Some notes on Noldorin and Sindarin past-tense verbs
I. Eldarin past-tense formation classes:
Since at least the publication of _The Lost Road_ in 1987, it has been
widely recognized that the Eldarin tongues, at least as exhibited from
the latter half of the 1930s through the publication of _The Lord of
the Rings_, had two primary classes of past-tense verb formation, which
we can call (using Tolkien's own terminology in "Quendi and Eldar",
itself borrowed from Indo-European linguistics): _strong_ (indicating
internal modification) and _weak_ (indicating modification by affixes).
The strong past tense, which class includes chiefly basic verbs (that
is, verbs arising directly from simple bases with verbal meaning, such
as KAT- 'shape', BAT- 'tread', SUK- 'drink', etc.), is characterized by
lengthening or otherwise strengthening the root syllable and suffixion
of original final *_-ê_ (I employ the circumflex to indicate a macron).
Within the strong past tense there are, again, two chief means of root
strengthening: a) lengthening or fortification of the root vowel, and
b) nasal infixion. The weak past tense, which class includes chiefly
derived verbs (e.g., verbs formed from a root by suffixion of some
derivational element, such as *_-tâ_, *_-yâ_, *_-â_, etc.), is
characterized by the lack of root syllable strengthening and the
suffixion of some past-tense marker, very often *_-nê_ These two main
classes are neatly exemplified by the two Qenya past-tense verbs
arising from the base ONO- 'beget' in _Etymologies_, _óne_ (strong) and
_ontane_ (weak), the former arising directly from the base and the
latter from the derived verb _onta-_. For Quenya of _The Lord of the
Rings_, the same mechanisms are exemplified by the past-tense verbs
found in Galadriel's Lament, which includes, on the one hand,
_unduláve_ 'down-licked', where _-láve_ '-licked' is a strong
past-tense verb formed on LAB- 'lick'; and, on the other hand, _ortane_
'uplifted', a weak past-tense verb (in _-ne_) based on the derived verb
_orta-_ 'raise' (cf. ORO-).
(It is interesting to note that, like the strong past tense, the
Eldarin present tense is characterized by strengthening of the root
syllable, but further with suffixion of *_-â_ rather than *_-ê_; thus
Q. _sîla_ 'shines, is shining'; while the aorist is characterized by
the absence of strengthening and the suffixion of *_-i_; thus Q _sile_
II. Loss of original final long *_-ê_ in Noldorin and Sindarin:
In considering the reflexes of these past-tense formation mechanisms in
Noldorin and Sindarin, it turns out that we can disregard the suffixion
of original final long *_-ê_ that is found in the strong formation (all
types) and as the last phoneme of the weak formation ending *_-nê_,
since in the development of Noldorin and Sindarin this final vowel was
uniformly lost without a trace. Indeed, we can only presume that it was
part of the history of Noldorin and Sindarin past-tense verbs because
of the evidence for them in the other Eldarin languages (chiefly
Q(u)enya); from strictly the viewpoint of Noldorin and Sindarin, this
original long vowel can be regarded as superfluous, and in what follows
I have not bothered to indicate it, though for completeness it can be
III. Metathesis of *_-tn-_ to _-nt_ in Noldorin and Sindarin:
As in Q(u)enya, it appears that the original contact *_tn_ (such as
that formed by the suffixion of derivative elements beginning in _n_ to
bases ending in _t_) in Noldorin and Sindarin underwent metathesis to
_nt_ (and then often later simply to _nn_, especially medially); as,
for example, N _pann_ 'wide' < *_patnâ_ < PAT- (q.v.). The upshot of
this is that it is not always able to be said with certainty whether a
Noldorin or Sindarin past-tense formation in _-nt_ (or other nasal +
stop cons. contact) is due to nasal infixion or to suffixion of *_-nê_
with subsequent methathesis (the same is true of Q(u)enya as well,
actually). Nonetheless, noting the past-tense formations in the other
Eldarin languages (particularly Q(u)enya), it appears that basic verbs
with past-tense in final _-nt_ or other nasal + stop combinations most
likely arose by nasal infixion (and thus are strong), not by suffixion
of *_-nê_, and we so assume here. But the blurring by normal
phonological development of the actual historical development of
past-tense forms in _-nt_ has significant consequences for Noldorin and
Sindarin, and will be remarked upon again below.
IV. Eldarin personless verb forms:
From the various examples of sentences in the Eldarin tongues that
Tolkien has supplied, it is evident that the Eldarin verb conjugations,
at least as exhibited during the composition of _The Lord of the Rings_
and beyond, could inflect for the person of the subject by suffixion of
suitable pronominal endings (e.g., _hiruva-lye_ 'thou shalt find' in
Galadriel's Lament, and _linnatho-n_ 'I will sing' in the Hymn to
Elbereth); but it is equally clear that such personal inflection was
not _required_, at least when an explicit subject noun or separate
pronoun preceded the verb: thus we see _Nai elye hiruva_ 'Be it that
even thou shalt find [it]' in Galadriel's Lament, where the emphatic
form of the subject pronoun, _elye_ 'thou', precedes the verb; but _nai
tiruva-nte-s i hárar_ *'be it that they will guard it, [those] who sit'
in Cirion's Oath, with 3rd pl. subject inflexion _-nte_ preceding the
plural subject _i hárar_. Nonetheless, personless verbs still inflect
to agree with the number of their subjects: in Quenya, this takes the
form of an endingless singular (_elye hiruva_ 'thou shalt find', or
_elen síla_ 'a star shines' in Frodo's greeting to Gildor) and a plural
in _-r_ (as, for example, in _lassi lantar_ 'leaves fall' or _yéni
avánier_ 'years have passed away', both found in the version of
Galadriel's Lament put into more normal word order, R:66-67).
