On pre-nominal classifying adjectives in Polish 243
okazjach powinny być bez skorupek,
occasions should be.inf without shells
podzielone na cząstki.
divided on parts
‘Hard-boiled eggs are served unpeeled only at Easter breakfast, on
other occasions they should be (served) peeled, sliced.’
The correlation between the right edge of a prosodic constituent (e.g. the Intonational
Phrase, which typically coincides with the end of a clause or a sentence) and unmarked
nuclear stress is a well-attested phenomenon in various Indo-European languages
(discussed for English by Selkirk 1995, or for Spanish by Zubizarreta 1998). Given the
sentence final position of the phrase pusty ciężarowy samochód (lit. empty cargo.adj
car) in (35a) or the clause-final position of the phrase kreskowym kodem (lit. line.adj.
inst code.inst) in (35b) and (36a), it is the head noun (samochód ‘car’ or kodem ‘code.
inst’) which will bear the pitch accent and represent the neutral (non-contrastive)
focus. Thus, the pre-nominal placement of the classifying adjective will result in the
greater discourse prominence of the head noun.
- Conclusion
This paper investigated the relatedness between the position and interpretation of
attributive adjectives in Polish. It was shown that the model of the Polish noun phrase,
proposed in the work by Rutkowski & Progovac (2005) and later work by Rutkowski
(2007, 2009 , 2012 ) is too restrictive. It cannot account for the pre-head occurrence
of Polish classifying adjectives since it predicts them to surface post-nominally. Any
pre-head adjective in a Polish noun phrase is treated in their theoretical approach as
a qualifying (i.e. describing) attribute, which is expected to differ considerably in its
semantic reading and morphosyntactic properties from a classifying attribute. The data
considered in this paper did not confirm the latter expectation. As in the previous work
with my colleagues, Helen Trugman & Agnieszka Pysz, I employed the representational
theory which was postulated in Bouchard (2002) on the basis of data from French and
English. The representational model assumes distinct modes of semantic composition
for Polish adjectival attributes which occur either pre-nominally or post-nominally.
Post-nominal classifying adjectives in Polish are merged with non-atomized (number-
less) nouns whereas adjectival attributes which occur post-nominally combine with
nouns to which semantic Number has been assigned. The distinct composition modes
and linearization patterns can result in meaning differences between N+CA and CA+N
combinations, yet such differences are not as drastic as to warrant the recategoriza-
tion of the pre-head adjective as a qualifying attribute. They involve stylistic differences