Treebank Construction Step 1: Transcription, Time-Stamping and Syntax Tag
How to use the files?
In the following, the transcribed recordings of the fieldwork in Tehran and the fieldwork in Paris can be viewed, together with the corresponding audio files. The transcriptions of the Paris fieldwork include, in addition, a syntactic analysis, coded by means of the syntax tag.
The transcription files can be viewed directly in the browser window. However, I suggest to open them in the Clan-Editor (which recognizes files with the extension .cha). The original uncompressed audio recordings (wav-format, 48KHz, 24Bit) have been recoded in mp3 (128KBit/s), in order to obtain tolerable download times. Because of the compression, acoustic-phonetic analyses with mp3 files are not recommended.
The time-stamping can be "experienced" if a transcription file and its corresponding audio file are downloaded (mouse right-click and save command) into the same folder, and the transcription file then opened in the Clan-Editor - note that sometimes the browser adds the second extension .txt; if this happens, you might want to consider to remove the .txt extension by renaming the downloaded file. The free Clan-Editor can be obtained here.
General Remarks
All utterances are transcribed, including marked pauses within a sentence and marked omissions of natural intonational breaks between sentences. However, specifics of spontaneous speech that are usually irrelevant for the tree structure, e.g. false starts, hesitations, or repetitions, are specially marked by curly braces. The segmentation is always syntactically motivated, i.e. each matrix clause and its subordinate clauses, is one unit. Each sentence is time-stamped, i.e. its start- and end-time in the audio-file is coded by a special tag. This information will allow the user to listen to any chosen sentence. In case of complex sentences, the CP structure is annotated by bracketing.
French
Stylebook and Transcription Guidelines for the French Question-Elicitation Interviews
Part I: Procedure and Non-Syntactic Tags
Time-Stamped and Coded Transcriptions of 10 Subjects
|
Subject
|
Transcription File
|
Audio File
|
|
30
|
 |
 |
|
31
|
|
|
|
32
|
|
|
|
39
|
|
|
|
40
|
|
|
|
41
|
|
|
|
44
|
|
|
|
45
|
|
|
|
46
|
|
|
|
47
|
|
|
Persian
Stylebook and Transcription Guidelines for the Persian Question-Elicitation Interviews
The Persian recordings are transliterated in an IPA-like code, in which non-ASCII symbols have been substituted by ASCII symbols, namely:
(stylebooks coming soon)
Time-Stamped and Coded Transcriptions of 14 Subjects
|
Subject
|
Transcription File
|
Audio File
|
|
1
|
|
|
|
2
|
|
|
|
3
|
|
|
|
4
|
|
|
|
5
|
|
|
|
6
|
|
|
|
7
|
|
|
|
8
|
|
|
|
9
|
|
|
|
10
|
|
|
|
11
|
|
|
|
12
|
|
|
|
13
|
|
|
|
14
|
|
|