Introduction

Superframes is a scheme for frame-semantic annotation of text. Every predicate is assigned a frame, and every argument is assigned a role. Superframes annotations can help search corpora for phenomena of interest, enable quantitative cross-lingual comparisons, and support semantic parsing. Superframes aims to be easy to annotate. In particular,

Superframes does not require a lexicon. Frames are coarse and small enough in number to learn them by heart.
Superframes is language-independent.
Superframes ia annotated atop Universal Dependencies, thus delegating many difficult syntax-related decisions to an established framework.

Superframes defines a taxonomy of frames, each of which denotes a relation between two entities and thus defines two roles. Annotating a text comes down to assigning frame labels to content words and role labels to UD dependency edges between them.

Annotating Static Verbs

Consider the following sentence (relevant UD edges shown):

Here, we say that the (verbal) predicate owns evokes the POSSESSION frame, which defines the possessum and possessor roles. The possessum roles is filled by a house and the possessor roles is filled by Kim. So we annotate as follows:

We call the arguments that fill the predicate’s frame’s roles the core arguments. We will turn to non-core arguments later.

Note that semantic roles abstract away from syntactic alternation. As an example, consider the following passive sentences and its Superframes annotation:

Note also that there are only a few dozen superframes, so the meaning they specify is necessarily coarse. Different predicates with similar meaning can invoke the same superframe. This is a feature, as it makes explicit similar predicates and roles.

Verbs with One Argument

There are, of course, also predicates with only one argument, in which case only one of the two roles is used (most often the first one).

INTERNAL-STATE is one of those superframes whose second role state is rarely filled by an argument. It denotes an abstract entity (here: the state of being asleep) that is already expressed by the predicate.

Verbs with Zero Arguments

It also happens that predicates occur with no core arguments. One example is weather verbs such as rain. For consistency, there is a rule that says predicates with no core arguments must always be annotated with the most general superframe ENTITY. So that is what happens for example for weather verbs such as rain (which has a syntactic, but semantically empty argument):

We will turn to what NONCOMP means later.

Annotating Dynamic Verbs

Many predicates, especially verbs, do not simply denote a state. Rather, they denote a process working towards the coming about of a state, or a process working against the coming about of a state, or the continuation of a state, or a process working towards the end of a state, or a habitual repeated coming about of a state, or the transitioning through a state, or even a process that transitions through a multitude of states.

Using Aspectual Frames

Superframes inidicates this kind of dynamism through a number of special frames that we call aspectual frames: INIT, PREVENTION, CONTINUATION, DEINIT, HABIT, TRANSIT, DYN. They are combined with ordinary frames through a mechanism called frame stacking. For example:

Here, the predicate got has been annotated with the POSSESSION frame, with Kim as possessor and the house as possessum, as we have seen before. What is new is that the INIT frame has been stacked onto the POSSESSION frame, indicating that the verb denotes not the state of possession, but a process working towards it.

Note

On a technical level, frame stacking indicates that the first frame – here: POSSESSION – fills the first role of the second frame – here: INIT.

Further examples of INIT include:

The DEINIT frame indicates a process working towards the end of a state:

The TRANSIT frame indicates a process that transitions through a state where the state does not necessarily exist at the beginning or the end:

The CONTINUATION frame indicates that a state persists:

The PREVENTION frame indicates that a state fails to come about:

The HABIT frame indicates that a state comes about repeatedly, or exists normally:

Finally, the DYN frame indicates a process that is not straightforwardly described in terms of one or more specific states but instead is a complex process transitioning through a multitude of states. Because these states can be very heterogeneous, we usually stack DYN on top of the very general SITUATION frame.

Stacking Aspectual Frames Onto Roles

Some predicates take multiple arguments wrt. different phases of the process. For these cases, there is a special notation where you can stack an aspectual frame atop a role. For example:

Using Aspectual Frames by Themselves

Some verbs take another predicate (more precisely: a phrase headed by another predicate) as an argument and contribute an aspectual meaning. We call such predicates matrix predicates, taking an embedded predicate. In such cases, use aspectual frames by themselves, without stacking.