Because these personless forms of verbs are so frequently found in the
corpus in conjunction with 3rd person subjects (sg. or pl.), it is
(unfortunately) very common to find them described as: _-0_ (zero /
endingless) 3rd pers. sg., and _-r_ 3rd pers. pl. However, strictly
speaking, these are _not_ personal forms; they _only_ indicate
_number_. This is demonstrated by the use of the personless forms with
other than 3rd pers. subjects: for example, _elye hiruva_ 'thou shalt
find" (where the subject is 2nd sg.), or _emme avatyarir_ 'we forgive'
(where the subject is 1st. pl.; VT43:20). We can therefore refer to
these as impersonal forms, or to avoid confusion with the term
"impersonal" as applied to subjectless (often stative) verbs ("it is
raining", "it is snowing", etc.), as simply personless.
Noldorin past-tense verbs from _The Etymologies_:
With these preliminaries in mind, let's take a look at the exemplars of
the strong past-tense formations in Noldorin of _The Etymologies_. A
survey turns up the following:
Strong past I: Root vowel lengthening or A-infixion:
†_daul_ *'concealed' (s.v. DUL- 'hide, conceal'). This is listed as a
poetic past tense, against two derived verbs: _doelio_ / _delio_
(*_-yâ_-verb, _doelio_ < *_dulyâ-_) and _doltha_ (*_-tâ_-verb, _-ltha-
< *_-ltâ_). The past-tense form itself, however, shows derivation
directly from the base, presumably by A-infixion; thus < *_daul_ by
A-infixion < DUL-.
_trenor_ (also _trener_) *'told' (< NAR2- 'tell, relate'). The prefix
_tre-_ 'through' (< TER-, q.v.) is unrelated to the past-tense
formation, which concerns only the root _-nor_. In unstressed
word-final syllables, N _o_ can result, _inter alia_, from original
long *_â_, which given the base form is most likely the case here; the
lengthening of the root vowel can here be due either to A-infixion or
to simply lengthening, as they would both produce the same result; thus
< *_nâr_ by root vowel strengthening < NAR2-.
Note that we find only basic verbs in this category.
Strong past II: Nasal infixion:
_trevant_ *'traversed' (BAT- 'tread'). Again, the prefix _tre-_
'through' (< TER-, q.v.) is unrelated to the past-tense formation,
leaving us with the root _-vant_. The development appears to be by
lenition from *_bant_ < *_ba-n-t_ by nasal infixion < BAT-.
_echant_ *'fashioned' < *_et-kat-_ (KAT- 'shape'). Tolkien's own
etymological note shows that formation contains an original prefix
*_et-_ (< ET- 'out, forth', q.v.), leaving us with the root *_kant_ <
*_ka-n-t_ by nasal infixion < KAT-.
_hant_ *'hurled' (KHAT- 'hurl') < *_ha-n-t_ by nasal infixion < KHAT-.
_nestanc_ *'inserted' (STAK- 'split, insert'). The prefix _ne-_ *'in'
is perhaps related to the base NÉD 'middle, centre', and/or to the
(admittedly later) N/S word _ned_ 'in' found in "The King's Letter";
but the root form _-stanc_ can be isolated, and explained as <
*_sta-n-k_ by nasal infixion < *_STAK-_.
_sunc_ *'drank' (SUK- 'drink') < *_su-n-k_ by nasal infixion < SUK-.
_gwend_ (also_gwenn_) *'bound' (WED- 'bind') < *_we-n-d_ by nasal
infixion < WED-.
Note that half of this (admittedly small, but in terms of _Etymologies_
relatively populous) class of strong, basic past-tense verbs ends in
the combination _-nt_.
There is in addition a smaller class of strong nasal-infixion verbs,
which however are formed from _derived_ verbs, including:
_lhimmint_ *'moistened' (pa.t. of _lhimmid_ 'moisten') (LINKWI- *'wet')
< *_linkwi-n-tâ_ < *_linkwi-tâ_ < LINKWI-.
_nimmint_ *'whitened' (pa.t. of _nimmid_ 'to whiten') (NIK-W-) <
*_ninkwi-n-tâ_ < *_ninkwi-tâ_ < NIK-W-.
The prevalence (within this admittedly small sample space) of strong
pa.t. verbs (both basic and derived) formed by nasal infixion and, as a
result of Noldorin phonological development, coming to end in _-nt_, is
significant in light of a class of pa.t. verbs ending in _-nt_ that
cannot be similarly explained as due to regular development; namely:
Weak past I: Suffixion of _-nt_:
_orthant_ *'raised' (pa.t. of _ortho_ 'raise') (ORO- 'up; rise; high')
_asogant_ (also _sunc_) *'drank' (pa.t. of _sogo_ *'drink') (SUK-
_tiriant_ *'watched, guarded' (pa.t. of _tiri_ or _tirio_) (TIR-
_gwedhant_ *'bound' (pa.t. of _gwedi_) (WED- 'bind')
(* The apparent prefix _a-_ of _asogant_ is problematic, but for our
purposes irrelevant, since we are concerned only with the root *_suk-_
and its extension.)
These past-tense verbs _appear_ to have been derived by nasal-infixion
from earlier forms in *_-tâ_, _themselves_ formed on _already_
derivative stems (sc., *_ortâ-n-tâ_, *_sukâ-n-tâ_, *_tiryâ-n-tâ_,
*_wedâ-n-tâ_), and thus to be _doubly_ derivative; but there is aside
from these forms no evidence whatsoever for such double-derivative verb
forms in any of the Eldarin tongues. How to account for this?
The answer appears to be that the _-nt_ ending of the past-tense of
these verbs is _not_ historical, but instead was extended to this class
of verbs by analogy with the (relatively) frequent strong KAT-type
verbs and derived verbs in _-itâ_ exhibiting _-nt_ by regular
phonological development.* A significant motivation for this analogical
extension might have had to do with the results of the means of
past-tense formation that, based on the parallel development in
Q(u)enya, would be expected to have been applied to weak verbs, namely
suffixion of *_-nê_; e.g., *_ortâ-nê_ > N *_orthon_, *_tiryâ-nê_ > N
*_tirion_, etc., which look like 1st pers. sg. present-tense verbs (cf.
N/S _linnon_ 'I sing', III:354). (It should be noted that _asogant_ and
_gwedhant_ are not necessarily derivative verbs (i.e., historically
formed from stems *_suk-â_, _wed-â_), but instead may be formed
directly from basic stems, *_suk-_ and *_wed-_, with the past-tense
ending _-ant_ arising, like simple _-nt_, by analogy with the _-ant_
found in the past-tense of basic, KAT-type verbs.)
