- i Héru aselye (VT43:28) - Dominus Tecum - the Lord is with Thee.
A short inquiry into a prepositional complement.
Let's look at the English sentence first. _The Lord is with Thee_ is a
nominal sentence. _Is_ here seems to represent the copula, not a full
verb. More likely so, because we have probably to think of the
association not as strictly speaking spatial or temporal but (also)
metaphorical. The subject of the sentence is clearly _The Lord_. The
occurrence of an object in a sentence is governed by the predicate of
that sentence. The copula doesn't allow for an object. What, then, is
_with Thee_ ? Well, I think it's used predicatively just as _great_ in
_God is great_ or _grass is green_. As it stands with the copula it is
by necessity a subject predicate, just as in e.g. _Jon is a student_.
It describes God.
Now for the grammatical encoding.
How then is it, that in most modern European languages prepositions
demand inflected nouns after them? The answer to that, I believe, is
that all those languages carry historical vestiges in their innards
that survive, with the *function* of the prepositions having
(gradually) changed over time. What are prepositions and how did they
work historically ?
As can be gleaned from information provided by the oldest attested
I.-E. languages (namely Vedic, Hittite and Greek) PIE did not have
preposition strictu sensu. Rather those (as it appears now) little
words were used adverbially or adnominally rather freely. The important
thing, however, is, that they were originally *inflected nouns". Cf.
*_(s)up-o_ (with *-o as directive), *_(s)uper(i) (with *-er(i) as
locative) and *_per-i_ (locative *-i, cf. also Hittite _per_ "house").
Combining one inflected noun (later: preposition) with another demands
(more or less) complex syntax. Now, how did this system work ? Many
notions that are expressed by prepositions in many modern European
languages, weren't in PIE. Rather, they were expressed by means of
(case) endings. Modern examples will be from English:
To say "with" (sociative) e.g. Vedic would use the instrumental, cf.
Rigveda 1.1.5b _devó devébhir aá gàmat "the god (Agnih) shall come
(here) *with the gods*. "With" is expressed by the instrumental pl.
_-bhih_. Similarly Rigveda 1.1.4b: _sá íd devésù gacchati_ "It is
verily he (Agnih) who walketh amongst (the) gods. With _devesu_ loc.
pl. of _devah_ "god".
These two examples should illustrate what I mean. Now, when PIE used
"prepositions" more complex things resulted, as they were themselves
inflected. Let's take an English sentence: _He walked out of the room_.
If PIE would have used a "preposition" in such a context, the English
equivalent would have been something like: "He walked from the room
(gen./abl.) to the outside". Likewise, _with you_ didn't require a
preposition at all. With a preposition something like _together you-at_
== "with you" can be imagined. Now we can see two things 1) Modern
prepositional phrases are *over-characterized". Case endings do
indicate differences in usage (although not typically in English but
cf. German _im Haus_ vs _ins Haus_ "in the house vs into the house")
but, at any rate, they work differently than they did then. 2) An
accusative or dative must also be interpreted as directional cases.
Case endings are much smaller units in PIE. Case endings themselves
seem to (at least some) derive from postposition. To be very brief,
these endings were more like Quenya prepositions (it would seem to me).
They are always attached to previously inflected forms.
So, we see that in Quenya, with its prepositions _ó_ and _as_
apparently not being inflected for case, there is no reason whatsoever
to suppose the (pro)nouns governed by them were necessarily in an
oblique case (unless double marking occurred or a special movement was
to be indicated). _-elye_ in _aselye_ has neither the function (syntactically)
of an object in the sentence, nor seems there to be a reason for _as-_ or
_ó_ to take a specific case.
