occurs in other parallel examples from the }{ \i\lang1033 Etymologies}{\lang1033 (see below concerning }{\b\lang1033 tuo}{\lang1033 , }{\b\lang1033 ri\'eb}{\lang1033 ). \par \par In a monosyllabic word, a final }{\b\lang1033 w}{\lang1033 merging into a preceding }{\b\lang1033 u}{\lang1033 does produce a long }{\b\lang1033 \'fb}{\lang1033 , still surviving as Quenya }{\b\lang1033 \'fa}{\lang1033 : \par \par \tab (\'a4}{\i\lang1033 ku3}{\lang1033 >) \'a4}{\b\lang1033 kuw}{\lang1033 > }{\b\lang1033 k\'fb}{\lang1033 "bow" (= Q }{\i\lang1033 c\'fa}{\lang1033 ) \par \par Older }{\b\lang1033 w}{\lang1033 , where it still survived after the various changes discussed above, in very many cases became }{\b\lang1033 v}{\lang1033 . It seems that the change of }{\i\lang1033 medial}{\lang1033 }{\b\lang1033 w}{\lang1033 to }{ \b\lang1033 v}{\lang1033 occurred before the same change happened initially (the initial change }{\b\lang1033 w}{\lang1033 > }{\b\lang1033 v}{\lang1033 is rather a feature of Exilic Quenya \endash see below). But post-vocalically, }{\b\lang1033 w}{ \lang1033 normally became }{\b\lang1033 v}{\lang1033 very early, and that this change would have taken effe ct already during the Classical Valinorean period. See further discussion below (under "Exilic Quenya"). Also notice that while Tolkien thought of }{\b\lang1033 Ava\'fear}{\lang1033 as an Elvish name meaning "the Shadows" (before he turned it into a Valarin word), he stated that it was "ancient Quenya" (MR:284). The }{\b\lang1033 v}{\lang1033 of }{\b\lang1033 Ava\'fear}{\lang1033 certainly reprensents earlier }{\b\lang1033 w}{ \lang1033 (this goes whether }{\b\lang1033 Ava\'fear}{\lang1033 is derived from a root like }{\i\lang1033 WAT}{\i\lang1033\super H}{\lang1033 "shade", cf. LR:397, or from *}{\i\lang1033 GWAT}{\i\lang1033\super H}{\lang1033 > }{\b\lang1033 3wa\'fe}{ \lang1033 - > }{\b\lang1033 3awa\'fe}{\lang1033 - as we theorized above). Hence, the change of post-vocalic }{\b\lang1033 w}{\lang1033 to }{\b\lang1033 v}{\lang1033 had already occurred in "ancient Quenya", perhaps meaning pre-record Quenya. \par \par Examples of post-vocalic }{\b\lang1033 w}{\lang1033 becoming }{\b\lang1033 v}{\lang1033 : \par \par \tab (\'a4}{\i\lang1033 sr\'e2w\'ea}{\lang1033 > *}{\i\lang1033 hr\'e2w\'ea}{\lang1033 >) *}{\b\lang1033 hr\'e1w\'eb}{\lang1033 > }{\b\lang1033 hr\'e1v\'eb}{\lang1033 "flesh" \par \tab (\'a4}{\i\lang1033 kuiw\'ea}{\lang1033 >) *}{\b\lang1033 cuiw\'eb}{\lang1033 > }{\b\lang1033 cuiv\'eb}{\lang1033 "awakening" }{\i\lang1033 (for the most primitive form, cf. the prefixed variant }{\lang1033 \'a4et-kuiw\'ea}{\i\lang1033 listed in the }{\lang1033 Etymologies}{\i\lang1033 , entry KUY}{\lang1033 ) \par \tab (*}{\i\lang1033 r\'e2w\'ee}{\lang1033 >) *}{\b\lang1033 r\'e1wi}{\lang1033 > }{\b\lang1033 r\'e1vi}{\lang1033 pl. "lions" (cf. root }{\i\lang1033 RAW}{\lang1033 ) \par \tab (\'a4}{\i\lang1033 ter\'eaw\'e2}{\lang1033 > *}{\i\lang1033 ter\'eawa}{\lang1033 >) }{\b\lang1033 terewa}{\lang1033 > }{\b\lang1033 tereva}{\lang1033 "fine, acute" \par \par However, it would seem that }{\b\lang1033 w}{\lang1033 remained unchanged following the vowel }{\b\lang1033 o}{\lang1033 ; cf. }{\b\lang1033 n\'f3wi}{\lang1033 as the plural form of }{\b\lang1033 n\'f3}{\lang1033 "conception" (the pl. form more closely reflecting the original root }{\i\lang1033 NOWO}{\lang1033 ). \par \par For some reason, }{\b\lang1033 w}{\lang1033 also persists unchanged following the diphthong }{\b\lang1033 ai}{\lang1033 . In the }{\i\lang1033 Etymologies}{\lang1033 , Tolkien has \'a4}{\i\lang1033 slaiw\'e2}{\lang1033 yielding Quenya }{\b\lang1033 laiwa}{\lang1033 "sick" (not **}{\b\lang1033 laiva}{\lang1033 ). Contrast the related word \'a4}{\i\lang1033 sl\'eew\'ea}{\lang1033 , producing Quenya }{\b\lang1033 l\'edv\'eb}{\lang1033 "sickness"; here there is no diphthong }{\b\lang1033 ai}{ \lang1033 to hinder the development }{\b\lang1033 w}{\lang1033 > }{\b\lang1033 v}{\lang1033 . Other examples: \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\b\lang1033 aiw\'eb}{\lang1033 "bird" (< \'a4}{\i\lang1033 aiw\'ea}{\lang1033 , given as an entry-head in Etym)}{\b\lang1033 \par maiw\'eb}{\lang1033 "gull" (< *}{\i\lang1033 maiw\'ea}{\lang1033 , with an A-infixed variant of the root }{\i\lang1033 MIW}{\lang1033 ) \par }{\b\lang1033 vaiwa}{\lang1033 "wind" (< *}{\i\lang1033 waiw\'e2}{\lang1033 , cf. root }{\i\lang1033 WAIWA}{\lang1033 ) \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par }{\fs20\lang1033 NOTE: Since the Quenya possessive ending -}{\b\fs20\lang1033 va}{\fs20\lang1033 evidently descends from *-}{\b\fs20\lang1033 w\'e2}{\fs20\lang1033 , one may wonder why we see }{\b\fs20\lang1033 Eldaiva}{\fs20\lang1033 rather than *}{ \b\fs20\lang1033 Eldaiwa}{\fs20\lang1033 as the pl. possessive of }{\b\fs20\lang1033 Elda}{\fs20\lang1033 (WJ:368). The explanation seems to be that -}{\b\fs20\lang1033 va}{\fs20\lang1033 had originally only occurred in the singular (}{ \b\fs20\lang1033 Eldava}{\fs20\lang1033 , with no preceding diphthong }{\b\fs20\lang1033 ai}{\fs20\lang1033 ); the plural form }{\b\fs20\lang1033 Eldaiva}{\fs20\lang1033 was a later, analogical formation based on the singular }{\b\fs20\lang1033 Eldava}{ \fs20\lang1033 . See WJ:407. \par }{\lang1033 \par Depending on its position, the }{\i\lang1033 original}{\lang1033 back spirant }{\b\lang1033 3}{\lang1033 had either turned into }{\b\lang1033 h}{\lang1033 (or }{\b\lang1033 ch}{\lang1033 ) or been lost, as described above. Now Quenya had re-acquired }{ \b\lang1033 3}{\lang1033 (developed from original }{\b\lang1033 g}{\lang1033 ), but this sound was evidently not destined to be a lasting part of the Quenya sound-system: This "new" }{\b\lang1033 3}{\lang1033 , whether initial, medial or final, also disappeared: \par \tab \par \tab (*}{\i\lang1033 weg\'f4}{\lang1033 > \'a4}{\i\lang1033 we3\'f4}{\lang1033 >) *}{\b\lang1033 we3o > }{\lang1033 *}{\b\lang1033 w\'ebo}{\lang1033 "man" (in later Q }{\i\lang1033 v\'ebo}{\lang1033 ) \par \tab (\'a4}{\i\lang1033 Denwego}{\lang1033 > *}{\i\lang1033 Denweg}{\lang1033 >) *}{\b\lang1033 Lenwe3}{\lang1033 > }{\b\lang1033 Lenw\'eb}{\lang1033 (personal name, WJ:412) \par \tab }{\b\lang1033 (}{\lang1033 \'a4}{\i\lang1033 gass\'e2}{\lang1033 > *}{\i\lang1033 3ass\'e2}{\lang1033 >) *}{\b\lang1033 3assa}{\lang1033 > }{\b\lang1033 assa}{\lang1033 "opening" \par \tab (\'a4}{\i\lang1033 grauk[\'f4]}{\lang1033 > *}{\i\lang1033 3rauk\'f4}{\lang1033 >) *}{\b\lang1033 3arauko}{\lang1033 > }{\b\lang1033 arauco}{\lang1033 "terrible creature" (WJ:455) \par \tab }{\b\lang1033 (}{\lang1033 \'a4}{\i\lang1033 gard\'e2}{\lang1033 > *}{\i\lang1033 3ard\'e2}{\lang1033 >) *}{\b\lang1033 3arda}{\lang1033 > }{\b\lang1033 Arda}{\lang1033 (a name of the world, WJ:402) \par \tab (*}{\i\lang1033 gond\'f4}{\lang1033 > *}{\i\lang1033 3ond\'f4}{\lang1033 >) *}{\b\lang1033 3ondo}{\lang1033 > }{\b\lang1033 ondo}{\lang1033 "rock" \par \tab (\'a4}{\i\lang1033 galad\'e2}{\lang1033 > *}{\i\lang1033 gald\'e2}{\lang1033 > *}{\i\lang1033 3ald\'e2}{\lang1033 >) }{\b\lang1033 3alda}{\lang1033 > }{\b\lang1033 alda}{\lang1033 "tree" \par \tab \par In some sources Tolkien only speaks of how "primitive initial }{\i\lang1033 g}{\lang1033 - was lost in Q" (Letters:410), but }{\b\lang1033 g}{\lang1033 did not disappear "directly"; it first turned into }{\b\lang1033 3}{\lang1033 . We know this because } {\b\lang1033 3}{\lang1033 survived long enough to be explicitly recorded in R\'familian writing: In R\'familian orthography, the word }{\b\lang1033 alda}{\lang1033 was still spelt }{\b\lang1033 3alda}{\lang1033 \endash that is, it had an initial letter "which was known by tradition among the loremasters to have represented the open back spirant" (VT39:7). Hence we need not asterisk the form }{\b\lang1033 3alda}{\lang1033 . (If we knew more about R\'fa milian writing, we could probably have removed the asterisk from words like *}{\b\lang1033 3assa}{\lang1033 , *}{\b\lang1033 3arda}{\lang1033 , *}{\b\lang1033 3ondo}{\lang1033 as well.) \par \par Between vowels, and following }{\b\lang1033 l}{\lang1033 , the spirant }{\b\lang1033 3}{\lang1033 is lost with no trace: \par }\pard \qj\nowidctlpar\widctlpar\adjustright {\lang1033 \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \tab (*}{\i\lang1033 k}{\i\lang1033\super h}{\i\lang1033 ugan}{\lang1033 >) *}{\b\lang1033 chu3an}{\lang1033 > *}{\b\lang1033 chuan}{\lang1033 "hound" (Exilic Quenya }{\i\lang1033 huan}{\lang1033 ) \par \tab }{\b\lang1033 (}{\lang1033 \'a4}{\i\lang1033 ulgund\'f4}{\lang1033 >) *}{\b\lang1033 ul3undo}{\lang1033 > }{\b\lang1033 ulundo}{\lang1033 "monster" \par \par Where a postvocalic }{\b\lang1033 3}{\lang1033 was followed by a consonant, the preceding vowel was lengthened to "make up" for its disappearance (maintaining prosodic length \endash see SD:420 for Ad\'fbnaic parallels): \par \par \tab (\'a4}{\i\lang1033 lugni}{\lang1033 > *}{\i\lang1033 lugne}{\lang1033 >) *}{\b\lang1033 lu3ne}{\lang1033 > }{\b\lang1033 l\'fan\'eb}{\lang1033 "blue" \par \tab (*}{\i\lang1033 rign\'e2}{\lang1033 > *}{\i\lang1033 rigna}{\lang1033 >) *}{\b\lang1033 ri3na}{\lang1033 > }{\b\lang1033 r\'edna}{\lang1033 "crowned" \par \tab (\'a4}{\i\lang1033 sagr\'e2}{\lang1033 > *}{\i\lang1033 sagra}{\lang1033 >) *}{\b\lang1033 sa3ra}{\lang1033 > }{\b\lang1033 s\'e1ra}{\lang1033 "bitter" \par \tab (\'a4}{\i\lang1033 magr\'e2}{\lang1033 > *}{\i\lang1033 magra}{\lang1033 >) *}{\b\lang1033 ma3ra}{\lang1033 > }{\b\lang1033 m\'e1ra}{\lang1033 "good" \par \par It would be interesting to know whether R\'familian orthography represented these words in a spelling similar to the (bold) asterisked forms above.