THRC YAML source

code: THRC
mfte_code:
  - THRC
  - THRCother
name: That relative clauses on subject position
definition: >-
  Relative clauses introduced by "that" where "that" functions as the subject
  of the relative clause (e.g., "the man that came", "the ring that was lost").
normalization: finite_verbs
parent: THRC_ALL
detection:
- requires:
  - word
  - pos
  - dep
  semgrex: '{word:/[Tt]hat/; pos:WDT}=that <nsubj {pos:/VB.*/}=verb'
  description: >-
    "That" as WDT functioning as subject of the relative clause verb.
- source: pybiber
  requires:
  - word
  - pos
  - dep
  cql: '[pos="NN|NNS|NNP|NNPS|CD|DT"] [word="that" & dep="nsubj|nsubjpass"]'
  description: >-
    "that" with dep nsubj or nsubjpass, preceded by noun, cardinal, or
    determiner. pybiber uses dep_rel.str.contains("nsubj") to match
    both active and passive subjects, and requires the preceding token
    to be N*/CD/DT to filter out demonstrative "That" at sentence start.
- requires:
  - word
  - pos
  cql: '[word="that" & pos="WDT"]'
  description: >-
    All "that" as relative pronoun. Requires post-filter to distinguish
    subject from object position.
- source: mfte
  requires:
  - word
  - pos
  cql: '[word="[Tt]hat" & pos="WDT"]'
  description: >-
    MFTE tags all that_WDT as THRC (line 762). Does not distinguish
    subject from object position. Pure POS-tag based.
examples:
- text: You must be very clever to find a use for something _that_ costs nothing.
  source: le_foll_2024
- text: I'll just run a cable _that_ goes from here to there.
  source: le_foll_2024
sources:
- biber_1988
- mfte
- pybiber
notes: >-
  D6 loading .46. MFTE merges subject and object positions. Biber counted
  them separately.