Using Aspectual Frames with the `META` Frame

Some matrix verbs take over the syntactic subject of their embedded predicate in what is known as a control construction. In such cases, we stack an aspectual frame over the special META frame. Its scene role is filled by the embedded predicate, and its participant role is filled by the controlled subject. It is then required to add a semantic dependency from the embedded predicate to the controlled subject, even though no such syntactic dependency exists in (basic) UD.

Annotating Non-core Arguments

So far, we have only seen predicates with at most two arguments, all of whose roles are defined by the the superframe that the predicate is annotated with (the core frame). But many verbs have more than two arguments, or otherwise arguments with roles from different frames. One very common example is causative subjects, denoting an entity presented as causing the situation denoted by the core frame. For example, consider the following:

In both sentences, the core frame is INTERNAL-STATE-INIT, with the vase as the has-state argument. In the second sentence, Kim is added as another argument, presented as the causer of the situation denoted by the core frame (the breaking of the vase). We thus annotate it with the causer role from the CAUSATION frame. To distinguish them from core arguments, non-core arguments are written in all upper case.

Note

On a technical level, the non-core CAUSATION frame is interpreted as being added “around” the core INTERNAL-STATE-INIT frame so that the latter fills the former other role (the effect role). This makes sense: the vase breaking is the effect here.

Similar examples include:

Two other very common non-core arguments are the sender and recipients (experiencers) of messages.

Other non-core arguments are usually rather predicate-specific.

Annotating Nouns and Pronouns

So far, we have only looked at verbal predicates. But nouns can have (core) arguments too. Specifically, this is true for event nouns and relational nouns.

Event nouns are annotated much like verbs:

Relational nouns are not much different, except that they additionally refer to one of the participants in the relation. But we are only concerned with evocation, not with reference.

When used in predicative function in copula clauses, relational nouns assign a role to the subject:

Many nouns, of course, are neither event nouns nor relational nouns, and thus do not have any (core) arguments. The includes proper nouns and is also true for most pronouns. Thus, the rule that we mentioned earlier applies: an occurrence of a predicate without any core arguments is always annotated with the special superframe ENTITY.

Occasionally, pronouns do have arguments. In that case, frame them like you would the antecedent:

When used with a non-event, non-relational noun, the copula construction assigns INSTANCE to the subject if predicative, and SAME if equational (cf. Stassen 1997, p. 100–6; Croft 2022, § 10.1.2):

With adpositions, the copula construction can assign different roles still:

Annotating Nominal Modifiers, Oblique Nominals, and Adverbial Clauses

Both nominal modifiers (nmod, nmod:poss) and oblique nominals (obl) can be annotated as part of the argument structure of their head, as core or non-core arguments:

If they are not part of the argument structure, they should be annotated just like non-core arguments – or, as it were, as non-core arguments, because we do not attempt to dinstiguish between the two cases. The same goes for adverbial clauses (advcl):

Annotating Adjectives, Adverbs, and Relative Clauses

Predicative Adjectives

Predicative adjectives are annotated much like verbs:

Attributive Adjectives and Adverbs

Attributive adjectives and adverbs behave syntactically like modifiers, but they are also predicates assigning the modifiee a role. For the modification relation, we use the META frame, assigning the modifying predicate the scene role and – implicitly – the modifiee the participant role. We then introduce an additional semantic dependency (we call it a backlink) from the modifying predicate to the modifiee to annotate the role that the former assigns to the latter. For example:

Dynamic Adverbs

Dynamic adverbs, like dynamic verbs, are annotated using aspectual frames with stacking:

Relative Clauses

Relative clauses work very similarly: they modify a noun but also act as predicates assigning that noun a role. Again, we use SCENE for the acl:relcl relation, and a backlink to the modified noun. This is in addition to the dependency to any relative pronoun, which has the same role.