(* Historically, strong past-tense verbs were probably even more
frequent in the Eldarin tongues than they came to be by the period
present to _The Etymologies_, just as strong verbs in English, such as
"stride / strode", "think / thought", "buy / bought", etc., are
remnants of a once much more populous class of past-tense verb
formation, and shrinking remnants at that (again, just as in English),
as indicated by the poetic status of _daul_, and the several coexistent
(and most likely later) strong and weak past-tense forms of verbs, such
as weak _gwedhant_ (which Tolkien notes was a later form) against
(original) strong _gwend_ / _gwenn_ (cf. English present tense "shine"
vs. coexistent pa.t. "shone" and "shined", where "shone" is the older
and historically correct form, but is now considered quaint-sounding).
So much for Noldorin of _The Etymologies_, save for a small number of
other past-tense verbs that possibly or apparently employ other
mechanisms, differing in detail if not in spirit: namely _drammen_
(also †_dramp_), pa.t. of _dravo_ 'to hew' < DARÁM- 'beat, hew';
_hennin_, pa.t. of _hedi_ < KHAT- 'hurl'; _dammint_, pa.t. of _damna-_
'to hammer' < NDAM- 'hammer, beat'; and _mudas_, pa.t. of _mudo_
*'labour, toil' < MÔ-. (It should be noted that if _drammen_ is in fact
a 1st pers. sg. form, _dramme-n_ 'I hewed', as has been suggested by,
e.g., Helge Fauskanger, the apparent stem *_dramme-_ would in fact be a
typical strong, nasal-infix formation, i.e., < *_dra-m-m-ê-_.)
But it is interesting to note that even the small corpus of Sindarin
verbs found in _The Lord of the Rings_ exhibits of all three of the
chief mechanisms of past-tense formation that we have identified for
Strong past I: Root vowel lengthening or A-infixion:
_óne-n_ 'I gave' < *_ân-ê_ (LR:1036, probably derived from ANA1-
*'give'; note that the long initial _ó_ of the verb is in accordance
with the earliest editions and printings of _The Lord of the Rings; in
later printings, the accent was dropped -- according to Douglas
Anderson, most likely inadvertently).
Strong past II: Nasal infixion:
_echant_ 'made' (in the phrase "_Im, Narvi, echant_", 'I, Narvi, made',
LR:298), presumably like the Noldorin verb < *_et-kant_ < *_et-ka-n-t_
by nasal infixion < KAT-.
Weak past I: Suffixion of _-nt_:
_teithant_ 'drew' (in the phrase "_Celebrimbor ... teithant_"
'Celebrimbor ... drew', LR:298), as though < *_tek-tâ-n-tâ_, but
actually an analogical past-tense in _-nt_.
I conclude this survey with a few resultant observations on the
treatment and analysis of Noldorin and Sindarin past-tense verbs by
Helge Fauskanger, in his articles "Sindarin - the Noble Tongue"
<http://www.uib.no/People/hnohf/sindarin.htm> and "Reconstructing the
Sindarin Verb Conjugation"
<http://www.uib.no/People/hnohf/sverb-rec.htm> on his "Ardalambion" web
A. The use and status of evidence:
It is to be remarked that Fauskanger's treatment of the Sindarin verb
(by which he evidently means, Sindarin as exemplified in and after _The
Lord of the Rings_, and those parts of the evidence for Noldorin verbs
that he judges to be consistent with his notions of Sindarin) makes _no
mention_ of fully one fourth of the actual Noldorin past-tense forms
given in _The Etymologies_; and moreover provides two lists of
"Sindarin" past-tense verb forms -- one of basic verbs that he calls
"3rd person singular" (without support, see below); and another which
he calls "mixed conjugation", extrapolated on the basis of a _single_
past-tense form *_dram_, which _itself_ is an unattested hypothetical
form -- in which there is only _one_ actual, authentic, attested
past-tense verb-form listed, the rest being entirely hypothetical forms
constructed by Fauskanger himself.
B. So-called "3rd pers. sg." past-tense verbs:
Throughout these two articles, Fauskanger repeatedly refers to
past-tense verbs in _-nt_ as "3rd pers. sg." (in some cases
"transparently" so). In light of this, it is noteworthy that _none_ of
the Noldorin verbs of this form found in _Etymologies" is translated
with specifically 3rd-person sg. glosses; short of an explicit
statement by Tolkien that all Noldorin past-tense verbs in _-nt_ are
specifically 3rd pers. sg., there is in fact no way to prove what
Fauskanger silently asserts. In fact, given what we know about
personless verb forms in the Eldarin tongues (see the Preliminaries
above), and based on the evidence we actually have, it is far more
likely that these forms are singular personless forms, and therefore
would be used in any case where an explicit, singular subject precedes
the verb. Indeed, the same holds true of the Sindarin past-tense verbs
in _-nt_, for although _teithant_ is indeed used by Tolkien with a 3rd
pers. sg. subject (_Celebrimbor ... teithant_ 'Celebrimbor ... drew'),
this does not _in itself_ necessarily preclude its possible usage as a
personless verb with 1st and 2nd person subjects, any more than the use
of endingless present-tense verbs in Quenya with 3rd pers. sg. subjects
precludes their use with other singular persons, such as in _elye
hiruva_ 'thou shalt find" (where the subject is 2nd sg.). Indeed,
another such verb, _echant_, is used by Tolkien with a 1st pers. sg.
subject, and translated as such: _Im, Narvi, ... echant_ 'I, Narvi, ...
made'. (This assumes, of course, that _im_ is in fact a 1st pers. sg.
pronoun, and not, say, a reflexive form ('myself') that acts
grammatically as 3rd pers. sg. -- and its use in "Lúthien's Song" with
an explicitly 1st pers. sg. present tense verb: _le linnon im Tinúviel_
*'to thee I sing, I, the Nightingale' (III:354) may argue against its
being 3rd pers. But again, even if it is regarded as 3rd pers., the
mere fact that the two examples we have of Sindarin past-tense verbs in
_-nt_ happen to occur with 3rd pers. sg. subjects does _not_
necessarily preclude its usage with other singular persons.)