[But as Pat Wynne has already shown
the attested form _óni_ (with optional full, and apparently nominative, form
_onye_; VT43:29), if the result of regular phonological development, must
be due to the old accusative form (or else we would expect **_óne_). And
other forms in the same chart -- e.g., _óle_ (with optional nominative form
_olye_), _óme_, _óte_, etc. -- could also have originally arisen from
accusative forms with lengthened vowels (and, in the cases of _óme_ and
_óte_, the pronominal part corresponds formally with attested, specifically
accusative, independent forms, sc. _me_ (VT43:8) and _te_ (LR:932),
repsectively). That being said, the optional full forms may indicate a
tendency towards later levelling of the prepositional pronouns to
nominative forms; and in this regard it is noteworthy that other
prepositional charts on the same page (VT43:36n.1) do show full,
apparently nominative, pronominal forms in positions where the first
chart has short forms. CFH]
It remains, however, possible that Tolkien simply mimicked the
situation in I.-E. languages of the stage of Latin, Early Germanic
etc. without really giving much considerations to the syntactical
nature of prepositions. Still, there is a good chance he did. Lastly,
to be able to really judge matters in Quenya, it would be important to
know whether the accusative in Quenya served any "directional" purpose
Grammatical encoding is ultimately motivated and triggered by
semantics, thematic roles. However, every system once established,
focuses on certain aspects, while not marking others. Ironically, my (I
think wrong) point to equate subject == agent, so booed at, is the only
way, I think, to describe _with Thee_ as an object. Note, however, that
this is not correct. Rather, on a thematic level _with Thee_
accumulates a substantial number of (proto-) patientoid
characteristics, while _The Lord_ is clearly a (Proto-) Agent. Typical
features of an agent are, among others, +control, +instigation, while
the typical patient lacks those features. While it may be argued that
Saint Mary "instigated" God's being with her because of her comport and
pureness, she certainly didn't control it. While God clearly is the one
who wills, "instigates" and controls the action.
P.S.: I would like to ask those concerned to forgive me the somewhat
eruptive style of previous replies.
P.S.2: For those interested in that sort of craziness: Some time ago
(when I was less busy) I translated the ring poem into Syriac:
Note there is a mistake in line 3: It should be _labnainosho_. Also,
the Syriac font is less than perfect.
- David Kiltz <dkiltz@...> wrote:
> i Héru aselye (VT43:28) - Dominus Tecum - the Lord is with Thee.There is a difference between an _appositum ex recto_ like:
> The copula doesn't allow for an object. What, then, is _with Thee_ ? Well,
> I think it's used predicatively just as _great_ in _God is great_ or
> _grass is green_. As it stands with the copula it is by necessity a subject
> predicate, just as in e.g. _Jon is a student_. It describes God.
_God_ in _The Lord is God_
_thou_ in _The Lord art thou, O Jesu!_
_great_ in _The Lord is great_
_the Father, the Son and the Holy Ghost_ in
_The Lord is the Father, the Son and the Holy Ghost_
and an _appositum ex obliquo_ like:
_with thee_ in _The Lord is with thee_
_of Abraham, Isaac and Jacob_ in
_The Lord is the God of Abraham, Isaac and Jacob_
The former describe God in himself, the latter in relation to others, and the _apposita ex oblquo_ are oblique cases of precisely those others.
Hans Georg Lundahl
Gå före i kön och få din sajt värderad på nolltid med Yahoo! Express
[Non-text portions of this message have been removed]
- I'm sorry, but I think the moderators are right: we should return to JRRT. In fact,
Russian has FAR MORE passive constructions, than English! The reflexive
construction for passive IS used often, and there are two passive participles
(past and present), not one as in English or German. But that's off topic,
So let's resume after David's "What are prepositions and how did they work
historically ?", looking at JRRT's notions. He often called them (or the particles
from which they evolved) "adverbial elements". Just a few examples:
Q&E, root *AWA. Adverbs _oa_, _oar_, survived in Sindarin as a preposition
_o_ (from, of) (XI:366).
*HO: "ancient adverbial element, occurring principially as a proclitic or enclitic",
it was the origin of a verbal prefix or a case-marker _-o_ (XI:368).
*WO: the origin of a verbal prefix, not surviving as a word according to Q&E
(XI:367), but obviously the origin of the preposition _ó/o_ mentioned in the earlier
translations of prayers (VT43:29).
From Adunaic (at the same time that Tolkien played with pronominal prefixes
in Quenya): the normal case of nouns (zero marking) is also "the base to which
certain adverbial 'prepositional' affixes are added; such as _ô_ 'from', _ad_, _ada_
'towards', _mâ_ 'with', _zê_ 'at'." (IX:429). This is extremely interesting, because
all of it has obvious parallels in Quenya. Remember the suffix _-ma_ for
instruments with which you do something, or _má_ 'hand'. Remember the
preposition _sê_ 'at' (VT43:30), or the locative ending _-sse_, which appeared
as _-ze_ occasionally.
Since prepositions modify the relation of a verb to its object, they ARE adverbial.