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 in the previous chapter with attributive adjectives, adverbs, and relative clauses: these elements are attached to their head using the META frame (manifesting as a SCENE modifying relation), and then a nonlocal dependency (a “backlink”) is added from the modifying element back to its head.

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:

Object control:

Non-obligatory control:

Non-obligatory control with parasitic gap:

Raising:

Secondary predicate (see also the section on secondary predicates below):

Tough construction:

Relative clause:

Reduced relative clause:

Non-finite reduced relative clause:

Relative clause with complex extracted element:

Relative clause with extraction across clause boundaries:

In cleft sentences (including wh-cleft/pseudocleft), annotate the clausal subject with the SCENE role:

Coordination:

Secondary Predicates

Secondary predicates are modifiers that syntactically attach to a (primary) predicate, but semantically predicate over one of the primary predicate’s arguments, or even something more deeply embedded. The semantic relation between the primary and the secondary predicate can be one of simple accompaniment (these secondary predicates are also called depictives), effect (these secondary predicates are also called resultatives), or something else.

Note that in the resultative example You’re talking me silly, the primary predicate talking does not assign a role to its syntactic object me, only the secondary predicate silly does. But the annotation scheme still forces us to label the dependency from talking to me. We solve this by labeling it EFFECT too, just like the dependency to the secondary predicate.

Principles for Annotation

Prefer Core Roles over Non-core Roles

Thus:

Rather than:

Try Adding Arguments

When not sure how to frame a predicate, it can help to think of additional arguments that the predicate could have and that are just not realized in this instance. For example, what is the role of It in It is not important? If we add an argument, we might get It is not important for the process, where INFLUENCE seems to fit with It as influence and the process as influenced. Thus:

Prefer to Avoid Stacking

If something could be framed with either a single frame or two stacked frames, prefer using a single frame.

Thus:

Rather than:

Prefer More Specific Frames

If not sure which of two frames to use, tend to prefer the more specific one. For example, if TIME and CONDITION would both fit and you feel that CONDITION is more specific, prefer CONDITION:

Similarly, if an argument can be seen as assigning a role to another argument, it is better to use META than, e.g., SITUATION:

Use Comments if Not Sure

You are encouraged to use comments to note down any doubts and thoughts. Comments can be added to the end of any line using the # symbol:

[TIME] generally (19) # ?
[has-time] fair (20)
[has-time] to - middling (24)

Obliqueness

Some annotation rules make reference to obliqueness, a concept used to rank dependents by the syntactic “coreness” of their role. For our purposes, nsubj dependents are least oblique, followed by obj, followed by iobj, followed by obl.

Terminology

This is a list of terms and how they are used in this manual (not necessarily in linguistics in general, though I try to be compatible with widespread usage).

Term	Explanation
adjunct	D is an adjunct of H if D is a dependent of H and not an argument, thus is assigned a role not by H but only by the adjunct construction – this means it must have a non-core role
argument	D is an argument of H if D is a dependent of H and is assigned a role by it – this can be a core role or a non-core role
dependent	if D, H are predicates, D is said to be a dependent of H if there is an edge from D to H in the annotation (either syntactic or non-syntactic)
event	a synonym for eventuality; sometimes also used as a synonym for process; I try to avoid the term due to its ambiguity
eventuality	a state or process
head	H is a head of D iff D is a dependent of H
predicate	a content word (noun, verb, adjective, or adverb) that evokes a frame
process	a dynamic eventuality, i.e., a transition over time through two or more states
scene	a synonym for eventuality
situation	a synonym for eventuality
state	a static eventuality, thus a state of affairs that does not include change over time

Extended Examples

Here we show the annotation of some full sentences:

ENTITY

This is the top of the superframe hierarchy, i.e., the most general frame. It is also the only frame that defines 0 rather than 2 roles. It is used for all those and only those occurrences of content words that do not have any core arguments, including all pronouns and all proper names.

Note that we go strictly by occurrence: the same word in another context may well have core arguments and thus be framed differently.

