Folder: data/xml

This version of the corpus consists of TEI compliant XML files. Each file contains the text of one sample and the annotations of the project Redewiedergabe.

For a more detailed explanation of the annotation structure see Annotation structure.

• TEI header with general information
• <fs> tag with metadata in the format defined by the Redewiedergabe project (see Metadata)
• The sample text contains the following XML tags: <said> (STWR annotation), <seg> (frame/speaker/intExpr annotation), <note> (footnote text), <p> (paragraph; always covers the full sample text)

We created a RELAX-NG syntax schema that adapts the TEI Module for Linguistic corpora to the annotation schema of the project Redewiedergabe. The following paragraph lists the extensions we defined.

Schema and Documentation: data/xml/resources

Attribute	Values	Description
id	number	ID
mode	direct, freeIndirect, indirect, reported, indirect freeIndirect	STWR type
content	speech, thought, writing (and combinations, separated by whitespace)	STWR medium
level	number (starts with 1)	nesting depth of this STWR annotation
nonFact	yes/no	non-factual STWR, e.g. negated
prag	yes/no	STWR with a different pragmatic function, e.g. rhetorical figures
metaph	yes/no	metaphorical STWR
border	unspec/state/percept	borderline cases of STWR (typically for thought representation), e.g. perceptions

Attribute	Values	Description
id	number	ID
pos	start, mid, end	if type=frame: position of the frame relative to its STWR

The attribute type of the <seg> tag is used to specify which information the tag is used for. Options are: frame, speaker, intExpr.