}{\b\lang1033 \par }{\lang1033 \par }{\fs20\lang1033 NOTE: The noun }{\b\fs20\lang1033 huan}{\fs20\lang1033 "hound" (descended from \'a4}{\b\fs20\lang1033 k}{\b\fs20\lang1033\super h}{\b\fs20\lang1033 ugan}{\fs20\lang1033 via *}{\b\fs20\lang1033 chu3an}{\fs20\lang1033 ) is said to appear as }{\b\fs20\lang1033 h\'fan}{\fs20\lang1033 - when endings are added. This would seem to come from *}{\b\fs20\lang1033 chu3n}{\fs20\lang1033 -, exemplifying the same phenomenon of vocalic lengthening. It may be that soon after }{ \b\fs20\lang1033 g}{\fs20\lang1033 became }{\b\fs20\lang1033 3}{\fs20\lang1033 , an unstressed non-final syllable }{\b\fs20\lang1033 3a}{\fs20\lang1033 lost its vowel: *}{\b\fs20\lang1033 chu3an}{\fs20\lang1033 - > *}{\b\fs20\lang1033 chu3n}{ \fs20\lang1033 - (> ultimately }{\b\fs20\lang1033 h\'fan}{\fs20\lang1033 -). This did not happen where no ending followed so that the combination }{\b\fs20\lang1033 3a}{\fs20\lang1033 was part of the }{\i\fs20\lang1033 final}{\fs20\lang1033 syllable; hence *}{\b\fs20\lang1033 chu3an}{\fs20\lang1033 by itself produces Quenya }{\b\fs20\lang1033 huan}{\fs20\lang1033 with no loss of }{\b\fs20\lang1033 a}{\fs20\lang1033 . \par }{\lang1033 \par It should be noted that a }{\b\lang1033 w}{\lang1033 that became intervocalic only by the loss of }{\b\lang1033 3}{\lang1033 , "missed" the normal change of intervocalic }{\b\lang1033 w}{\lang1033 > }{\b\lang1033 v}{\lang1033 ; this change had ceased to operate when }{\b\lang1033 3}{\lang1033 disappeared: \par \par \tab (\'a4}{\i\lang1033 jagw\'ea}{\lang1033 > *}{\i\lang1033 jagwe}{\lang1033 >) *}{\b\lang1033 ya3w\'eb}{\lang1033 > }{\b\lang1033 y\'e1w\'eb}{\lang1033 "ravine" }{\i\lang1033 (NOT}{\lang1033 > **}{\b\lang1033 y\'e1v\'eb}{\lang1033 ) \par \par According to the }{\i\lang1033 Etymologies}{\lang1033 , a long vowel is shortened if, by the loss of }{\b\lang1033 3}{\lang1033 , it is brought into direct contact with another vowel: \par \tab \par \tab (\'a4}{\i\lang1033 t\'fbgu}{\lang1033 > *}{\i\lang1033 t\'fbgo}{\lang1033 >) *}{\b\lang1033 t\'fb3o}{\lang1033 > }{\b\lang1033 tuo}{\lang1033 "muscle" (not *}{\b\lang1033 t\'fao}{\lang1033 ) \par \tab (\'a4}{\i\lang1033 r\'eeg\'ea}{\lang1033 > *}{\i\lang1033 r\'eege}{\lang1033 >) *}{\b\lang1033 r\'ee3\'eb}{\lang1033 > }{\b\lang1033 ri\'eb}{\lang1033 "crown" (not *}{\b\lang1033 r\'ed\'eb}{\lang1033 ) \par \par But according to a post-LotR example (PM:347), such a vowel remains long; this is actually a variant of }{\b\lang1033 ri\'eb}{\lang1033 above: \par \par \tab (\'a4}{\i\lang1033 r\'eeg\'e2}{\lang1033 > *}{\i\lang1033 r\'eega}{\lang1033 >) *}{\b\lang1033 r\'ee3a}{\lang1033 > }{\b\lang1033 r\'eda}{\lang1033 "wreath" (not *}{\b\lang1033 ria}{\lang1033 ) \par \par Whether we should emend Quenya }{\b\lang1033 tuo}{\lang1033 , }{\b\lang1033 ri\'eb}{\lang1033 above to *}{\b\lang1033 t\'fao}{\lang1033 , *}{\b\lang1033 r\'ed\'eb}{\lang1033 is unclear. Perhaps we may rather assume that they "actually" descend from *}{ \i\lang1033 tugu}{\lang1033 , *}{\i\lang1033 rig\'ea}{\lang1033 instead, the stem-vowel being short from the beginning. \par \par From one example it may seem that two identical vowels brought into direct contact by the loss of }{\b\lang1033 3}{\lang1033 merge into one (long) vowel: (\'a4}{\i\lang1033 ugu}{\lang1033 - >) *}{\b\lang1033 u3u}{\lang1033 - > }{\b\lang1033 \'fa}{ \lang1033 -, the Quenya negative prefix "un-" (in the }{\i\lang1033 Etymologies}{\lang1033 derived from a stem }{\i\lang1033 UGU}{\lang1033 ). However, it may be that the second }{\b\lang1033 u}{\lang1033 was lost already in the syncope, and that *}{ \b\lang1033 u3}{\lang1033 -, usually occurring before a consonant, turned into }{\b\lang1033 \'fa}{\lang1033 - by same process that turns *}{\b\lang1033 y}{\b\ul\lang1033 a}{\b\lang1033 3w\'eb}{\lang1033 into *}{\b\lang1033 y}{\b\ul\lang1033 \'e1}{ \b\lang1033 w\'eb}{\lang1033 (see above). In other examples of two vowels of the same quality coming into contact by the loss of }{\b\lang1033 3}{\lang1033 , the first of them is dissimilated: }{\b\lang1033 \'f33o}{\lang1033 > }{\b\lang1033 o'o}{ \lang1033 > }{\b\lang1033 uo}{\lang1033 , and }{\b\lang1033 e3e}{\lang1033 > }{\b\lang1033 e'e}{\lang1033 > }{\b\lang1033 ie}{\lang1033 . (Compare }{\b\lang1033 e'\'ea}{\lang1033 becoming }{\b\lang1033 i\'ea}{\lang1033 in Common Eldarin, after the loss of intervocalic }{\b\lang1033 \'f1}{\lang1033 or }{\b\lang1033 3}{\lang1033 .)}{\b\lang1033 \par }{\lang1033 \par \tab (\'a4}{\i\lang1033 K}{\i\lang1033\super h}{\i\lang1033 \'f4-gor\'ea}{\lang1033 "heart-vigour" >) *}{\b\lang1033 Ch\'f33or\'eb}{\lang1033 > *}{\b\lang1033 Cho'or\'eb }{\lang1033 > *}{\b\lang1033 Chuor\'eb}{\lang1033 (masc. name, Exilic Quenya }{ \i\lang1033 Huor\'eb}{\lang1033 , Sindarin }{\i\lang1033 Huor}{\lang1033 ) \par \tab (*}{\i\lang1033 weg\'ea}{\lang1033 > \'a4}{\i\lang1033 we3\'ea}{\lang1033 >) *}{\b\lang1033 we3e}{\lang1033 > *}{\b\lang1033 we'e}{\lang1033 > *}{\b\lang1033 wie}{\lang1033 "manhood" (Q *}{\i\lang1033 wi\'eb}{\lang1033 > }{\i\lang1033 vi\'eb}{ \lang1033 ) \par \par (In the first of these examples we again a long vowel being shortened before another vowel after the loss of }{\b\lang1033 3}{\lang1033 ; the long }{\b\lang1033 \'f4}{\lang1033 evidently becomes short }{\b\lang1033 o}{\lang1033 and is then dissimilated to }{\b\lang1033 u}{\lang1033 before another }{\b\lang1033 o}{\lang1033 . According to the post-LotR example \'a4}{\b\lang1033 r\'eeg\'e2}{\lang1033 > *}{\b\lang1033 r\'ed3a}{\lang1033 > }{\b\lang1033 r\'eda}{\lang1033 quoted above, we might expect the vowel to remain long: *}{\b\lang1033 Ch\'f33or\'eb}{\lang1033 > *}{\b\lang1033 Ch\'faor\'eb}{\lang1033 , later *}{\b\lang1033 H\'faor\'eb}{\lang1033 .) \par \par By a change that first occurred in Telerin, the original bilabial }{\b\lang1033 f}{\lang1033 became a }{\i\lang1033 labio-dental}{\lang1033 }{\b\lang1033 f}{\lang1033 , similar to English }{\b\lang1033 f}{\lang1033 (pronounced with friction between the lower lip and the upper teeth). This change made it easier to distinguish }{\b\lang1033 f}{\lang1033 from }{\b\lang1033 hw}{\lang1033 , so the Noldor adopted labio-dental }{\b\lang1033 f}{\lang1033 from Telerin (VT41:7). The Vanyar did not, and the Noldor accused them of confusing }{\b\lang1033 f}{\lang1033 and }{\b\lang1033 hw}{\lang1033 . \par \par }{\i\lang1033 Secondary syncope:}{\lang1033 The last recorded syncope operating in Quenya involved the omission of the second of two identical short vowels where a wor d occurs in "long compound names" (VT41:9). We call this secondary syncope because it is unrelated to the much older, primary syncope that produced, say, }{\b\lang1033 al'da}{\lang1033 "tree" from original \'a4}{\b\lang1033 gal}{\b\ul\lang1033 a}{ \b\lang1033 d\'e2}{\lang1033 . The primitive word \'a4}{\b\lang1033 ar\'e2t\'e2}{\lang1033 had produced Quenya }{\b\lang1033 arata}{\lang1033 "exalted, noble" with the second }{\b\lang1033 a}{\lang1033 surviving the primary syncope because it had originally been long (PM:363, VT41:9). Yet a word like }{\b\lang1033 arata}{\lang1033 could be shortened to }{\b\lang1033 arta}{\lang1033 (PM:354) when it occurred as part of a longer name (e.g. in }{ \b\lang1033 Artaher}{\lang1033 "noble lord", PM:346, or masculinized -}{\b\lang1033 arto}{\lang1033 in }{\b\lang1033 Ambarto}{\lang1033 , PM:354). We must assume that the }{\b\lang1033 silma}{\lang1033 - part of the word }{\b\lang1033 Silmaril}{ \lang1033 has also undergone secondary syncope, since this element represents "the name }{\i\lang1033 silima}{\lang1033 that F\'ebanor gave to the substance from which [the Silmarils] were made" (}{\i\lang1033 Silmarillion}{\lang1033 Appendix, entry }{ \i\lang1033 sil}{\lang1033 -). \par \par }{\fs20\lang1033 NOTE: In the earlier scenario recorded in the }{\i\fs20\lang1033 Etymologies}{\fs20\lang1033 , the initial element of }{\b\fs20\lang1033 Silmaril}{\fs20\lang1033 is suggested to be the adjective }{\b\fs20\lang1033 silma}{\fs20\lang1033 "silver, shining white", which is referred to a primitive adjectival form \'a4}{\b\fs20\lang1033 silim\'e2}{\fs20\lang1033 that would have lost its second }{\b\fs20\lang1033 i}{\fs20\lang1033 already in the }{\i\fs20\lang1033 primary}{\fs20\lang1033 syncope (LR:385 s.v. }{\i\fs20\lang1033 SIL}{\fs20\lang1033 ). Indeed a primitive form of the word }{\b\fs20\lang1033 Silmaril}{\fs20\lang1033 is provided, namely \'a4}{\b\fs20\lang1033 silimarille}{\fs20\lang1033 (LR:383 s.v. }{\i\fs20\lang1033 RIL}{ \fs20\lang1033 , the final -}{\b\fs20\lang1033 \'e2}{\fs20\lang1033 of }{\b\fs20\lang1033 silim\'e2}{\fs20\lang1033 regularly being shortened when the word occurs as the first part of a compound). But the actual existence of such a primitive form is problematic for historical reasons: Surely the primary syncope was long since over when F\'eb anor invented the Silmarils? It must be placed in the unrecorded past long before he was even born. Perhaps the form \'a4}{\b\fs20\lang1033 silimarille}{\fs20\lang1033 is to be seen as an entirely hypothetical "back-formation" even within the fictional context? But if we accept the new scenario set out in the }{\i\fs20\lang1033 Silmarillion}{\fs20\lang1033 Appendix \endash that the first element of }{ \b\fs20\lang1033 Silmaril}{\fs20\lang1033 represents a contemporary noun }{\b\fs20\lang1033 silima}{\fs20\lang1033 coined by F\'ebanor himself \endash we can appeal to secondary syncope regarding its reduction to }{\b\fs20\lang1033 silma}{ \fs20\lang1033 - in a compound. Thus the linguistic "anachronisms" are avoided. \par }{\lang1033 \par In the dialect of the Noldor, unlike that of the Vanyar, }{\b\lang1033 nd}{\lang1033 is simplified to }{\b\lang1033 n}{\lang1033 before }{\b\lang1033 y}{\lang1033 (WJ:360, 361): \par \par \tab (\'a4}{\i\lang1033 k}{\i\lang1033\super w}{\i\lang1033 endj\'e2}{\lang1033 >) }{\b\lang1033 Quendya}{\lang1033 > }{\b\lang1033 Quenya}{\lang1033 \par \par }{\fs20\lang1033 NOTE: A double form like }{\b\fs20\lang1033 endya}{\fs20\lang1033 , }{\b\fs20\lang1033 enya}{\fs20\lang1033 "middle" (Etym, entry }{\i\fs20\lang1033 \'c9NED}{\fs20\lang1033 ) may be taken as the Vanyarin/original form and the later Noldorin form, respectively. However, this would mean re-interpreting the }{\i\fs20\lang1033 Etymologies}{\fs20\lang1033 material in the light of Tolkien's later ideas, for when he wrote Etym, there was no such thing as a Noldorin dialect of Quenya (the Noldor were still conceived as speaking something very much like the language Tolkien later called }{\i\fs20\lang1033 Sin darin}{\fs20\lang1033 ). \endash In the case of }{\b\fs20\lang1033 indyo}{\fs20\lang1033 "grandson" (entry }{\i\fs20\lang1033 \'d1GY\'d4}{\fs20\lang1033 ), no later form *}{\b\fs20\lang1033 inyo}{\fs20\lang1033 is mentioned. \par }{\lang1033 \par }{\i\fs32\lang1033 Changes producing }{\fs32\lang1033 EXILIC QUENYA \par }{\lang1033 \par The noon-time of Valinor is drawing to an end. Soon, F\'ebanor will lead the Noldor into exile, and Quenya will become known in Middle- earth as well as in Aman. In Middle-earth it will become a language of lore, largely shielded from further changes by the fact that it is learnt from writing by each new generation. But before the language to some extent becomes fixed like this, a few mor e changes will occur \endash some of them greatly controversial, intertwined with Valinorean politics! They will define the shape of }{\i\lang1033 Exilic Quenya}{\lang1033 , by definition the only form of Quenya known in Middle-earth. These changes also defined the }{\i\lang1033 Noldorin}{\lang1033 dialect of Que nya, as opposed to the Vanyarin dialect (though it is possible that Vanyarin did not share with Noldorin all the changes so far listed). \par \par An important source of information about the changes affecting Exilic Quenya is the list of Tengwar names in LotR Appe ndix E. The original name of all the relevant Tengwar includes the sound this letter originally represented. But its precise pronunciation was later altered, as when }{\b\lang1033 \'fe\'fal\'eb}{\lang1033 (in LotR Appendix E spelt "th\'fa le") came to be pronounced }{\b\lang1033 s\'fal\'eb}{\lang1033 instead. In some cases, a Tengwa could be renamed altogether because of the phonological changes.}{\b\lang1033 \par }{\lang1033 \par Wrote Tolkien regarding the Tengwar names, "Where there are variants this is due to the names being given before certain changes that affected Quenya as spoken by the Exiles." This should not be taken to mean that all of these changes necessarily occurred }{\i\lang1033 during}{\lang1033 the Exile, after the Noldor had actually left Valinor. We know that Tolkien imagined the shift reflected in the change }{\b\lang1033 \'fe\'fal\'eb }{\lang1033 > }{\b\lang1033 s\'fal\'eb}{\lang1033 to have occurred in Valinor during F\'ebanor's lifetime. The shift }{\b\lang1033 \'fe}{\lang1033 > }{\b\lang1033 s}{\lang1033 is indeed the most celebrated of all Quenya sound-changes, and Tolkien was able to develop an entire story around it: It was underway in Valinor during F\'ebanor's lifetime, but he and all the loremasters deplored it, since "it cau sed a great deal of homophony, and confused the derivatives of originally quite distinct stems." (VT41:8 \endash actually there are very few words in our corpus that would merge in form; *}{\b\lang1033 \'feamna}{\lang1033 > }{\b\lang1033 samna}{ \lang1033 "wooden post" vs. original }{\b\lang1033 samna}{\lang1033 "diphthong" may be one example, but these are hardly difficult to distinguish in context!) However, F\'eb anor's "advice" was presented with such arrogance and haughtiness that he rather accelerated the merger of }{\b\lang1033 \'fe}{\lang1033 and }{\b\lang1033 s}{\lang1033 : "Into the strife and confusion of loy alties in that time this seemingly trivial matter, the change of }{\i\lang1033 \'fe}{\lang1033 to }{\i\lang1033 s}{\lang1033 , was caught up to its embitterment, and to lasting detriment to the Quenya tongue. Had peace been maintained there can be no doubt that the advice of F\'ebanor, with which all the other l oremasters privately or openly agreed, would have prevailed. But an opinion in which he was certainly right was rejected because of the follies and evil deeds into which he was later led. He made it a personal matter; he and his sons adhered to }{ \i\lang1033 \'fe}{\lang1033 , and they demanded that all those who were sincere in their support should do the same. Therefore those who resented his arrogance, and still more those whose support later turned to hatred, rejected his shibboleth" (PM:333-334). Hence: \par \par \tab }{\b\lang1033 \'fe\'fal\'eb}{\lang1033 > }{\b\lang1033 s\'fal\'eb}{\lang1033 "spirit" \par \tab }{\b\lang1033 \'feerind\'eb}{\lang1033 > }{\b\lang1033 Serind\'eb}{\lang1033 (the name of F\'ebanor's mother; the change annoyed both mother and son, PM:333) \par \tab }{\b\lang1033 \'feind\'eb}{\lang1033 > }{\b\lang1033 sind\'eb}{\lang1033 "grey" \par \tab }{\b\lang1033 \'feelma}{\lang1033 > *}{\b\lang1033 selma}{\lang1033 "will" (WJ:319) \par \tab *}{\b\lang1033 chi\'few\'eb}{\lang1033 > *}{\b\lang1033 chisw\'eb}{\lang1033 "fog" (later }{\b\lang1033 hisw\'eb}{\lang1033 ) \par \tab }{\b\lang1033 ri\'feil}{\lang1033 > *}{\b\lang1033 risil}{\lang1033 "circle" \par \tab }{\b\lang1033 nau\'fe\'eb}{\lang1033 > }{\b\lang1033 naus\'eb}{\lang1033 "imagination"}{\b\fs20\lang1033 \par }{\lang1033 \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 According to the }{\i\lang1033 Etymologies}{\lang1033 , final -}{\b\lang1033 s}{\lang1033 might also represent earlier -}{\b\lang1033 \'fe}{\lang1033 sometimes: The name }{\b\lang1033 Tulkas}{ \lang1033 is apparently derived from }{\b\lang1033 Tulkat}{\b\lang1033\super h}{\b\lang1033 o}{\lang1033 in LR:395 s.v. }{\i\lang1033 TULUK}{\lang1033 ; if so its older Quenya form might be *}{\b\lang1033 Tulka\'fe}{\lang1033 . (However, Tolkien would later explain the name of this Vala as a borrowing from Valarin: WJ:499, 404.) \par In Etym, there even seems to be an example of double }{\b\lang1033 ss}{\lang1033 representing earlier double *}{\b\lang1033 \'fe\'fe}{\lang1033 ; the word }{\b\lang1033 nessa}{\lang1033 "young" is ultimately derived from \'a4}{\b\lang1033 net}{ \b\lang1033\super h}{\b\lang1033 -r\'e2}{\lang1033 , presumably via *}{\b\lang1033 ne\'fer\'e2}{\lang1033 > *}{\b\lang1033 ne\'fe\'fea}{\lang1033 . In the }{\i\lang1033 Etymologies}{\lang1033 , this is presented as the origin of the name of the Vali\'eb }{\i\lang1033 Nessa}{\lang1033 . But in a later source Tolkien suggested that this name comes from \'a4}{\b\lang1033 neres\'e2}{\lang1033 instead, or explained it as a borrowing from Valarin (WJ:416, 404). However, this \'a4}{\b\lang1033 neres\'e2}{ \lang1033 does not mean "young" as did \'a4}{\b\lang1033 net}{\b\lang1033\super h}{\b\lang1033 -r\'e2}{\lang1033 , so perhaps the two must now be considered wholly separate words, the later forms }{\b\lang1033 Nessa}{\lang1033 and *}{\b\lang1033 ne\'fe \'fea}{\lang1033 coinciding only in late Quenya, because of the }{\b\lang1033 \'fe}{\lang1033 /}{\b\lang1033 s}{\lang1033 merger. \par The change }{\b\lang1033 \'fe}{\lang1033 > }{\b\lang1033 s}{\lang1033 eventually became universal in Noldorin Quenya (in Middle-earth, even the sons of F\'eb anor seem to have adopted the new pronunciation after his death), but the distinction was never dropped in writing (PM:332). According to LotR Appendix E, the letter now called }{\b\lang1033 s\'fal\'eb}{\lang1033 continued to be used for }{\b\lang1033 s} {\lang1033 derived from earlier }{\b\lang1033 \'fe}{\lang1033 \endash though we have an example of Tengwar calligraphy where Tolkien himself seems to have forgotten this, using the normal }{\b\lang1033 s}{\lang1033 -Tengwa (}{\b\lang1033 silm\'eb}{ \lang1033 ) even where he "ought" to have used }{\b\lang1033 s\'fal\'eb}{\lang1033 . \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par }{\fs20\lang1033 NOTE 1: As discussed earlier, }{\b\fs20\lang1033 \'fe}{\fs20\lang1033 was itself derived from the aspirate }{\b\fs20\lang1033 t}{\b\fs20\lang1033\super h}{\fs20\lang1033 , which in turn was sometimes original and sometimes changed from the initial cluster }{\b\fs20\lang1033 st}{\fs20\lang1033 -. However, it seems that when Tolkien wrote the }{\i\fs20\lang1033 Etymologies}{\fs20\lang1033 , he imagined the change from }{\b\fs20\lang1033 \'fe}{ \fs20\lang1033 to }{\b\fs20\lang1033 s}{\fs20\lang1033 to have occurred in the remote past \endash not as late as during F\'ebanor's lifetime. This is evident from a note in the entry }{\i\fs20\lang1033 STAR}{\fs20\lang1033 , where Tolkien observed that the "Noldorin" word }{\b\fs20\lang1033 tharn}{\fs20\lang1033 (meaning "sapless") was "not in Q" \endash that is, had no Quenya cognate \endash "since it would coalesce with }{\i\fs20\lang1033 sarn\'e2}{\fs20\lang1033 ". }{ \b\fs20\lang1033 Tharn}{\fs20\lang1033 comes from *}{\b\fs20\lang1033 starn\'e2}{\fs20\lang1033 (the "Old Noldorin" form }{\b\fs20\lang1033 st}{\b\fs20\lang1033\super h}{\b\fs20\lang1033 arna}{\fs20\lang1033 is given), whereas \'a4}{\b\fs20\lang1033 sarn\'e2}{\fs20\lang1033 comes from a distinct root }{\i\fs20\lang1033 SAR}{\fs20\lang1033 and produces the Quenya adjective }{\b\fs20\lang1033 sarna}{\fs20\lang1033 "of stone". According to the later scenario, *}{\b\fs20\lang1033 starn\'e2}{ \fs20\lang1033 would become *}{\b\fs20\lang1033 t}{\b\fs20\lang1033\super h}{\b\fs20\lang1033 arn\'e2}{\fs20\lang1033 and then *}{\b\fs20\lang1033 \'fearna}{\fs20\lang1033 , remaining distinct from (\'a4}{\b\fs20\lang1033 sarn\'e2}{\fs20\lang1033 >) }{ \b\fs20\lang1033 sarna}{\fs20\lang1033 until the merger }{\b\fs20\lang1033 \'fe}{\fs20\lang1033 /}{\b\fs20\lang1033 s}{\fs20\lang1033 occurred. Since Tolkien did not say that the words }{\i\fs20\lang1033 did}{\fs20\lang1033 coalesce so that one of them was suppressed, but implied that one had never been in Quenya because it }{\i\fs20\lang1033 would}{\fs20\lang1033 have coalesced with the other, we can infer that the merger }{\b\fs20\lang1033 \'fe}{\fs20\lang1033 /}{ \b\fs20\lang1033 s}{\fs20\lang1033 was originally imagined to have taken place already in the pre-record period. One remark in the entry }{\i\fs20\lang1033 SIL}{\fs20\lang1033 , to the effect that this stem and }{\i\fs20\lang1033 T}{ \i\fs20\lang1033\super H}{\i\fs20\lang1033 IL}{\fs20\lang1033 "in Q cannot be distinguished normally", points to the same conclusion. Later, Tolkien moved the merger }{\b\fs20\lang1033 \'fe}{\fs20\lang1033 /}{\b\fs20\lang1033 s}{\fs20\lang1033 forward in the imagined time-line and had it occurring in F\'ebanor's day. \par \par NOTE 2: In one word, }{\b\fs20\lang1033 \'fe}{\fs20\lang1033 may seem to behave