SITUATION

Roles: situee, situator

This superframe is the ancestor of all other superframes except ENTTIY. It denotes a situation in the most general sense: some entity (the situee) is in a situation, whatever the nature of that situation may be. If there is another core argument, it is labeled situator and the situation is then some sort of relation between two entities. There are four usual ways to use this frame:

Stacked with `-DYN`

Used for predicates that describe processes that cannot easily be framed in terms of a state but that involve complex transitions through multiple hetereogeneous states:

This often occurs with predicates that focus on aspect or made, leaving the precise nature of the relation between the two arguments underspecified.

As a Default Frame

Use SITUATION if an entity is described as being in a state, but it is not internal and there is not other frame that seems to fit better.

ACCOMPANIMENT

Roles: accompanied, accompanier

accompanier accompanies accompanied, meaning that it occurs together with it or participates equally in the same event.

Often, the accompanier denotes not the accompanying event but an entity participating in it, and must be metonymically understood as the event.

COMPARISON

Roles: compared, reference

compared is characterized with respect to reference.

reference need not be an entity similar to the compared, it can also be an abstract constraint:

CONCESSION

Roles: asserted, conceded

Special case of COMPARISON, indicating a contrast between something that’s the case (conceded) and something that seems to contradict it but is the case anyway (asserted).

IDENTITY

Roles: same, same-as

same is the same entity as same-as.

REGARD

Roles: has-regard, regard

Used for dependents that are in a semantic relation not so much with their head but with one or more of its dependents (or a descendant of one). A backlink is required from the former to the latter. When unsure which of two dependents to annotate with REGARD, choose the more oblique one.

IDENTIFIER

Roles: identified, identifier

identified is identified by identifier. identifier can be a name, an ID, a logo, or some other type of identifier.

INFLUENCE

Roles: influenced, influence

influence is or exerts an influence over influenced.

CAUSATION

Roles: effect, causer

effect is brought about by causer.

CREATION

Roles: creation, creator

Special case of CAUSATION that is used in connection with EXISTENCE-INIT events, when the causer is the creator.

MEANS

Roles: end, means

means is an intermediary causer of or is destined to serve an end (i.e., a purpose).

ASSET

Roles: has-asset, asset

has-asset is an exchange or wager or a good or service subject to an exchange or wager, and asset is given or offered in return.

MEDIUM

Roles: transmission, medium

medium is used to effect a transmission.

REACTION

Roles: reaction, trigger

trigger triggers a reaction in some other, more direct causer.

SENDING

Roles: sent, sender

sender originates a message, sent, that can be perceived.

PERFORMANCE

Roles: work, performer

performer performs a work (of art).

CONDITION

Roles: has-condition, condition

condition is presented as a (necessary or sufficient) condition for has-condition.

EXCEPTION

Roles: has-exception, exception

exception is an exception to has-exception.

LIMIT

Roles: limit, limited

The denotation of the limited dependent is restricted to the denotation of the limit dependent. With words like only, only one role is usually realized as a syntactic dependency. In the following examples, we nevertheless add the other role for clarity.

PROPORTION

Roles: numerator, denominator

Used to indicate proportional amounts.

EXPERIENCE

Roles: experiencer, experienced

experiencer experiences experienced. Can be used in a very general sense to label a non-core participant that are influenced by an eventuality, including beneficiaries and maleficiaries as well as impassive and even unconscious observers and “bystanders”. Especially used for sensory and mental perception as well as addressees in communication, often in combination with the MESSAGE and/or SENDING frame.

REASON

Roles: has-reason, reason

reason is a reason for has-reason. Useful if the reason is too vague or indirect or epistemic to call it a cause.

INTERNAL-STATE

Roles: has-state, state

The entity with the has-state role is in a state that is construed as internal to it, rather than consisting in its relation to some other entity. Typical examples include things such as being open or closed, whole or broken, asleep or awake.

