WebLicht/TüNDRA search syntax

Basics

  • WebLicht relies on the TüNDRA searching app and the TIGERSearch search ‘syntax’.
  • Searching is case-sensitive
  • You may name your search—this ‘variable name’ can be any string that begins with a hashtag and ends in a colon, e.g., #query: or #q1:
  • Search with this model: [attribute=”value”]
  • Criteria may be combined with the logical operators:
    • and: &
    • or: |
    • not: !
  • Searching for sequences: Put a full stop between several queries to search for consecutive words. Put squared brackets around each word separately. You may add a variable name to each slot – it may describe the contents (#noun:) but need not (#q1:). This makes the Statistics function easier to use and interpret.
  • Put all criteria within one set of squared brackets to look for one word that fulfils them all.

Examples

#q1:[word=”going”]   finds all instances of the form going

#q2:[token=”has”]  finds all instances of the form has

#q3:[word=”had”].#q4:[word=”gone”] finds all instances of had gone

Searching by annotation type

POS/Lemma annotation

…uses the Penn Treebank tags

You can search for parts of speech (pos=”tag”) or lemmas (lemma=”lemma”)

Examples

#q1:[pos=”JJ”]  finds all (positive) adjectives

#q2:[lemma=”be”]  finds all forms of the verb be

#q3:[word=”left”&pos=”JJ”]  finds all adjectival usages of left

#q4:[pos=”JJ”|pos=”JJR”|pos=”JJS”]  finds all adjectives (also comparatives and superlatives)

#q5:[lemma=”be”&word!=”been”]  finds all forms of the verb be, except for been

#q6:[lemma=”student”].#q7:[pos=”VBP”|pos=”VBZ”] finds all forms of student followed by a singular or plural verb

#noun:[pos=”NNP”].#possessive:[pos=”POS”]  finds all (singular) proper nouns followed by a possessive ending

Morphology annotation

…uses MorphAdorner

Searchable morphological categories:

  • word/token for a word form
  • morphmajorwordclass for 17 major word classes
  • morphwordclass for 34 sub-word classes
  • morphsyntax for a fine-grained classification into 241 word classes
  • morphnumber, morphcase, morphtense, morphmood, and morphdegree for searching with a value of number, case, tense, mood, or degree, respectively

Examples

#q1:[word=”going”] finds all instances of the form going

#q2:[token=”has”] finds all instances of the form has

#q3:[morphwordclass=”np”] finds all proper nouns

#q4:[morphsyntax =”pix”] finds all indefinite pronouns

#q5:[morphtense=”past”] finds all past tense forms

#q6:[morphperson=”third”] finds all 3rd person forms

#q7:[morphmood=”inf”] finds all words of infinitive mood

#q:[word=”left”&morphmajorwordclass=”adjective”] finds all adjectival usages of left

#q:[morphdegree=”comp”|morphdegree=”sup”] finds all comparatives and superlatives

#q:[morphwordclass =”vm”&word!=”will”] finds all modal verbs, except for will

#neg:[morphwordclass=”xx”].#any:[word=”any”] finds all negation words followed by any

#aux1:[morphwordclass=”va”].#aux2:[morphwordclass=”va”] finds all instances of two consecutive auxiliary verbs

Constituent parsing

…uses Penn Treebank

Examples

#query:[cat=”SINV”] finds inverted declarative sentences

#q1:[cat=”NP”].#q2:[cat=”NP”].#q3:[cat=”NP”] finds any 3 consecutive noun phrases