pine trees

[Alexei Lebedev]

Dear colleagues:

The following is my summary of the results from the query I posted to
SEELangs list (and George Fowler's list of Slavic linguists) a few days
ago.  I heard from quite a few people.  Below I show the names of those
who actually responded with judgments.  There were, of course, some who
either didn't tolerate any of the data, complained about the Russian-
only nature of the query, and even one who said, "I think they are all
pretty bad, and should not even be studied." [From a person who claims to
be a linguist!]  The following respnded with judgments on at least one of
the three data pairs:

Tania Avgustinova
Larry Bogoslaw
Edward Dumanis
Sharon Flank
an acquanintance of Genevra Gerhart
Natasha Kondrashova
Alexei Lebedev
Slava Paperno
Ari Solovyova
Kim Wertz
Misha Yadroff

I begin by repeating the query, in somewhat enhanced form.  If anyone wishes
add anything to this, please write me at billings at princeton.edu anytime:

In the first example I asked respondents to choose between 1a and 1b below:

(1a)  Otnositel'no ochkov tridtsati i govorit' ne stoit.
(1b)  Ochkov otnositel'no tridtsati i govorit' ne stoit.
[both from ex. 25 in Mel'chuk (1985:153); the latter marked as ungrammatical]

I glossed this example as follows:  'It is not even worth having a discussion
regarding thirty points.'  I did this because I wanted to avoid complications
with the English words _discuss about_ (i.e., the two meaning of _about_).

In the second example I showed the following small fragments:

(2a)  okolo sosen desiati
(2b)  sosen okolo desiati

with the proviso that it mean 'approximately ten pine trees'.

In the last set I gave the same strings:

(3a)  okolo sosen desiati
(3b)  sosen okolo desiati

but with the meaning of 'NEAR approximately ten pine trees'.

Both (2a-b) and (3a-b) are attempts to determine how to apply approximative
inversions to the phrase _okolo desiati sosen_, which is two-way ambiguous
in Russian, meaning either 'near ten pine trees' or 'approximately ten pine
trees' [= exx. 14a-b, respectively, in Babby (1985:98); references below].

In several messages back and forth to Alexei Lebedev I decided that I had
not provided enough material for (2) and (3).  He had assessed all six
word-order permutations and I added the preceding words to make them at
least full sentences.  He also preferred a smaller number, so I replaced
_desiati_ 'ten' with _trekh_ 'three'.  Example (2a) corresponds, then, to
(b); (2b) corresponds to (e); (3a) to (h); and (3b) to (k).

All of (a) through (f) should mean '(my) uncle planted about ten pine trees':

     a.  Diadia posadil okolo trekh sosen.
     b.    "       "    okolo sosen trekh.
     c.    "       "    trekh sosen okolo.
     d.    "       "    trekh okolo sosen.
     e.    "       "    sosen okolo trekh.
     f.    "       "    sosen trekh okolo.

As for the other meaning (of 'near'), try the following structure only with
the meaning of '(my) uncle built a house NEAR approximately three pine trees':

     g.  Diadia postroil dom okolo trekh sosen.
     h.    "       "      "  okolo sosen trekh.
     i.    "       "      "  trekh sosen okolo.
     j.    "       "      "  trekh sosen okolo.
     k.    "       "      "  sosen okolo trekh.
     l.    "       "      "  sosen trekh okolo.

Now, to the results:

There are two distinct groupings of responses:  The larger group of the two
prefers (1a) over (1b) [the same judgments as Mel'chuk (1985), incidentally],
(2b) over (2a), and (3a) over (3b).  The smaller group prefers (1b) over (1a)
and (2b) over (2a), but could not really provide a judgment on (3a-b).

I should qualify these results with the following:  In the first (larger)
group while each respondent prefers 1a, 2b, and 3c, not all consider these
three examples to be perfect.  Some added certain things to each sentence
(for example, E. Dumanis finds 1a to be perfect as _otnositel'no
ochkov PORJADKA tridtsati_; N. Kondrashova prefers 3a with clausal emphasis
on _sosen_).  Three respondents of the first group considered both of 3a-b
to be equally bad and one could not choose between 2a-b for the same reason.
In the second (smaller) group both (1a) and (2b) were at least _dopustimo_
and both both (3a-b) were too bad to judge.