This relation could as well be described by an inflection of the object (and the
role of objects IS usually marked by inflections), so they are redundant, in principle.
BUT redundancy is built into language naturally. Moreover, inflections have a tendency
of fading away in languages like the modern Germanic or Romance languages or
Sindarin. Then, the burden lays more and more heavily on the adverbial particles
placed between verb and object, and so they become PREPOSITIONS, i.e. words
PLACED BEFORE objects. Of course, they can have this role before, in heavily
inflected languages like Russian or Latin or Quenya. They are connected with
objects, however, with inflected nouns or pronouns, and since they (for semantic
reasons) combine only with certain inflections, a synchronic view may be that
they "govern" a case, or that the object is the object of a preposition. No, they
don't/aren't, the defining factor is the verb and its relation to the object.
- On Freitag, Mai 16, 2003, at 10:39 Uhr, Hans wrote:
> *HO: "ancient adverbial element, occurring principially as a procliticThis refers probably to the genitive in _-o_, pl _-on_ and probably
> or enclitic",
> it was the origin of a verbal prefix or a case-marker _-o_ (XI:368).
also the "ablative" _-Vllo_. In the latter case, _3o_ or _ho_ is
assimilated to the shorter form _-Vl_ (cf. Entu, Ensi, Enta Declension
VT36:8). _-o(n)_ is attached to the noun either in singular or plural
respectively. But is it attached to a specific case ? The question,
perhaps, can be answered by looking at the chart of the Entu, Ensi,
Enta declension. The first three lines run thus:
1) entu ensi enta en
2) ento ente enta en
3) ento ente etta enda.
The authors of VT36 take the first line to represent the nominative on
the ground of general considerations. I.e., conventionally, the nom. is
written first in any chart. The second row, then, is taken to represent
the forms of the acc. Lastly, the third row is considered as
representing an alternative genitive.
While I agree on the first row, I do not think the interpretation of
the 2. and 3. row likely.
Firstly, the difference between 1 and 2 in masc, and fem., namely
_-tu_, _-si(<ti)_ vs _-to_, _-te_ finds its exact correspondents in the
demonstratives _su/so_. _si/se_ (cf. Et:385 sub S-). The entry in the
Etymologies indicates that the distinction between _su/so_ is not one
of case. Hence the same is likely for _entu_ and _ento_. A possible
distinction might be that one form refers to actual males/females while
the other is used more metaphorically. But that's pure speculation at
this point. It may also be suggestive that pure _en_ occurs in the
first two rows, whereas an (old) accusative is normally characterised
by V+ (e.g. nom. _cirya, vs ac. _ciryá_). We find such a + form in the
third row (_enda_). Note, however, that (en) is given as variant (?).
Lastly, the assumption of an "uninflected" genitive for Quenya seems a
little flawed to me.
Things such as _airetári-lírinen_ or _ambar-metta_ should rather be
treated as compounds. The hyphen seems to indicate that, while such
compounds could apparently be formed rather freely in Quenya. Two
elements of a compound can have various relations. Where two nouns are
involved the most frequent (virtual) case relation is that of a
genitive. That means simply that, the first element functions as an
attribute to the second. So, the fact that _ambar-metta_ can, of
course, be paraphrased as _metta i ambaro_ doesn't allow for the
assumption of an endingless genitive. Note that in Indo-European
languages compounds where the first member is indeed inflected occur.
However, they aren't in a majority. The older the language the less
likely this is to happen.
The authors also note a difference in English between _Here is her
book_ and _the book he has is hers_(VT36:20). I don't see how this is
related to the issue. Suffice it to say that both _her_ and _hers_ are
inflected forms. Also, they don't normally appear in compounds. (Unless
there is a "her-cat" next to a "she-cat"). So I take it the latter
examples or English pronouns strictly refer to the use of the pronoun
_enta_ etc.. Still, how would such a distinction (attributive vs
predicative) work with a *demonstrative* pronoun. Note that _his_ and
_her_ are possessive pronouns. That is, they are inherently
"genitival". The formal distinction _her_ vs _hers_ is a very late
innovation to avoid confusion with _her_ (=¬c/dat.). E.g. **the book
is her == she is the book. There is no distinction for _his_ or _its_
since they're unambiguous. Cf. also _I see the colour of that yonder_
vs _this is the colour of that yonder_. I don't see how you would fit
_enta_ vs _entan_ in here.
