Deck 13: Speech Perception

Question 1

Computer speech recognition is

A) better than human speech recognition in all conditions.
B) better than human speech recognition in accuracy.
C) equal in all respects to human speech recognition.
D) worse than human speech recognition.

Answer 1

D

Question 2

The vowel sound /ae/ (as in "had") has

A) a single formant.
B) two formants.
C) three formants.
D) no formants.

Answer 2

C

Question 3

Spectrograms of sentences show

A) that clear pauses occur between each spoken word.
B) that formant transitions account for the breaks between words.
C) no clear pauses or breaks between words.
D) segmentation of words is easily accomplished by listeners in any context.

Answer 3

C

Question 4

The voice onset time (VOT) for the sound /da/ is 17 ms, and the VOT for the sound /ta/ is 91 msec. When a computer produces a sound with a VOT of 65 msec, listeners are likely to report hearing

A) the /da/ sound
B) the /ta/ sound
C) the /ja/ sound
D) a combination of /ta/ and /da/

Question 5

The McGurk effect is most similar to which of the following performers?

A) ventriloquists
B) hockey players
C) ballet dancers
D) gymnasts

Question 6

Listening to someone speak a foreign language you are not familiar with can lead to

A) the correspondence problem.
B) the segmentation problem.
C) the formant transition effect.
D) acoustic signaling.

Question 7

Which of the following can be considered an articulator?

A) a sound spectrogram
B) a running spectral display
C) the soft palate
D) an articulation agreement

Question 8

Humans perceive the sound /b/ to be the same, even when the coarticulation of the sound is be different. This phenomenon is an example of

A) perceptual constancy.
B) acoustic consistency.
C) phonemic transitions.
D) the segmentation problem.

Question 9

The ______ is the shortest segment of speech that, if changed, changes the meaning of the word.

A) formants
B) phonemes
C) tadomas
D) morphemes

Question 10

There are _____ phonemes for vowel sounds in the English language.

A) two
B) five
C) six
D) thirteen

Question 11

The McGurk effect illustrates the importance of ___________ on speech perception.

A) the motor cortex
B) articulators
C) formants
D) vision

Question 12

Rubin et al. (1976) asked participants to respond when they heard a word that started with the /b/ sound. The average response time when real words were used was ____; and ______ when non-words were used.

A) 100 msec; 57 msec
B) 580 msec; 631 msec
C) 995 msec; 900 msec
D) 1.87 sec; 1.88 sec

Question 13

The ubiquitous "Whazzup!" is a sloppy pronunciation of "What's up?" The spectrograms of each of these two spoken phrases would indicate

A) no difference in the spectrograms between the two phrases.
B) only a difference in the frequency axis between the two phrases.
C) that there is a pause in the middle of "What's up?".
D) major differences between the two, especially in the middle of the spectrograms.

Question 14

The existence of phonetic boundaries

A) is currently debated among speech perception researchers.
B) only occurs at VOTs of greater than 250 ms.
C) has been demonstrated using discrimination experiments.
D) shows that categorical perception does not occur in speech perception.

Question 15

Jessica looks at Ashlee on a videotape. Ashlee's lips are making the movement for the sound /ga-ga/, but the sound that is actually presented is the acoustic signal for /ba-ba/. What sound is Jessica most likely to report hearing?

A) /ga-ga/
B) /ba-ba/
C) /da-da/
D) /pa-pa-joe/

Question 16

If a listener is asked to pay attention to speech provided by familiar voices, the _____ is activated, as shown by fMRI studies.

A) FFA
B) STS
C) both the FFA and STS
D) none of these

Question 17

The problem of variability from the way different people speak can be demonstrated by

A) the pitch differences in different people's voices.
B) the different accents of different speakers.
C) the speed at which the speaker talks.
D) all of these.

Question 18

When you say "bat" and "boot," the /b/ sound is articulated differently. This is an example of

A) phoneme contiguity.
B) phoneme incontiguity.
C) coarticulation.
D) alveolar context.

Question 19

A sound spectrogram is a plot of ______________ as a function of _____________, with darker areas representing greater intensity.

A) frequency; time
B) amplitude; frequency
C) time; amplitude
D) time; spatial location of sound source

Question 20

The consonant sound "____" is produced by placing your bottom lip against your upper front teeth and then pushing the air between the lips and the teeth.

A) d
B) g
C) r
D) f

Question 21

Brain scanning research has shown that the ________ is responsible for identifying sounds, and the _____ is responsible for locating sounds.

A) "where" (dorsal) stream; what" (ventral) stream
B) "what" (ventral) stream; "where" (dorsal) stream
C) corpus callosum; "where" (dorsal) stream
D) pacinian area; what" (ventral) stream

Question 22

Palmeri et al. (1993) had participants listen to a word list that was spoken by (1) the same speaker, or (2) different speakers. In a subsequent recognition memory test, participants were

A) more accurate when one speaker said all of the words.
B) more accurate when different speakers said the words.
C) faster in responding when different speakers said the words.
D) the same in accuracy in both conditions.

Question 23

Saffron et al. (1996) found that the ability to use transitional probabilities to segment sounds develops around

A) 2 months old.
B) 8 months old.
C) 2 years old.
D) 5 years old.

