Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Lexical frequency effects on word accent processing in Swedish

Söderström, Pelle LU ; Horne, Merle LU orcid and Roll, Mikael LU (2016) Fonologi i Skandinavien
Abstract
Lexical frequency effects on word accent processing in SwedishSeveral studies have investigated the neurophysiological underpinnings of the Swedish word accents, “accent 1” and “accent 2” ([1]-[5]), which are known to have a strong association with morphology ([6]-[8]). For example, if the singular noun suffix -en is attached to the word stem bil (‘car’), the resulting word is bil1-en (‘the car’) where the subscript indicates which word accent is attached. However, if the plural suffix -ar is attached to the stem, the word will be associated with accent 2 (bil2-ar, ‘cars’). The same is true for loan words: the word chatt ‘chat room’ takes accent 1 together with the singular suffix (chatt1-en) and accent 2 with the plural suffix... (More)
Lexical frequency effects on word accent processing in SwedishSeveral studies have investigated the neurophysiological underpinnings of the Swedish word accents, “accent 1” and “accent 2” ([1]-[5]), which are known to have a strong association with morphology ([6]-[8]). For example, if the singular noun suffix -en is attached to the word stem bil (‘car’), the resulting word is bil1-en (‘the car’) where the subscript indicates which word accent is attached. However, if the plural suffix -ar is attached to the stem, the word will be associated with accent 2 (bil2-ar, ‘cars’). The same is true for loan words: the word chatt ‘chat room’ takes accent 1 together with the singular suffix (chatt1-en) and accent 2 with the plural suffix (chatt2-ar). Furthermore, in Central Swedish, accent 2 is associated with compound words. Compounding is highly productive in Swedish, allowing the formation of novel compounds such as trädkrig2 (‘tree war’) or forskningsministerskola2 (‘research minister school’). Even if all constituents making up a compound would individually be accent 1 words, the compound will still have accent 2. For example, although neder1 ‘downwards’ and länderna1 ‘the lands’ are accent 1 words, the compound Nederländerna2 (‘the Netherlands’, example taken from [12]) has accent 2. What is interesting from the point of view of predictive models of language processing is that initial word stems with accent 2 (as opposed to stems with accent 1) could then potentially cue related accent 2 suffixes (e.g. plural, past tense) as well as a potentially infinite number of compounds. Investigations using the electroencephalography/event-related potentials (EEG/ERP) technique have found that words with accent 1 stems elicit larger ERP negativity effects as compared to accent 2 stems. This effect has previously been viewed as a positivity effect driven by accent 2’s prosodic salience ([1]). However, recent investigations have shown that accent 1 stems lead to increased neural activity, meaning that the ERP effect is more likely to be driven by some feature associated with accent 1 stems instead ([4]-[5]). We propose that the accent 1 stem negativity reflects a process by which upcoming suffixes are pre-activated by the word accent. Furthermore, evidence ([3]-[4]) now indicates that accent 1 stems pre-activate their associated suffixes more strongly than accent 2 stems. In the present contribution, we present results that suggest that the stem negativity is modulated by both the token and type frequency of lexical items that are possible continuations of a particular stem.ReferencesM. Roll, M. Horne and M. Lindgren, “Word accents and morphology—ERPs of Swedish word processing,” Brain Research, vol. 1330, pp. 114–123, 2010.M. Roll, P. Söderström and M. Horne, “Word-stem tones cue suffixes in the brain.” Brain Research, vol. 1520, pp. 116–120, 2013. P. Söderström, M. Horne and M. Roll, “Stem tones pre-activate suffixes in the brain,” (submitted).M. Roll, P. Söderström, P. Mannfolk, Y. Shtyrov, M. Johansson, D. van Westen and M. Horne, “Word tones cueing morphosyntactic structure: neuroanatomical substrates and activation time course assessed by EEG and fMRI,” Brain & Language, vol. 150, pp. 14–21, 2015.M. Roll, “A neurolinguistic study of South Swedish word accents: Electrical brain potentials in nouns and verbs,” Nordic Journal of Linguistics, vol. 38, no. 2, pp. 149–162, 2015.G. Bruce, Swedish word accents in sentence perspective. Lund: Gleerup, 1977.T. Riad, The Phonology of Swedish. Oxford: Oxford University Press, 2014.J. Rischel, “Morphemic Tone and Word Tone in Eastern Norwegian,” Phonetica, vol. 10, no. 3–4, pp. 154–164, 1963. (Less)
Please use this url to cite or link to this publication:
author
; and
organization
publishing date
type
Contribution to conference
publication status
published
subject
conference name
Fonologi i Skandinavien
conference location
Gothenburg, Sweden
conference dates
2016-02-19 - 2016-02-20
language
English
LU publication?