In other Germanic languages we find strong and weak forms of the
adjective in these functions but that doesn't bear on the matter. That
means, in my eyes, there is no base for the assumption of a third
In my view, the third row must be regarded as representing the
accusative. If this is correct, we can also answer the question to what
"case" (if any) Quenya inflectional suffixes are attached to. If,
indeed, the third form in the third row _etta_ is an accusative, then
it seems clear that this is not a derivational basis since the
inflected forms of _enta_ all show the stem _ent-_ (with one noticeable
So it would seem likely, that also _-o_ was not attached to the
accusative. But what about _-ho_ ? It was apparently attached to _-Vl_.
This _Vl_ must, it would seem, contain an element _l_ since a final d
would have > r in auslaut. Or maybe *_Vd+3V_ > _VllV_ can be assumed?
Maybe via *_VddV_. What do you think ?
> Since prepositions modify the relation of a verb to its object, theyUnless, of course, we're dealing with a NOMINAL SENTENCE in which case
> ARE adverbial.
they are ADNOMINAL.
- In message 438, David Kiltz discusses the so-called "_Entu_, _Ensi_,
_Enta_ Declension", first presented and analyzed by Christopher Gilson
in _Vinyar Tengwar 36. A few comments:
> The authors of VT36 take the first line to represent the nominative onThis last is a very flat statement, concealing a considerable amount of
> the ground of general considerations. I.e., conventionally, the nom.
> is written first in any chart. The second row, then, is taken to
> represent the forms of the acc.
the actual process of arriving at the proposal that the second row
reflects the accusative inflection, which in fact involves (VT36:18-19)
a detailed formal and structural comparison between the two halves of
the chart in question; i.e. between the inflection of _entu_, _ensi_,
_enta_, _en_, and that of the apparently dual forms of a stem _as-_
(perhaps from demonstrative *_sa-_ 'that' or 'it'; cf. N _ha_ 'it'
(V:385), Q _san_ 'then', i.e. *'at that (time)' (MC:216), and, perhaps,
the conjunction _sa_ 'that' in the so-called "_Merin_ Sentence", for
which see: <http://www.elvish.org/elm/merin.html>).
> Lastly, the assumption of an "uninflected" genitive for Quenya seems aWere those the only examples of an uninflected genitive cited by Chris
> little flawed to me. Things such as _airetári-lírinen_ or
> _ambar-metta_ should rather be treated as compounds.
in forming his argument, I might agree with you. However, Chris also
cited two other, very clear, examples (VT36:20): _Valinóre Yénie_ 'the
Annals of Valinor' (X:200), and _Coron Oiolaire_ 'Mound (of)
| 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." |
[Non-text portions of this message have been removed]
- --- In firstname.lastname@example.org, David Kiltz <dkiltz@g...> wrote:
> On Freitag, Mai 16, 2003, at 10:39 Uhr, Hans wrote:No, this refers to the page I quoted, where we further read that the ancient
>> *HO: "ancient adverbial element, occurring principially as a proclitic
>> or enclitic", it was the origin of a verbal prefix or a case-marker _-o_
> This refers probably to the genitive in _-o_, pl _-on_ and probably
> also the "ablative" _-Vllo_.
adverbial element occurred also as "enclitic, as attached to noun stems (the
usual place for the simpler 'prepositional' elements in PQ)." (XI:368) Since
medial _h_ was lost, this became _-ô_, a genitive inflexion, properly a
partitive genitive. I didn't have in mind the ablative.
> In the latter case, _3o_ or _ho_ is assimilated to the shorter form _-Vl_VT36 doesn't say so, and I doubt it, honestly speaking. The shorter
> (cf. Entu, Ensi, Enta Declension VT36:8).
form with _Vl_ is more likely the result of loss of a final vowel, imho.
> _-o(n)_ is attached to the noun either in singular or plural respectively.In the sentence quoted above, JRRT spoke of "noun stems", not nouns
> But is it attached to a specific case ?
having an inflectional suffix already.
> 1) entu ensi enta enThat's not entirely true: Chris Gilson compared the chart against others,
> 2) ento ente enta en
> 3) ento ente etta enda.
> The authors of VT36 take the first line to represent the nominative on
> the ground of general considerations.
like the Bodleian Declensions and the Plotz Letter, and made a rather
convincing observation concerning the first row of the lower, obviously
dual part. By the way, the conclusion was that the first row is a subjective
case, that is not necessarily the unmarked case, the one which is the base
for inflexions. It was not unmarked in the Bodleian Declensions, but had a
suffix _-n_ in the singular.
