Annotating Nonlocal Dependencies
Nonlocal dependencies are dependencies from one clause into another. They do
not exist as syntactic dependencies in (basic) UD, so for Superframes
annotation, we add them when needed. We have already seen nonlocal dependencies
with the META frame in the chapter on Annotating Dynamic
Verbs and with attributive adjectives, adverbs,
relative clauses, and relation nominal adjuncts in the chapter on Annotating
Adjuncts. In both cases, the element filling the
scene role has an additional dependency (a “backlink”) pointing back at the
element filling the participant role.
There are a number of other constructions that systematically introduce nonlocal dependencies, which we also annotate. We give examples of each in the following.
Subject Control
Dependency graph for sentence: Kim promised Sandy to come. Token promised is labeled META-NORM-INIT and has an edge labeled participant to token Kim and an edge labeled EXPERIENCER to token Sandy and an edge labeled scene to token come. Token come is labeled LOCATION-INIT and has an edge labeled has-location:hl to token Kim.
Dependency graph for sentence: Kim used a hammer to smash the vase. Token used is labeled META and has an edge labeled participant to token Kim and an edge labeled MEANS to token hammer and an edge labeled scene to token smash. Token smash is labeled STATE-INIT and has an edge labeled CAUSER:hl to token Kim and an edge labeled has-state to token vase.
Dependency graph for sentence: Kim wants to swim. Token wants is labeled META-NORM and has an edge labeled participant to token Kim and an edge labeled scene to token swim. Token swim is labeled LOCATION-DYN and has an edge labeled has-location:hl to token Kim.
Dependency graph for sentence: Kim is hard to love. Token hard is labeled META-PREVENTION and has an edge labeled participant to token Kim and an edge labeled scene to token love. Token love is labeled EVALUATION and has an edge labeled evaluee:hl to token Kim.
Dependency graph for sentence: Kim is loath to leave. Token loath is labeled META-NEG-NORM and has an edge labeled participant to token Kim and an edge labeled scene to token leave. Token leave is labeled LOCATION-DEINIT and has an edge labeled has-location:hl to token Kim.
Dependency graph for sentence: This is all you need. Token all is labeled LIMIT and has an edge labeled limit to token This and an edge labeled limited to token need. Token need is labeled RELATION-NECESSITY and has an edge labeled related-to:hl to token This and an edge labeled related to token you.
Object Control
Dependency graph for sentence: Kim persuaded Sandy to come. Token persuaded is labeled META-INIT and has an edge labeled SENDER to token Kim and an edge labeled participant to token Sandy and an edge labeled scene to token come. Token come is labeled LOCATION-INIT and has an edge labeled has-location:hl to token Sandy.
Dependency graph for sentence: Kim helped Sandy smash the vase. Token helped is labeled META-INIT and has an edge labeled CAUSER to token Kim and an edge labeled participant to token Sandy and an edge labeled scene to token smash. Token smash is labeled STATE-INIT and has an edge labeled CAUSER to token Sandy and an edge labeled has-state to token vase.
Dependency graph for sentence: Kim wants Sandy to swim. Token wants is labeled META-NORM and has an edge labeled EXPERIENCER to token Kim and an edge labeled participant to token Sandy and an edge labeled scene to token swim. Token swim is labeled LOCATION-DYN and has an edge labeled has-location:hl to token Sandy.
Light Verbs
From the perspective of Superframes, light verbs are just control verbs whose
scene role happens to be filled by a noun. For details and examples, see
META.
“Non-obligatory” Control
Here, the control relationship is contributed by an adjunct construction rather than by the predicate.
Dependency graph for sentence: Kim left after trashing the room. Token left is labeled LOCATION-DEINIT and has an edge labeled has-location to token Kim and an edge labeled TIME to token trashing. Token trashing is labeled STATE-INIT and has an edge labeled CAUSER:hl to token Kim and an edge labeled has-state to token room.
Dependency graph for sentence: Kim has come to stay. Token come is labeled LOCATION-INIT and has an edge labeled has-location to token Kim and an edge labeled END to token stay. Token stay is labeled LOCATION-CONTINUATION and has an edge labeled has-location:hl to token Kim.
Dependency graph for sentence: The little prince asked me abruptly , as if seized by a grave doubt .... Token asked is labeled MESSAGE-INIT and has an edge labeled SENDER to token prince and an edge labeled REFERENCE to token seized. Token seized is labeled META-INIT and has an edge labeled participant:hl to token prince and an edge labeled scene to token doubt.
