On middle vs passive in Polish and Russian (and, if possible,

[Zbigniew Kanski]

Dear SEELangs readers.  I recently posted a couple of queries to this
list, one about middles vs passives in R, Ukr, and P; the other, a plea
for glosses of some Pol-dialect data.  Thanks to Z. Kan'ski (whose reply
to both these questions appears below), H. Baran, O. Bobrowski, and E.
Ladna for their considered responses.  I am happy to report that the -no/
-to bibliography that Joan Maling and I have been preparing is on its
way to _JSL_ for publication.  Thanks again to all who have replied over
the past few frenetic weeks (I hope I've acknowledged everyone).  Look for
the first half (A-M) to come out in the spring issue (vol. 3, no. 1).

Best,  (for Joan as well) --Loren Billings (billings at princeton.edu)
----------------------------Original message----------------------------
Loren,

Please find the glosses under your cited examples (I am not quite certain
about the meaning of "ni,s" in (2), the first and only guess is that it's
standard Polish ni.z 'than'(sic!) which in turn might be extended in this
dialect to  "before" (???).

The answer to your "passive" query follows, but I am not quite certain
whether it really clarifies your doubts. Quite a lot depends on terminology
here, too. I am trying to be as terminology neutral as possible.

On Sun, 9 Apr 1995 20:28:05 EDT,
Loren A. Billings  <BILLINGS%PUCC.bitnet at plearn.edu.pl> wrote:

>Zbyszek,
>Enclosed below is the query I sent to SEELangs on passives and middles.
>
>First, however, could you gloss the following data forms (the ^ means the
>preceding high vowel isa glide; it appears to be in phonetic transcr'n):
>
>1.  I^ak'e tu bu^oto, pevno   byu^a        voda           vylana
     what  here mud,  sure   was-f.sg.pst  water-NOM.f.sg. spilt-pst.prt.f.sg.
   'All this mud here, some water must have been spilt.'
>
>2.  (krova) ni,s ,se o,celi,             ne byn.ze ,se,s,c ,ne.zel doi^uno
>     cow  before(?) REFL calve-fut.prf.3sg., not will-be six  weeks  milked
    'a/the cow will not be milked for six weeks before it bears calves.'

>3.  gotowano woda (sic.) (no citation given)
     boil-NO  water-f.sg.ACC(?)  (probably Silesian dialect where ACC=NOM
                 even with feminine nouns distinguished in other dialects)
>    'the water was boiled/has been boiled' OR 'one/people boiled water.'

>Incidentally, for the other diacritics I use the system you used before.
>Let me know what these mean if possible.  They're from an article by
>Bartnicka in =Prace filologiczne= 1969.  Best,  --Loren
>----------------------------Original message----------------------------
>Can anyone enlighten me on the crucial distinctions between passives and
>middles in Pol and Rus (and Ukr).  As I understand the situation (based
>primarily on Babby's framework) for Russian, the passive is expressed in
>Rus using eith an imperfective verb with -sja or a perfective verb in its
>past-passive participial form.  Imperfective verbs in past-prt form are
>all but ungrammatical in CSR.  Perfective verb stems with -sja can only be
>middles (testable because an INST-case 'by' phrase is bad with these).
>The middle is expressed with either a perfective or and imperfective verb
>stem with -sja (with no middles can there be an INST 'by' phrase).  Middles,
>of course, often have adverbials:  _Dver' legko otkryvaetsja/otkryvalas'._
>Is all this also the case with Pol and Ukr?  It would appear to be the
>case with Ukr, but seems to be quite the opposite in Pol (as one comment
>leads me to believe).  Again, I'd appreciate your comments by midday or so
>on Monday, if possible.  Best,  --Loren Billings (billings at princeton.edu)
>
Answer:

1. In Polish, no agent PP can appear with ANY -sja(Pol. si,e) form, whether
you read it as passive or middle and whether the verb (or rather its stem)
is perfective or im-. Therefore you can't really use it as a criterion to
distinguish betw passives and middles. Thus while (i) is fine in Russian,
its Pol equivalent (ii) is * on all counts:
(i) Eta kartina pisalas' Titianom-INSTR 'This picture was painted by
Titian.'
(ii) *Ten obraz (na)malowa,l si,e Tycjanem-INSTR/przez Tycjana-by Titian.
The only grammatical way of saying (i) in Polish is a participial pass.
(iii) Ten obraz zosta,l namalowany przez Tycjana.

2. The perfective/imperfective distinction is relevant only to the extent
that middles as generic/gnomic sentences naturally prefer imperfectives.
Hence you can have good "si,e" middles with imperfective forms of those
verbs which otherwise cannot co-occur with "si,e" and agree with NOM
subjects (deep objects), like "paint" in the above examples or "read",
"write", "eat", each taking a human subject in the active. However, it's
very difficult to generalize even on these, because you encounter lots of
idiosyncracies with particular verbs of the same or related semantic
classes. I have no idea why, for instance, it is much easier to get a
middle with "read" ("Ludlum dobrze si,e czyta." = "Ludlum reads well.")
than with "write" or "paint" (??"Wiersze ,latwo si,e pisz,a" = "Poems
write easily."   ??"Martwa natura ,latwo si,e maluje" = "Still life paints
easily.") In the latter cases Polish prefers impersonal refl. with poems
and still life in ACC and no NOM (and therefore unmarked agr 3sg.neut.).
These preferences seem to coincide with English prefering the
tough-constructions or expl. IT with these verbs rather than simple middles
(It's easy to write poems/poems are easy to write, etc.).

Conclusion: The passive-like refl. construction in Modern Pol. (with
semantic object in NOM subj. and subj.-verb agr.) is good only with
imperfective forms inasmuch as they get a middle interpretation supported
by some adverbial (and conversely, the imperfectives are allowed because
middles semantically feed on imperfectives). Thus if you want to
distinguish passives from middles, the sad conclusion is that you don't
get the former at all in Mod Pol (that was my point in my comments on
Jill Christensen's example - a good middle).

3. Two apparent exceptions to this generalization: (A) verbs of the
"break", "open", "close" class which do take "si,e" and agree with NOM
subjects like "the glass", "the door", respectively, and allow for both
perfective and im- forms (and can form middles with the latter); (B)
SPRZEDA,C 'sell-perf.' and SPRZEDAWA,C 'sell-imperf.' which (God knows
why) unlike BUY allow for both passive and middles with both perfective
and imperfective forms (I guess there is a similar assymetry in English).

But refl. forms of verbs of class A can hardly be classified as true
passives  as they all have passive past participial forms, too and
sentences with the latter are not semantically (and pragmatically)
equivalent to the refl. forms in that the former entail human agency and
can be modified by "active" adverbials, while the latter don't and can't.
The former have their natural English equivalents in the past participial
"be broken", "be opened" forms, while the latter in the intransitive
"break" and "open" forms (I tried to explain the semantic differences
between the two forms in my paper on the semantic difference between
participial and reflexive passives).

As for (B) - SELL, it might be easier to consider it a passive in (iv):

(iv) Sprzeda,ly       si,e ju.z dwa obrazy Dwurnika.
     sold-pl.perf.    sja  already two pictures-NOM of Dwurnik's

for there is very little difference between it and the PPass (v):

(v) Zosta,ly sprzedane       ju.z      dwa obrazy   Dwurnika.
    were      sold-pl.pst.prt. already two pictures-NOM Dwurnik's

but still, you can't add an agentive PP "przez NP" to (iv) while you can to
(v), and there IS a subtle semantic difference between the two coinciding
with the agency entailments.

I hope you will find these remarks of some help.

Zbyszek