Re: Middle voice of intransitive?

Hi Henrik,

First message to the list. Greek, yay! Please all, be gentle.

I'm not sure I would really agree with your analysis of the Greek Middle, though
it's a perfectly reasonable interpretation of what *A* middle could be or do,
and perhaps is and does elsewhere. The function of the middle in Greek is
really only loosely connected to the valency considerations you raise, and is
really as much of a functional semantic colouring as a formal category. Its
meaning is more often (a) reflexive, reciprocal, or self-benefactive (i.e.,
essentially functional and inevitably idiosyncratic), than (b) intransitive
from transitive or concerning self-working vs. externally-powered or -incited
motion (i.e. essentially formal), though quite often (c) it is none of the
above, just how it turned out to be through force of habit, alignment of the
planets, PIE inheritance, etc. I include in this category all cases where (a)
could be argued on semantic grounds, but where the verb is deponent -- meaning
no contrast is being discretely marked as in classic (a) or (c)!
  cases, though one generally see, semantically, why it has middle morphology

The archetypal (b) is the verb (isthmi, "to stand", where as an active it is
causative "to make stand" and as middle it is "to stand". (You're a German
speaker, I presume? Just like the split between strong and weak forms of hangen
and others.) There are a lot of verbs that can do this, like louw, for example,
"to wash". There are many verbs in this category but not all the verbs you
would expect to be able to do this actually do it. I expect this is the
category you are primarily thinking of from your discussion below. Under your
analysis, you reasonably might imagine "swim" as a motion verb to behave this
way, although I have a hard time imagining a contrasting "active/causative"
semantics in the real world. ("Well, the bugger drowned on us. Whadda we do
now?" "I dunno, Phil, get out Smyth's grammar.")

Well, Greek is Greek.... This verb quite often does go into the middle, and
increasingly over time (though it appears in Homer, probably for metrical
reasons), but probably not for the reason you suspect. In fact, I would say
that there are very few cases of pure motion verbs acting as (a). The basic
verb bainw ("to go") has no middle in the present (though the future bhsw vs.
bhsomai obeys perfectly), and a specially derived causative bibazw, while the
basic verb "to come", erchomai, is always middle in the present and always
active in the aorist. It has no causative that I can think of.

I haven't reviewed all the citations in the LSJ for "to swim", but I would bet
that the meaning in the middle was more along the lines of (a) at first -- "I'm
swimming for some personal, self-imposed reason, perhaps for exercise or to
escape the Phoenecian marines who are hurling spears into the water trying to
skewer me", and then later switched to category (c), simply conventionalized as
middle through use...indeed, in late prose, the verbs for swim appear to be
almost deponent. (Again, I haven't done a study of this, just a quick peruse
through the lexicon.)

This (a) use is quite large, and provides quite a lot of opportunity for
hair-splitting pedantry in undergraduate intermediate reading courses....
Sometimes you just have to throw up your hands: it's Greek, there are no rules
here. ("The Greeks and the Irrational" was not a grammar but might have been!)
Ssometimes, however, these (a) uses are quite typified to the point of becoming
(c): the great example here is lyw/lyomai. In the active, it means "to loosen,
to untie, to dissolve, etc."; in the middle, it means "to ransom someone".
Parallel is graphw "to write" vs. graphomai "to indict". No explaining that
through formal analysis!

There are, by the way, a large number of intransitive middle deponents (or near
deponents...again, no rules!) in Greek. I'll finish this long, long message
with a bit of data from an old vocab list for you to enjoy, including some
interesting examples of this class, which I will call (d).

(amphi)penomai -- to toil, to work for one's living (c/d)
(kata)koimaw -- to lull to sleep vs. (kata)koimaomai -- to fall asleep (b)
(ap)arneomai -- to deny utterly (c/d)
sebomai -- to feel awe, to revere (c)
stochazomai -- to aim at, to guess (transitive, btw.) (c)
agoraomai -- to hold assembly, to speak (c/d)
anainomai -- to refuse (c)
anachazomai -- to withdraw, retreat in battle (a if you are a coward, but otherwise c but
as a motion verb, also perhaps b)
atyzomai -- to be distraught or confused (c/d)
(amillaomai -- to compete (c/d)
(aptomai -- to grab (c)
boulomai -- to want (c)
gignomai -- to become (a?/c/d)
dechomai -- to take, receive (c)
dhleomai -- to hurt, be hurtful (c)
hgeomai -- to be leader, to tell (c)
(hdomai -- to enjoy (c, perhaps formally d as it takes a dative object)
(ikneomai -- to come, to arrive (c/d)
ktaomai -- to acquire (c)
manteuomai -- to consult an oracle (c/d)
memphomai -- to blame (a if you are human, but for our purposes c)
mhdomai -- to devise, to contrive, to conspire (c)
orcheomai -- to dance (c/d)
petomai -- to fly (c/d)
politeuomai -- to act as a citizen in public life, to govern (c)
wneomai -- to buy (c)

Evan

P.S. I leave out the famous and interesting class of verbs, usually about sensory
perception, which are active in the present and middle in the future. But
discuss!)


On 9/25/08 7:24 PM, "Henrik Theiling" <ht@...> wrote:

> Hi!
>
> I was wondering whether Ancient Greek had middle voice forms of
> intransitive verbs.
>
> The question arose when I asked myself whether the English sentence
>
>   'I don't swim.'
>
> which is translated as 'I cannot swim' into German (it the right
> context), was middle voice.
>
> English has not overt marker for middle voice, of course.  But to me,
> it feels like that, because 'swim' in the active voice has an
> implicit, here reflexive causation/volition/agent, namely the subject
> itself.  Because middle voice deletes the semantic agent (usually
> promiting the object to subject, which would be impossible for
> intransitives), I was wondering whether 'I don't swim' could be
> analysed as middle voice, and whether Ancient Greek had cases where it
> was overtly marked.
>
> Bye,
>   Henrik