poster1

1
The Mouse Was Chased by the Hat: The Influence of Semantic Context on Phonetic Processing Cristen Sullivan Department of Psychology, Salem State University The purpose of this experiment is to serve as a behavioral pilot study exploring the connection between semantic context and phonetic processing. Using a phonetic morphing algorithm we created pairs of phonetically ambiguous words, which have two possible interpretations. Participants listened to sentences that were biased towards a particular target word. (ex.. Her piano was never in TUNE vs. He climbs the DUNE). After hearing the sentence participants reported weather or not they heard a particular consonant (T vs D) in the sentence they just heard. This research posits that semantic context creates a bias towards context appropriate solutions when subjects interpret ambiguous stimuli in constraining compared to non constrained conditions. Abstract Semantic Influences on Speech Perception Conclusions Prospective Scanning Experiment References Subjects Stimuli Task Results Borsky, S,Tuller, B, & Sharpio, L.P.(1998) “How to milk a coat:”The effects of semantic and acoustic information on phoneme categorization. Journal of Acoustic Society of America, 103(5),2670-2676. Gow,D.W.(2012)The cortical organization of lexical knowledge: A dual lexicon model of spoken language processing.Brain and Language, 121, 273-288. Gow,D.W, & Caplan, D.N. (2012) New levels of language processing complexity and organization revealed by granger causation. Frontiers In Psychology. 506, 1-11. Warren,R,& Warren R. Auditory Illusions and Confusions. Scientific American,223, 30-36. This work was supported by NICDC grant R013108 (PI David Gow) through a subcontract to Salem State University. My sincere thanks to Prof. Gow and A. Conrad Nied for their assistance with this work. This study was approved by Institutional Review Boards at Salem State University, Massachusetts General Hospital, and MIT. 25 participants were used 4 male and 21 female. All participants were native speakers with no uncorrected auditory or visual defects. Task effects: more distractors eliminate participant memorization and predictability of the experiment. A clear effect was needed to justify running the prospective scanning experiment. Stimuli can be made stronger by eliminating the K target. The behavioral pilot test was conducted to determine that our test produces the most robust result. Granger causation MEG and EEG scanning. Allows for causal interactions between brain regions to be analyzed. Does sentence context influence speech perception? Warren and Warren (1970) observed that people tend to utilize sentence context to interpret ambiguous speech. Borsky (1997) , using goat or coat as target stimulus finds participants are more likely to identify target words placed in contextually biased sentences. However only a small effect was noted. Top down and bottom up effects. This experiment seeks to create a more robust effect towards context biased stimuli and serves as a behavioral pilot test for a prospective brain scanning experiment. Figure 2 from Borsky et al. (1998) JASA P Voiceless Voiced 0 10 20 30 40 50 60 70 80 90 100 Consistent Inconsistent Sentential/Semantic Bias % Responses Our results based on 25 subjects (Overall accuracy 93% on fillers) Voiced “Dip” Unvoiced “Tip” Time Visual fixation stimulus (500ms) Auditory stimulus 300ms delay Visual probe Stimulus with VOT cutback Bad Items 12 t You can come here by TRAIN 21 t She joined a brownie TROOP sound? 57 d She says her horoscope is DIRE sound 68 d The jungle growth was quite DENSE 101 p Eat after the drinks are POURED 116 p The man losing his job PLED awkward 178 b He never paid his electric BILL 201 k The officer searched for a CLUE 202 k The door is about to CLOSE 203 k Winter in Massachusetts is very COLD 204 k You should invite her to COME 205 k Needlepoint is my favorite new CRAFT 206 k The raft floated down the CREEK sound 207 k The pirate ship needs a CREW sound? 208 k The bird was probably a CROW sound? 209 k The little prince was just CROWNED 210 k Her hair has a natural CURL 211 k He slice dthe apple to its CORE 212 k There are 30 students in the CLASS 213 k Our state fish is the COD 214 k Fall evenings are very COOL 215 k I would do it if I COULD 216 k They lift the beams by CRANE 219 k The mad man was totally CRAZED 220 k He bent down into a CROUCH 263 g Neptune was the Roman sea GOD sound Top down effects Bottom up effects Sound Sound Decision VOT=61ms VOT=17ms VOT=28ms Voiced “Dip” (BDG) Unvoiced “Tip” (PTK) A large number of stimuli were used. 150 word pairs with 350 distractor stimuli. No nonsense words or continuously repeated stimuli. Prevents memorization as well as participant confusion. Participants stay more task focused. Meaning Meaning Sound

Upload: cristen-sullivan

Post on 18-Aug-2015

6 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Poster1

The Mouse Was Chased by the Hat:The Influence of Semantic Context on Phonetic Processing