C. Coexistence vs. "replacement":
In his article "Sindarin - the Noble Tongue", Fauskanger asserts that
the strong past-tense form _gwend_ / _gwenn_ "was later replaced by
_gwedhant_", and that _gwend_ / _gwenn_ "had come to be regarded as
archaic or poetic". There is nothing in Tolkien's statement about these
verb forms to suggest any such "replacement" or "archaism". It is not
uncommon for strong and weak forms of past-tense verbs to coexist (see
the Eldarin and English examples above), and the unsupported assertion
that one form "replaced" the other seems an unjustifiable attempt to
remove a feature shared by "real" languages from Tolkien's
D. A so-called "totally abnormal", "unjustifiable" past-tense verb:
In his article titled "Reconstructing the Sindarin Verb Conjugation",
Fauskanger calls the Noldorin past-tense verb _mudas_ *'laboured,
toiled' (pa.t. of _mudo_, < MÔ-) "totally abnormal" and asserts that
"Nowhere in the entire published Tolkien-linguistic corpus is there any
past tense formation even resembling this, nor any hint of how such a
past tense form could possibly be historically justified." I would
direct Fauskanger's attention to the information on the Noldorin verb
found in the "Early Noldorin Grammar" published in _Parma Eldalamberon_
13 (in particular, pp. 126-32), wherein we find numerous examples of
past-tense verbs in _-th-_; this is not so very far removed from the
_-s_ of _mudas_. In any event, simply because most derived verbs found
in the Noldorin of _The Etymologies_ came by phonology or analogy to
form their past tense in _-nt_ in no way precludes the use and survival
of other past-tense formants, such as _-s_, besides (analogical) _-nt_
or (historical) suffixion of *_-nê_. Again, Fauskanger's
characterization of other, less prevalent means of derivation as
aberrations that should be ignored is an assertion of his own notions
of the shape, mechanisms, "regularity", and history of Tolkien's
languages over the actual evidence of the corpus, and thus of Tolkien's
=========================================================================================Carl F. Hostetter Aelfwine@... http://www.elvish.org
ho bios brachys, he de techne makre.
Ars longa, vita brevis.
The lyf so short, the craft so long to lerne.
"I wish life was not so short," he thought. "Languages take such
a time, and so do all the things one wants to know about."
- Carl F. Hostetter wrote, regarding Helge Fauskanger's statement
that the Noldorin pa.t. _mudas_ (< _mudo_ 'labour, toil', < MÔ-)
is "totally abnormal" and that "Nowhere in the entire published
Tolkien-linguistic corpus is there any past tense formation even
resembling this, nor any hint of how such a past tense form could
possibly be historically justified":
> I would direct Fauskanger's attention to the information on theMoreover, such pa.t. Noldorin verbs in _-th_ (e.g. _mad- 'to eat',
> Noldorin verb found in the "Early Noldorin Grammar" published in
> _Parma Eldalamberon_ 13 (in particular, pp. 126-32), wherein we
> find numerous examples of past-tense verbs in _-th-_; this is not
> so very far removed from the _-s_ of _mudas_.
past indicative singular _madath_, PE13:131) also date back to
Goldogrin. In the Gnomish Lexicon there are a large number of
verbs that form a weak pa.t. by addition of a suffix _-thi_. The
following list, only a small sample, demonstrates the addition of
this pa.t. ending to a wide variety of verbal stems:
_belu-_ 'runroll, unfurl', pret. _beluthi_.
_haw-, hau_ 'to lie', pret. _hui_ or _hauthi_.
_ilta-_ 'to stick in, prod, prick', pret. _iltathi_.
_îr-_ 'am willing', pret. _airi_ or _irthi_.
_ol-_ 'appear, seem', pret. _ûli_ or _olthi_.
_pâ-_ 'put', pret. _pôthi_ or _pâthi_.
_rô-_ 'remain, stay; stand', pret. _rûthi_ or _rôthi_.
_sana-_ 'can, know how to', pret. _sôni_ or _santhi_.
Etymologically, this weak pa.t. ending _-thi_ is probably
identical to _thi_, pret. of the verb _na(1)-_ 'is' (described
as "quite irregular", as further demonstrated by its
participle _ol-_ (PE11:58).
This would make the Goldogrin weak pa.t. forms in _-thi_ quite
similar in etymology to the Latin imperfect tense, marked by an
element _-ba-_ (e.g., _amabam_ 'I was loving, I used to love')
that is thought to derive from the Indo-European root BHEU@-,
BHEU- 'to be, exist, grow', which is also the source of English
This parallel with the formation of the Latin imperfect also
appears in the early Noldorin material published in Parma 13.
In the "Noldorin Dictionary", various Noldorin pa.t. forms in
_-th_ are labeled as "past imperfect": _amró_ *'rise' - "p. impf."
_amro(a)th_; _mad_ 'eats' - _madath_ "past imperfect, was
eating"; _dadnú_ *'sink' - "impf." _dadnovath_ (PE13:159, 163,
164). This early Noldorin pa.t. ending _-th_ might be cognate
with _tha-_ 'to make, cause to be' (< *_s'ta-_), given in the
"Noldorin Word-lists", a verb that is also said to be the source
of the causative suffix _-tha_ (PE13:153).
The forms cited above amply demonstrate that Fauskanger's
assertion that the Noldorin pa.t. _mudas_ is "totally abnormal"
and does not resemble any past tense formation found "in the
entire published Tolkien-linguistic corpus" is simply wrong, as
is his claim that there is no "hint of how such a past tense form
could possibly be historically justified". Exactly how the etymol-
ogies of Gn. _-thi_ and N. _-th_ proposed above relate to the
etymology Tolkien envisioned for N. _-s_ in _mudas_ is unclear,
to say the least -- but it _is_ clear that N. _mudas_ cannot
be simply dismissed as an isolated anomaly.
