Re: Subject vs. object vs. agent
- On Mittwoch, Mai 14, 2003, at 06:03 Uhr, Carl Hostetter wrote:
> When you used "subject" as a shorthand for the "logical subject" ofThe distinction logical vs grammatical subject is, IMHO, in general
> the sentence, others (quite naturally, I think) took it to mean the
> _grammatical_ subject of the sentence. This was not really a matter
> of form, but of grammatical vs. (your view of) logical function. But I
> still haven't been persuaded that the predicate prepositional phrases
> in the sentences under discussion are in any way a subject, logical or
> otherwise. CFH]
The definition of "grammatical subject" is apparently based solely on
[Not at all. I think of the grammatical subject of a sentence as that
part of the sentence filling the role or function of the subject,
regardless of its form. You are quite right to have noted that form and
function must be kept separate (no matter how much they might coincide
at times); but that is not the failure of distinction here. As I see it,
you are taking a predicate prepositional phrase == prep. + grammatical obj.
of prep., and calling this the "subject" of the sentence, by which you mean
"logical subject". I have yet to be convinced that this has any validity,
but even if it does, it was your use of "subject" for (whatever you mean by)
"logical subject" that caused the confusion. CFH]
I think this view doesn't do the language justice. If, syntactically,
the subject is the first participant of a verb (or the only one of the
copula) then it is most prominently encoded as nominative but it
doesn't have to be. As you accepted, "to be" doesn't take an object.
What then is "thee", the object of a preposition ? I never heard
anybody call it that.
[Oh, but I have, many, many times, from many different sources,
spoken and written. I daresay anyone educated in grammar in an
English-speaking school will have learned to call such the "object of
the preposition". CFH]
The term object is ambiguous here since it is applied to two
syntactically (not formally) completely different phenomena. What I'm
aiming it, is establishing a relation between the syntactical terms subject
and object vis-à-vis case endings. Unlike Ivan Derzhanski I don't think
this obscures anything because there is no one-to-one relation between
those categories. I still don't see how it could make sense to say a
preposition takes the accusative and dative, hence we give it the same
name as a syntactical relation to a verb.
[And yet we do, and for very long now have. CFH]
Why bother with syntax anyway, if we don't make a distinction? I'm not
trying to obscure an equation that doesn't exist but to illustrate a
Out of curiosity: How would you define the function of subject and
object in a sentence?
[I don't think I would attempt to; I'm not that kind of linguist. But I
knows 'em when I sees 'em. CFH]
- [This discussion is drifting far afield from Tolkien's languages -- whatever
the linguistic merits of the discussion, we can be pretty certain that Tolkien
never concerned himself with "pivots" or "binding theory prominence" while
practising his language-making Art. If anyone wishes to continue this
discussion, it will have to be brought 'round to Tolkien again. CFH]
Please pardon me if I crash this thread.
Yehuda Falk of the Hebrew University has written a very nice (short,
clear, to the point) paper on the distinction between different notions
of "subject." He summarizes varies threads of research going back to at
least the early 1970's, and motivates a basic distinction between the
Pivot (grammatical subject, discourse-functional subject), and Logical
Subject (most 'salient' thematic participant). The distinction is
backed up by a battery of criteria drawing on a wide range of
languages. It's available online as a PDF:
Distinguishing between the Pivot and Logical Subject might go a long
way toward clearing up some of the terminological issues addressed in
the preceding posts. Whether or not one chooses to adopt the
terminology, the evidence clearly supports the distinction between a
grammatical subject and a semantic subject.
Frederick M. Hoyt
University of Texas at Austin
LIN 312 - The Linguistics of Middle-earth website:
- David Kiltz wrote:
> A functional object, as I understand it, is the second participant of aPlease excuse me to intrude on your learned discourse, but it seems
> verbal sentence. Cf. "I (subj.) see you (obj.)". In a nominal sentence
> there is no object (functionally or logically). "The lord is with thee"
> is functionally the same as "Thou art with the Lord" or "The Lord and
> thou art together" (with semantic nuances, but that's irrelevant at
> this point).
to me that the differences are more than "slight". It seems to me that
arguing that these three sentences are "logically the same" is to make
the same mistake as many logicians do when they claim that since a
proposition A1 is equivalent to A2, they are one and the same (and
according to the logician's preferences, either A1 is claimed to be
only an inadequate expression of which the true logical form is A2,
or vice versa).
In my opinion, at least on the logical level there is a clear distinction
between the proposed three sentences. In "The Lord is with thee",
"The Lord" is the logical subject, of which "being with thee" is
predicated. In "You are with the Lord", the subject is "You",
and in the last one the subject is "The Lord and thou".
Although these sentences are _logically_ equivalent, i.e. they imply
from each other mutually, they are in fact essentially different:
they have different subjects and predicates. In the first you say
about the Lord that he has some relation of proximity to "thee", in
the second you say that "thou" have some relation of proximity to the
Lord. As theologians would claim, these relations are not the same and
differ radically in their ontological nature.
I am not exactly aware of what you mean by "functional", so I will
refrain from arguing whether they can be said to be "functionally" the
same. However, my belief is that grammar serves to express in some way
the logical structure of thought (or language, if you will). Therefore
it seems to me that if there is a logical distinction between two
sentences, and if there is some distinction in their grammatical structure,
that can be seen as corresponding to this logical distinction, then it
is unsubstantiated to deny that the grammatical structure reflects directly
the actual logical strucxture of these sentences, and can therefore
not be dismissed as mere surface variation in expressing one and the
same thing (thought or proposition).
> Maybe, it boils all down to a question of terminology. Just as in aI think I have to disagree from the (philosophical-)logician's point of
> sentence "the window was hit by a bullet" "bullet" is the logical
> subject, so is "with the" in the sentence in question.
view. In this sentence the subject indeed is "the window". It is
the window that is the object of the mental act of characterizing,
known as judgement, therefore it is the window what is the subject of
the judgement from the logical point of view; and the characteristic
ascribed to the window is certain passion it suffers from the bullet,
i.e. it is the characteristic "that which has been hit by a bullet" is
the predicate of the judgement. Of course, this proposition or
judgement implies a _distinct_ judgement about the bullet, ascribing
it an _action_, namely that of having hit the window. And it also
implies many other distinct judgements, e.g. a judgement about the
existence of the action of the bullet's hitting the window, etc. They
are logically equivalent, but that does not mean identical.
I think I can trace why it seems that the "logical" subject of passive
sentences is the agent. It is because the real agent is the
ontological subject of the action. But the real patient is also
the ontological subject - of its passion. And regardless of all that, the
question of what the _logical_ subject of a judgement or proposition is
does not in any way depend on what the ontological subject is and what
the ontological property is. Nothing hinders you from making an
ontological property, such as action or passion (see the abovementioned
example) the logical subject of your proposition, since as long as
anything is capable of becoming an (ontological) object of human thought,
it is capable of becoming the logical subject of the judgements made
by humans, since to think about something means to ascribe it some
characteristics in judgements, in which it is the subject and the
characteristics the predicate.
Lukas, a would-be philosopher, but not a linguist,
who is nevertheless deeply enjoying this extremely
- David Kiltz wrote:
> Why bother with syntax anyway, if we don't make a distinction? I'm notThere was a rule in scholasticians' disputations that it was the task
> trying to obscure an equation that doesn't exist but to illustrate a
of the one who denied a distinction to prove the identity, not vice
versa. The reason is that a false distinction does not produce
any false implications (it only hinders from inferring some true ones);
whereas false identity does.
- Patrick Wynne wrote in response to Hans' wish "to see some evidence
why _-ni, -le_ are accusative in form":
> One reason for assuming _óni_ contains acc. _-ni_ is phonological,**However, it must be noted that development of monosyllabic
> and a key piece of evidence occurs in the very sentence Hans cites:
> _tye-meláne_ 'I love thee' (V:61). The Etymologies gives the base
> of 'I' as NI- (2), and a consistent phonological rule throughout the
> external history of Q(u)enya is that original short final _*-i_
> becomes _-e_, e.g. *_liñwi_ 'fish' > Q _lingwe_ (V:369 s.v. LIW-).
words (MWs) differed from development of polysyllabic ones.
For instance, it seems that certain short vowels were lost finally
(cf. _abaro_ > CE _abar_, WJ/XI:371). This could not happen in MW,
because the very word would be lost then. Also, we know that long
vowels were reduced to short ones finally. However, MWs seem not
to reduce them, cf. _ní_ "woman" < NÍ (Etym) or _vá_, apparently
from _bá_ (WJ/XI:370). The Plotz Letter informs us that "all long
vowels were reduced to short vowels finally" in Spoken Quenya.