Because there is no “other entity”, the second role of this frame and all its subframes is rarely used. It denotes the state itself, which is normally fully expressed by the predicate and does not need another argument. If the state was fully expressed by another argument, the predicate should have META as its base frame. However, there are some cases (rare in English) where a predicate and an argument jointly express a state, illustrated below with the QUANTITY frame. In such cases, INTERNAL-STATE and its subframes can appear with a second core argument.

POSE

Roles: has-pose, pose

Special case of INTERNAL-STATE desribing someone or something’s internal spatial configuration, e.g., body pose or posture.

PROPERTY

Roles: has-property, property

A special case of INTERNAL-STATE where the state is construed as stable, i.e.,

a concept that describes something that persists over time, and is construed to be a relatively inherent characteristic of the object over its lifetime. Example: being smart or being heavy (of a stone slab) are stable states. The stability of states is a matter of construal; a state construed as stable may change – for example, a tall tree whose top is broken off.

(Croft 2002)

LITERALITY

Roles: has-literality, literality

Property of predicate occurrences, describing the relationship between the predicate used and the eventuality it describes in terms of how literal (vs. figurative) it is.

PRECISION

Roles: has-precision, precision

Property of predicate occurrences, describing the relationship between the predicate used and the eventuality it describes in terms of how precise (vs. vague) it is.

QUANTITY

Roles: has-quantity, quantity

quantity is the quantity, degree, or extent of has-quantity.

For counts or lengths of eventualities, use TIME instead:

RANK

Roles: has-rank, rank

rank indicates the order that has-rank has in some sequence.

REALITY

Roles: has-reality, reality

Property of predicate occurrences, describing the relationship between the predicate used and the eventuality it describes in terms of how true or certain (vs. false or speculative) it is.

EXISTENCE

Roles: existent, existence

Used to indicate that something exists. Use EXISTENCE only for non-eventualities like objects. For eventualities, use REALITY or a bare aspectual frame instead.

NECESSITY

Roles: necessary, necessity

Used to indicate that something is necessary, either in an epistemic or deontic sense.

NEG

Roles: negated, negation

This frame denotes negation. It can be used by itself or stacked onto META or other frames.

POSSIBILITY

Roles: possible, possibility

Used to indicate that something is possible, or someone is capable of something.

LOCATION

Roles: has-location, location

Describes has-location as located with respect to location.

CONTACT

Roles: ornament, surface

Special case of LOCATION where ornament is on the surface of something.

CONTAINMENT

Roles: contained, container

Describes contained as contained in container. Also used for events of ingestion and excretion.

ORIENTATION

Roles: oriented, landmark

oriented is oriented in space with respect to landmark.

WRAPPING-WEARING

Roles: worn, wearer

Used for events of wrapping and wearing, where the borders between containment and contact are fluid.

MESSAGE

Roles: topic, message

A message about topic with content message exists in perceived, measured, or recorded form. When a message is created through expression or observation, use MESSAGE-INIT. When message and topic are both realized, content must assign a role to topic. Combine with SENDER and EXPERIENCER.

Use MESSAGE for predicates of expression if there is a topic and/or message argument (otherwise, use SENDING):

Gesture is a kind of expression, too:

What is depicted or talked/thought/etc. about gets the topic role:

The result of recording something gets the message role:

Use MESSAGE for predicates of perception if there is a topic and/or message argument (otherwise, use EXPERIENCE):

Predicates that denote the initiation of perception (e.g., by acquiring knowledge, or observation, or reasoning, use MESSAGE-INIT):

Predicates that denote the deinititation of perception use MESSAGE-DEINIT:

EVALUATION

Roles: evaluee, evaluation

Special case of MESSAGE that involves a value judgment or emotion about evaluee.

NORM

Roles: norm-topic, norm

A special case of MESSAGE that does not make a descriptive claim, but is about how things should be. This includes wishes, orders, goals, etc.

NONCOMP

Roles: has-noncomp, noncomp

Pseudoframe used to mark syntactic dependents that are part of a multiword predicate.

PERTINENCE

Roles: belongs, belongs-with

belongs belongs to or with belongs-with, typically in a vague sense. Use this only if none of the subframes really fit.