Cristen SullivanDepartment of Psychology, Salem State University

The purpose of this experiment is to serve as a behavioral pilot study exploring the connection between semantic context and phonetic processing. Using a phonetic morphing algorithm we created pairs of phonetically ambiguous words, which have two possible interpretations. Participants listened to sentences that were biased towards a particular target word. (ex.. Her piano was never in TUNE vs. He climbs the DUNE). After hearing the sentence participants reported weather or not they heard a particular consonant (T vs D) in the sentence they just heard. This research posits that semantic context creates a bias towards context appropriate solutions when subjects interpret ambiguous stimuli in constraining compared to non constrained conditions.

Abstract

Semantic Influences on Speech Perception

Conclusions

Prospective Scanning Experiment

References

Subjects

Stimuli Task

Results

Borsky, S,Tuller, B, & Sharpio, L.P.(1998) “How to milk a coat:”The effects of semantic and acoustic information on phoneme categorization. Journal of Acoustic Society of America, 103(5),2670-2676.

Gow,D.W.(2012)The cortical organization of lexical knowledge: A dual lexicon model of spoken language processing.Brain and Language, 121, 273-288.

Gow,D.W, & Caplan, D.N. (2012) New levels of language processing complexity and organization revealed by granger causation. Frontiers In Psychology. 506, 1-11.

Warren,R,& Warren R. Auditory Illusions and Confusions. Scientific American,223, 30-36.

This work was supported by NICDC grant R013108 (PI David Gow) through a subcontract to Salem State University. My sincere thanks to Prof. Gow and A. Conrad Nied for their assistance with this work.

• This study was approved by Institutional Review Boards at Salem State University, Massachusetts General Hospital, and MIT.• 25 participants were used 4 male and 21 female. All participants were native speakers with no uncorrected auditory or visual

defects. • Task effects: more distractors eliminate participant memorization and predictability of the experiment.

• A clear effect was needed to justify running the prospective scanning experiment.

• Stimuli can be made stronger by eliminating the K target.

• The behavioral pilot test was conducted to determine that our test produces the most robust result.

• Granger causation MEG and EEG scanning. • Allows for causal interactions between brain regions

to be analyzed. • Does sentence context influence speech perception?• Warren and Warren (1970) observed that people tend to

utilize sentence context to interpret ambiguous speech.• Borsky (1997) , using goat or coat as target stimulus finds

participants are more likely to identify target words placed in contextually biased sentences. However only a small effect was noted.

• Top down and bottom up effects.• This experiment seeks to create a more robust effect towards

context biased stimuli and serves as a behavioral pilot test for a prospective brain scanning experiment.

Figure 2 from Borsky et al. (1998) JASA

P

Voiceless Voiced0

10

20

30

40

50

60

70

80

90

100

Consistent Inconsistent

Sentential/Semantic Bias

% R

espo

nses

Our results based on 25 subjects(Overall accuracy 93% on fillers)

Voiced “Dip” Unvoiced “Tip”

Time

Visual fixation stimulus (500ms)

Auditory stimulus

300ms delay

Visual probe

Stimulus with VOT cutback

Bad Items12 t You can come here by TRAIN21 t She joined a brownie TROOP sound?57 d She says her horoscope is DIRE sound68 d The jungle growth was quite DENSE

101 p Eat after the drinks are POURED116 p The man losing his job PLED awkward178 b He never paid his electric BILL201 k The officer searched for a CLUE202 k The door is about to CLOSE203 k Winter in Massachusetts is very COLD204 k You should invite her to COME205 k Needlepoint is my favorite new CRAFT206 k The raft floated down the CREEK sound207 k The pirate ship needs a CREW sound?208 k The bird was probably a CROW sound?209 k The little prince was just CROWNED

210 k Her hair has a natural CURL211 k He slice dthe apple to its CORE212 k There are 30 students in the CLASS213 k Our state fish is the COD214 k Fall evenings are very COOL215 k I would do it if I COULD216 k They lift the beams by CRANE219 k The mad man was totally CRAZED220 k He bent down into a CROUCH263 g Neptune was the Roman sea GOD sound

Top down effects Bottom up effects

Sound

Sound

Decision

VOT=61msVOT=17ms

VOT=28ms

Voiced “Dip” (BDG) Unvoiced “Tip” (PTK)

• A large number of stimuli were used.

• 150 word pairs with 350 distractor stimuli.

• No nonsense words or continuously repeated stimuli.

• Prevents memorization as well as participant confusion.

• Participants stay more task focused.

MeaningMeaning

Sound