5 The Hearing Brain and Voice Perception Flashcards
what does voice reveal about you
Voice hints at how social you are - introverts and extroverts
Your trustworthiness - deep pitch at end of word in men
Higher pitch at end of word in woman
When saying hello
Trust worthy
Stressed - vocal cords tighten - disorganised speech
Speaking slowly - in control - indicator keep emotional in check
what is voice dependent on
Voice is dependent on vocal tract
Parameters of voice dependent on anatomy
Dependent on physical state - vocal tract vibrates differently
study about voice revealing a lot about you
N = 320 participants rated social traits on “hello” spoken by 64 people
High consistency
If voices convey no information, ratings on these social traits would be random.
But Cronbach Alpha (a measure for interrater agreeableness) suggests that people consistently agree on whether a voice sounds attractive or confident.
We seem to associate voices with social traits on a collective level.
Many social traits are related and can be mapped to two principle
components, what are they?
In both genders,
The 1st Principle Component is highly correlated with Trustworthiness;
The 2nd Principle Component is highly correlated with Dominance.
they are independent of eachother
what can a hello tell us
From a simple “hello” we can quickly assess the speaker’s personality and decide whether and how to approach them.
how are voices transmitted
• Voices are transmitted via vibrations of the air.
what are the 3 parameters voices can be decribed as
pitch
intensity
tempo
what is pitch
high or low
fundamental frequency
Female or male Stressed or sad Vibrations of energy Vibrates a lot - higher frequency - high pitch Vibrates slowly - low How fast energy oscillates
pitch - vibrates alot
high voice
pitch - vibrates slow
low voice
what is intensity
amplitude
How far the peaks and troughs are in the voice
Loud - high
intenisty- loud
high peaks low troughs
intensity - quiet
small peaks and troughs
what is tempo
speed
how fast you talk
when you enjoy something
pitch up
intensity down
tempo down
when you experience elation
pitch up
intensity up
tempo down
when experience sadness
pitch down
intensity down
tempo down
when experience grief
pitch up
intensity up
tempo up
different emotions are experiences differently
distinct acoustic profiles
Software to analyse voice
Different emotions have different acoustic profiles
Elations increase different amount in comparison to grievances
can all emotions be reliably dinstinguished from each other
no
german people speaking - can the vocal emotional expressions be recognised by people from a different culture?
Record saying pseudosentences in different emotions
Emotion is not very clearly distinguished from other emotions
Their performance in detecting German emotions are low in comparison to European counterparts
Measures how consistent they’ve judged these emotions between pps
Indonesia - confused by German emotions
Indonesian and german
Some vocal emotions are shared cross cultures
Others may be influenced by culture- and language-specific factors
joy is most ambiguous emotion
what do voices convey information
Voices convey important socio-emotional information
can information be accurately decoded by the brain
This information can be accurately decoded by the brain:
• There must be specialised neural architecture to process voice
• And different processing pathways for different aspects of voice
fMRI assumptions
Increased neural activity needs more oxygen and glucose
fMRI detects changes in blood oxygen levels, which indirectly reflects the changes in neural activity
fMRI allows us to study cognitive activity in vivo, without causing any harm to the brain
based on assumption when brain working hard need more oxygen and glucose
what does the temporal voice area (TVA) show in voice perception
Listening to voices vs. non-vocal sounds elicits increased bilateral neural activity in the temporal cortices
It provides objective evidence that there are brain regions dedicated for voice processing
what is neural adaptation in speaker adaptation
Neural Adaptation: a gradual decrease over time in the responsiveness of the sensory system to a constant stimulus
Some brain regions will adapt to repeating speaker identity in the voice
some brain areas will adapt to repeating syllable in the voice
belin, zatorre , 2003
how to undersrand which part of brain responds to speaker identity
Understands which part of brain area responds to speaker identity
Repeat over and over - brain gets adapted to it and it no longer bothers you - e.g. constant fan
Exhaust the responsiveness of sensory system
Speaker identity - respond less when same speaker
Repeat syllables - brain area that responds to syllables will adapt - different speakers but same syllables
which brain area adapts to the repeating speaker identity in the voice
Speaker identity
Right side of brain - anterior part of temporal lobe
Part of the brain that responds to the person
what did grandjean et al 2005 find
participants pay attention to the angry voices
and the left audio
right voice area selectivey responses to angry speech
Are the Voice Areas necessary for voice processing?
Correlation is not causation
cannot say whether brain area is causally libked to voice processing
Is voice processing correlated with or caused by the activity in the Temporal Voice Areas?