Dependency graph for sentence: Stop drawing sheep , whether small or big. Token Stop is labeled META-DEINIT and has an edge labeled scene to token drawing and an edge labeled CONDITION to token small and an edge labeled CONDITION to token big. Token drawing is labeled MESSAGE-INIT and has an edge labeled content to token sheep. Token small is labeled PROPERTY and has an edge labeled has-property:hl to token sheep. Token big is labeled PROPERTY and has an edge labeled has-property:hl to token sheep.
Dependency graph for sentence: Some people , though short , reach amazing heights. Token short is labeled PROPERTY and has an edge labeled has-property:hl to token people. Token reach is labeled META-INIT and has an edge labeled participant to token people and an edge labeled CONCEDED to token short and an edge labeled scene to token heights. Token amazing is labeled PROPERTY and has an edge labeled has-property to token heights. Token heights is labeled PROPERTY and has an edge labeled has-property to token people.
Non-obligatory Control with Parasitic Gap
Dependency graph for sentence: the question we raised without answering. Token question is labeled ENTITY and has an edge labeled SCENE to token raised. Token raised is labeled MESSAGE-INIT and has an edge labeled message2 to token question and an edge labeled SENDER to token we and an edge labeled ACCOMPANIER-NEG to token answering. Token answering is labeled MESSAGE-INIT and has an edge labeled content:hl to token question and an edge labeled SENDER:hl to token we.
Depictives
Depicitves are like non-obligatory control, but without a conjunction and with
role ACCOMPANIER.
Dependency graph for sentence: Kim entered the room singing. Token entered is labeled LOCATION-INIT and has an edge labeled has-location to token Kim and an edge labeled location to token room and an edge labeled ACCOMPANIER to token singing. Token singing is labeled EXPRESSION and has an edge labeled sender:hl to token Kim.
Dependency graph for sentence: Kim ate the radish raw. Token ate is labeled CONTAINMENT-INIT and has an edge labeled container to token Kim and an edge labeled contained to token radish and an edge labeled ACCOMPANIER to token raw. Token raw is labeled STATE and has an edge labeled has-state:hl to token radish.
Resultatives
Resultatives are like non-obligatory control, but without a conjunction and
with role EFFECT, and sometimes AFFECTED if there is no other role assigned
by the matrix predicate to the controlled argument.
Dependency graph for sentence: He took a whole herd of elephants away with him. Token took is labeled LOCATION-DEINIT and has an edge labeled CAUSER to token He and an edge labeled has-location to token herd and an edge labeled EFFECT to token away and an edge labeled ACCOMPANIED to token him. Token away is labeled LOCATION-DEINIT and has an edge labeled has-location:hl to token herd.
Dependency graph for sentence: I would bring myself down to his level. Token bring is labeled META-INIT and has an edge labeled participant to token myself and an edge labeled scene to token down. Token down is labeled PROPERTY-INIT and has an edge labeled has-property to token myself and an edge labeled EFFECT to token level. Token level is labeled PROPERTY and has an edge labeled has-property:hl to token myself and an edge labeled has-property to token his.
Dependency graph for sentence: You ’re talking me silly. Token talking is labeled EXPRESSION and has an edge labeled sender to token You and an edge labeled AFFECTED to token me and an edge labeled EFFECT to token silly. Token silly is labeled STATE and has an edge labeled has-state:hl to token me.
Dependency graph for sentence: Kim sneezed the napkin off the table. Token sneezed is labeled CONTAINMENT-DEINIT and has an edge labeled container to token Kim and an edge labeled AFFECTED to token napkin and an edge labeled EFFECT to token table. Token table is labeled ENTITY and has an edge labeled HAS-LOCATION-DEINIT:hl to token napkin.
Dependency graph for sentence: She went out into the world all rumpled. Token went is labeled LOCATION-INIT and has an edge labeled has-location to token She and an edge labeled EFFECT to token out and an edge labeled location to token world and an edge labeled ACCOMPANIER to token rumpled. Token out is labeled LOCATION-INIT and has an edge labeled has-location:hl to token went.
Raising
Dependency graph for sentence: Kim seemed to fly. Token seemed is labeled META-MESSAGE and has an edge labeled participant to token Kim and an edge labeled scene to token fly. Token fly is labeled LOCATION-DYN and has an edge labeled has-location:hl to token Kim.
Attributive Adjective/Adverb/Relative Clause/Relation Nominal Adjunct
See the chapter on Annotating Adjuncts.
Pseudocleft
Dependency graph for sentence: This is what you need. Token what is labeled META and has an edge labeled participant to token This and an edge labeled scene to token need. Token need is labeled RELATION-NECESSITY and has an edge labeled related-to:hl to token This and an edge labeled related to token you.
