Webbläsaren som du använder stöds inte av denna webbplats. Alla versioner av Internet Explorer stöds inte längre, av oss eller Microsoft (läs mer här: * https://www.microsoft.com/en-us/microsoft-365/windows/end-of-ie-support).

Var god och använd en modern webbläsare för att ta del av denna webbplats, som t.ex. nyaste versioner av Edge, Chrome, Firefox eller Safari osv.

A brain potential signalling linguistic pre-activation? : an analysis of the pre-activation negativity (PrAN)

In English

Författare

Summary, in English

Introduction: It has previously been found that strongly predicted words can be pre-activated in constraining contexts (e.g. DeLong, Urbach & Kutas, 2005). Furthermore, it has been suggested that an anticipated continuation from one syllable to another can be pre-activated before it is present in the input (Roll, Söderström, Mannfolk, Shtyrov, Johansson, van Westen & Horne, 2015). We describe a proposed event-related potential (ERP) effect – the ‘pre-activation negativity’ (PrAN) – which is hypothesised to index the degree of pre-activation anticipation and lexical competition on the previously found brain potential. The effect, occurring between 136 and 280 ms following spoken word onset, was of highly likely possible continuations of word-initial fragments (WIFs). ERP data from three previous studies was re-visited in order to investigate the effect of found to increase as lexical competitors associated with a word-initial fragment decreased. Methods: ERP item data from three previous studies was used (Roll, Horne & Lindgren, 2010; Roll et al., 2015; Roll, 2015) to investigate the relationship between the amplitude of the proposed PrAN and degree of lexical competition, defined as the “number of words which could complete a given word-initial fragment”, where “word-initial fragment” is defined as the first 2-3 phonemes of a word, including word prosodic features. The hypothesis was that the PrAN amplitude would increase as the number of upcoming potential lexical competitors decreased. The language studied was Swedish, in which tones on word-initial fragments can function as cues to possible continuations of a word (i.e. whether a suffix will follow or whether the initial fragment is the first constituent of a compound word). The lexical competition data was taken from a Swedish lexicon database, which listed the number of words which began with a particular WIF. Results: As a first step, an analysis of variance with the factor Competitors (“few”, “many”, as determined by a median split) was conducted in order to determine which electrode sites to focus the subsequent regression analysis on. Following this, a linear regression analysis was carried out on z-scores of ERP amplitude and lexical competition on the dataset of each study separately. Significant regression equations were found for each dataset, showing that as the number of potential lexical competitors decreased, the PrAN amplitude on the word-initial fragment increased. Conclusion: The results indicate that brain potentials elicited by word-initial fragments become more negative as a function of the number of lexical competitors that are potential continuations of a given fragment. We suggest that this pre-activation negativity (PrAN) is an index of the certainty as to how a word will end.

Publiceringsår

2016-08-19

Språk

Engelska

Dokumenttyp

Affisch

Ämne

  • General Language Studies and Linguistics

Nyckelord

  • ERP
  • PrAN
  • pre-activation
  • prediction
  • speech recognition

Conference name

Annual Meeting of the Society for the Neurobiology of Language

Conference date

2016-08-17 - 2016-08-20

Conference place

London, United Kingdom

Status

Published

Projekt

  • SWE-CLARIN: Svensk språkteknologi för humaniora och samhällsvetenskap