24/10/2022

SPEECH PRODUCTION
• Competing goals: fluency vs speed • Famous quotes from British Formula 1
commentator Murray Walker:
• System needs to be flexible to act
appropriately in situations where fluency is • Tambay’s hopes, which were nil before, are
vital, or speed is vital absolutely zero now
• Self-monitoring ability • There’s nothing wrong with the car except that
it’s on fire
• Preformulated phrases: (e.g., ‘_____ 2: Electric • With half the race gone, there’s still half the race
Boogaloo’) to go
• Underspecification (e.g., ‘you know, stuff like • The battle is well and truly on if it wasn’t on
that’) before, and it certainly was
• Schumacher has made his final stop three times!
• Anything can happen in Grand Prix racing, and it
usually does

SLIPS OF THE TONGUE

• Slips of the tongue • Shifts – “…he was d unk in prublic…”

• Occur at phoneme, morpheme or word
level
• Errors mostly occur within the same level • Perseveration “…I will come to the cym with you…”
• Similar positions of exchanged material in
utterances
• Switched speech segments have similar
properties (e.g. consonants, prosody)
• Errors based on phonological, not
semantic similarity
• Exchange “…would you like tea in your milk?”
• Anticipation “…twitch on the television….”
• Deletion “…I have to __verse the car…” (re
deleted)
• Addition “…the number of options are infinite…” (in
added)
• Substitution “…let’s play some TV…” (watch
substituted by play)
• Blend “…my gooshness…” (blend of goodness and
gosh)

ACCOUNTS OF SPEECH PRODUCTION

Stage 1 – conceptualisation, determination of stress patterns and determination of syntactical structure

• Serial account The students prepared for the exam

_______ _______ _______ _______ _______ _______
Stage 2 – content words and free morphemes added

_______ _______ _______ _______ _______ _______

Stage 3 – bound morphemes added

_______ _______ _______ _______ _______ _______

Stage 4 – function words added

_______ _______ _______ _______ _______ _______

Stage 5 – overt articulation

From Robinson-Riegler & Robinson-Riegler (2012)

1
 24/10/2022

SPEECH PRODUCTION

• Self-monitoring
• Overseeing what we say
• Correcting when we detect errors online – self repair
• Levelt (1983) perceptual-loop theory
• Detect error and pause speech Meaning

• Editing expression…..e.g. “umm, err….”

Language Language
• Correct error…. “I mean…..” comprehension production
Inner loop Inner speech

Audition Articulation
Outer loop
Outer speech

From Robinson-Riegler & Robinson-Riegler (2012)

GLOSSARY

• Orthography • Morpheme • Lexicon

• Phonology • Cohort model • Aphasias (types)
• Grammar • Phonotactics • Slips of the tongue (and
types)
• Phone • Metrical segmentation
• Stages of speech
• Phoneme • Transitional probability
production
• Coarticulation • Mondegreens
• Serial/parallel processing
• Phonemic restoration • Word frequency of speech planning
effect
• Polysemy

WK13: COGNITIVE
BIASES AND REASONING

2
 24/10/2022

THIS LECTURE WILL BE DIFFERENT!

• Flux.qa polls! (bear with me…)
• Log in at https://flux.qa/NFJDXE
• (Read Chapters 13-14 afterwards!)
• QI-style Klaxons for the answers people
commonly get wrong…like this:

YOU’LL SEE
SOMETHING LIKE
THIS IF YOU GET
IT WRONG

LEARNING OUTCOMES
• Define judgement, decision-making, and reasoning
• Identify different cognitive biases, e.g.,
• The availability heuristic
• Representativeness heuristic
• Anchoring-and-adjustment heuristic
• Confirmation bias
• Name and describe models that attempt to explain
our decision-making and judgement processes

3
 24/10/2022

DECISIONS, JUDGEMENT,
PROBLEM-SOLVING ETC
• Homo economicus?
• Judgement: deciding on the likelihood of
various events using incomplete information
– how accurate is it?
• Decision-making: selecting an option – did
we make the right decision?

ZOOM POLL #1: MONET VS HEDGEHOG

HEDGEHOGS VS MONET
• If you acted like the average person in previous
research (Wilson et al., 1993), you’ll select the
hedgehog if you go on your gut feeling, and the
Monet if you think about it.
• But when people with Monets or hedgehogs on
their walls in 6 months are asked how much they
like it, people with hedgehogs are happier.
• Gut feelings can be a better guide to behaviour
than deliberate thinking!

4
 24/10/2022

POLL #2: THE BAT-AND-BALL PROBLEM

• A cricket bat and a cricket ball together cost
$220. The bat costs $200 more than the ball.
How much does the ball cost?
• If it takes 5 machines 5 minutes to make 5
widgets, how long would it take 100
machines to make 100 widgets?