My preliminary analysis of the first group:
This is based on Mel'chuk (1985), which mentions that _ne-pervoobraznye_
'non-primary' or 'derived' prepositions, like _otnositel'no_ 'regarding',
if they undergo approximative inversion, must not appear between the noun
and numeral (i.e., 1a is the correct form; 1b is not).  Mel'chuk explains this
difference between large prepositions and light ones in terms of obligatory
contact between the noun and numeral, with only certain constituents that can
intervene (_pervoobraznye_ 'primary' prepositions, words expressing
approximation, etc.).  Since _okolo_ is a non-primary preposition, but
expresses approximation, I decided to test how it inverts.  Mel'chuk does
provide other examples like 2b, incidentally.  My informal elicitations
(prior to this query) showed that only 2b is allowed.  In a discussion with
Steve Franks (Indiana Univ.) last week, he suggested that the two meanings
of _okolo_, 'near' and 'approximately' might invert differently.  This was
confirmed in the first group, which requires locative _okolo_ and other
heavy, non-approximative prepositions to be uttered before both the noun
and the preposition when there is approximative inversion.  Only
approximative _okolo_ among the heavy prepositions can appear between the
noun and numeral under approximative inversion.  My explanation for this is
the following:  locative _okolo_ (without inversion) has the following phrase
structure according to Babby (1985:98), his example 14a:

Locative reading:  [ okolo [ desiati sosen ]] 'NEAR ten pine trees'

while the approximative/quantificational reading has the following phrase
structure (his ex. 14b):

Approximative reading:  [[ okolo desiati ] sosen ] 'APPROX. 10 pine trees'

Assuming that approximative inversion requires the quantified noun _sosen_
to be uttered before the constituent that quantifies it, then the predicted
orders for these two readings is (3a) and (2b), respectively, repeated here:

Locative reading with inversion:  [ okolo [ sosen desiati ]]
                                  'near approximately ten pine trees'

Approximative reading with inversion:  [ sosen [ okolo desiati ]]
                                       'approximately ten pine trees'

Note that the last example has two separate approximative devices.  This is,
prima facie, redundant looking.  Note that only one of the twelve people who
responded with judgments found this odd (this person supplied _okolo desiati
sosen_); all others IN BOTH GROUPS prefer this order (2b), to the one in (2a).
Mel'chuk (1985:363) even supplies the following example [= his ex. 5a], in
which there are three approximative devices (_'etak_ 'about', _s_ (+ACC)
'about', and approximative inversion):

Lenia poluchal togda rublei 'etak s piat'sot.
'Lenia was receiving (a monthly salary of) about 500 roubles then.'

This, then, is empirical confirmation of Babby's (1985) proposals about the
separate phrase structures associated with the two meanings of _okolo_.

A final comment on the smaller group, who preferred (1b) and (2b), and
rejected both of (3a-b):  Apprently this group just requires approximative
inversion in which the noun precedes the preposition if there is one,
regardless of the the preposition's size (i.e., whether it is derived).
This accounts for (1b) and (2b), but predicts that (3b) should be acceptable.
What I think is going on is that this would produce two identical orders--
both _sosen okolo desiati_--which would mean two different things.  The
meaning of 'approximately ten pines' blocks the interpretation of 'near
approximately ten pines'.  In any event, not one of the respondents from
either group found either of (3a-b) to be fully acceptable.

I end with a comment to the respondent who considers this stuff not worth
studying:  Not only is this data worth studying, it was also study-able.

I would appreciate all your comments, as usual, even from detractors.

References:

Babby, Leonard H. (1985)  "Prepositional quantifiers and the Direct Case
     Condition in Russian."  In _Issues in Russian morphosyntax._  M.S.
     Flier & R.D. Brecht, eds.  Columbus, OH:  Slavica.  91-117.

Mel'chuk, Igor' A. (1985)  _Poverkhnostnyi sintaksis russkikh chislovykh
     vyrazheniia._  (= Wiener slawistischer Almanach Sonderband 11.)  Wien:
     Institut fuer Slawistik der Uniwersitaet Wien.

Sincerely,
(c) Loren A. Billings

billings at princeton.edu
billings at pucc.bitnet