Abstract
This paper focuses on the position of three most frequent Czech subordinating conjunctions že ('that'), když ('when'), and až ('when') within the stress group. It compares two types of data, namely spontaneous speech and read texts.
Both data contain segmentation into stress groups (phonetic words) besides traditional segmental transcription. The spontaneous speech, collected within the ORTOFON corpus (Kopřivová et al., 2017), was recorded in a private environment with fluctuating sound quality.
On the other hand, the read speech was recorded in a professional studio for 146detailed phonetic research.The position of subordinating conjunctions is traditionally (e.g. Hrbáček, 1967) described as initial within the subordinate clause.
According to the orthographical rule, it is necessary to write a comma before subordinating conjunctions in Czech. The rhythmical structuring of read texts does not need to be dependent on particular words rather on punctuation (by extension, the intonation patterns are more closely related to syntax, cf.
Wichmann, 2000: 123). It means that the subordinating conjunctions couldstay at the beginning of the stress group more often and can, but does not have to be stressed.
On the contrary, Wichmann (2000: 21, 39) claims that the prosodic boundary occurs after rather than before a conjunction more often within spontaneous, unprepared speech. The online syntax theory (Auer, 2000) offers pragmatic reasons for that.
The prosodic boundary occurring after a conjunction enables the speaker to signal her/his intent to continue the speech with some kind of new or known but rephrased information within the following subordinate clause. At the same time, the speaker indicates that s/he needs more time in order to think her/his next statement through.
Therefore, the other speakers do not start to talk and wait for another opportunity for turn-taking.However, Wichmann's claim also invites some doubts about the speaker's ability to predict what s/he will say. Does the meaning of the conjunction correspond to the information within main and subordinate clauses, or does it rather freely connect the clauses as a kind of discourse marker? Concerning the issue of discourse markers, the position within the stress group could be taken as a cue to determine the semantic bleaching of conjunctions (cf. the subordinate conjunction protože, 'because', in Čermáková et al., 20