Question 24

The joke in the textbook with the punch line "Yeah, right" is used as an example of

A) sarcasm as a type of indexical characteristic.
B) the effect of aphasia on inappropriate social interaction.
C) shadowing in naturalistic environments.
D) multimodal processing differences between Russian and English speakers.

Question 25

Micelli et al. (1980) found that brain damage to the parietal lobe caused the patient to have difficulty discriminating between syllables. Micelli et al found that

A) all these patients could not understand words.
B) all these patients had "word deafness."
C) some of these patients could not hear pure tones.
D) some of these patients could still understand words.

Question 26

Saffron et al. (1996) found that 8-month-old infants listened to ______ test stimuli longer, providing evidence that infants are capable of __________ learning.

A) whole word; vicarious
B) whole word; transitional probability
C) part word; statistical
D) part word; formal operational

Question 27

Top-down processing can help

A) segment acoustic signals.
B) recognize phonemes.
C) recognize words.
D) all of these.

Question 28

Using fMRI, Belin et al., (2000) were able to reveal that, in humans, the superior temporal sulcus (STS) is

A) activated more for human voices than for other sounds.
B) activated for any sound that falls in the same frequency range as the human voice.
C) responsible for coordinating conflicting visual and auditory information.
D) responsible for phoneme segmentation.

Question 29

Japanese children at the age of _______ can tell the difference between the /r/ sound and the /l/ sound just as well as American children.

A) 6 months old
B) 1 year old
C) 2 years old
D) 5 years old

Question 30

Damage to Broca's area in the frontal lobe results in difficulty

A) in speaking.
B) in understanding speech.
C) in lip reading.
D) in determining phonetic boundaries.

Question 31

Masakazu is a 4-month-old Japanese infant. When presented the phonemes /r/ and /l/,
He will

A) not be able to discriminate between these two phonemes.
B) be able to discriminate between these two phonemes.
C) need the VOT changed to 10 msec to discriminate between the two phonemes.
D) spontaneously mimic the /r/, but not the /l/.

Question 32

Warren showed that when a cough sound replaced the sound of the first /s/ in the word "legislatures," listeners reported hearing

A) just the cough sound where the /s/ was originally.
B) just the cough sound because it masked the whole word.
C) just the /s/ sound.
D) the cough and the /s/ sound, but the cough position was not correctly identified.

Question 33

The discovery of ______ is used as support for the motor theory of speech perception.

A) Broca's area
B) Wernicke's area
C) audiovisual mirror neurons
D) simple cells in area V1

Question 34

The ability to categorize speech sounds is found in infants as young as

A) one-month-old.
B) three-months-old.
C) six-months-old.
D) 10-months-old.

Question 35

Link et al. (2003) studied the perceived meaning of listeners' to the phrase "Let's do lunch sometime," based on the speaker's inflection and emotional state. These are examples of

A) indexical characteristics.
B) segmentation effects.
C) lip reading effects.
D) speech shadowing techniques.

Question 36

A fan of science fiction television shows would be more likely to make sense of the phrase "Start Wreck In Tore Prize." This demonstrates the importance of

A) the McGurk effect on speech perception.
B) bottom-up processing on phomenic boundaries.
C) Tadoma on speech production.
D) meaning on segmentation.

Question 37

Miller and Isard presented listeners with grammatical sentences ("Gadgets simplify work around the house"); ungrammatical word strings ("Between gadgets highways passengers the steal"), and anomalous sentences ("Gadgets kill passengers from the eyes"). The results showed that the listener's ability to accurately report the phrase was

A) highest for the grammatical condition, followed by ungrammatical, and then anomalous.
B) highest for the grammatical condition, followed by anomalous, and then ungrammatical.
C) the same for grammatical and anomalous, which were both better than ungrammatical.
D) the same for all three conditions.

Question 38

A person with Wernicke's aphasia

A) has damage to an area of the occipital cortex.
B) can comprehend words, but can't produce speech.
C) can easily isolate phonemes, but have trouble with word segmentation.
D) produces fluent speech, but in nonsensical "word salads."

Question 39

Your ability to read the sentence "H*V* A N*C* D*Y" is used as an example of the importance of

A) bottom-up processing.
B) top-down processing.
C) audiovisual speech perception.
D) sideways processing.

Question 40

Watkins, using transcranial magnetic stimulation, found that motor-evoked potentials were highest when the participant

A) listened to speech sounds.
B) listened to non-speech sounds.
C) watched someone else's lip movements make speech sounds.
D) both heard speech sounds and watched someone else's lips make speech sounds.

Question 41

Discuss the methods and results of two studies of the phonemic restoration effect.

Question 42

Discuss two sources of the variability problem. Provide examples for each.

Question 43

What is the McGurk effect? What evidence exists for the physiological basis for this effect?

Question 44

Discuss what information is used by listeners to accomplish speech segmentation.

Question 45

Describe how voice onset times (VOTs) have been used to study categorical perception.

Question 46

(a) What are indexical characteristics?
(b) Describe examples of indexical characteristics used in your everyday life.

Question 47

The research by Rivera-Gaxiola on the effect of Spanish speech sounds on the electrical potentials of American infants shows that speech perception involves

A) experience-dependent plasticity.
B) the genetic basis to speech perception.
C) the genetic basis for speech production.
D) multimodal stimulation.

Question 48

Discuss the "dual-stream model of speech perception." Briefly describe research that supports this model.