> demonstratives _su/so_. _si/se_ (cf. Et:385 sub S-). The entry in theIt doesn't indicate that in my copy. Instead, there are listed three N
> Etymologies indicates that the distinction between _su/so_ is not one
> of case
forms _ho_, _hon_, _hono_ for "he", three _he_, _hen_, _hene_ for "she"
and two _ha_, _han_ for "it". Alternative forms or cases? They are followed
by three plurals _huin_, _hîn_, _hein_. Are they alternative forms of "they"?
Are they the respective plurals of masculine, feminine, neuter gender? Or
are they cases of "they"? The last two forms, _hîn_ and _hein_, look familiar:
_Narvin hain echant_, "Narvi made them", _teithant i thiw hin_ "drew these
signs" (LR: 297) and here _hain_ is certainly accusative, "them", and I
wouldn't be too much surprised if it's literally "the signs their", genitive.
Now if the last three forms can be different cases, why not _su_/_so_ or
_si_/_se_? Which cases, that would be the question. Certainly the first one
would be a specifically subjective one, denoting a subject, i.e. an animate.
> Lastly, the assumption of an "uninflected" genitive for Quenya seems aOf course, there's an uninflected (without quotation marks!) genitive in
> little flawed to me.
Quenya. The same page XI:368 says "though 'possession was indicated by
the adjectival suffix -va, or (especially in general descriptions) by a 'loose
compound'... Orome róma would mean 'an Orome horn', sc. one of Orome's
horns". The flaw in the argument in VT36 is: if genitive is expressed by
word order here, it doesn't need an inflexion, i.e. it isn't a case at all, much
as accusative vanished from spoken Quenya. Now that doesn't mean that
the genitive was always uninflected and expressed by word order alone,
and it would be reasonable if the remains of those inflexions would show
up in pronouns, not in nouns. So the suggestion in VT36:20 is a
possibility... one of several. My guess would be, that the first two lines, much
as in the Bodleian Declensions, denote subjective and normal cases, which are
not exactly like nominative and accusative: subjective case was marked only
when needed. Quenya became a language of nominative/accusative type only
later, and the subjective/normal case reappeared in Adunaic.
> So it would seem likely, that also _-o_ was not attached to the accusative.Of course, it wasn't, see above. Those enclitics attached to noun stems, much
as in Adunaic the normal (uninflected) case was (among other things) "the base
to which certain adverbial 'prepositional' affixes are added; such as _ô_ 'from',
_ad_, _ada_ 'to, towards', _mâ_ 'with', _zê_ 'at'." (IX:429) It's very interesting
how many of the early concepts of Quenya entered into Adunaic, and how
many of them resurfaced after its abandoning.
- On Samstag, Mai 24, 2003, at 01:10 Uhr, Hans wrote:
>> demonstratives _su/so_. _si/se_ (cf. Et:385 sub S-). The entry in the"S- demonstrative stem _su, so_ "he"; _si, se_ "she". What else would
>> Etymologies indicates that the distinction between _su/so_ is not one
>> of case
> It doesn't indicate that in my copy.
_he_ and _she_ be than nominatives ?
In Sindarin, _hain_ and _hin_ (_i thiw hin_ is certainly "the signs
these") can be used for the accusative. But one shouldn't adduce
Sindarin forms as if they were Quenya, I think.
> Of course, there's an uninflected (without quotation marks!) genitiveOf course there isn't. And you just gave the line where Tolkien says
> in Quenya. The same page XI:368 says "though 'possession' was
> indicated by the adjectival suffix _-va_, or (especially in general
> descriptions) by a 'loose compound'... _Orome róma_ would mean
> 'an Orome horn', sc. one of Orome's horns".
so: 'Loose compound'. That's exactly it. So there is no *case*. The
nominative can be used and the function is indicated by word order.
That means, there are no *formal* markings in such constructions, hence
it is not a case. Unlike the accusative, it was never marked in such
constructions. Please, make a distinction between form and function, I
think it's vital here.