Again MWs seem not follow the rule, because we have _sí_ "now"
in "_Namárie_", for instance.
> Thus it is likely that the subject pronoun _-ne_ 'I' in _tye-**That is certainly possible (though the subject pronoun for
> meláne_is from earlier *_-ni_.
"I" could have been simply *_-ne_ then (i.e. with the original
_e_, not from _i_)).
I think we do not have explicit examples of development of final
short _-i_ in MWs, but it is possible that even in them the _-i_
turned to _-e_. Or at least in the Etym (and _Lost Road_) era.
We may also note that Etym gives _no_ "under", apparently from
*_nu_. This would then show change of final _-u_ to _-o_
(parallel to change of _-i_ to _-e_) in a MW. Nevetheless, "_Namárie_"
gives _nu_ "under" instead. If we suppose that the "_Namárie_"
version of the preposition has the same origin, i.e. *_nu_, then it
might be that Tolkien changed his mind and decided that vowels
(or at least _u_ and _i_) did not undergo any change. Of course, the
origin of _nu_ might have been *_nó_, but we should then ask why
the long vowel was shortened if another long vowel in _sí_ was not.
> Note that the various versions of the Átaremma consistentlyIn a reply to Beregond Patrick wrote:
> maintain the distinction between nom. _emme_ 'we' and
> acc. _me_ 'us', e.g. _emme avatyarir_ 'we forgive' versus _úa
> mittanya me_ 'do not lead us' in At. I (VT43:8).
> I have shown that _-ni_ in _óni_ must derive from *_-nî_, theHowever, _me_ seems to be a nominative form as well, cf. _men_
> lengthened vowel strongly suggesting that it is accusative
> ["because *_ô-ni_ would regularly yield **_óne_"]; I have shown
> that _-me_ in _óme_ is identical in form with accusative _me_
> 'us' in the Átaremma and elsewhere, and that _te_ in _óte_
> appears to be identical to accusative _te_ 'them' in _a laita
> te_ 'praise them'. It is not unreasonable then to suppose that
> the other forms in this same chart, _óle_ et al., are based on
> accusative forms as well, with of course the exception of
> _onye, olye_, in which the endings _-nye, -lye_ are attested
> as nominative.
in the same text. The form is not *_mén_, so it points rather to
nominative. I have mentioned that the form _sí_ "now" did not
undergo the shortening. Now consider _sín_ in SD/IX:310: the
vowel is not shortened when an ending _-n_ is added (whatever
its function). In Etym such a shortening is seen, because the base
SI- lists _sin_ besides _sí_. This is therefore another piece that
suggests that Tolkien changed his mind as regards the development
and behavior of MWs, because the form _sin_ appears in an
earlier version of the _Atalante_ Fragments (see LR/V:46).
Moreover, it is usually nominative (the least marked form) that
acquires case ending (more precisely, it is the least marked form
serves as nominative).
The fact that the reflex of final short CE _-i_ in _-e_ in Quenya is
not, in my view, sufficient for assuming that _-ni_ in an accusative
form, because as I have tried to show the behavior of CE
monosyllabic words is slightly different to the behavior of CE
The question is whether_-ni_, _-le_ etc. in the _ó-_ chart (VT43:29)
are suffixes or whether the _ó-_ is a prefix. What I want to say is
which of the segments could stand alone, that is, which of them is
a separate word -- if any of them.
If _-ni, -le_ etc. are only suffixes and cannot stands as separate words,
I would not speak of them as of nominative or accusative forms but
rather as subject and object forms. Nominative does not necessarily
means subject and accusative does not necessarily mean object. Now
the question is of course whether they are subject or object forms.
It may be they are both (with _nye_ and _lye_ as alternatives).
However, if _ni, le_ etc. are separate words, then we can speak
about nominative and accusative, because the least marked forms
would be nominative from which accusative could be formed.
Nominatives would act as subjects and accusative as (direct) objects
in most cases. Yet here again I do not think we can say whether
they are the former or the latter, since the accusative as a case did
not exist in Spoken Quenya. Now as regards the forms _-s_ and
_-t_ (in _ós_ and _ót_, being variants of _ósa_ and _óta_), they
are hardly separate words. They may be reduced forms of _-sa_
and _-ta_ or plain suffixes, perhaps like _-nye_ and _-lye_, but
these could also perhaps stand alone, cf. _tye_ and _lye_ in _lyenna_.