EXPLAINING THE BAT AND BALL

• $210 + $10 = $220, and $210 - $10 = $200
• Frederick (2005): on average, naïve participants at
American universities asked questions like this got them
right 41% of the time
• …when people are asked hard questions, they often lazily
try to answer an easier question:
• A cricket bat and a cricket ball together cost $220. The bat costs
$200 more than the ball. How much does the ball cost?
VS
• A cricket bat and a cricket ball together cost $220. The bat costs
$200. How much does the ball cost?

KAHNEMAN & TVERSKY

• Nobel Prize in Economics
• There are cognitive illusions just like
perceptual illusions, and for the same
reasons: tradeoffs in efficiency vs accuracy
• Dual Process Theory (Kahneman, 2003)
• System 1 (focused on efficiency) vs System 2
(focused on accuracy)

5
 24/10/2022

SYSTEM 1 VS SYSTEM 2
1. Focuses on efficiency i. Focuses on accuracy
2. Fast, automatic, ii. Slow, deliberate,
effortful
unconscious, effortless
iii. Feels like ‘making a
3. Feels like ‘intuition’ choice’, like
concentration
4. Actually does most of
the work iv. Thinks it does most of
the work, but is lazy
5. Good mood v. Bad mood

POLL #3
• You won a prize on a TV game show, and the
host gives you two options. Which would
you pick?
• Answer 1: $800
• Answer 2: An 80% chance of winning $1000
and a 20% chance of winning $100.

RISK AVERSION
• If in this one you chose the $800, you’re risk
averse; on average, Answer 2 was the better
option (if you did made that choice over and
over again, you’d make $820 per choice)
• Humans tend to be risk-averse in this way
(this is why there’s insurance companies)

6
 24/10/2022

POLL #4
• You have two options:
• a) I'm going to toss a coin 5 times. For each heads,
you win $200, for each tails, you lose $100
• b) I could just give you $50.
• Which would you choose?

LOSS AVERSION
• If you toss a coin five
times, you could win
$1000, you could lose
$500
• On average, you’ll win
$100 per toss – but plenty
of you still chose $50
• Loss aversion – how we
are not homo economicus

POLL #5
• You meet a somewhat shy woman who wears
glasses and has an eye for detail. Which of
these professions is she most likely to have?
• Librarian
• Anaestheologist
• Retail
• Hairdresser

7
 24/10/2022

REPRESENTATIVENESS HEURISTIC
• There are 7000 librarians in
Australia, and 1.2 million retail
workers. She’s way more likely to
be a retail worker, even if she
does wear glasses.
• But she’s more representative of
our mental image of librarians
than our mental image of retail
workers or farmers
• We confuse which profession does
she best fit the stereotype? for
which profession is she most likely
to be?
http://www.flickr.com/photos/webchicken/2519590928/sizes/l/in/photostream/

POLL #6
• There are 328 million people in the USA. In a
relatively close 2-way election, about how many
votes will Joe Biden end up with once they finish
counting, according to pollsters? (If they ever finish
counting, that is...)
a. 80 million
b. 100 million
c. 120 million
d. 140 million
e. 160 million

THE ANCHORING EFFECT

• Because I put in the overall
population of the USA, it may have
anchored your estimates (e.g., ‘if half
of the population voted for him, it’s
probably about 160 million?’)
• This is the anchoring effect - giving
you such numbers affects your
estimates of likelihood by giving you
something to compare it to
• It also works for totally unrelated
numbers (the anchoring-and-
adjustment heuristic mentioned in
the textbook)

http://www.flickr.com/photos/8829172@N02/2763895688/

8
 24/10/2022

POLL #7
• Which word most accurately reflects how
you felt at the start of your first PSYC236
lecture with me a few weeks back?
1. Calm
2. Scared/Anxious
3. Happy
4. Bored/Sad

POLL #8
• Which cities have the larger population?
• Wuhan, China vs Chongqing, China
• Nowra-Bomaderry, NSW vs Alice Springs, NT
• Wollongong, NSW vs Hobart, TAS

THE AVAILABILITY HEURISTIC

• The rule of thumb that the frequencies of
events can be estimated by the subjective
ease with which they can be retrieved
• E.g., “I’ve heard of that one, it must be
bigger”
• Or “This is how I’m feeling now, it’s probably
how I was feeling then”

9
 24/10/2022

THE 2-4-6 TASK (WASON, 1960)

• ‘Subjects were told that three numbers – 2, 4, 6 – confrmed to
a simple relational rule, and that their task was to discover it by
making up successive sets of three numbers…” and then
hearing whether it fit the rule or not
• That rule was ‘three numbers in increasing order of magnitude’
• Examples that fit the rule: ‘1, 2, 3’, ‘5, 10, 15’ or ’99, 109, 119’
• Examples that don’t fit the rule: ‘1, 2, 4’, ‘4, 2, 0’, ‘1, 4, 9’
• Only 6 of 29 undergrads got the rule right on their first
try…why?
• Confirmation bias

