Morehead State University
Accents Can Affect Attachment
Institution
Morehead State University
Faculty Advisor/ Mentor
Katy Carlson
Abstract
This project explored the effects of prosody, specifically the placement of accents, on the meaning of sentences. Given the sentence “John claimed that Mary had arrived last week,” the listener could interpret “last week” as modifying “claimed,” meaning that the claiming happened last week (the high attachment meaning); or could interpret “last week” as modifying “arrived,” so Mary arrived last week (low attachment). We know that a prosodic boundary, a pause, after “arrived” makes it more likely that people will consider high attachment, even though low attachment is generally preferred. In this project, we placed accents on either “claimed” or “arrived,” hypothesizing that emphasis on a particular verb might also draw attachment to that verb. The presence of a prosodic boundary was also varied. After hearing each of the 20 sentences, listeners chose between two visually-presented paraphrases. With 24 subjects so far, we found a significant effect of the prosodic boundary, with a break after “arrived” increasing high attachments to 26%, vs. 13% without a boundary. We also found a marginally significant effect of accent position, with the accent on the higher verb “claimed” also increasing high attachments. This finding is novel because accents are usually thought to affect processes outside the sentence, such as pronoun resolution. But this shows accents affecting the basic structure of the sentence. This accent effect may be due to increased memorability of a particular verb or to increased semantic processing.
Accents Can Affect Attachment
This project explored the effects of prosody, specifically the placement of accents, on the meaning of sentences. Given the sentence “John claimed that Mary had arrived last week,” the listener could interpret “last week” as modifying “claimed,” meaning that the claiming happened last week (the high attachment meaning); or could interpret “last week” as modifying “arrived,” so Mary arrived last week (low attachment). We know that a prosodic boundary, a pause, after “arrived” makes it more likely that people will consider high attachment, even though low attachment is generally preferred. In this project, we placed accents on either “claimed” or “arrived,” hypothesizing that emphasis on a particular verb might also draw attachment to that verb. The presence of a prosodic boundary was also varied. After hearing each of the 20 sentences, listeners chose between two visually-presented paraphrases. With 24 subjects so far, we found a significant effect of the prosodic boundary, with a break after “arrived” increasing high attachments to 26%, vs. 13% without a boundary. We also found a marginally significant effect of accent position, with the accent on the higher verb “claimed” also increasing high attachments. This finding is novel because accents are usually thought to affect processes outside the sentence, such as pronoun resolution. But this shows accents affecting the basic structure of the sentence. This accent effect may be due to increased memorability of a particular verb or to increased semantic processing.