Two further points should be made. In "Reconstructing the Sindarin
Verb System", Fauskanger writes: "If the verb _mudo_ '[to] labour,
toil' is the source of a form _mudas_, the latter would therefore be
expected to mean 'labour, toil' as a noun. It seems entirely possible
that Tolkien distractedly wrote 'pa.t.' intending 'noun'." Possible,
perhaps -- but not probable, given that _mudas_ appears simply
to be part of a conceptual continuum of Goldogrin/Noldorin
pa.t. forms in _-thi_, _-th_, _-s_. That Tolkien envisioned irregu-
larities in the Elvish languages is well attested, whether it be the
"irregular vocalism" of Q. _málo_ 'friend' < base MEL- (V:372),
the "irreg. pret." _(e)halle_ of Q _halta-_ 'to leap' (PE12:39), the
"quite irregular" Gn. verb _na-_ 'is', or the "irregular pret."
_tîli_ of Gn. _teltha-_ 'cover in' (PE11:70), to cite but a few
examples. Such deliberate irregularities were a conscious part
of Tolkien's linguistic artistry, adding an element of realism to
his languages that would otherwise be lacking were they perfectly
regular. To attempt to dismiss _mudas_ as a pa.t. simply
because it is irregular or "abnormal" fails to recognize this
aspect of Tolkien's creative process.
Fauskanger also proposes an alternative means of explaining away
the troublesome form _mudas_ as a pa.t. verb: "Alternatively, what
Christopher Tolkien transcribed as 'pa.t.' may well be a highly
tentative reading of an illegible scrawl in the original manuscript;
he describes the M-stems of his father's manuscript as 'faint and
difficult to interpret, and some are very confused' (LR:370)." This
is, fortunately, not the case with this particular entry -- I can
confirm, having examined my photocopy of the entry for the
base MÔ-, that the reading "N _mudo_ (pa.t. _mudas_)" is clear
-- Patrick H. Wynne
- Patrick H. Wynne wrote:
>This early Noldorin pa.t. ending _-th_ might be cognate...
>with _tha-_ 'to make, cause to be' (< *_s'ta-_), given in the
>"Noldorin Word-lists", a verb that is also said to be the source
>of the causative suffix _-tha_ (PE13:153).
>Exactly how the etymologiesCertainly not, especially since there is some evidence that a Noldorin
>of Gn. _-thi_ and N. _-th_ proposed above relate to the
>etymology Tolkien envisioned for N. _-s_ in _mudas_ is unclear,
>to say the least -- but it _is_ clear that N. _mudas_ cannot
>be simply dismissed as an isolated anomaly.
word with suffix _-s_ actually may be the cognate of a Quenya word
with suffix _-sta_: That would be _pennas_ , "history" (cf. KWET- in
Etymologies). The Quenya cognate is _quentasta_, obviously, cf.
VT39:16. The explanations given there show that _-sta_ means
some collection or particular grouping of records (_quenta_). That
_pennas_ is indeed the Noldorin cognate is confirmed by the variant
_gobennas_: the prefix _go-_ means "together" (cf. WO- in
Etymologies), apparently followed by a lenited form of _pennas_. So
this is a collection, too. This shows that the Noldorin reflex of _-sta_
may be not only _-th_ (as was suggested in VT 39:20), but also _-s_.
The difference seems to be merely phonological: in the
same entry in Etymologies, there's also a word _gobennathren_
"historical". By the way, it seems likely that _certhas_ has the same
origin, it's a particular grouping or collection of runes.
The fact that most Noldorin words ending in _-as_ are nouns isn't
very conclusive: it doesn't mean they were nouns always. German
nouns like "Hieb, Tat, Schliff, Stand" are identical with pa.t.sg. of the
verbs "hauen, tun, schleifen, stehen". Indeed, a blow/stroke is the
result of hewing, a deed is the result of doing, etc. As I pointed out
in message #197, some Quenya nouns are formed from verbal roots
in the same way as what Carl called "Strong past I" in his recent
post. Those nouns have the same semantics as the results of the
corresponding actions. Of course, both the German and the
(speculative) Quenya examples would be strong past tense forms.
Whether a Noldorin noun _caras_ (KAR- in Etym) really just
meant "built", originally, is another question and would need
much more evidence. The gloss is "a city (built above ground)".
As I said, that's just speculation without more evidence, because
paradigms changed: if an early weak past tense of the verb _gala-_
"thrive, prosper" (GALA- in Etym) was _galas_, that could become
the word for "plant, growth". At the time of Etymologies, the
interpretation could have changed, already: _faras_ "hunt" is said
to derive from ON _(s)pharasse_ (cf. SPAR-).
- On Samstag, Juli 12, 2003, at 06:07 Uhr, Patrick H. Wynne wrote:
> That Tolkien envisioned irregularitiesIn full knowledge that such 'irregularities' derive from earlier
> in the Elvish languages is well attested, whether it be the
> "irregular vocalism" of Q. _málo_ 'friend' < base MEL- (V:372),
> the "irreg. pret." _(e)halle_ of Q _halta-_ 'to leap' (PE12:39), the
> "quite irregular" Gn. verb _na-_ 'is', or the "irregular pret."
> _tîli_ of Gn. _teltha-_ 'cover in' (PE11:70), to cite but a few
regularities (a point not in dispute, I know), which only reinforces
the point made on _mudas_, namely to take the form seriously.
A few comments on the above forms:
1) Q. _málo_ may well be a Telerin form, from MEL- with 'a'- infixion
2) Q. _halta_, pret. _(e)halle_. Leaving aside the anlauting 'e',
_halle_ is a regular past tense form, but not for a _-ta_ verb (be it <
sundóme+t or _-tâ_). Possibly a new present stem was formed in _-t_ but
the old past tense retained. It would seem harder to explain it the
other way round. Maybe the present was reformed to avoid homophony with
Q. _halla_ 'tall' [LR3:507].
3) The 'irregularity' of _na_ is, of course, in line with what is known
from many real languages. In the case of Q(u)enya, we're dealing mostly
4) The preterite of Gn. _teltha_ seems to be in the same line as #2.
