Tid: Torsdagen, 27 januari kl. 15-17
Plats: C307

Sammanfattningen av seminariet kommer nedan.
Postseminarium med förfriskningar äger rum direkt efter seminariet i institutionens pentry.

Hjärtligt välkomna!

Ljuba Veselinova & Maria Koptjevskaja-Tamm


Relative clauses in spoken Russian and elsewhere

The paper addresses the problem of systematic discrepancy between syntactic and prosodic grouping in postnominal relative clauses. In Russian, as in many other right-branching languages, the prosodic break occurs inside NP – between the head noun and the relative clause, no matter restrictive or non-restrictive. Typically, this prosodic break appears to be at the same time the border between the topic and the comment in the sentence, cf. (1) with square brackets showing syntactic constituents, bold curly brackets showing intonation units, and arrows showing phrasal pitch accents. This sort of discrepancy usually doesn’t occur in left-branching languages with prenominal relative clauses, like Japanese, where each head noun is grouped prosodically with the preceding relative clause, the prosodic break thus occurring after the head noun and, hence, after the whole NP rather than inside it.
The paper argues that in Russian the discrepancy emerges as a cumulative effect of the syntactic right-branching, on the one hand, and, on the other hand, of the discourse tendency for “ideal delivery”
(Clark&Clark 1977, inter alia) with intonation units matching syntactic clauses. Basing on the first-hand natural spoken discourse data (oral corpora systematically annotated for prosodic details, incl. pausation and pitch movements) we attempt to demonstrate the “autonomy” of relative clauses from their heads. We will introduce the following structural and prosodic symptoms of this “autonomy”, which previously remained underestimated or simply unnoticed in the literature on relativization based mainly on written data:

1. In spoken Russian, the head noun can be separated from the relative clause. This detached head noun, cf. underlined plat’je ‘dress’ in (2), can retain the main phrasal accent (typically, with the rising pitch movement) responsible for signaling the topical status of the first intonation unit and its non-final status in the sentence, i.e. signaling the continuation of the unfolding discourse.

2. Previous studies (Zaliznjak, Padučeva 1979, Ljutikova 2009) have convincingly demonstrated the effect of pied-piping in Russian relative clauses: if a relative pronoun appears inside a smaller constituent within the relative clause, the whole this constituent is fronted rather than the single relative pronoun. What remained unnoticed so far is that the fronted constituent and the rest part of the relative clause can be articulated as separate intonation units (shown with angle brackets in (3)). Within the fronted constituent the pied-piped relative pronoun (underlined in (3)), otherwise strictly atonic, acquires the main phrasal accent. The accent (typically, with the rising pitch movement) signals the topical status of the first intonation unit and its non-final status within the relative clause, which thus acquires its internal information structure.

3. Pauses between head nouns and relative clauses appear to be registered much more often than on the border of complement clauses and even slightly more often than on the border of adverbial clauses.
The length of pauses shows the same tendency (see Korotaev 2009 for detailed quantitative results). The intonation unit that includes the head noun can be articulated with the falling pitch movement in main phrasal accent, i.e. as an independent clause projecting no continuation in the unfolding discourse. Hence, the relative clause is produced autonomously as a suspended clause (the term coined in in Ohori 1996), added after the speaker recognizes the just produced portion of discourse as insufficient and requiring elaboration, cf.

4. In our data up to 25% of relative clauses are produced as suspended – much more often than other types of subordinate clauses.

Finally, relative clauses in spoken Russian will be compared with other heavy NPs, first and foremost with coordinate NPs, which also demonstrate behavior cumulatively affected by the syntactic right-branching and the discourse tendency for “ideal delivery”.


{Mne nravitsja [­plat’je}T(opic) {[kotoroe ona iz ¯sitca
sšila ]]}C(omment)
I like dress REL she of printed.cotton made
‘I like the dress that she made of printed cotton ’

{Mne ­plat’je nravitsja} {kotoroe ona iz ¯sitca sšila}
I dress like REL she of printed.cotton made
Lit.‘I dress like that she of printed cotton made’

{Mne nravitsja ­plat’je}T1 {árukava ­kotorogoñT2 ášjity iz ¯sitcañC2}C1
I like dress sleeves REL made of printed.cotton
Lit. ‘I like dress sleeves [of] which are made of printed cotton ’

Mne nravitsja ¯plat’je. Kotoroe ona iz ¯sitca sšila.
I like dress REL she of printed.cotton made
‘I like the dress. [I mean the one] that she made of printed cotton ’


Zaliznjak Andrej A., Padučeva Elena V. 1979. Sintaksičeskie svojstva mestoimenija kororyj // Kategorija opredeljonnosti-neopredeljonnosti v slavjanskix i balkanskix jazykax. Moscow: Nauka, 289-329.

Korotaev Nikolai. A. 2009. Sintaksis i prosodija v sisteme sredstv diskursivnoj svjaznosti teksta. PhD dissertation. Russian State University for the Humanities.

Ljutikova Ekaterina A. 2009.Otnositel’nye predloženija s sojuznym slovom kororyj // Korpusnye issledovanija po russkoj grammatike. Eds.
K.Kiselёva et al. Moscow: Probel, 436-511.

Ohori Toshio. 1996 Remarks on suspended clauses: a contribution to Japanese phraseology // Essays in Semantics and Pragmatics: In honor of Charles J. Fillmore. Eds. Masayoshi Shibatani and Sandra A.
Thompson. John Benjamins, 201-218.