The role of word accents in semantic processing in South Swedish

Summary, in English

Prosodic cues can aid speech processing by adding semantic information in lexical tones or functional information in intonational tones. Swedish word accents are considered to have stronger grammatical functions than semantic roles, although they are shaped by both lexical and intonational information structure. This has been repeatedly observed by studies using event-related potentials (ERPs), functional magnetic resonance imaging (fMRI) and reaction times, confirming that a maximal prosodic word containing a word accent is processed by decomposition of the word into partaking morphemes. The present thesis investigated another aspect of the contrast, in which the tonal representation is associated with different semantic features. A perception experiment was conducted, and contextual constraints were used to elicit predictions of semantic features of the upcoming words. Word accent incongruency caused a longer reaction time, implying that word accents play a role in the comprehension of an utterance. An N400 effect was observed, indicating that the accent tones and the word forms can be stored together in one unit in the mental lexicon. A difference in ERP was also found between the two word accents. Accent 1 also had a greater negativity compared to accent 2, starting already at 230 ms after the word onset. This was interpreted as a pre-activation negativity (PrAN), indicating that accent 1 gives a stronger predictive certainty even when different contexts are involved. It is concluded that word accents have a semantic function, in parallel with the previously supported morphological function.

Summary, in English

Prosodic cues can aid speech processing by adding semantic information in lexical tones or functional information in intonational tones. Swedish word accents are considered to have stronger grammatical functions than semantic roles, although they are shaped by both lexical and intonational information structure. This has been repeatedly observed by studies using event-related potentials (ERPs), functional magnetic resonance imaging (fMRI) and reaction times, confirming that a maximal prosodic word containing a word accent is processed by decomposition of the word into partaking morphemes. The present thesis investigated another aspect of the contrast, in which the tonal representation is associated with different semantic features. A perception experiment was conducted, and contextual constraints were used to elicit predictions of semantic features of the upcoming words. Word accent incongruency caused a longer reaction time, implying that word accents play a role in the comprehension of an utterance. An N400 effect was observed, indicating that the accent tones and the word forms can be stored together in one unit in the mental lexicon. A difference in ERP was also found between the two word accents. Accent 1 also had a greater negativity compared to accent 2, starting already at 230 ms after the word onset. This was interpreted as a pre-activation negativity (PrAN), indicating that accent 1 gives a stronger predictive certainty even when different contexts are involved. It is concluded that word accents have a semantic function, in parallel with the previously supported morphological function.

• Author: Jinhee Kwon
• Department/s: Master's Programme: Language and Linguistics; Phonetics
• Publishing year: 2023
• Language: English
• Full text: Available as PDF - 6 MB; Download statistics
• Document type: Student publication for Master's degree (two years)
• Topic: Languages and Literatures
• Keywords: Word accents; spoken word recognition; prediction; ERP; N400
• Supervisor: Mikael Roll