Example 2 and 4 apparently show that past tense forms could end up 'in
the wrong category' or rather, that present tense stems exhibit a
tendency to 'renovate' (something which is beautifully mirrored by
Indo-European languages, especially at earlier stages, abounding in
(often concurrent) present stem formations)(1).
So maybe the pa.t. _mudas_ was actually formed from an (earlier)
_mó-_. Thus the 'd' could be interpreted as corresponding to the 'th'
seen in Early
Noldorin verbs (cf., e.g. PE13:131).
Neither th > s nor th > d are trivial in Sindarin/Noldorin be it
internally or externally. So, the interpretation of pa.t. mudas <
mú-da-s as a gender specific past tense form like N. _madathas_
(PE13:131) is just an alternative hypothesis.
We may still be dealing with a writing error, as the forms _mudo-_ and
_mudas_ [V:373] seem to lack the vowel length seen in N. _mûl_
(<_*mól-_) [ibidem] but this is always an awkward assumption.
Lastly it might be considered that we have here a transitory concept
(externally) for a Noldorin/Sindarin past tense, as the _-(a)th_ was
taken by the future/prospective at the time of the Lord of the Rings
(cf. _linnathon [LR1:114]).
(1) Innovation in the field of the present stem/tense can also be seen
in other language families, e.g. Uralic, Altaic etc.)
- There has been further discussion of the Noldorin past-tense verb
_mudas_ *'laboured, toiled', continuing topics first broached on this
list, on the Elfling discussion list (with a further contribution from
myself made on the Elfling-d discussion list, since David Salo banned
me from Elfling). Here is a list of links to the relevant posts, in the
order they appeared:
From Patrick H. Wynne:
From Helge K. Fauskanger:
From Patrick H. Wynne:
From David Salo:
From Patrick H. Wynne:
From Carl F. Hostetter:
From Helge K. Fauskanger:
From Patrick H. Wynne:
- Carl F. Hostetter posted a slew of links to posts from an Elfling thread about
_mudas_ as the somewhat unexpected past tense of _mudo-_ in Etym. I do not
at the moment have anything relevant to say on that topic, but I'd like to
comment on one of Carl's tangential comments in said thread, namely:
> For that matter, what "historical explanation" can David offer for the pluralI assume Carl is talking about the nominal pl. ending _-r_, because as Carl is
> ending in _-r_ in Quenya? None, in fact, because it is a Quenya innovation.
perfectly aware the verbal pl. _-r_ is well attested in both Quenya and
Sindarin, strongly suggesting that is inherited from Common Eldarin.
I have repeatedly suggested that the Q nominal -r is not an innovation "out of
thin air", but simply the verbal ending applied also to nouns. Whether JRRT
actually imagined this way is now, as far as I am aware, impossible to say,
but one might well think it represents an "historical explanation".
[Andreas is right that I was referring to the Quenya nominal general plural ending
_-r_. And his suggestion that this _-r_ might have arisen from the verbal
personless plural ending is indeed a strong possibility (and has in fact been
bandied about by Tolkienian linguists for decades now). But in the specific context
in which I wrote my comment, even this hypothesis does not seem to represent
sufficient "historical explanation" for this _-r_ of the sort David Salo requires for
the Noldorin pa.t. ending _-as_; for if it did then he could, for example, similarly
suppose that _-as_ arose as a verbal application of the ending *_-ssê_ evidenced
in Eldarin abstract nouns, or that it represents a remnant of a long form in *_-ss-_
of the apparent 3rd sg. ending *_-s_ seen in ON _persôs_ 'it affects, concerns'
(< PERES-). No such verbal application of *_-ssê_ or application or long-form 3rd sg.
pronominal ending is evidenced in Quenya or elsewhere in Noldorin (at least, not
that I can think of at the moment, please correct me if I'm wrong), but that in no way
exlcudes the possibility that such existed in Eldarin or arose independently in
Noldorin. The point being, and remaining, that mere absence of an obvious or secure
"historical explanation" evidenced by more than one language does not render a
grammatical form or feature anomalous, and certainly not erroneous, despite David's
apparent argument that it does. CFH]
- Carl F. Hostetter wrote:
> [Andreas's] suggestion that this _-r_ might have arisen from the verbalThat's interesting to know - when I first brought up the topic on Elfling a
> personless plural ending is indeed a strong possibility (and has in fact been
> bandied about by Tolkienian linguists for decades now).
few years ago, I did, as far as I can recall, not get any indication the idea
wasn't new. (The original context, BTW, was what ending Amanya Telerin may
use on verbs - Helge's Ardalambion article at the time suggested _-i_, which I
found unlikely given that Q and S both have _-r_. Has any evidence on this
come to light in the years since?)
[You'll have to remember that there have been at least three distinct
generational "waves" of Tolkienian linguists, starting with those centered
around _Parma Eldalamberon_, Robert Foster's _Guide to Middle-earth_,
and Jim Allan's _An Introduction to Elvish_ back in the '70s, including such
still-active scholars as Christopher Gilson and Bill Welden; then joined by
those participating in _Quettar_ and (later) _Vinyar Tengwar_ in the '80s
and '90s, including myself, Arden Smith, and Patrick Wynne; and finally
those participating primarily on the Internet in the latter half of the '90s
until the present, including Helge Fauskanger and (to a much lesser extent,
at least overtly) David Salo. Most of those who joined the endeavor only with
the rise of the Internet seem quite unaware of their predecessors, the true
pioneers of the field; a blindered view unfortunately fostered by the most
vocal participants and founders of the main Internet fora. CFH]
Regarding possible "historical explanations" of _-as_: Since no explanation
not coming from JRRT can be regarded as certain, the issue is, or ought to be,
whether we can offer a probable historical explanation. While nominal _-r_ <
verbal _-r_ seems a convincing enough explanation to me, I can't think of any
convincing one for a past ending _-as_. Now I, unlike David apparently, do
not see this as much of a problem - as you've mentioned there's quite enough
Sindarin endings of whose origins we can say very little - but I do think there's
[Agreed on all counts. I ought to have noted that I didn't offer those ideas as
real proposals, only as illustrative examples of the sorts of explanations one
might offer for consideration. CFH]
> The point being, and remaining, that mere absence of an obvious or secureI certainly agree on that. I'd still consider _mudas_ rather 'anomalous' -
> "historical explanation" evidenced by more than one language does not render
> a grammatical form or feature anomalous, and certainly not erroneous, despite
> David's apparent argument that it does.
despite Patrick's listing of more-or-less similar forms, it remains an isolate
within the Noldorin of _The Etymologies_.
