lecture 7 - Phonology Flashcards
how much frequency do we hear
variation in vocal chords
we do have natural variation between the length of vocal chords (eg females and males , female smaller vocal chords have higher voices)
how does sensory transduction in the cochlea work
Cochlear breaks down/filters incoming sounds into different frequencies
‘Amplifies’ key frequencies and converts this information into neural code.
This is referred to as Transduction - energy from outer word to the code
readable by the inner world (the brain!)
We know a lot about the initial stages (cochlea → brain) and little about more complex stages (auditory signal/code → phonemes → words)
what is the cochlear
●ear composed of three bits Outer, middle, inner ear
air pressure hits eardrum
● Vibration of the the Eardrum
transfers sounds into cochlea
-
● Cochlea is a complex organ -
but we are interested in how it
‘filters’ the frequencies
describe the process of frequency decopisition in cochlea
● Fluid in the cochlea moves
when pressure is applied,
● Causes the Basilar Membrane
to respond in a bell-like manner
● From base to apex - different
frequencies
● Acts like a filter
only takes 5 milliseconds to decode outside of ear to inside
What does the world sound like if parts of cochlear stop working?
?
tonotopy
ability of cochlea
the systematic arrangement of neurons based on their response to tones of different frequencies (different nerves sensitive to certain frequencies)
Auditory cortex - recognising speech features
tonotopic organisation
certain parts of pac fire for certain frequencies
whole process of sound being processed through ear
Cochlea breaks down sound into a
frequency patterns and converts it into the neural code
● Cochlea is organised tonotopically and this organisation is preserved through the auditory pathway from cochlea to PAC
● Auditory pathways in the brainstem and thalamus future ‘clean’ and ‘amplify’ the
code send from cochlea
● PAC is a complex area where patterns of sound are bound together into auditory
features
● PAC has tonotopic organisation but it is highly variable within individuals
phonolgy
into the brain..
what is a phoneme
phoneme a unique combination of sound features - spectral fingerprint
● Phoneme - a unit of sound that can distinguish
one word from another
phonemes sounding differently
● Phonemes → words; Atoms → elements
● We all sound differently, our phonemes sound differently, yet they are all similar enough for us
to understand each other (think visual objects)
● Babies learning phoneme ‘blueprints’ of their
first language
what is phonological perception in babies
baby born with fully grown cochlea, hairs in cochlea grow accordingly and learn and adjust
At 6-8 months babies can
perceive changes in phonetic
boundaries in their native and
in non-native languages
● At this age babies are ‘perfect
sound learners’
● At 10-12 they only pay
attention to sound changes
relevant to their native
language
phonemes are learnt _____ words
before
Learning individual
phonemes and syllables
are prerequisite to
word production
● Babies learn individual
phonemes and their
simple combinations
before being able to put
them together into
word
how are phonemes categorised according to features
some phoneme sound similar /b/ - /p/ because they share features
● Features are grounded in the spectral characteristics of the sound
Vowels - periodic
Consonants - aperiodic sound burst
Phonology Primer - Features to Phonemes
the international phonetic alphabet
Each features is linked to property of the sound spectrum
summary of phonemes / phonology
● To recognise a word we need to recognise its consistent sounds
● Sounds of the word - phonemes
● A phoneme a unique combination of sound auditory features - a fingerprint
● Spectrogram shows how some sounds are similar based on shared features
● Phonology - gives us descriptive labels for features, groups them
● Brain recognises words by mapping phonemes
how does phoneme recognition happen in the brain
● Tonotopic organisation of auditory areas helps to recognise complex frequency patterns map them to learned phonemes
Phoneme recognition in the brain
kumar et al
● New sound patterns were
repeated over trials.
people were given sudophonemes , phonemes that sounds like phoneme but not in their language
● All sounds were not phonemes
but matched on complexity
● fMRI activity patterns could
very quickly recognise and
learn repeated sound patterns
fast mapping in children
form phonemes to words
*Children learn very quickly in early childhood.
*First word between 10–14 months of age, but mapping words begins earlier
*300 words by their second birthday
is fast mapping and its rate unique to humans ?
arguments for both humans and animals fast mapping
Fast Mapping” and its rate is thought to be unique to humans
* Children use context and visual clues to infer the meaning of words.
* 1-2 exposure to a new word sufficient for mapping (ages 1-2)
* Show retention after the word has been mapped (short→ long term memory)
summary
● Sound travels down the ear to cochlea where it is filtered decomposed
into different frequencies
● In the cochlea the physical signal is first converted into the neural code
● When this code arrives into the brain, different parts of PAC process
different frequency ranges - tonotopy
● Different spectral features are quickly mapped onto phonemes
● Children have to lean phonemes of their language to be able to repeat
that they hear and learn to speak