yes
id
2ffb7aae-f64c-45e3-a4c6-81fde2dca167
date added to LUP
2019-05-22 03:26:07
date last changed
2019-05-29 02:19:02
@misc{2ffb7aae-f64c-45e3-a4c6-81fde2dca167,
  abstract     = {{Lexical frequency effects on word accent processing in SwedishSeveral studies have investigated the neurophysiological underpinnings of the Swedish word accents, “accent 1” and “accent 2” ([1]-[5]), which are known to have a strong association with morphology ([6]-[8]). For example, if the singular noun suffix -en is attached to the word stem bil (‘car’), the resulting word is bil1-en (‘the car’) where the subscript indicates which word accent is attached. However, if the plural suffix -ar is attached to the stem, the word will be associated with accent 2 (bil2-ar, ‘cars’). The same is true for loan words: the word chatt ‘chat room’ takes accent 1 together with the singular suffix (chatt1-en) and accent 2 with the plural suffix (chatt2-ar). Furthermore, in Central Swedish, accent 2 is associated with compound words. Compounding is highly productive in Swedish, allowing the formation of novel compounds such as trädkrig2 (‘tree war’) or forskningsministerskola2 (‘research minister school’). Even if all constituents making up a compound would individually be accent 1 words, the compound will still have accent 2. For example, although neder1 ‘downwards’ and länderna1 ‘the lands’ are accent 1 words, the compound Nederländerna2 (‘the Netherlands’, example taken from [12]) has accent 2. What is interesting from the point of view of predictive models of language processing is that initial word stems with accent 2 (as opposed to stems with accent 1) could then potentially cue related accent 2 suffixes (e.g. plural, past tense) as well as a potentially infinite number of compounds. Investigations using the electroencephalography/event-related potentials (EEG/ERP) technique have found that words with accent 1 stems elicit larger ERP negativity effects as compared to accent 2 stems. This effect has previously been viewed as a positivity effect driven by accent 2’s prosodic salience ([1]). However, recent investigations have shown that accent 1 stems lead to increased neural activity, meaning that the ERP effect is more likely to be driven by some feature associated with accent 1 stems instead ([4]-[5]). We propose that the accent 1 stem negativity reflects a process by which upcoming suffixes are pre-activated by the word accent. Furthermore, evidence ([3]-[4]) now indicates that accent 1 stems pre-activate their associated suffixes more strongly than accent 2 stems. In the present contribution, we present results that suggest that the stem negativity is modulated by both the token and type frequency of lexical items that are possible continuations of a particular stem.ReferencesM. Roll, M. Horne and M. Lindgren, “Word accents and morphology—ERPs of Swedish word processing,” Brain Research, vol. 1330, pp. 114–123, 2010.M. Roll, P. Söderström and M. Horne, “Word-stem tones cue suffixes in the brain.” Brain Research, vol. 1520, pp. 116–120, 2013. P. Söderström, M. Horne and M. Roll, “Stem tones pre-activate suffixes in the brain,” (submitted).M. Roll, P. Söderström, P. Mannfolk, Y. Shtyrov, M. Johansson, D. van Westen and M. Horne, “Word tones cueing morphosyntactic structure: neuroanatomical substrates and activation time course assessed by EEG and fMRI,” Brain & Language, vol. 150, pp. 14–21, 2015.M. Roll, “A neurolinguistic study of South Swedish word accents: Electrical brain potentials in nouns and verbs,” Nordic Journal of Linguistics, vol. 38, no. 2, pp. 149–162, 2015.G. Bruce, Swedish word accents in sentence perspective. Lund: Gleerup, 1977.T. Riad, The Phonology of Swedish. Oxford: Oxford University Press, 2014.J. Rischel, “Morphemic Tone and Word Tone in Eastern Norwegian,” Phonetica, vol. 10, no. 3–4, pp. 154–164, 1963.}},
  author       = {{Söderström, Pelle and Horne, Merle and Roll, Mikael}},
  language     = {{eng}},
  month        = {{02}},
  title        = {{Lexical frequency effects on word accent processing in Swedish}},
  year         = {{2016}},
}