Anaphoric structure emerges between neural networks

Nick Edwards, Hannah Rohde, Henry Conklin

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract / Description of output

Pragmatics is core to natural language, enabling speakers to communicate efficiently with structures like ellipsis and anaphora that can shorten utterances without loss of meaning. These structures require a listener to interpret an ambiguous form—like a pronoun—and infer the speaker’s intended meaning—who that pronoun refers to. Despite potential to introduce ambiguity, anaphora is ubiquitous across human language. In an effort to better understand the origins of anaphoric structure in natural language, we look to see if analogous structures can emerge between artificial neural networks trained to solve a communicative task. We show that: first, despite the potential for increased ambiguity, languages with anaphoric structures are learnable by neural models. Second, anaphoric structures emerge between models ‘naturally’ without need for additional constraints. Finally, introducing an explicit efficiency pressure on the speaker increases the prevalence of these structures. We conclude that certain pragmatic structures straightforwardly emerge between neural networks, without explicit efficiency pressures, but that the competing needs of speakers and listeners conditions the degree and nature of their emergence.
Original languageEnglish
Title of host publicationProceedings of the 45th Annual Meeting of the Cognitive Science Society
EditorsMicah Goldwater, Florencia Anggoro, Brett Hayes, Desmond Ong
PublisherThe Cognitive Science Society
Number of pages7
Publication statusPublished - 29 Jul 2023
Event45th Annual Conference of the Cognitive Science Society: Cognition in Context - ICC Sydney, Sydney, Australia
Duration: 26 Jul 202329 Jul 2023

Publication series

NameProceedings of the Cognitive Science Society
PublisherThe Cognitive Science Society
ISSN (Electronic)1047-1316


Conference45th Annual Conference of the Cognitive Science Society
Abbreviated titleCogSci 2023
Internet address

Keywords / Materials (for Non-textual outputs)

  • language emergence
  • pronouns
  • ellipsis
  • pragmatic
  • neural networks


Dive into the research topics of 'Anaphoric structure emerges between neural networks'. Together they form a unique fingerprint.

Cite this