quite differently. The Quenya form corresponding to Sindarin }{\i\fs20\lang1033 Morgoth}{\fs20\lang1033 is said to be }{\b\fs20\lang1033 Mori\'f1gotto}{\fs20\lang1033 , yet it is also given as "Mori\'f1gotho", presumably = }{\b\fs20\lang1033 Mori\'f1go\'feo}{\fs20\lang1033 . The latter form is explicitly said to be the one originally used by F\'ebanor (MR:294), whereas }{\b\fs20\lang1033 Mori\'f1gotto}{\fs20\lang1033 is merely called a "more ancient form" than Morgoth (MR:194). Are we to assume that }{\b\fs20\lang1033 Mori\'f1go\'feo}{\fs20\lang1033 later }{\i\fs20\lang1033 became}{\fs20\lang1033 }{\b\fs20\lang1033 Mori\'f1gotto}{\fs20\lang1033 (instead of **}{ \b\fs20\lang1033 Mori\'f1goso}{\fs20\lang1033 as we might expect), so that we must reckon with a strange development }{\b\fs20\lang1033 \'fe}{\fs20\lang1033 > }{\b\fs20\lang1033 tt}{\fs20\lang1033 ? Or did Tolkien simply change his mind about what the original form of the name Morgoth really was, so that one is not to be derived from the other? There are a few other }{\i\fs20\lang1033 possible}{\fs20\lang1033 examples of a development }{ \b\fs20\lang1033 \'fe}{\fs20\lang1033 > }{\b\fs20\lang1033 tt}{\fs20\lang1033 . The words }{\b\fs20\lang1033 latta}{\fs20\lang1033 and }{\b\fs20\lang1033 motto}{\fs20\lang1033 "blot" are derived from roots }{\i\fs20\lang1033 LAT}{ \i\fs20\lang1033\super H}{\fs20\lang1033 , }{\i\fs20\lang1033 MBOT}{\i\fs20\lang1033\super H}{\fs20\lang1033 . If the most primitive forms were *}{\b\fs20\lang1033 lat}{\b\fs20\lang1033\super h}{\b\fs20\lang1033 \'e2}{\fs20\lang1033 and *}{ \b\fs20\lang1033 mbot}{\b\fs20\lang1033\super h}{\b\fs20\lang1033 \'f4}{\fs20\lang1033 , these words might have appeared as *}{\b\fs20\lang1033 la\'fea}{\fs20\lang1033 and *}{\b\fs20\lang1033 mo\'feo}{\fs20\lang1033 in Valinorean Quenya, becoming }{ \b\fs20\lang1033 latta}{\fs20\lang1033 and }{\b\fs20\lang1033 motto}{\fs20\lang1033 in the Exilic dialect. (Interestingly, this would mean that *}{\b\fs20\lang1033 la\'fea}{\fs20\lang1033 > }{\b\fs20\lang1033 latta}{\fs20\lang1033 "strap" was not originally a homophone of }{\b\fs20\lang1033 latta}{\fs20\lang1033 "pit", itself < \'a4}{\b\fs20\lang1033 datt\'e2}{\fs20\lang1033 and therefore showing }{\b\fs20\lang1033 tt}{\fs20\lang1033 all along.) However, it is eminently possible that the double }{\b\fs20\lang1033 tt}{\fs20\lang1033 of }{\b\fs20\lang1033 latta}{\fs20\lang1033 and }{\b\fs20\lang1033 motto}{\fs20\lang1033 is "original", these words descending from *}{ \b\fs20\lang1033 latt\'e2}{\fs20\lang1033 , *}{\b\fs20\lang1033 mbott\'f4}{\fs20\lang1033 despite the final aspirate of the roots }{\i\fs20\lang1033 LAT}{\i\fs20\lang1033\super H}{\fs20\lang1033 , }{\i\fs20\lang1033 MBOT}{\i\fs20\lang1033\super H}{ \fs20\lang1033 . (Double }{\i\fs20\lang1033 aspirated}{\fs20\lang1033 consonants were evidently not tolerated in the primitive language; cf. "Noldorin" }{\b\fs20\lang1033 lhoch}{\fs20\lang1033 "ringlet" being derived from \'a4}{\b\fs20\lang1033 lokko}{ \fs20\lang1033 rather than **}{\b\fs20\lang1033 lo}{\fs20\lang1033 (}{\b\fs20\lang1033 k}{\fs20\lang1033 )}{\b\fs20\lang1033 k}{\b\fs20\lang1033\super h}{\b\fs20\lang1033 o}{\fs20\lang1033 even though the root is }{\i\fs20\lang1033 LOK}{ \i\fs20\lang1033\super H}{\fs20\lang1033 .) All things considered, there is little solid evidence for a development }{\b\fs20\lang1033 \'fe}{\fs20\lang1033 > }{\b\fs20\lang1033 tt}{\fs20\lang1033 in Quenya, though it cannot be wholly ruled out. The regular development in any case remains }{\b\fs20\lang1033 \'fe}{\fs20\lang1033 > }{\b\fs20\lang1033 s}{\fs20\lang1033 . \par }{\lang1033 \par It should be noted that the change from }{\b\lang1033 \'fe}{\lang1033 to }{\b\lang1033 s}{\lang1033 did }{\i\lang1033 not}{\lang1033 occur in the Vanyarin dialect of Quenya, but only in the Noldorin dialect (thus }{\b\lang1033 sind\'eb}{\lang1033 rather than }{\b\lang1033 \'feind\'eb}{\lang1033 is said to be a "\'d1[oldorin] dialect" form in WJ:384). When Indis, herself a Vanya, married Finw\'eb, she made a conscious decision to change her pronunciation: "I have joined the people of the \'d1 oldor, and I will speak as they do" (PM:336). F\'ebanor was not amused. \par \par Another change that seems to have occurred already in Valinor, but very late, was the shift }{\b\lang1033 z}{\lang1033 > }{\b\lang1033 r}{\lang1033 in nearly all positions. (One exception: according to WJ:413, medial }{\b\lang1033 z}{\lang1033 escaped this change if there already was an }{\b\lang1033 r}{\lang1033 in the same or an adjacent syllable. If }{\b\lang1033 z}{\lang1033 thus did not become }{\b\lang1033 r}{\lang1033 , it apparently became }{\b\lang1033 s}{\lang1033 instead, for }{ \b\lang1033 z}{\lang1033 was never used in Exilic Quenya.) The Tengwa that F\'ebanor had originally named }{\b\lang1033 \'e1z\'eb}{\lang1033 ("sunshine") thus came to be called }{\b\lang1033 \'e1r\'eb}{\lang1033 , but since there already existed another letter for }{\b\lang1033 r}{\lang1033 , }{\b\lang1033 \'e1r\'eb}{\lang1033 ceased to have any clear function: It would seem that the distinction between original }{\b\lang1033 r}{\lang1033 and }{\b\lang1033 r}{ \lang1033 changed from }{\b\lang1033 z}{\lang1033 was not upheld in writing. (In Middle-earth, the now "vacant" letter }{\b\lang1033 \'e1r\'eb}{\lang1033 was eventually assigned a new value, double }{\b\lang1033 ss}{\lang1033 , and the letter itself was renamed }{\b\lang1033 ess\'eb}{\lang1033 to reflect this.) As in the case of the shift }{\b\lang1033 \'fe}{\lang1033 > }{\b\lang1033 s}{\lang1033 , the Vanyar were conservative; Vanyarin never shifted }{\b\lang1033 z}{ \lang1033 to }{\b\lang1033 r}{\lang1033 , so the Vanyar would perceive this change as a dialectal feature of Noldorin Quenya. A place-name like }{\b\lang1033 Ezellohar}{\lang1033 indicates that }{\b\lang1033 z}{\lang1033 remained a living sound in Amanya Elvish, though the Noldor probably came to say *}{\b\lang1033 Erellohar}{\lang1033 instead. In WJ:363, the forms }{\b\lang1033 Aurel}{\lang1033 and }{\b\lang1033 Oarel}{\lang1033 (evidently Noldorin Quenya) are }{ \i\lang1033 contrasted}{\lang1033 with }{\b\lang1033 Auzel}{\lang1033 and }{\b\lang1033 Oazel}{\lang1033 in the Vanyarin dialect; these Vanyarin forms also reflect the older, common Vanya-Noldorin pronunciation. This also seems to confirm that the Noldorin change of }{\b\lang1033 z}{\lang1033 > }{\b\lang1033 r}{\lang1033 occurred already in Val inor, or there would be little point in using it to contrast the Vanyarin and Noldorin dialects of Quenya. \par \par The }{\b\lang1033 z}{\lang1033 's that now become }{\b\lang1033 r}{\lang1033 may derive from either }{\b\lang1033 d}{\lang1033 or }{\b\lang1033 s}{\lang1033 in the primitive language: \par \par \tab (*}{\i\lang1033 olos\'ee}{\lang1033 >) }{\b\lang1033 olozi}{\lang1033 > }{\b\lang1033 olori}{\lang1033 pl. "dreams" (sg. }{\i\lang1033 olos}{\lang1033 , UT:396) \par \tab (\'a4}{\i\lang1033 nidw\'f4}{\lang1033 > *}{\i\lang1033 nidw\'e2}{\lang1033 > *}{\i\lang1033 ni\'f0wa}{\lang1033 >) *}{\b\lang1033 nizwa}{\lang1033 > }{\b\lang1033 nirwa}{\lang1033 "cushion" \par }{\b\lang1033 \par }{\lang1033 With this change, the already very limited number of consonants that Quenya can have finally is further reduced by one; after this change, only -}{\b\lang1033 t}{\lang1033 , -}{\b\lang1033 l}{\lang1033 , -}{\b\lang1033 n}{\lang1033 , -}{ \b\lang1033 s}{\lang1033 and -}{\b\lang1033 r}{\lang1033 (original or derived from -}{\b\lang1033 z}{\lang1033 ) remain. \par \par The letter }{\b\lang1033 harma}{\lang1033 F\'ebanor originally meant to represent }{\i\lang1033 ach-Laut}{\lang1033 , the spirant }{\b\lang1033 ch}{\lang1033 , and undoubtedly the original pronunciation of its name was *}{\b\lang1033 charma}{\lang1033 . However, LotR Appendix E records how }{\b\lang1033 ch}{\lang1033 "became breath }{\i\lang1033 h}{\lang1033 initially (though remaining [}{\b\lang1033 ch}{\lang1033 ] medially)". Hence: \par \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 k}{\i\lang1033\super h}{\i\lang1033 er\'fb}{\lang1033 > *}{\i\lang1033 cher\'fb}{\lang1033 >) *}{\b\lang1033 cheru}{\lang1033 > }{\b\lang1033 heru}{\lang1033 "lord" (cf. Letters:282) \par (\'a4}{\i\lang1033 k}{\i\lang1033\super h}{\i\lang1033 ugan}{\lang1033 > *}{\i\lang1033 chu3an}{\lang1033 >) *}{\b\lang1033 chuan}{\lang1033 > }{\b\lang1033 huan}{\lang1033 "hound" \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \tab (\'a4}{\i\lang1033 skarw\'ea}{\lang1033 > *}{\i\lang1033 k}{\i\lang1033\super h}{\i\lang1033 arw\'ea}{\lang1033 > *}{\i\lang1033 charw\'ea}{\lang1033 >) *}{\b\lang1033 charw\'eb}{\lang1033 > } {\b\lang1033 harw\'eb}{\lang1033 "wound" \par \par The early Quenya forms in }{\b\lang1033 ch}{\lang1033 - are implied to have existed, but no such form is directly quoted in the published material, so we must asterisk them. In the }{\i\lang1033 Etymologies}{\lang1033 , only the Exilic Quenya forms in }{ \b\lang1033 h}{\lang1033 - are recorded (that is, what would be Exilic Quenya forms in Tolkien's later scena rio). It may be that Tolkien originally held this change to have occurred further back in history, in the pre-record period. But in the linguistic scenario as it appears in the LotR appendices, this weakening of initial }{\b\lang1033 ch}{\lang1033 to }{ \b\lang1033 h}{\lang1033 can perhaps be ascribed to Sindarin influence on Quenya as spoken by the Exiles in Middle-earth: Sindarin had also turned initial }{\b\lang1033 ch}{\lang1033 into }{\b\lang1033 h}{\lang1033 . However, Tolkien noted that this change occurred in "all Eldarin tongues" (VT41:9), which would include the Telerin of Aman; hence this change was not unheard of in Valinor either. Perhaps it was already beginning in Valinor shortly before the Exile, but occurring too late to be reflected in the original F\'ebanorian orthography. \par \par According to LotR Appendix E, the sound }{\b\lang1033 ch}{\lang1033 remained }{\i\lang1033 medially}{\lang1033 even after it had become breath }{\b\lang1033 h}{\lang1033 initially. The Tengwa }{\b\lang1033 harma}{\lang1033 was therefore renamed }{ \b\lang1033 aha}{\lang1033 ("wrath"), since in this word the middle consonant was still pronounced }{\b\lang1033 ch}{\lang1033 . But later, intervocalic }{\b\lang1033 ch}{\lang1033 was also weakened; in a post-LotR source, Tolkien wrote: "In Quenya and Telerin medial [}{\b\lang1033 ch}{\lang1033 ] eventually became }{\i\lang1033 h}{\lang1033 also in most cases" (VT41:9). Perhaps this is why *}{\b\lang1033 acha}{\lang1033 , as it must originally have been pronounced, is nonetheless spelt }{\b\lang1033 aha}{\lang1033 in Appendix E: In the end, medial }{\b\lang1033 ch}{\lang1033 also came to be pronounced }{\b\lang1033 h}{\lang1033 \endash "in most cases". The ex ceptions are probably the combinations normally spelt }{\b\lang1033 aht}{\lang1033 , }{\b\lang1033 oht}{\lang1033 , }{\b\lang1033 uht}{\lang1033 , where orthographic "h" ever remained a full spirant }{\b\lang1033 ch}{\lang1033 . In the combinations }{ \b\lang1033 eht}{\lang1033 , }{\b\lang1033 iht}{\lang1033 it turned into }{\i\lang1033 ich-Laut}{\lang1033 , but in these groups it was quite possibly still represented by the letter }{\b\lang1033 aha}{\lang1033 in Tengwar orthography \endash though we have no examples. \par \par When initial }{\b\lang1033 ch}{\lang1033 became }{\b\lang1033 h}{\lang1033 , it seems that it }{\i\lang1033 merged}{\lang1033 with the initial breath-}{\b\lang1033 h}{\lang1033 that already existed in Valinorean Quenya. This breath-}{\b\lang1033 h}{ \lang1033 was in most cases derived from primitive }{\b\lang1033 3}{\lang1033 (or, if Tolkien was in that mood, it had been }{\b\lang1033 h}{\lang1033 already in the primitive language and never changed \endash e.g. Primitive Quendian \'a4}{ \b\lang1033 hekla }{\lang1033 "outcast" > Quenya }{\b\lang1033 hecil}{\lang1033 in WJ:361, 365). Exceptionally }{\b\lang1033 h}{\lang1033 was also derived (via }{\b\lang1033 hw}{\lang1033 -) from the aspirate }{\b\lang1033 p}{\b\lang1033\super h}{ \lang1033 before }{\b\lang1033 u}{\lang1033 , as in }{\b\lang1033 huin\'eb}{\lang1033 "gloom". Initial }{\b\lang1033 ch}{\lang1033 was on the other hand derived from the aspirate }{\b\lang1033 k}{\b\lang1033\super h}{\lang1033 - (original or changed from even older }{\b\lang1033 sk}{\lang1033 -). The change }{\b\lang1033 ch}{\lang1033 - > }{\b\lang1033 h}{\lang1033 - in Exilic Quenya lead to some cases of homophony. The word }{\b\lang1033 harw\'eb}{\lang1033 "treasury", representing primitive *}{\b\lang1033 3arw\'ea}{\lang1033 (root }{\i\lang1033 3AR}{\lang1033 , LR:360) would have had breath }{\b\lang1033 h}{\lang1033 already in Valinorean Quenya. But the word }{\b\lang1033 harw\'eb}{\lang1033 "wound" was not originally a homophone; this comes from primitive \'a4}{\b\lang1033 skarw\'ea}{\lang1033 , which (after first becoming *}{\b\lang1033 k}{\b\lang1033\super h}{\b\lang1033 arw\'ea}{\lang1033 ) yielded Old Quenya *}{\b\lang1033 charw\'eb}{ \lang1033 . Only with the change }{\b\lang1033 ch}{\lang1033 > }{\b\lang1033 h}{\lang1033 did the words coincide in form. \par \par }{\fs20\lang1033 NOTE: There is one problem; in the }{\i\fs20\lang1033 Etymologies}{\fs20\lang1033 Tolkien derived the word }{\b\fs20\lang1033 harma}{\fs20\lang1033 from a root }{\i\fs20\lang1033 3AR}{\fs20\lang1033 , and then it should have had breath }{\b\fs20\lang1033 h}{\fs20\lang1033 already in Valinorean Quenya. Yet Tolkien picked this word as the name of the Tengwa originally denoting }{\b\fs20\lang1033 ch}{\fs20\lang1033 , which would imply that it was earlier pronounced *}{\b\fs20\lang1033 charma}{\fs20\lang1033 instead. We }{\i\fs20\lang1033 could}{\fs20\lang1033 assume that earlier breath }{\b\fs20\lang1033 h}{\fs20\lang1033 merged with }{\b\fs20\lang1033 ch}{\fs20\lang1033 in Valinorean Quenya, later reverting to }{\b\fs20\lang1033 h }{\fs20\lang1033 in Exilic Quenya (when original }{\b\fs20\lang1033 ch}{\fs20\lang1033 also became }{\b\fs20\lang1033 h}{\fs20\lang1033 ). However, LotR Appendix E indicates that even the original F\'ebanorian system had distinct signs for breath }{ \b\fs20\lang1033 h}{\fs20\lang1033 (originally represented by the sign }{\b\fs20\lang1033 halla}{\fs20\lang1033 , a raised stem with no bow) and the spirant }{\b\fs20\lang1033 ch}{\fs20\lang1033 (represented by }{\b\fs20\lang1033 harma}{\fs20\lang1033 ). This would suggest that }{\b\fs20\lang1033 h}{\fs20\lang1033 and }{\b\fs20\lang1033 ch}{\fs20\lang1033 coexisted in Valinorean Quenya. Perhaps Tolkien simply changed his mind about the derivation of }{\b\fs20\lang1033 harma}{\fs20\lang1033 "treasure", deriving it from a stem *}{\i\fs20\lang1033 K}{\i\fs20\lang1033\super H}{\i\fs20\lang1033 AR}{\fs20\lang1033 instead of }{\i\fs20\lang1033 3AR}{\fs20\lang1033 , so that it would originally have had initial }{\b\fs20\lang1033 ch}{ \fs20\lang1033 . If the word }{\b\fs20\lang1033 halla}{\fs20\lang1033 ("tall") contained breath }{\b\fs20\lang1033 h}{\fs20\lang1033 already in Valinorean Quenya, Tolkien may have altered its derivation as well: It is not actually listed in the }{ \i\fs20\lang1033 Etymologies}{\fs20\lang1033 , but it would probably have to be referred to the root }{\i\fs20\lang1033 K}{\i\fs20\lang1033\super H}{\i\fs20\lang1033 AL}{\i\fs20\lang1033\super 2}{\fs20\lang1033 as a cognate of "Noldorin"/Sindarin }{ \b\fs20\lang1033 hall}{\fs20\lang1033 "exalted, high" (older form }{\b\fs20\lang1033 k}{\b\fs20\lang1033\super h}{\b\fs20\lang1033 alla}{\fs20\lang1033 given). But }{\b\fs20\lang1033 halla}{\fs20\lang1033 with original breath }{\b\fs20\lang1033 h}{ \fs20\lang1033 would require a stem *}{\i\fs20\lang1033 3AL}{\fs20\lang1033 (or, *}{\i\fs20\lang1033 HAL}{\fs20\lang1033 ). Conceivably, }{\b\fs20\lang1033 halla}{\fs20\lang1033 "tall" simply describes the form of the letter (a raised stem) rather than exemplifying the sound it denoted: If the root is to remain }{\i\fs20\lang1033 K}{\i\fs20\lang1033\super H}{\i\fs20\lang1033 AL}{\fs20\lang1033 , this word }{\i\fs20\lang1033 must }{\fs20\lang1033 have been pronounced *}{\b\fs20\lang1033 challa}{\fs20\lang1033 in Valinorean Quenya.}{\lang1033 \par \par The name of the Tengwa }{\b\lang1033 wilya}{\lang1033 was pronounced }{\b\lang1033 vilya}{\lang1033 in Exilic Quenya. It would seem, then, that initial }{\b\lang1033 w}{\lang1033 became }{\b\lang1033 v}{\lang1033 (merging with }{\b\lang1033 v}{ \lang1033 derived from earlier }{\b\lang1033 b}{\lang1033 ). Above we have discussed how }{\b\lang1033 w}{\lang1033 became }{\b\lang1033 v}{\lang1033 }{\i\lang1033 medially}{\lang1033 , and one may well ask why we cannot let initial and post-vocalic }{ \b\lang1033 w}{\lang1033 become }{\b\lang1033 v}{\lang1033 at the same time \endash but there is good reason to believe that post-vocalic }{\b\lang1033 w}{\lang1033 normally became }{\b\lang1033 v}{\lang1033 much earlier t han the initial shift occurred. Consider this: The loss of }{\b\lang1033 3}{\lang1033 must be placed very }{\i\lang1033 early}{\lang1033 in the historical period, since it was only by the "tradition among the loremasters" anyone knew what sound had once been represented by the initial letter of such a R\'familian spelling as }{\b\lang1033 3alda}{\lang1033 (VT39:7). Yet the medial shift } {\b\lang1033 w}{\lang1033 > }{\b\lang1033 v}{\lang1033 must }{\i\lang1033 predate}{\lang1033 the loss of }{\b\lang1033 3}{\lang1033 : In the case of (\'a4}{\i\lang1033 jagw\'ea}{\lang1033 >) *}{\b\lang1033 ya3w\'eb}{\lang1033 > }{\b\lang1033 y\'e1w \'eb}{\lang1033 "ravine" a }{\b\lang1033 3}{\lang1033 that was later lost hindered the development of post-vocalic }{\b\lang1033 w}{\lang1033 to }{\b\lang1033 v}{\lang1033 . A later shift }{\b\lang1033 w}{\lang1033 > }{\b\lang1033 v}{\lang1033 , }{ \i\lang1033 after}{\lang1033 the loss of }{\b\lang1033 3}{\lang1033 , would have turned *}{\b\lang1033 ya3w\'eb}{\lang1033 > }{\b\lang1033 y\'e1w\'eb}{\lang1033 into }{\b\lang1033 y\'e1v\'eb}{\lang1033 (producing a homophone with the word for "fruit", itself < *}{\b\lang1033 j\'e2b\'ea}{\lang1033 ). \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 So the evidence might seem to be contradictory: One line of reasoning leads to the conclusion that the shift }{\b\lang1033 w}{\lang1033 > }{\b\lang1033 v}{\lang1033 cannot be later than the earl iest parts of recorded history, whereas Tolkien elsewhere seems to imply that }{\b\lang1033 w}{\lang1033 and }{\b\lang1033 v}{\lang1033 remained fully distinct at the time F\'ebanor devised the Tengwar \endash the sounds merging only in Exilic Quenya. The only way of reconciling all the evidence is to distinguish between a shift of }{\i\lang1033 post-vocalic}{\lang1033 }{\b\lang1033 w}{\lang1033 > }{\b\lang1033 v}{\lang1033 , which took place very early before the loss of }{\b\lang1033 3}{\lang1033 (and which *}{\b\lang1033 ya3w\'eb}{\lang1033 > }{\b\lang1033 y\'e1w\'eb}{\lang1033 thus escaped), and an }{\i\lang1033 initial}{\lang1033 shift }{\b\lang1033 w}{\lang1033 > }{\b\lang1033 v}{\lang1033 , which took place only }{\i\lang1033 after}{\lang1033 F\'ebanor invented the Tengwar and is reflected in the change }{\b\lang1033 wilya}{\lang1033 > }{\b\lang1033 vilya}{\lang1033 . Since E\'e4 rendil is said to have named his ship }{\b\lang1033 Wingal\'f3t\'eb}{\lang1033 or }{\b\lang1033 Wingel\'f3t\'eb}{\lang1033 (PM:370, 371) with initial }{\b\lang1033 w}{\lang1033 still intact, this change may have occurred after the end of the First Age. It would seem, then, that changes like the following happened among the Exiles in the Second Age or later: \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\b\lang1033 w\'e9ra}{\lang1033 > }{\b\lang1033 v\'e9ra}{\lang1033 "private, personal, own" (PM:340, where it is made clear that this change occurred after the stage Tolkien called "Old Quenya") \par }{\b\lang1033 waia}{\lang1033 "envelope" > }{\b\lang1033 vaia \par w\'e1n}{\lang1033 > }{\b\lang1033 v\'e1n}{\lang1033 "goose" \par }{\b\lang1033 waiwa}{\lang1033 > }{\b\lang1033 vaiwa}{\lang1033 "wind" \par }{\b\lang1033 wend\'eb}{\lang1033 > }{\b\lang1033 vend\'eb}{\lang1033 "maiden" \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par Where Tolkien quotes Quenya forms in }{\b\lang1033 w}{\lang1033 - only, as he sometimes does, we must assume that he intends them to be "Old Quenya", First Age Quenya. The }{\i\lang1033 Etymologies}{\lang1033 is quite inconsistent in this regard: Quenya words derived from bases in }{\i\lang1033 W}{\lang1033 - are sometimes quoted with initial }{\b\lang1033 w}{\lang1033 -, sometimes with initial }{\b\lang1033 v}{\lang1033 -, and sometimes both. All contradictions are resolved when we recognize the forms in }{\b\lang1033 w}{\lang1033 - as Old Quenya, whereas the forms in }{\b\lang1033 v}{\lang1033 - are later. (In one case, }{\b\lang1033 vilin}{\lang1033 "I fly" derived from the root }{\i\lang1033 WIL}{\lang1033 , Christopher Tolkien interprets his father's notes to mean that he "changed" it to }{\b\lang1033 wilin}{\lang1033 : LR:399. However, there can be no question of }{\b\lang1033 vilin}{\lang1033 being }{\i\lang1033 rejected}{\lang1033 in favour of }{\b\lang1033 wilin}{\lang1033 ; for some reason, Tolkien decided to quote this verb as an Old Quenya form rather than the "modern Quenya" variant.) \endash It is not clear whether the distinction between }{\b\lang1033 w}{\lang1033 and }{\b\lang1033 v}{\lang1033 was upheld in Tengwar writing (the letter }{\b\lang1033 vilya}{\lang1033 still being used for initial }{\b\lang1033 v}{\lang1033 even after it had been changed from }{\b\lang1033 w}{\lang1033 ), or whether all initial }{\b\lang1033 v}{\lang1033 's came to be spelt with the letter }{\b\lang1033 vala}{\lang1033 regardless of derivation. (}{\b\lang1033 Vala}{\lang1033 originally represented the }{\b\lang1033 v}{\lang1033 's that were present already in Valinorean Quenya; when initial they were always derived from older }{\b\lang1033 b}{\lang1033 -.) In his RGEO Tengwar transcript of }{\i\lang1033 Nam\'e1ri\'eb}{ \lang1033 , Tolkien did use the letter }{\b\lang1033 vala}{\lang1033 to spell the initial }{\b\lang1033 v}{\lang1033 of the word }{\b\lang1033 vanwa}{\lang1033 "lost", though this is understood to represent primitive *}{\b\lang1033 wanw\'e2}{\lang1033 and must have been *}{\b\lang1033 wanwa}{\lang1033 in Old Quenya. (But then he also used }{\b\lang1033 silm\'eb}{\lang1033 to spell the }{\b\lang1033 s}{\lang1033 of }{\b\lang1033 h\'edsi\'eb}{\lang1033 "mist", though this }{\b\lang1033 s}{\lang1033 represents earlier }{\b\lang1033 \'fe}{\lang1033 and should have been spelt with the letter }{\b\lang1033 s\'fal\'eb}{\lang1033 according to the rules he set out elsewhere!) \par \par Another change implied by the Tengwar names involves initial }{\b\lang1033 \'f1}{\lang1033 becoming normal }{\b\lang1033 n}{\lang1033 (this includes labialized }{\b\lang1033 \'f1}{\b\lang1033\super w}{\lang1033 becoming }{\b\lang1033 n}{ \b\lang1033\super w}{\lang1033 , normally spelt }{\b\lang1033 nw}{\lang1033 in Roman Quenya orthography). In LotR Appendix E, Tolkien records how the Tengwa originally called "ngoldo" (alternative spelling of }{\b\lang1033 \'f1oldo}{\lang1033 ) had its name changed to }{\b\lang1033 noldo}{\lang1033 , and "ngwalme" (that is, }{\b\lang1033 \'f1walm\'eb}{\lang1033 ) similarly became }{\b\lang1033 nwalm\'eb}{\lang1033 . In the same Appendix, Tolkien noted that the sound of }{\b\lang1033 \'f1}{ \lang1033 (or "ng") "also occurred initially in Quenya, but has been transcribed }{\i\lang1033 n}{\lang1033 (as in }{\i\lang1033 Noldo}{\lang1033 ), according to the pronunciation of the Third Age". The }{\i\lang1033 Etymologies}{\lang1033 hints at a similar change. In the case of }{\b\lang1033 \'f1armo}{\lang1033 "wolf" from the root }{\i\lang1033 \'d1GAR(A)M}{\lang1033 , the later form }{\b\lang1033 narmo}{\lang1033 is also listed. In othe r cases, Tolkien is quite inconsistent about whether he lists the "older" form in }{\b\lang1033 \'f1}{\lang1033 - or the "later" form in }{\b\lang1033 n}{\lang1033 -; he may even be inconsistent within the same entry in Etym, as when the root }{ \i\lang1033 \'d1GAW}{\lang1033 yields }{\b\lang1033 \'f1auro}{\lang1033 "werewolf" next to }{\b\lang1033 naul\'eb}{\lang1033 "wolf-howl": When }{\b\lang1033 naul\'eb}{\lang1033 had reached this form (from older *}{\b\lang1033 \'f1aul\'eb}{\lang1033 ), }{\b\lang1033 \'f1auro}{\lang1033 had likewise become *}{\b\lang1033 nauro}{\lang1033 ! In the case of some words that are obviously related to }{\b\lang1033 \'f1walm\'eb}{\lang1033 > }{\b\lang1033 nwalm\'eb}{\lang1033 "torment" in LotR Appendix E, namely }{\b\lang1033 nwalya}{\lang1033 - "to pain" and }{\b\lang1033 nwalca}{\lang1033 "cruel" from the root }{\i\lang1033 \'d1G}{\i\lang1033\super W}{\i\lang1033 AL}{\lang1033 , only these "later" forms are recorded in Etym: Given the form of the root, the words must earlier have appeared as *}{\b\lang1033 \'f1walya}{\lang1033 - and *}{\b\lang1033 \'f1walca}{\lang1033 . \par \par In the Plotz letter, Tolkien mentioned certain changes that occurred before Quenya ceased to be a birth tongue amon g the Noldor, hence early in their exile. To the extent Quenya was spoken in Middle-earth, it reflected these changes: At the end of polysyllabic words, all long vowels became short. This affected the long final vowels Quenya had reacquired }{\i\lang1033 after}{\lang1033 the pre-historic shortening of the long final vowels of primitive Elvish. It included -}{\b\lang1033 \'ed}{\lang1033 as the nominative plural ending of words in -}{\b\lang1033 \'eb}{\lang1033 (this long -}{\b\lang1033 \'ed}{\lang1033 being reduced from -}{\b\lang1033 ei}{\lang1033 at an older stage) as well as -}{\b\lang1033 \'f3}{\lang1033 in the genitive forms of nouns in -}{\b\lang1033 a}{\lang1033 (reduced from earlier -}{\b\lang1033 ao}{\lang1033 ). Thus "Book Quenya" forms like }{\b\lang1033 lass\'ed}{\lang1033 "leaves" or }{\b\lang1033 ciry\'f3}{\lang1033 "of a ship" now became }{\b\lang1033 lassi}{\lang1033 , }{\b\lang1033 ciryo}{\lang1033 . This shortening of final vowels also resulted in the loss of distinct accusative forms, that in "Book Quenya" were formed by making a short final vowel long (e.g. }{\b\lang1033 ciry\'e1}{\lang1033 as the accusative of }{\b\lang1033 cirya}{\lang1033 "ship", but now }{\b\lang1033 ciry\'e1}{\lang1033 was reduced to }{\b\lang1033 cirya}{\lang1033 and merged with the nominative form). \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 Another change mentioned by Tolkien in the Plotz letter affects final -}{\b\lang1033 ai}{\lang1033 at the end of polysyllabic words. This diphthong first became a long -}{\b\lang1033 \'e6}{ \lang1033 (with the same quality as the vowel of English }{\i\lang1033 cat}{\lang1033 , but longer). Then, like other long vowels at the end of polysyllabic words, it was shortened and turned into a regular final -}{\b\lang1033 \'eb}{\lang1033 . In this way adjectives in -}{\b\lang1033 a}{\lang1033 came to have plural forms in -}{\b\lang1033 \'eb}{\lang1033 , e.g. }{\b\lang1033 lint\'eb}{\lang1033 as the plural form of *}{\b\lang1033 linta}{\lang1033 "swift", }{\b\lang1033 lint\'eb}{ \lang1033 representing earlier }{\b\lang1033 lintai}{\lang1033 . Adjectives in -}{\b\lang1033 \'eba}{\lang1033 , that originally had plural forms in -}{\b\lang1033 \'ebai}{\lang1033 , came to have plural forms in -}{\b\lang1033 i\'eb}{\lang1033 (obviously dissimilated from *-}{\b\lang1033 \'ebe}{\lang1033 ). \par In Plotz, Tolkien also mentioned another late c hange: In polysyllabic words, a long vowel occurring before a final consonant might also be shortened. Some of these long vowels had occurred in a final syllable since ancient times, others had ended up in the final syllable by the loss of some final vowe l. Shortened forms of names, such as }{\b\lang1033 Valinor}{\lang1033 (for *}{\b\lang1033 Valin\'f3r}{\lang1033 as a shorter form of }{\b\lang1033 Valin\'f3r\'eb}{\lang1033 ) would also belong on this list: \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par \tab \'a4}{\b\lang1033 an\'e2r}{\lang1033 - }{\b\lang1033 > Anar}{\lang1033 "sun" \par \tab \'a4}{\b\lang1033 ter\'ean}{\lang1033 }{\b\lang1033 > teren}{\lang1033 "slender" \par \tab }{\b\lang1033 (}{\lang1033 \'a4}{\i\lang1033 gaj\'e2r}{\lang1033 > *}{\i\lang1033 3ay\'e2r}{\lang1033 > *}{\i\lang1033 ay\'e2r}{\lang1033 >) }{\b\lang1033 \'eb\'e1r}{\lang1033 > }{\b\lang1033 \'ebar}{\lang1033 "sea" \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 (\'a4}{\i\lang1033 Mbelek\'f4ro}{\lang1033 > *}{\i\lang1033 Mbelek\'f4r}{\lang1033 >) *}{\b\lang1033 Melk\'f3r}{\lang1033 > }{\b\lang1033 Melkor}{\lang1033 \par (*}{\i\lang1033 Eruk}{\i\lang1033\super h}{\i\lang1033 \'een\'e2}{\lang1033 > *}{\i\lang1033 Eruk}{\i\lang1033\super h}{\i\lang1033 \'eena}{\lang1033 > *}{\i\lang1033 Eruk}{\i\lang1033\super h}{\i\lang1033 \'een }{\lang1033 > *}{\i\lang1033 Eruch\'edn}{ \lang1033 >) *}{\b\lang1033 Eruh\'edn}{\lang1033 > }{\b\lang1033 Eruhin}{\lang1033 "Child of Eru" (but still }{\b\lang1033 Eruh\'edn}{\lang1033 - before endings)}{\i\lang1033 \par }{\lang1033 *}{\b\lang1033 Valat\'e1r}{\lang1033 "Vala-king" > }{\b\lang1033 Valatar}{\lang1033 (but still }{\b\lang1033 Valat\'e1r}{\lang1033 - before endings) \par *}{\b\lang1033 Valin\'f3r}{\lang1033 }{\i\lang1033 (shortened form of }{\b\i\lang1033 Valin\'f3r\'eb}{\i\lang1033 ) }{\lang1033 > }{\b\lang1033 Valinor}{\lang1033 \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par Where a word occurs compounded or with endings, an originally long vowel might stay long because according to the normal stress rules it happens to receive the accent. The singular *}{\b\lang1033 Eruh\'edn}{\lang1033 > *}{\b\lang1033 Eruhin}{\lang1033 thus retains the plural }{\b\lang1033 Eruh\'edni}{\lang1033 with the long vowel intact. (But the plural of }{\b\lang1033 \'ebar}{\lang1033 is }{\b\lang1033 \'ebari}{\lang1033 rather than *}{\b\lang1033 \'eb\'e1ri}{\lang1033 ; as we noted above, three-syllable words tend to move the stress to the first syllable and shorten the vowel in the second-to-last syllable.) \par \par Quenya as spoken in Valinor possessed voiceless }{\b\lang1033 l}{\lang1033 and }{\b\lang1033 r}{\lang1033 , spelt }{\b\lang1033 hl}{\lang1033 and }{\b\lang1033 hr}{\lang1033 . LotR Appendix E indicates that by the Third Age, }{\b\lang1033 hl}{\lang1033 was "usually pronounced as [normal, voiced] }{\i\lang1033 l}{\lang1033 ". It may be assumed that }{\b\lang1033 hr}{\lang1033 was likewise normally pronounced as a voiced }{\b\lang1033 r}{\lang1033 in Frodo's day, though there is no explicit statement to this effect. (However, it seems that the digraphs }{\b\lang1033 hl}{\lang1033 - and }{\b\lang1033 hr}{\lang1033 - persisted in the established }{\i\lang1033 spelling}{\lang1033 of Quenya \endash these }{\b\lang1033 h}{\lang1033 's being spelt with the Tengwa }{\b\lang1033 halla}{\lang1033 .) \par \par Between }{\b\lang1033 i}{\lang1033 and another vowel, a glide consonant }{\b\lang1033 y}{\lang1033 could develop, breaking up the hiatus. Citing a word for "shingle" as }{\b\lang1033 sarni\'eb}{\lang1033 , Tolkien also listed }{\b\lang1033 sarniy\'eb}{ \lang1033 as a parenthetical alternative (VT42:11). This glide consonant may also develop between a }{\i\lang1033 diphthong}{\lang1033 in -}{\b\lang1033 i}{\lang1033 and a following vowel, as when older }{\b\lang1033 waia}{\lang1033 , }{\b\lang1033 vaia}{\lang1033 "envelope" could also appear as }{\b\lang1033 waiya}{\lang1033 , }{\b\lang1033 vaiya}{\lang1033 (Etym, entry }{\i\lang1033 WAY}{\lang1033 ). These examples suggest that it is difficult to pinpoint exactly }{\i\lang1033 when}{\lang1033 the glide consonant developed: was it before or after initial }{\b\lang1033 w}{\lang1033 - became }{\b\lang1033 v}{\lang1033 -, a change we tentatively assume to have occurred in the Second Age? As suggested by the example }{\b\lang1033 sarni}{\lang1033 (}{\b\lang1033 y}{\lang1033 )}{\b\lang1033 \'eb}{\lang1033 , the pre sence of such glide consonants seems to be an "optional" feature of Exilic Quenya, and their representation in writing is apparently rather irregular. There is no trace of glide consonants in }{\i\lang1033 Nam\'e1ri\'eb}{\lang1033 in LotR; we have forms like }{\b\lang1033 lauri\'eb}{\lang1033 and }{\b\lang1033 nam\'e1ri\'eb}{\lang1033 rather than *}{\b\lang1033 lauriy\'eb}{\lang1033 , *}{\b\lang1033 nam\'e1riy\'eb}{\lang1033 . On the other hand, Frodo speaking in tongues in Cirith Ungol employs the word }{\b\lang1033 aiya}{\lang1033 for "hail!", this form representing older }{\b\lang1033 aia}{\lang1033 (so in Tolkien's Quenya translation of the }{\i\lang1033 Hail Mary}{ \lang1033 ). \par \par Lastly, certain assimilations seen to occur in contact may be listed; it is difficult to say at what stage these assimilations "happened". Perhaps the combinations in question were assimilated already in Valinorean Quenya, but were treated likewise in younger compounds. \par \par The groups }{\b\lang1033 nl}{\lang1033 , }{\b\lang1033 nm}{\lang1033 , }{\b\lang1033 nr}{\lang1033 , }{\b\lang1033 ns}{\lang1033 are all assimilated to double consonants }{\b\lang1033 ll}{\lang1033 , }{\b\lang1033 mm}{\lang1033 , }{\b\lang1033 rr}{ \lang1033 , }{\b\lang1033 ss}{\lang1033 : \par \par }{\b\lang1033 N\'famen}{\lang1033 "West" + }{\b\lang1033 l\'f3t\'eb}{\lang1033 "flower" = }{\b\lang1033 N\'famell\'f3t\'eb}{\lang1033 "Flower of the West" \par }{\b\lang1033 elen}{\lang1033 "star" + }{\b\lang1033 macil}{\lang1033 "sword" = }{\b\lang1033 Elemmacil}{\lang1033 *"Star-sword" \par }{\b\lang1033 elen}{\lang1033 "star" + }{\b\lang1033 r\'edna}{\lang1033 "crowned" = }{\b\lang1033 Elerr\'edna}{\lang1033 "Star-crowned" \par }{\b\lang1033 elen}{\lang1033 "stone" + }{\b\lang1033 sar}{\lang1033 "stone" = }{\b\lang1033 Elessar}{\lang1033 "Star-stone" (or "Elf-stone") \par \par The cluster }{\b\lang1033 ll}{\lang1033 may also arise from }{\b\lang1033 rl}{\lang1033 , e.g. }{\b\lang1033 Casalli}{\lang1033 as the partitive plural of }{\b\lang1033 Casar}{\lang1033 "Dwarf" (}{\b\lang1033 Casar}{\lang1033 + -}{\b\lang1033 li}{ \lang1033 ). \par \par }{\i\fs32\lang1033 APPENDIX: Some Special Cases}{\fs32\lang1033 \par }{\lang1033 I will briefly discuss certain words and phonological features which do not readily agree with the system sketched above. \par \par In the }{\i\lang1033 Etymologies}{\lang1033 , Tolkien derived the name }{\b\lang1033 Melko}{\lang1033 (Melkor) from \'a4}{\b\lang1033 Mailik\'f4}{\lang1033 , an }{\i\lang1033 A}{\lang1033 -infixed variant of the extended base }{\i\lang1033 MIL-IK}{ \lang1033 supplied with the masculine ending -}{\b\lang1033 \'f4}{\lang1033 . The name is apparently here meant to suggest *"Greedy One". This seems like a strange ad hoc etymology. It is difficult to get from \'a4}{\b\lang1033 Mailik\'f4}{\lang1033 to }{\b\lang1033 Melko}{\lang1033 . The shortening of final -}{\b\lang1033 \'f4}{\lang1033 to -}{\b\lang1033 o}{\lang1033 is of course regular enough, but Tolkien here has the second }{\b\lang1033 i}{\lang1033 undergoing syncope in a word that does }{ \i\lang1033 not}{\lang1033 have two identical short vowels following one another \endash the environment normally triggering such syncope. Or are we to assume that the final element of the diphthong }{\b\lang1033 ai}{\lang1033 in the initial syllable constitutes the first of this pair of identical vowels? The intermediate form *}{\b\lang1033 Mailko}{\lang1033 is then supposed to become }{\b\lang1033 Melko}{\lang1033 , the diphthong }{\b\lang1033 ai}{\lang1033 collapsing into a monophthong }{ \b\lang1033 e}{\lang1033 . This development is apparently meant to be triggered by the following consonant cluster; yet this does not agree with other material from the }{\i\lang1033 Etymologies}{\lang1033 , cf. a word like }{\b\lang1033 aista}{\lang1033 - (entry }{\i\lang1033 G\'c1YAS}{\lang1033 ) failing to become **}{\b\lang1033 esta}{\lang1033 -. In any case, Tolkien adandoned the etymology here proposed for the name }{\b\lang1033 Melko}{\lang1033 . In a post-LotR source, this alternative nam e of Melkor is said to mean simply "Mighty One", derived from a stem }{\b\lang1033 melk}{\lang1033 - or }{\b\lang1033 melek}{\lang1033 - having to do with force and strength (MR:350, 360). This }{\b\lang1033 mel}{\lang1033 (}{\b\lang1033 e}{\lang1033 )}{ \b\lang1033 k}{\lang1033 - may be taken as the Classical Quenya form of the original }{\b\lang1033 mbelek}{\lang1033 - seen in the old form }{\b\lang1033 Mbelek\'f4ro}{\lang1033 "Melkor" (WJ:402). So according to Tolkien's later ideas, the shorter name } {\b\lang1033 Melko}{\lang1033 is certainly meant to come from *}{\b\lang1033 Mbelek\'f4}{\lang1033 , and the troublesome evolution \'a4}{\b\lang1033 Mailik\'f4}{\lang1033 > }{\b\lang1033 Melko}{\lang1033 presupposed in the }{\i\lang1033 Etymologies}{ \lang1033 does not have to be considered part of the "classical" linguistic scenario. \par \par The same goes for }{\b\lang1033 malarauco}{\lang1033 as the Quenya word for "Balrog". It is listed in the }{\i\lang1033 Etymologies}{\lang1033 , entry }{\i\lang1033 RUK}{\lang1033 , and explicitly referred to a primitive form \'a4}{\b\lang1033 \'f1g}{ \b\lang1033\super w}{\b\lang1033 alarauk\'f4}{\lang1033 . However, the idea that initial \'a4}{\b\lang1033 \'f1g}{\b\lang1033\super w}{\lang1033 - is to produce Quenya }{\b\lang1033 m}{\lang1033 - is blatantly contradicted by the entry }{\i\lang1033 \'d1G }{\i\lang1033\super W}{\i\lang1033 AL}{\lang1033 (though there is actually a cross-reference to it in the same entry }{\i\lang1033 RUK}{\lang1033 !) From }{\i\lang1033 \'d1G}{\i\lang1033\super W}{\i\lang1033 AL}{\lang1033 - Tolkien derived words like }{ \b\lang1033 nwalya}{\lang1033 - "to pain" and }{\b\lang1033 nwalca}{\lang1033 "cruel" \endash not **}{\b\lang1033 malya}{\lang1033 - and **}{\b\lang1033 malca}{\lang1033 , though the primitive forms are obviously meant to be *}{\b\lang1033 \'f1g}{ \b\lang1033\super w}{\b\lang1033 alj\'e2}{\lang1033 -, *}{\b\lang1033 \'f1g}{\b\lang1033\super w}{\b\lang1033 alk\'e2}{\lang1033 . We can be certain that the words }{\b\lang1033 nwalya}{\lang1033 - and }{\b\lang1033 nwalca}{\lang1033 reflect Tolkien's final decision on how primitive initial *}{\b\lang1033 \'f1g}{\b\lang1033\super w}{\lang1033 - is to come out in Quenya, for what is obviously another derivative from this stem occurs as the name of a Tengwa letter in LotR Appendix E: } {\b\lang1033 Nwalm\'eb}{\lang1033 "torment" (earlier "ngwalme" = }{\b\lang1033 \'f1walm\'eb}{\lang1033 ). \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 So why does \'a4}{\b\lang1033 \'f1g}{\b\lang1033\super w}{\b\lang1033 alarauk\'f4}{\lang1033 produce }{\b\lang1033 malarauco}{\lang1033 instead of the expected form *}{\b\lang1033 nwalarauco}{ \lang1033 in the }{\i\lang1033 Etymologies}{\lang1033 ? Comparison with earlier material reveals that }{\b\lang1033 malarauco}{\lang1033 as the High-elven word for Balrog is influenced by such early "Qenya" forms as }{\b\lang1033 Malcarauke}{\lang1033 (LT1:250). However, the }{\b\lang1033 m}{\lang1033 -forms can be ignored altogether, for Tolkien later decided that in Quenya, a Balrog is called a }{\b\lang1033 Valarauco}{\lang1033 (WJ:415, VT39:10). This is clearly meant to be derived from *}{ \b\lang1033 balarauk\'f4}{\lang1033 , and the phonological evolution would be quite regular. \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par The word }{\b\lang1033 palant\'edr}{\lang1033 is somewhat puzzling. In the Plotz letter, Tolkien explicitly stated that a long vowel in the final syllable of a polysyllabic word became }{\i\lang1033 short}{\lang1033 ; yet the long }{\b\lang1033 \'ed}{ \lang1033 of }{\b\lang1033 palant\'edr}{\lang1033 apparently persisted as late as in the Third Age. On the other hand, }{\b\lang1033 Palantir}{\lang1033 with a short }{\b\lang1033 i}{\lang1033 is mentioned as the name of a N\'fa menorean king in the LotR appendices. \par \tab In Letters:427, Tolkien refers }{\b\lang1033 palant\'edr}{\lang1033 to a primitive form \'a4}{\b\lang1033 palant\'eer\'e2}{\lang1033 or \'a4}{\b\lang1033 palant\'eera}{\lang1033 ; he marked the final -}{\b\lang1033 a}{\lang1033 with a diacritic indicating that it could be either long or short. (This "primitive" form must be seen as a hypothetical back-formation, since the palant\'edri were made and presumably named by the Noldor }{\i\lang1033 after}{\lang1033 the end of the "pre-historic" period.) According to the rules