[I think it is generally unwarranted to assume that sparsely or even uniquely
attested formations _in languages that are themselves sparsely attested_, of
which the Noldorin of _Etymologies_ is one (and Sindarin of _The Lord of the
Rings_ even more so), are necessarily isolates. They may only appear to be
such due to the selective vagaries of records preservation (and, in the case of
invented art-languages, of records _production_). Moreover, the idea that such
things as linguistic isolation need to be decided and declared, one way or
another, arises only when one departs from language description, and begins
to construct rules purporting to prescribe what is "normal": itself a comically
absurd thing to do for any sparsely-attested language. CFH]
- On 14.11.2003, at 20:30, Andreas Johansson wrote:
> I have repeatedly suggested that the Q nominal -r is not an innovationTypologically that would, AFAIK, be unique. Glottogonically speaking
> "out of thin air", but simply the verbal ending applied also to nouns.
the reverse would be more likely. In many languages verbal inflection
is basically a nominal part + pronominal, or personal endings. Thusly,
_síla_ would originally mean *'shining one, a shiner' to which personal
forms are added: *'Shining-I', *'shining-you' etc... At least in the 3rd
persons we only have a specific (originally) pronominal ending when no
subject precedes the verb (cf. UT:317). That, of course, makes sense
when _síla/sílar_ are originally nominal forms: *'the star, a shiner',
*'the stars, shiners' but *'may be guarders', who? they! == _tiruva-nte_.
Conversely, there would be no apparent motivation for two sets of
endings if both were purely 'verbal'.
So, whatever its ultimate origin, the Q. plural marker _-r_ seems to be
entirely nominal in origin.
- Quoting David Kiltz <dkiltz@...>:
> On 14.11.2003, at 20:30, Andreas Johansson wrote:[snip]
> > I have repeatedly suggested that the Q nominal -r is not an innovation
> > "out of thin air", but simply the verbal ending applied also to nouns.
> Typologically that would, AFAIK, be unique.
> So, whatever its ultimate origin, the Q. plural marker _-r_ seems to beI'm not about to question your superior expertise in these matters, but early
> entirely nominal in origin.
Quenya was apparently happy to use verbs as nouns; "Quendi and Eldar" informs
us that _Vala_ was originally a verb _vala-_ "has power", and offers the
translation "they have power" for _valar_ (XI:403). Could this not represent a
way in which a verbal ending might have sneaked into nominal inflection?
[Perceived "superior expertise" should never be an issue on this list. The
only one with superior expertise is J.R.R. Tolkien, and arguments should
stand or fall based on the evidence in Tolkien's writings, not on the
authority of the scholar proposing a particular theory.
The passage Andreas refers to above also cites _eques_ as a Q. verb
form that also came to be used as a noun. Earlier in Q&E Tolkien writes:
"In Quenya the form _eques_ originally meaning 'said he, said someone'
(see Note 29) was also used as a noun _eques_, with the analogical
plural _equessi_, 'a saying, dictum, a quotation from someone's
uttered words', hence also 'a saying, a current or proverbial dictum'."
(XI:392) -- PHW]
- Quoting Andreas Johansson <andjo@...>:
> Quoting David Kiltz <dkiltz@...>:[snip]
> > On 14.11.2003, at 20:30, Andreas Johansson wrote:
> > > I have repeatedly suggested that the Q nominal -r is not an innovation
> > > "out of thin air", but simply the verbal ending applied also to nouns.
> > Typologically that would, AFAIK, be unique.
> > So, whatever its ultimate origin, the Q. plural marker _-r_ seems to be
> > entirely nominal in origin.
> I'm not about to question your superior expertise in these matters
> [Perceived "superior expertise" should never be an issue on this list. The
> only one with superior expertise is J.R.R. Tolkien, and arguments should
> stand or fall based on the evidence in Tolkien's writings, not on the
> authority of the scholar proposing a particular theory.
> -- PHW]Clarification: I meant superior expertise as regards what is and what is not
found in primary-world languages, not Tolkienian ones.
I, however, see that my snipping above made Patrick's misinterpretation pretty
much inevitable, for which I apologize.
[No apology is necessary -- my comments regarding "superior expertise"
were not meant to _admonish_ you, but to _encourage_ you to not indimidate
yourself into abandoning a theory purely on the assumption that others have
a broader knowledge. And this is as true regarding references to primary-
world languages on this list as it is to Tolkien's languages. -- PHW]
- It will be best to refer to JRRT himself for an answer... even though
there will be more than one. Unfortunately, I don't own PE 11, so I
have to quote after a secondary source, VT40, which used the early
lexicons to analyse Narqelion. Here's a comparision of case endings
(genitive-ablative) in Qenya and Goldogrin, singular and plural: "with
-ion cp. Q -ion, both being double plural -i + ô + n; with -a cp. Q
-o, [from] ô; with -thon cp. Q -ron, where -r- is from the
nom[inatives,] for -son; with -n cp Q -n" (VT40:9/10).
This is supposed to mean that both -i and -n were plural markers, and
that -r is a nominative (plural, obviously) coming from rhotacism and
compares to Goldogrin -th. So it isn't an innovation at all: "-th is
original and [the] same as Q -r".
Obviously, JRRT hesitated whether this is was the right way, and
"The existence in G. of an -r plural sign in verbs has given rise to
the conjecture (coupled with [the] Q. form gen.pl. -ron) that G -th
does not represent Q -r[,] but that -r is a true plural ending (i.e. r
liquid) and -tt == Q -t dual". (both VT40:22, from PE11:10)
This would mean three original plural markers -i, -n (from former -m,
as the entry 3O- in Etymologies suggests) and -r. We are told it is
also a plural marker in G verbs, but that seems to be secondary.
