Annotation Model of morposyntactic, morphological and syntactic annotations of the TüBa-D/Z corpus (v5) following Telljohann et al. (2009), unless marked otherwise
Heike Telljohann, Erhard W. Hinrichs, Sandra K¨ubler, Heike Zinsmeister, Kathrin Beck (2009),
Stylebook for the T¨ubingen Treebank of Written German (T¨uBa-D/Z), Tech. rep. Universität Tübingen, Seminar für Sprachwissenschaft, version of November 2009 (TüBa-D/Z v5)
Generally, the left conjunct of a split coordination is located in MF, in rare cases in VF, and the right conjunct occurs
in NF. In order to express the relation between them, the left conjunct carries the label of
its grammatical function (ON, OA, OD, etc.) whereas the right conjunct carries a label
that denotes that it is the conjunct of this grammatical function (e.g. ONK, OAK, ODK,
etc.). In asymmetric coordination, the syntactic category of the second split conjunct
determines the syntactic category one level higher up: "Jedes Ja-Wort zieht [Applaus]_OA nach sich, [Unterschriften, K?sse, H?ndesch?tteln]_OAK."
(Telljohann et al. 2009, p.124f)
In a German clause, the finite verb can appear in three different positions: verb-second,
verb-initial, and verb-final. Only in verb-final clauses the verb complex consisting of the
finite verb and non-finite verbal elements forms a unit. The discontinuous positioning
of the verbal elements in verb-first and verb-second clauses is the traditional reason
for structuring German clauses into fields. The positions of the verbal elements form
the Satzklammer (sentence bracket) which divides the sentence into a Vorfeld (initial
field), a Mittelfeld (middle field), and a Nachfeld (final field). The Vorfeld and the
Mittelfeld are divided by the linke Satzklammer (left sentence bracket), which is the finite
verb, the rechte Satzklammer (right sentence bracket) is the verb complex between the
Mittelfeld and the Nachfeld.
(Telljohann et al. 2009, p.13)
The C-Feld only occurs in verb-final clauses (exception: the conjunction als in
subordinated sentences of comparison als w?are es nie geschehen.). It is obligatorily
occupied in finite verb-final clauses if there is no conjunction in the Linke Klammer. In
non-finite verb-final clauses the C-position may be empty. This field can be occupied by
conjunctions of sentential objects (e.g. da?, ob) or sentence initial conjunctions like um,
obwohl, wenn and also by complex interrogative or relative phrases, e.g. ..., ?um wieviel
Geld? geht es dabei? / ..., ?an der? Max Daniel Professor f?r Klavier ist.
(Telljohann et al. 2009)
discourse marker
Generally, discourse markers are expressions or phrases of greeting, apologizing, thanking,
short emotional utterances, and interjections. Their node label is DM. ... Typical discourse markers are:
ja, nein, hallo, oh, aha, pst, nunja, gewi?, toll, nun ja, etc.
(Telljohann et al. 2009, p. 136)
phrase internal relation between two parts of a proper noun (Telljohann et al. 2009, p. 26)
The secondary edge label EN gives information about the relation of two parts of
a proper noun within a complex phrase consisting, for instance, of an article and/or
an attributive adjective which do not belong to the proper noun itself, e.g. der [zweite
Weltkrieg EN] (Telljohann et al. 2009, p.19).
If the original form of a proper noun (e.g. Zweiter Weltkrieg, Hamburger Reichssender)
is inflected and/or premodified by an article and/or attributive adjective, the included
proper noun is indicated by the secondary edge label EN. EN always points from the
dependend part of the proper noun to its head noun: "Er ist Zeitzeuge des Zweiten--EN-->Weltkrieges"
(Telljohann et al. 2009, p.50f)
The KOORD-field is the field for coordinating particles. In contrast to
the PARORD-field, it can optionally occur as the left-most element of all clause types
(Telljohann et al. 2009, p. 15)
The Linke Klammer is the position of the finite verb in verb-second and
verb-first clauses or a conjunction in verb-final clauses. It consists of exactly one element.
(Telljohann et al. 2009)
Apart from those units which are optionally located in other fields, any
non-verbal constituent may occur in the Mittelfeld. It consists of a sequence of any
number of constituents. The linear order of the constituents depends on the specific
word order principles for German and their interaction.
(Telljohann et al. 2009)
predicate
Typically, the complement type PRED (predicate) occurs with verbs like sein, haben,
scheinen, aussehen, sich anh?ren, klingen, etc. PRED is annotated, if the following
conditions apply:
? if it is not possible to determine the case of the constituent in question properly (e.g. gut in Das ist gut.)
? if the constituent in question actually predicates the subject, i.e. the subject is characterized as having the property expressed by PRED (e.g. in Die Ursache war unklar. Die Ursache is characterized by the property of being unclear)
? raising-verbs (subject without theta-role)
? if als-phrases are selected by the verb they are labeled as PRED (e.g. Unter dem Motto Kino-Extrem agiert der Regisseur als Filmjockey.)
(Telljohann et al. 2009, p.87)
The Verbkomplex is a sequence of verb forms. In verb-second and verb-first
clauses it consists of one or more non-finite elements or - depending on the verb - of
a separable prefix. In verb-final clauses it also contains the finite verb. The rule for
the linear order in general is: right determines left. If there is a finite verb in the verb
complex, it is usually the right-most element.
(Telljohann et al. 2009)
The Vorfeld consists of only one constituent. Usually it is the subject. But
because of the high degree of non-configurationality in German, the subject can also
occur in theMittelfeld, thus allowing almost every other constituent to occupy the Vorfeld.
(Telljohann et al. 2009, p.15)