set out above, \'a4}{\b\lang1033 palant\'eer\'e2}{\lang1033 should have produced Quenya *}{\b\lang1033 palant\'edra}{\lang1033 , whereas \'a4}{\b\lang1033 palant\'eera}{\lang1033 should have yielded }{\b\lang1033 palantir}{\lang1033 , attested as the personal name of Tar-Palantir of N\'famenor. (\'a4}{\b\lang1033 Palant\'eer\'e2}{\lang1033 could also have produced }{\b\lang1033 palantir}{\lang1033 if the primitive form was recognized as a compound, so that the long final vowel was shortened at an early stage.) So what are we do make of the form }{\b\lang1033 palant\'edr}{\lang1033 occurring in LotR? Perhaps we are to assume that since the Seeing Stones were so closely associated with "ancient lore" and were rarely mentioned otherwise, the word persisted in its older Valinorean form rather than undergoing the normal sound-changes? Of course, the long vowel may also have been re-in troduced into the singular by analogy with the plural form }{\b\lang1033 palant\'edri}{\lang1033 . \par }\pard \fi708\nowidctlpar\widctlpar\adjustright {\lang1033 Another example of a word with a long final vowel persisting in the final syllable is }{\b\lang1033 Erus\'e9n}{\lang1033 , translated "chil dren of God" in RGEO:74. This form is doubly perplexing, for judging by the translation, it should be a plural form. Yet there is no plural ending. Is this a collective of some sort? \par }\pard \nowidctlpar\widctlpar\adjustright {\lang1033 \par The noun "head" is given in the }{\i\lang1033 Etymologies}{\lang1033 as }{\b\lang1033 c\'e1r}{\lang1033 with stem }{\b\lang1033 cas}{\lang1033 -; the root is }{\i\lang1033 KAS}{\lang1033 . This makes little sense within the later phonological framework; it seems to be a form carried over from Tolkien's earliest "Qenya" (QL:45: notice that the }{\i\lang1033 K}{\lang1033 -stems are among the oldest parts of the }{\i\lang1033 Etymologies}{ \lang1033 , apparently never thoroughly revised). In the }{\i\lang1033 Etymologies}{\lang1033 , Tolkien lets a final postvocalic -}{\b\lang1033 s}{\lang1033 become -}{\b\lang1033 r}{\lang1033 , as when the root }{\i\lang1033 TELES}{\lang1033 produces Quenya }{\b\lang1033 Teler}{\lang1033 "Teler-elf". The root }{\i\lang1033 KAS}{\lang1033 , or rather a primitive form *}{\i\lang1033 kasa}{\lang1033 > Common Eldarin *}{\i\lang1033 k\'e2s}{\lang1033 , *}{\i\lang1033 kas}{\lang1033 -, might therefore produce }{\b\lang1033 c\'e1r}{\lang1033 via *}{\b\lang1033 c\'e1z}{\lang1033 . However, this would }{\i\lang1033 not}{\lang1033 have a stem-form }{\b\lang1033 cas}{\lang1033 -, e.g. plural **}{\b\lang1033 casi}{\lang1033 "heads" \endash no more than the plural of }{\b\lang1033 Teler}{\lang1033 was **}{\b\lang1033 Telesi}{\lang1033 even within the }{\i\lang1033 Etymologies}{\lang1033 scenario (the plural }{\b\lang1033 Teleri}{\lang1033 occurs in the entry }{\i\lang1033 TELES} {\lang1033 ). According to the system Tolkien employed in most of Etym, the word for "head" ought to be *}{\b\lang1033 c\'e1r}{\lang1033 with stem *}{\b\lang1033 car}{\lang1033 -. Later Tolkien apparently decided that a }{\i\lang1033 final}{\lang1033 original -}{\b\lang1033 s}{\lang1033 remained unchanged, though the development }{\b\lang1033 s}{\lang1033 > }{\b\lang1033 z}{\lang1033 > }{\b\lang1033 r}{\lang1033 does occur between vowels; thus we have }{\b\lang1033 olos}{\lang1033 "dream", pl. } {\b\lang1033 olori}{\lang1033 for older }{\b\lang1033 olozi}{\lang1033 , in UT:396. If the noun }{\b\lang1033 olor}{\lang1033 of Etym (entry }{\i\lang1033 \'d3LOS}{\lang1033 ) was later emended to }{\b\lang1033 olos}{\lang1033 , the noun }{\b\lang1033 c \'e1r}{\lang1033 "head" (}{\i\lang1033 KAS}{\lang1033 ) must also become *}{\b\lang1033 c\'e1s}{\lang1033 , though the stem-form remains (*}{\b\lang1033 caz}{\lang1033 - >) *}{\b\lang1033 car}{\lang1033 -. A possible attestation of "}{\b\lang1033 cas}{ \lang1033 " as a word for "head" occurs in the compound }{\b\lang1033 cas-raya}{\lang1033 "tressure" (literally *"head-net"), later becoming }{\b\lang1033 carr\'eba}{\lang1033 (VT41:12). Interestingly, the form }{\b\lang1033 cas-raya}{\lang1033 is not asterisked, and it is spelt with a }{\b\lang1033 c}{\lang1033 rather than }{\b\lang1033 k}{\lang1033 also in the primary source \endash suggesting that this is a "historical" or "attested" Quenya form. Perhaps, then, }{\b\lang1033 cas-raya}{ \lang1033 is meant to represent the form of the word in Valinorean Quenya of the early historical period, the "classical" form }{\b\lang1033 carr\'eba}{\lang1033 arising within recorded history. (The assimilation of medial }{\b\lang1033 sr}{\lang1033 to }{\b\lang1033 rr}{\lang1033 probably began already in the prehistoric period, but perhaps it continued to operate and was eventually applied even to more recent compounds.) \par \par According to the }{\i\lang1033 Etymologies}{\lang1033 , the stem }{\i\lang1033 \'d1OL}{\lang1033 yielded the Quenya word }{\b\lang1033 holm\'eb}{\lang1033 "odour". This must be derived from primitive *}{\b\lang1033 \'f1olm\'ea}{\lang1033 . So does initial *}{\b\lang1033 \'f1}{\lang1033 - produce Quenya }{\b\lang1033 h}{\lang1033 -? If so, this change must have occurred }{\i\lang1033 before}{\lang1033 the ancient initial combination }{\b\lang1033 \'f1g}{\lang1033 - was simplified to }{ \b\lang1033 \'f1}{\lang1033 -, since the latter persisted in Valinorean Quenya (becoming normal }{\b\lang1033 n}{\lang1033 - in the Third Age). However, in a much later source, Tolkien lets primitive }{\b\lang1033 \'f1}{\lang1033 - remain unchanged in Quenya: the word }{\b\lang1033 \'f1alta}{\lang1033 "radiance, glittering reflection" he referred to primitive \'a4}{\b\lang1033 \'f1alat\'e2}{\lang1033 . So why does *}{\b\lang1033 \'f1olm\'eb}{\lang1033 produce Quenya }{ \b\lang1033 holm\'eb}{\lang1033 instead of *}{\b\lang1033 \'f1olm\'eb}{\lang1033 ? Some think }{\b\lang1033 holm\'eb}{\lang1033 is simply a misreading for *}{\b\lang1033 nolm\'eb}{\lang1033 in Tolkien's manuscript. *}{\b\lang1033 Nolm\'eb}{\lang1033 could be the later pronunciation of *}{\b\lang1033 \'f1olm\'eb}{\lang1033 . \par \par The view presented above is that the shortening of the long final vowels of primitive Elvish took place at a relatively late stage (after Common Eldarin), whereas }{\i\lang1033 original}{\lang1033 short final -}{\b\lang1033 a}{\lang1033 , -}{\b\lang1033 e}{\lang1033 and -}{\b\lang1033 o}{\lang1033 disappeared during the Common Eldarin stage. Examples supporting this view are cited above. However, counterexamples can certainly be found. In some sources, Tolkien cites Common Eldarin words with short final vowels that survive unchanged into Quenya, as when \'a4}{\b\lang1033 sloga}{\lang1033 and \'a4}{\b\lang1033 netere}{\lang1033 is said to be the origin of Quenya }{\b\lang1033 hloa}{\lang1033 , }{\b\lang1033 nert\'eb}{\lang1033 (VT41:9, 26). According to the system here reconstructed, these C.E. words would rather produce Quenya **}{\b\lang1033 hl\'f3}{\lang1033 , **}{\b\lang1033 neter}{\lang1033 . To reach the "desired" outcome in Quenya we would have to assume that the C.E. forms were *}{\b\lang1033 slog\'e2}{\lang1033 , *}{\b\lang1033 neter\'ea}{\lang1033 with long final vowels. We could assume that after }{\i\lang1033 original}{\lang1033 short -}{\b\lang1033 a}{\lang1033 , -}{\b\lang1033 e}{\lang1033 , -}{\b\lang1033 o}{\lang1033 were lost, but still during the Common Eldarin stage, the original long final vowels were shortened so that *}{\b\lang1033 slog\'e2}{\lang1033 , *}{ \b\lang1033 neter\'ea}{\lang1033 turned into the forms \'a4}{\b\lang1033 sloga}{\lang1033 , \'a4}{\b\lang1033 netere}{\lang1033 mentioned by Tolkien. Some examples from the }{\i\lang1033 Etymologies}{\lang1033 may also seem to support such a view, as when Tolkien mentioned \'a4}{\b\lang1033 daio}{\lang1033 as the primitive form of }{\b\lang1033 l\'ebo}{\lang1033 "shade" (entry }{\i\lang1033 DAY}{\lang1033 ). The final -}{\b\lang1033 o}{\lang1033 must in all likelihood have been a long -}{\b\lang1033 \'f4}{\lang1033 at some earlier stage, but here it may seem to have been shortened quite early, before initial }{\b\lang1033 d}{\lang1033 - had become }{\b\lang1033 l}{\lang1033 -. The latter change must in turn have occurred before initial and post-vocalic }{\b\lang1033 b}{\lang1033 , }{\b\lang1033 d}{\lang1033 , }{\b\lang1033 g}{\lang1033 became spirants }{\b\lang1033 v}{\lang1033 , }{\b\lang1033 \'f0}{\lang1033 (> }{ \b\lang1033 z}{\lang1033 > }{\b\lang1033 r}{\lang1033 ), }{\b\lang1033 3}{\lang1033 , respectively, or primitive initial }{\b\lang1033 d}{\lang1033 - would have produced Quenya }{\b\lang1033 r}{\lang1033 - instead of }{\b\lang1033 l}{\lang1033 -. Yet this is totally contradicted by one attested example referred to above: }{\b\lang1033 Valin\'f4r\'ea}{\lang1033 as an early Quenya name of Valinor, with a long final vowel intact as late as }{\i\lang1033 after}{\lang1033 initial }{\b\lang1033 b}{ \lang1033 - has become }{\b\lang1033 v}{\lang1033 - (WJ:413). There is hardly any way all the "evidence" can be worked into a single system. We must probably conclude that Tolkien in some cases failed to mark some long final vowels as such. Of course, he may also have changed his mind (repeatedly!) about the precise order of certain phonological changes. \par \par In some sources, Tolkien seems to be toying with the idea that a }{\i\lang1033 double consonant}{\lang1033 is to be simplified before a long, accented vowel. Thus }{\b\lang1033 Elerr\'edna}{\lang1033 *"Star-crowned" (assimilated from *}{\b\lang1033 Elenr\'edna}{\lang1033 ) appears as }{\b\lang1033 Eler\'edna}{\lang1033 in LR:355. Likewise, the name of Tuor's ship }{\b\lang1033 E\'e4r\'e1m\'eb}{\lang1033 (WJ:352) is probably to be understood as }{\b\lang1033 E\'e4r}{\lang1033 + }{\b\lang1033 r\'e1 m\'eb}{\lang1033 , i.e. *"Sea-wing", rather than }{\b\lang1033 E\'e4}{\lang1033 + }{\b\lang1033 r\'e1m\'eb}{\lang1033 "Universe-wing". However, Tolkien may have abandoned this idea. }{\b\lang1033 Elerr\'edna}{\lang1033 is so spelt in the published }{ \i\lang1033 Silmarillion}{\lang1033 , and I think this form is to be preferred. \par }{ \par }{ \par }}