At that time, -r was not always a plural marker in Q verbs, as
examples in the Secret Vice poems show: "i lunte linganer... i súru
laustaner" (MC:216), the subjects (boat and wind) being singular.
The above quote seems to indicate that G -th was originally dual. It
may be that Noldorin -ath was interpreted as dual in origin, too, but
we know that this notion was dismissed, later. "ath: Though it cd. be
an S. form of Q. atta '2', it is not in fact related, nor a sign of
dual". (Letters: 427)
So, externally speaking, we have -r as a noun plural in Q (even in
Qenya) before it became a plural marker in Q verbs. There's also no
hint at an internal derivation devised later.
People seem to be surprised because this gives 3 original plural
markers in Quenya. The surprise may be provoked by the fact that most
modern European languages have only one (English has one and a half,
remember "geese" and "mice"). German has -e, -er, -(e)n, -s, depending
on the noun, and the occasional Umlaut, so why should Quenya have only
one or two?
- On 20.11.2003, at 07:40, Andreas Johansson wrote:
> <snip>Good point. Yet the question is, in my opinion, how frequent are this
> early Quenya was apparently happy to use verbs as nouns; "Quendi and
> Eldar" informs us that _Vala_ was originally a verb _vala-_ "has
> power", and offers the
> translation "they have power" for _valar_ (XI:403). Could this not
> represent a
> way in which a verbal ending might have sneaked into nominal
kind of derivations ? Of course, from a *synchronic* point of view
_vala_ and _valar_ are verbal forms (whatever their ultimate origin).
Note, however, that Tolkien says "...these words are from the point of
*Q* structure verbal in origin..." (emphasis mine). This doesn't, IMHO,
say anything about their *Eldarin* origin. And yes, in some cases a
'zero derivation' seems possible. _Ea_ is another such case and,
slightly different _eques_ cited by Patrick H. Wynne. Such direct
nominalizations do also, e.g. occur in English, cf. something like _a
caveat_. However, as far as I can see, such derivations are rare at
best in Quenya. Other agental construction show derivational morphology
and are attested much more amply (e.g. sundóma +r(o), -ô, -mo etc.).
The words _Vala_ and by all probability _Ea_ are translations of
Valarin words. I wouldn't be surprised if that played a role in their
peculiar derivation. _Eques_, on the other hand, was deliberately
re-interpreted with an analogical plural _equessi_ which exactly shows
*no* verbal morphology. So, at least in the case of _eques_ it is not
really correct to say that "Quenya uses verbs as nouns".It is
interesting in this context to ask why the plural of _Vala_ isn't
+_valante_. Possibly, in the case of _vala/Vala_ the same is true.
So, while your point on _valar/Valar_ is a very acute and enticing
observation, I still doubt that these, apparently few, forms could have
caused the creation of an entire plural paradigm. Moreover, if indeed,
the plural of the verbs would have been taken over by nouns, I wonder
why they didn't in the case of nouns in _-e_ as there must have been
lots of instances of past tense plurals in _-er_. ( _Tyeller_ [LR3:502]
might be interpreted in that way, but it is, as far as frequency is
concerned, an exception).
- Quoting Hans <gentlebeldin@...>:
> "The existence in G. of an -r plural sign in verbs has given rise toI'm not clear why you assume the verbal -r in G to be secondary? The passage
> the conjecture (coupled with [the] Q. form gen.pl. -ron) that G -th
> does not represent Q -r[,] but that -r is a true plural ending (i.e. r
> liquid) and -tt == Q -t dual". (both VT40:22, from PE11:10)
> This would mean three original plural markers -i, -n (from former -m,
> as the entry 3O- in Etymologies suggests) and -r. We are told it is
> also a plural marker in G verbs, but that seems to be secondary.
you quote does not appear to say either way.
> People seem to be surprised because this gives 3 original pluralWell, multiple pl formations are common enough in modern Europe, aren't they?
> markers in Quenya. The surprise may be provoked by the fact that most
> modern European languages have only one (English has one and a half,
> remember "geese" and "mice"). German has -e, -er, -(e)n, -s, depending
> on the noun, and the occasional Umlaut, so why should Quenya have only
> one or two?
Besides German, we've got the rest of Germanic family; Dutch has -en and -s,
Swedish has -ar, -er, -or, -n and -0 (zero), and so on. Italian has a couple,
as has Rumanian, if I remember correctly. And if you count 1.5 for English,
I figure you'd get something similar for French. I've heard Welsh has nineteen.
More on topic, there's of course no reason Quenya could not have had three or
more inherited nominal pl markers. It's just that that we know that in the
scenario as JRRT imagined it in later years, -r was a Quenya innovation, at
least as a pl marker on nominatives; we've for instance got _Banyai_ as an
early nom pl of _Vanya_ in PM:402.
I guess it's always possible that nominal pl -r is an innovation _only in
nominatives_ - there's to my knowledge no evidence to say whether the -r in
allative pl _-nnar_ and ablative pl _-llor_ is "original" or not. But since
these case forms are relatively infrequent, we'd rather expected the
nominative pl to spread to them rather than vice versa.
> People seem to be surprised because this gives 3 original plural markers inOf course, you have cited *three* English formations (-s, umlaut, and zero
> Quenya. The surprise may be provoked by the fact that most modern European
> languages have only one (English has one and a half, remember "geese" and
> "mice"). German has -e, -er, -(e)n, -s, depending on the noun, and the
> occasional Umlaut, so why should Quenya have only one or two?
ending). There is also the -en plural formation (ox-oxen, brother-brethren,
extended analogically to computers in VAX-VAXen), and the borrowed Latin -i or
-ii which is more often misused than used correctly. The -s formant has been
spreading through the vocabulary at the expense of the others for centuries,
but enough remnants exist for naive native speakers to have a feel for their
Why would you expect JRRT, a Germanic philologist, to stint on plurals in his
Rich Alderson | /"\ ASCII ribbon |
quenya@... | \ / campaign against |
"You get what anybody gets. You get a lifetime." | x HTML mail and |
--Death, of the Endless | / \ postings |