> My guess would be, that the first two lines, muchI think that is a very good guess. This would chime in with my
> as in the Bodleian Declensions, denote subjective and normal cases,
> which are not exactly like nominative and accusative: subjective
> case was marked only when needed. Quenya became a language
> of nominative/accusative type only later, and the subjective/normal
> case reappeared in Adunaic.
assumption that only row 3 represents the accusative. In the inner
history of Quenya, then, the accusative would get the _-t_ only later.
- On Samstag, Mai 24, 2003, at 01:10 Uhr, Hans wrote:
>> In the latter case, _3o_ or _ho_ is assimilated to the shorter formBased on what ? My suggestion seems to be phonetically plausible.
>> (cf. Entu, Ensi, Enta Declension VT36:8).
> VT36 doesn't say so, and I doubt it, honestly speaking. The shorter
> form with _Vl_ is more likely the result of loss of a final vowel,
Whereas loss of long -ô seems not.
- On Donnerstag, Mai 22, 2003, at 04:32 Uhr, Carl F. Hostetter wrote:
>> Lastly, the assumption of an "uninflected" genitive for Quenya seems aI think the answer is found in Hans post:
>> little flawed to me. Things such as _airetári-lírinen_ or
>> _ambar-metta_ should rather be treated as compounds.
> Were those the only examples of an uninflected genitive cited by Chris
> in forming his argument, I might agree with you. However, Chris also
> cited two other, very clear, examples (VT36:20): _Valinóre Yénie_ 'the
> Annals of Valinor' (X:200), and _Coron Oiolaire_ 'Mound (of)
> Ever-Summer' (S:357).
> "though 'possession was indicated by the adjectival suffix -va, orAs Hans notes, _Orome_ here has genitival function without (overt) case
> (especially in general descriptions) by a 'loose compound'... Orome
> róma would mean 'an Orome horn', sc. one of Orome's horns.
markings. Such constructions also occur in English: _Mount Doom_ ==
"Mount of doom" Hence, we do not have another genitive case here but
probably something that is formally == nominative but functionally a
genitive of possession or association.
[I agree with your statement regarding functionality, but I would myself be
more expansive than to write that the uninflected genitive "== nominative", as
that seems to imply more than I would commit to. Rather, I would simply
note that uninflected forms in Quenya are used for genitive and nomnative
functions (as well as accusative). Further, I do not agree that we cannot
speak of an uninflected genitive "case"; it seems to me that by your argument,
we can speak neither of nominative nor accusative _cases_ in Quenya, either;
which is clearly not the case (no pun intended). CFH]
- --- In email@example.com, David Kiltz <dkiltz@g...> wrote:
> As Hans notes, _Orome_ here has genitival function without (overt)I don't think this is _the_ answer: simply "case" is used in
> case markings. Such constructions also occur in English:
> _Mount Doom_ == "Mount of doom" Hence, we do not have another
> genitive case here but probably something that is formally ==
> nominative but functionally a genitive of possession or association.
different meanings. This is not a mistake, but general practice, as
the following quote from the entry "case" in the _American Heritage
Dictionary_ may show:
"11 b. Case In some varieties of generative grammar, the thematic or
semantic role of a noun phrase as represented abstractly but not
necessarily indicated overtly in surface structure. In such
frameworks, nouns in English have Case even in the absence of
inflectional case endings."
Consequently, "nominative" is sometimes used to denote a function,
namely being the subject (JRRT calls that "subjective" in Adunaic,
and it is inflected in that language), and sometimes it may denote
the absence of infectional endings, that's called "normal case" in
Adunaic (and may be used for subjects and direct objects).
The logical flaw in VT36, imho, is the conclusion that since an
*uninflected* genitive exists, the only slightly inflected (but
*inflected*!) third row in the chart could be a genitive. It's
possible, but I don't see sufficient evidence.
[Since as you say it _is_ possible, then it cannot be a logical flaw to
propose it. It would however be a fallacy to say that it was _proven_
to be so, but of course Christopher Gilson never said that. CFH]
As I wrote already, I think Quenya made a subjective/normal
distinction at that time (since even the later "Bodleian Declensions"
do so). This means marking of the subject in cases of ambiguity, so
the marking of direct objects would be superfluous, an accusative
inflexion simply wasn't needed.
The distinction _su_/_so_ may have been one of subjective/normal case
earlier, but at the time of the _Etymologies_, it may have been
reinterpreted already, and the final _-u_ was considered an older
form, replaced by _-o_ later. Cf. the entry ÓROK-: "*_órku_ goblin:
Q _orko_, pl. _orqi_."