We can test the effects of a disruption to the Voice Areas on the performance of voice recognition using Transcranial Magnetic Stimulation (TMS)
Are the Voice Areas necessary for voice processing- right temporal ?
Right temporal TMS impairs voice detection
TMS to individual Voice Areas impairs accuracy in voice/nonvoice discrimination compared to TMS to the Control Site
Low-level loudness discrimination is not affected by TMS to the Voice Areas.
what isnt affected bt TMS in voice areas
Low-level loudness discrimination is not affected by TMS to the Voice Areas.
what does TMS to individual voice areas impair
TMS to individual Voice Areas impairs accuracy in voice/nonvoice discrimination compared to TMS to the Control Site
what can voices tell us
• Voices can tell us who you are and how you feel
• Speaking identity and vocal emotion are processed in different regions of the
Temporal Voice Areas.
where are speaking identities and vocal emotions processed
• Speaking identity and vocal emotion are processed in different regions of the Temporal Voice Areas.
top down processing in speech perception
Speech is a continuous signal that does not have punctuation or spaces
brain can extract linguistic units even when speech is a continuous signal
what is top down processing in speech perception
Focuses on useful info while suppressing the noise
what is bottom up processing in speech perception
takes in all external stimuli
top down processing in brain processing
Brain - brain is predicting and directing behaviour in terms of how you interact with the environment - use what you know before about information from outside
vocoded speech demonstrating speech perception
Natural full voice - energy distribution into different bands - stripes coherent within different roles
Speech perception - eliminate all other noises - top down processing
Brain focuses on particular voice and suppresses others
is speech processing left lateralised
yes
brocas patients
two patients cant speak
Both patients - left frontal part of brain - damage
Important in speech production
wernikes patient
cant understand speech
Wernickes - cannot understand
Speech comprehension problem
Left posterior part of brain which connects temporal and parietal lobe
speech
spectral info - full
intellibigibility- full
rotated speech
spectral info - full
intelligiblity - bad
mirror image of first one
preserves energy levels but inverted
vocoded speech
intelligible
spectral - bad
intelligible - partial
rotated vocoded speech
spectral info - bad
intelligibility - bad
what does wernikes and brocas produce
double dissociation
what are the findings of vocoded intellibible speech in left temporal regions
Left temporal regions responded to intelligible speech more than rotated, unintelligible speech.
what are the findings of vocoded intellibible speech in right temporal regions
Right temporal regions responded to spectrally rich speech more than vocoded speech.
what is the acoustic uncertainty principle
time - trade off - frequecy
less precise in time more precise in frequency
more precise in time less precise in frewquency
is left side more specialised in processing temporal info
yes
is right slide more specialised in processing pitch and spectral
yes
in music whats more important
spectral/pitch more important than temporal
what more important in speech
temporal more important than spectral
Doesn’t matter whether female stressed
Only matters syllables, pauses whether makes sense
what processing does speech recognition depend on
top down processes
where is speech primarily processed
in the left hemisphere
what are right temporal regions more specialised in processing
time information
right temporal areas are more specialised in processing
spectral information
what is the mcgurk effect
ba da va
Sent to temporal lobe trying to predict the sound coming out
Brain trying to predict
if percieving ka and hearing pa
Combine the two - engage in auditory the most but also the motor region
Predict what the man is trying to say - motor system to simulate what man is trying to say
increased involvement of motor system in percieving mcgurk effects
is speech perception complemented by speech production
yes
How does the brain segment continuous speech signal into separate syllables and words?
Neural oscillations for speech segmentation
brain oscillations concide with the speed of talking
most human languages have the syballic rate of
4.5hz
Speech perception is informed by top-down predictions from the motor
system (predicting what the speech is likely to be)
true
The motor system may modulate neural oscillations which facilitate the
segmentation of speech signal
true
what is the optimal oscilation reflecting the average syllabic rate
Optimal oscillations = 4.5Hz which reflects the average syllabic rate across human languages
speech is processed in a network what are the two pathways
Speech is processed in a network
• Dorsal “how” pathway
• Ventral “what” pathway
what is the dorsal pathway
how
what is the ventral pathway
what
auditory motor connectivity
mcgurk illusions
neural oscillations
angular gyrus
phonological buffer
percieved gestures
posterior STS
heschls gyrus
primary auditory cortex
speech recognition
anterior STS
semantic knowledge
temporal pole
brocas area
planning of speech production