> Acc. _me_ 'us' occurs in the dual form _met_ 'us two' in _Namárie_ as**If _me_ is an accusative form, we should ask why the long vowel was
> the object of a preposition: _imbe met_ 'between us (two)'.
here shortened if long vowels seem not to be shortened in MWs. The
same with _met_ -- why is it not *_mét_?
> And I would propose that it is this same acc. _me_ that appears in_-te_ in _óte_ on the chart appears to be masculine, [...].
> _óme_ *'with us' in the chart cited in VT43:29. Similarly, the pl. pron.
Does it? I think you meant "personal" (_-ta_ being impersonal), at least
this is what is implied from what is said on p. 20 of VT43. But if you really
meant masculine, what would be the corresponding feminine form?
- --- In email@example.com, Ales Bican <ales.bican@s...> wrote:
> The Plotz Letter informs us that "all longIn fact, the Plotz Letter says so explicitly, the sentence you quoted
> vowels were reduced to short vowels finally" in Spoken Quenya.
> Again MWs seem not follow the rule, because we have _sí_ "now"
> in "_Namárie_", for instance.
continues: "... and before final cons. in words of two or more
syllables". This is obviously connected with stress, remember
that the prefix _ó-_ becomes _o-_ when unstressed (XI:367).
With the retraction of stress, final vowels became unstressed
always, and shortened. This did not necessarily (or never?)
happen in monosyllabic words.
> **That is certainly possible (though the subject pronoun forI can't see any reason to assume that. The _-ne_ in _meláne_ has a
> "I" could have been simply *_-ne_ then (i.e. with the original
> _e_, not from _i_)).
natural explanation, and in any other case I know of, the form is
_ni_ or derived from it. Let's analyze a few occurrences of the 1.
person sg. pronoun in the corpus:
_ni_ "I" (Arctic Sentence)
_Atarinya_ "my father" (V:61)
_meláne_ "I love" (same page)
_inye_ "I" (same page)
_indo-ninya_ "my heart" (V:72)
_nin_ "me" (same page)
_NI_2 "I" (V:378)
This shows a consistent picture up to Etymologies: _ni_ or _inye_ as
"I", _-nya_ or even _ni-nya_ as possessive suffix, _ni-n_ as dative.
The change *_-ni_ >_-ne_ in final position was purely phonological.
A short pronominal suffix _-n_ is found in numerous entries in
Etymologies, too. Again, this fits into the general picture: short final
vowels (since unstressed) were lost often. So we can see two
alternative developments: *_-ni_ > _-ne_ > _-n_, or instead
strengthening of the suffix _-ne_ > _-nye_. It seems likely that
the possessive suffix was formed by combining _ni_ with the adjectival
suffix _-ya_, *_-niya_ > _-nya_. The pronoun remained through all
stages of Quenya. It appeared as a prefix shortly:
_nilendie_ "I have come" (IX:56)
_nimaruvan_ "I shall dwell" (same page)
The dative form _nin_ "for me" appears in Namárie (LR:368) and in the
late notes on _óre_ (VT41:11). Some time between them, we have the
forms _ónye_ and _óni_. As I said already, the argument that the pronouns
are not nominative (or subjective) in form makes sense, in my opinion.
They shouldn't be, because a subject doesn't need prepositions.
> The fact that the reflex of final short CE _-i_ in _-e_ in Quenya isSure, but _-ni_ attached to anything are two syllables at least. Of
> not, in my view, sufficient for assuming that _-ni_ in an accusative
> form, because as I have tried to show the behavior of CE
> monosyllabic words is slightly different to the behavior of CE
> polysyllabic words.
course, Patrick's argument relies on the assumption that the
custom of attaching pronominal suffixes to prepositions (which
obviously did not exist in CE) occurred earlier than the change
of final short -i > -e.
It seems that _ni_ did not occur as a stand-alone word in the corpus
after the Arctic Sentence. _inye_ seems to be derived form an
augmented form *_i-ni_. At least, that would explain the difference
to _elye_ "you" (LR:368).
I'll return to "you" (and other pronouns) in other posts.
- On Freitag, Mai 23, 2003, at 09:00 Uhr, Ales Bican wrote:
> **If _me_ is an accusative form, we should ask why the long vowel wasOrthotone vs enclitic variants ?
> here shortened if long vowels seem not to be shortened in MWs. The
> same with _met_ -- why is it not *_mét_?