CONFIRMATION BIAS
• We are biased towards attempting to
confirm our hypotheses, rather than trying to
disconfirm our hypotheses
• E.g., in the 2-4-6 task, there’s a bunch of
other possible rules that 2-4-6 could fit into
• People had a hypothesis about what rule was
right: but they tried to prove it right rather
than try to prove it wrong

POPPER AND CONFIRMATION BIAS

• The philosopher of science Karl Popper (see
me in PSYC325 for more!) thought that
science was about trying to falsify or
disconfirm hypotheses, rather than trying to
confirm them
• …whether that’s true is complicated, but it’s
certainly useful for scientists to try to
disprove theories sometimes

10
 24/10/2022

PROBLEMS WITH KAHNEMAN’S DUAL

PROCESS THEORY
• Obviously explains some of these effects quite
well! But…
• Likely oversimplified:
• System 1 and System 2 are a bit vague and hard to
define
• System 1 as rapid and System 2 as slow isn’t borne out
by evidence – depends on the task
• Kahneman originally argued that System 1 was parallel
and System 2 was serial – not the case
• In general the system is more flexible than that

DE NEYS’ LOGICAL INTUITION MODEL

• Three processes involved in judgement:
• Rapid intuititive heuristic processing (System 1)
• Intuitive logical processing
• If heuristic and logical processing conflict,
then System 2 gets involved

WHY THIS MATTERS

• The Challenger disaster (1987)
• Engineers at the company that made the O-
Ring raised concerns to NASA about it being
brittle in cold temperatures
• NASA engineers discounted those concerns
and encouraged them to rethink, so they did
• People up the chain who okayed the flight
never knew there were concerns

11
 24/10/2022

FURTHER READINGS
• Daniel Kahneman, Thinking Fast and Slow
• Daniel Gilbert, Stumbling Into Happiness

MUSIC
PSYC236 Week 13 lecture b
No associated readings!

LEARNING OUTCOMES

• Understand sound waves and how they’re turned into neural signals
• Identify different aspects of music processing, e.g.,
• Relative pitch
• Musical expectancies
• Melodic contour
• Dynamic attending
• Name and describe the components of the ITPRA model

12
 24/10/2022

WAVEFORMS: A SINE WAVE AT 440HZ

SONOGRAM:
SINE WAVE FROM 440-880HZ

WAVEFORMS: C MAJOR 7 ON GUITAR

13
 24/10/2022

THE AUDITORY SYSTEM - THE EAR

• The outer ear: the shape subtly changes the
sounds that enter your ear depending on
where they come from.
• The tympanic membrane: effectively
converts sound waves into vibrations

• The malleus, incus and stapes: they amplify

the sound (and are evolved from bones that
once made reptile jaws very flexible)

• The cochlea’s ‘hair cells’ are

triggered by soundwaves
THE AUDITORY SYSTEM THE • Different pitches trigger ‘hair
COCHLEA cells’ in different parts of the
cochlea – the longer the wave,
the longer it gets down the coil
• The cochlea doesn’t always
perfectly capture sound
• Lower pitches • Usually, we’re happy to interpret
hit hair cells two notes sounding together as a
chord (e.g., that guitar you heard
further down earlier)
the cochlea • However, if the pitches of the
• Triggering hair notes are too close together, we
experience beating
cells activate
neural signals.

• There are mathematical/physical

relationships between two notes
CONSONANCE AND DISSONANCE together sounding ‘consonant’.
• Octave = e.g., 440Hz (A) +
880Hz (A)
• Perfect fifth = e.g., 440Hz (A) +
660Hz (E)
• In natural sounds, these pitches
often occur simultaneously, and
our perceptual systems quickly
group them (…)

14
 24/10/2022

MUSIC

TIME

• In some ways, the most basic thing in music is time

• If you pause a movie, you can still see the picture, but the sound stops.
• The pulse is the simplest organisation of time: it divides time into equal sized
chunks – pulses - at a particular speed
• More complicated than that and we get rhythms (which often repeat)
• This is useful for processing – it means we know when notes are likely to
happen

RELATIVE PITCH
• The song remains the same whether it starts on A or F#, so long as the
relationships between notes stays the same
• This is relative pitch because it’s about the mathematical relationships between the
pitches rather than the pitches themselves

15
 24/10/2022

RELATIVE PITCH

• E.g., it doesn’t matter what key this line from ‘Something’ is sung in:
• Phyllis Dillon (C# major)
• The Beatles (C major)
• Martha Reeves & The Vandellas (Eb major)
• Isaac Hayes (G major)
• Relative pitch is in many ways the fundamental basis of music perception and
cognition
• Music is abstract and recursive
• Music making us feel stuff (etc) is (mostly) about the relationship between the notes rather
than the notes themselves

ABSOLUTE PITCH
• ‘Absolute pitch’ or ‘perfect pitch’ generally • Levitin (1994) got unsuspecting students to
refers to musicians who can name musical sing the opening line of their favourite song
notes when they hear them
• Most were actually more or less singing the
• E.g., “that’s an A” right notes!
• However, while we might not be able to
name them, we have a pretty good sense of
pitch.

