From Event Representation to Linguistic Meaning

Abstract

A fundamental aspect of human cognition is the ability to parse our constantly unfolding experience into meaningful representations of dynamic events and to communicate about these events with others. How do we communicate about events we have experienced? Influential theories of language production assume that the formulation and articulation of a linguistic message is preceded by preverbal apprehension that captures core aspects of the event. Yet the nature of these preverbal event representations and the way they are mapped onto language are currently not well understood. Here, we review recent evidence on the link between event conceptualization and language, focusing on two core aspects of event representation, event roles and event boundaries. Empirical evidence in both domains shows that the cognitive representation of events aligns with the way these aspects of events are encoded in language, providing support for the presence of deep homologies between linguistic and cognitive event structure.

Roles for Event Representations in Sensorimotor Experience, Memory Formation, and Language Processing

“…The idea that the participants of a transitive event are perceived sequentially, in a canonical order, is supported by several other empirical studies, many of which are mentioned in Ünal, Ji, and Papafragou (2021). For instance, Wilson, Papafragou, Bunger, and Trueswell (2011) found that observers tasked with identifying agents or patients in a visually presented event can identify agents sooner than patients; observers also mention agents more frequently than patients, when reporting observed events, suggesting agents are more salient (Wilson, Ünal, Trueswell, & Papafragou, 2014).…”

“…For instance, Wilson, Papafragou, Bunger, and Trueswell (2011) found that observers tasked with identifying agents or patients in a visually presented event can identify agents sooner than patients; observers also mention agents more frequently than patients, when reporting observed events, suggesting agents are more salient (Wilson, Ünal, Trueswell, & Papafragou, 2014). However, other studies, also reviewed by Ünal et al (2021) indicate that the “gist” of an event—including information about its participants—can be perceived at very short latencies, which preclude serial allocation of attention to participants (see, e.g., Dobel, Gumnior, Bölte, & Zwitserlood, 2007; Hafri, Papafragou, & Trueswell, 2013; Hafri, Trueswell, & Strickland, 2018). This rapid gist‐identification mechanism is clearly real and clearly distinct from the serial mechanism that is our focus.…”

Tea With Milk? A Hierarchical Generative Framework of Sequential Event Comprehension

“…For example, the event, <Woman swallows tea>, terminates with the end state, [Tea consumed]. In Linguistics, end states are considered key to the conceptual structure of the so‐called bounded events (Comrie, 1976; Parsons, 1990; Vendler, 1957; see Ünal et al, 2021, this issue for discussion) 2 . In the study of goal‐relevant action, it has been proposed that the anticipation of end states is the key trigger of actions, ranging from simple motor movements (“ideomotor action,” see Hommel et al, 2001; James, 1890/1981; Lotze, 1852; Prinz, 1987) to complex sequences that define longer‐term goals and plans (e.g., Jones & Davis, 1965).…”

“… 2 2 As discussed by Unal, Ji, and Papafragou (2021, this issue), bounded events are distinguished from unbounded events in which the ending is not explicitly coded in the linguistic expression. For example, while the end state, [Tea consumed], is part of the conceptual structure of the event that is conveyed by the sentence, “The woman swallowed the tea,” the linguistic expression, “The woman drinks some tea” does not specify the end of the tea drinking.…”

Events, Event Prediction, and Predictive Processing

“…More is needed to fully characterize events and event coding, in addition to some notion of coherence, as indeed discussed by the papers in this issue: evolution of event inference, learning of events, context‐dependent inference of events, involvement of action and the self‐model, continuous versus discrete processing, recombinability, semantics, and linguistic properties (see Cooper, 2021; Paulin and Cahill‐Lane, 2021; Ünal and Papafragou, 2021). And things are, as always, more complicated when it gets to implementation and cognitive functions, with topics including temporal as well as hierarchical segmentation, involvement of the default network, relation to future thinking, overlaps (involving hippocampus) with episodic memory and spatial representation (see Bilkey and Jensen, 2019; Knott and Takac, 2021; Stawarczyk et al, 2021).…”