PART-WHOLE

Roles: part, whole

part is part of whole.

INSTANCE

Roles: instance, class

instance is an element or subset of the set denoted by class.

MATERIAL

Roles: material, object

A special case of PART-WHOLE where object does not just contain material, but is said to be made from it.

SUBSCENE

Roles: subscene, superscene

A part-whole relationship that holds between two scenes (i.e., eventualities, i.e., states or events). subscene is part of the superscene.

DYN

Roles: has-dyn, dyn

dyn is a dynamic process involving possibly multiple occurrences of has-dyn.

This frame and its subframes are used extensively in annotating dynamic predicates – they “convert” from the basic frames, which all denote states, to stacked frames denoting dynamic processes.

CONTINUATION

Roles: continued, continuation

continuation is an event with continued as an initial subeventuality that continues throughout continuation.

DEINIT

Roles: deinitiated, deinitiation

deinitiation is an event with deinitiated as an initial subeventuality, acting towards the end of deinitiated.

HABIT

Roles: habitual, habit

habit is an event with habitual as a frequent or usual subeventuality.

INIT

Roles: initiated, initiation

initiation is an event acting towards the bringing about of initiated, with initiated as a possible final subeventuality.

PREVENTION

Roles: prevented, prevention

prevention is an event as part of which the eventuality prevented fails to come about.

TRANSIT

Roles: transitory, transition

transition is an event during which the eventuality transitory comes about and ends, possibly repeatedly.

Roles: socially-related, socially-related-to

socially-related is socially related to socially-related-to with no clear hierarchy. Use BOND otherwise.

BOND

Roles: tied, tied-to

A SOCIAL-RELATION that can be framed as a bond, tie, or obligation of an individual to another individual, group, office, task, punishment, etc.

POSSESSION

Roles: possessum, possessor

possessor possesses possessum, i.e, has control over it through power and opportunity or through a social construct such as law.

TIME

Roles: has-time, time

Indicates when, how often, or for how long has-time takes place relative to time, which may be a point or span in time, or another event.

AGE

Roles: has-age, age

Special case of TIME that is about for how long somebody or something has existed.

SEQUENCE

Roles: follows, followed

follows and followed are two related eventualities – or, by metonymy, participants in related eventualities – where (barring unusual circumstances such as time travel) follows follows followed in time, but they are not only related temporally, but also logically, in the sense that followed sets up some kind of expectation of follows or its possibility.

COPY

Roles: copy, original

copy occurs similarly to or is created after the model of (but can differ from) original.

FULFILMENT

Roles: fulfiller, fulfilled

fulfilled is a promise or expectation or plan, etc., that is fulfilled by fulfiller.

SUBSTITUTION

Roles: substitute, replaced

replaced is replaced with substitute.

Annotating Multiword Predicates

Multiword predicates are predicates that consist of multiple words. They include light verb constructions, which we annotate with the META frame. Other dependents that do not fill a role of the head predicate but rather are part of the predicate – as in verbal idioms, weather verbs, existential there, names, other fixed expressions, or specialized sublanguages for such things as dates and times – are marked with NONCOMP:

In some cases, the arguments of a multiword predicate depend on different components. If the component heads a core argument of the multiword predicate, give it the same frame as the head component. Otherwise, use ENTITY.

Bound Pronouns

Bound pronouns that are part of multiword predicates (e.g., inherently reflexive verbs) are annotated with the same role as their antecedent:

Exocentric Predicates

Occasionally, there is no (direct) syntactic connection between the components of a multiword predicate. For example as … as … could be analyzed as a predicate with two components. We do not currently have a way to mark this, so we just annotate the parts as if each part were a full predicate.

Annotating Figurative Language

When a predicate is used in a figurative sense, you may optionally provide in addition the literal sense using the >> notation:

References

Leon Stassen (1997). Intransitive Predication. Oxford University Press.

William Croft (2022). Morphosyntax: constructions of the world’s languages. Cambridge University Press.

Keyboard shortcuts

Superframes