MELODIC CONTOUR
• A big part of what makes one melody sound different to another is the pattern of
‘ups and downs’
• Non-musicians can tell the difference between transposed melodies based purely
on the relationships inherent in ups and downs (Dowling, 1978)

16
 24/10/2022

MELODIC CONTOUR
• Infants recognise contour changes (Trehub
et al., 1997)
• Melodies with fewer contour changes
perceived as simpler (Boltz & Jones, 1986)
• We more easily notice notes when the up
changes to down (Dyson & Watkins, 1984)
• Essentially, contour is the basic
psychological framework of melody

• Q: So your brain has processed

the sound, figured out the
MUSICAL EXPECTANCIES contour and the key and
tempo…okay, how do we get
7.00 from there to smiling?
• A: When we process music, we
6.00
implicitly compare what we hear
5.00 to what we’ve heard in the past
4.00 • As a result, we expect certain
things to happen (at certain
3.00 times)
2.00

1.00

0.00
C C# D D# E F F# G G# A A# B

• This is the case whether you

know how to name what you’re
MUSICAL EXPECTANCIES hearing – it’s implicit
• Dminor9 G13 Cmajor9
7.00
• You might not know that the first
6.00 two chords in that sequence sets
up the expectation of the third
5.00
(in jazz, especially)
4.00
• But people appear to be
3.00 influenced by the statistical
properties of music, and are
2.00 surprised if it doesn’t happen!
1.00

0.00
C C# D D# E F F# G G# A A# B

17
 24/10/2022

SCHEMATIC VS VERIDICAL EXPECTATIONS

• Schematic Expectations – we have expectations about a song/piece based on

other music
• Veridical Expectations – we have expectations about a song based on the fact
that you’ve heard ‘Call Me Maybe’ before
• Even when we know every note in a song, it still triggers our schematic
expectations

TILLMAN ET AL. (2006)

• Musically untrained Ps heard a series of

MUSIC
chords Bright Dull
Reaction Time (msec)

• Ps had to judge if the last chord in the 800

series was played with a ‘bright’ or ‘dull’ 700
tone 600
500
• The last chord was either related (the tonic) 400
or less related (the subdominant) to the rest 300
of the series 200
100
• Ps took longer to judge the tone if the 0
chord was less related Related Less Related
Chord Type

TILLMAN ET AL. (2006)

• Regular people (or pop musicians) couldn’t
tell you what a subdominant or a tonic is Bright Dull
Reaction Time (msec)

• However that a subdominant or tonic can 800

affect reaction time in a priming task 700
suggests that regular people’s implicit 600
processes can discriminate between the two. 500
400
300
200
100
0
Related Less Related
Chord Type

18
 24/10/2022

TEMPO, PULSE & ENTRAINMENT

• ‘Dynamic attending’ • Large & Jones (1999) asked

MUSIC
• Attention is a limited resource
• If there is a beat, we know when best
to pay attention
people to judge if one tone
was the same length as the
others, or shorter or longer
• They found that if the tone
was in time with the others
(‘expected’), then people
made good judgements
• However, if the note was out
of time, people’s judgements
of length were influenced by
that.

ITPRA AND MUSIC

• David Huron’s (2006) ITPRA theory tries to explain musical emotion by interlocking
responses to these musical expectations
An unconscious
What we expect physiological response to
to happen in the a surprising musical event
song given our
(statistical)
knowledge of This integrates the
music in general reaction response and
prediction responses into
The extent to emotional reaction
which the music
diverges from our Did what we
expectations (unconsciously) predicted
causes a tension in the music actually
response happen?

WHY DOES THE CHORUS OF TINY DANCER BY

ELTON’JOHN DO IT FOR ME?

The chorus surprisingly resolves

Most pop songs some chordal tension (it goes
back to the chords you’d expect
have a big chorus in the key)
that happens
eventually, with
chords that do You weren’t sure when the
what you expect resolution would happen,
but knew that it would, so
that resolution feels good
It takes 2:30 in the
song before you
hear the chorus and
The tension in that section
right before the
before the chorus builds up
chorus does a big
a lot, so we’re expecting it
tension buildup
to resolve

19
 24/10/2022

THANKS FOR LISTENING!

20