Dependency graph for sentence: For what the king fundamentally insisted upon was that his authority should be respected. Token what is labeled META and has an edge labeled scene to token insisted and an edge labeled participant to token respected. Token insisted is labeled NORM and has an edge labeled SENDER to token king and an edge labeled SCENE to token fundamentally and an edge labeled norm-content:hl to token respected.
Dependency graph for sentence: all you need to do is move your chair. Token all is labeled LIMIT and has an edge labeled limited to token need and an edge labeled limit to token move. Token need is labeled META-NECESSITY and has an edge labeled participant to token you and an edge labeled scene to token do. Token do is labeled META-INIT and has an edge labeled participant to token you and an edge labeled scene:hl to token move. Token move is labeled LOCATION-INIT and has an edge labeled CAUSER to token you and an edge labeled has-location to token chair.
It-cleft
Dependency graph for sentence: It was through words that everything was revealed to me. Token words is labeled ENTITY and has an edge labeled SCENE to token revealed. Token revealed is labeled MESSAGE-INIT and has an edge labeled MEANS:hl to token words and an edge labeled content to token everything and an edge labeled EXPERIENCER to token me.
Inter-Argument Relations
Some arguments that are introduced by argument structure constructions are in a
semantic relation not so much with their head but with (or more) of the head’s
other dependents (or a descendant of one). Give such arguments the generic role
RELATED-TO and introduce a backlink from them to their semantic relate.
Dependency graph for sentence: Kim exceeds Sandy in height. Token exceeds is labeled COMPARISON and has an edge labeled compared to token Kim and an edge labeled reference to token Sandy and an edge labeled RELATED-TO to token height. Token height is labeled PROPERTY and has an edge labeled has-property:hl to token Kim and an edge labeled has-property:hl to token Sandy.
Dependency graph for sentence: That is great in terms of ROI. Token great is labeled PROPERTY and has an edge labeled has-property to token That and an edge labeled SCENE to token terms. Token terms is labeled RELATION and has an edge labeled related to token great and an edge labeled related-to to token ROI. Token ROI is labeled PROPERTY and has an edge labeled has-property:hl to token That.
Dependency graph for sentence: Kim ist auf den Kopf gefallen. Token Kopf is labeled ENTITY and has an edge labeled WHOLE:hl to token Kim. Token gefallen is labeled CONTACT-INIT and has an edge labeled ornament to token Kim and an edge labeled RELATED-TO to token Kopf.
Dependency graph for sentence: Kim hit Sandy on the head with a stick. Token hit is labeled CONTACT-INIT and has an edge labeled CAUSER to token Kim and an edge labeled surface to token Sandy and an edge labeled RELATED-TO to token head and an edge labeled ornament to token stick. Token head is labeled ENTITY and has an edge labeled WHOLE:hl to token Sandy.
Dependency graph for sentence: Something was broken in my engine. Token broken is labeled STATE and has an edge labeled has-state to token Something and an edge labeled RELATED-TO to token engine. Token engine is labeled ENTITY and has an edge labeled PART:hl to token Something and an edge labeled POSSESSOR to token my.
Dependency graph for sentence: Er wandte dem Televisor den Rücken zu. Token wandte is labeled ORIENTATION and has an edge labeled oriented to token Er and an edge labeled landmark to token Televisor and an edge labeled RELATED-TO to token Rücken. Token Rücken is labeled ENTITY and has an edge labeled WHOLE:hl to token Er.
Dependency graph for sentence: the thing that is so good about the box you have given me. Token good is labeled PROPERTY and has an edge labeled has-property to token thing and an edge labeled RELATED-TO to token box.
Dependency graph for sentence: One must not hold it against them. Token hold is labeled EVALUATION and has an edge labeled SENDER to token One and an edge labeled EVALUEE to token it and an edge labeled RELATED-TO to token them. Token them is labeled ENTITY and has an edge labeled RELATED:hl to token it.
Dependency graph for sentence: There is nothing sad about old shells. Token sad is labeled PROPERTY and has an edge labeled has-property to token nothing and an edge labeled RELATED-TO to token shells. Token shells is labeled ENTITY and has an edge labeled RELATED:hl to token nothing.
Coordination
Dependency graph for sentence: ein sogenannter Televisor oder Hörsehschirm. Token sogenannter is labeled IDENTIFIER and has an edge labeled identified to token Televisor and an edge labeled identified:hl to token Hörsehschirm. Token Televisor is labeled ENTITY and has an edge labeled SCENE to token sogenannter. Token Hörsehschirm is labeled ENTITY and has an edge labeled SCENE:hl to token sogenannter.