UD for Norwegian 
Tokenization and Word Segmentation
White space always indicates a token boundary and punctuation constitute separate tokens, except:
- numbers with periods, commas or colons, e.g. 1.3, 0,6, 10:13
- abbreviations, e.g. f.eks., Carl J. Hambro
- URLs, e.g. http://www.ifi.uio.no
The treebanks do not contain multiword tokens.
Morphology
Tags
- PART is used only for the infinitival marker å “to”.
- the VERB tag covers main verbs (content verbs) and copulas but it does not cover temporal, passive and modal auxiliary verbs, for which there is the AUX tag
- DET is used for possessives, demonstratives and quantifiers, which usually modify nominal heads. PRON is used for pro-words that substitute for nouns or noun phrases.
- Participles are tagged either as VERB or ADJ based on syntactic criteria.
Features
- Common nouns are associated with the features
Definite=Def/Ind,Gender=Masc/Fem/Neut,Number=Sing/Plurand, where applicable,Case=Gen - Non-finite verbs are distinguished by their
VerbFormfeature:VerbForm=PartandVerbForm=Inf. Finite verbs haveVerbForm=Fin. - Finite verbs have the following features:
Mood=Ind/Imp. Indicative verbs are marked for tense:Tense=Pres/Past - Passive verbforms are marked with
Voice=Pass
Syntax
- In sentences with matrix clause word order, subjects are either in the position right before or right after the finite verb. In sentences with subordinate clause word order, the subject is before the finite verb, but not necessarily immediately before: The finite verb and the subject can be separated by a sentence adverbial. Moreover, in relative clauses, the relativizer, som, ‘that’, cannot be omitted if the subject is relativized. Case-marked pronouns in subject position will have nominative case.
- The unmarked position of the object in clauses with subordinate clause word order is immediately after the verb of which it is a complement. Case-marked pronouns in object position will have accusative case.
- These subtype relations are used in the written language treebanks:
- acl:relcl is used for relative clauses
- compound:prt for verbal particles
- acl:cleft
- The LIA treebank, containing spoken language data, has these subtype relations in addition to those described above:
- discourse:filler
- parataxis:deletion
Treebanks
There are three Norwegian UD treebanks: