Week 7 Flashcards
Which animal is most intelligent?
Humans (most common answer; technology, language complexity, human niche modification)
If we restrict it to non-human animal’s people tend to think of dolphins, apes, elephants that are social animals, mammals, large brains and most similar to ourselves.
Some people may have thought of birds which are more divergent to humans than mammals?
Others included were bees and gecko’s.
Why do we have these biases when we think of TOM in non-human animals?
Why do we think that its the animals which are most like us that will be the most intelligent?
Great Greek philosophers are the first to introduce this idea: in the great chain of thinking (i.e., the scala naturae by Aristotle)
> Provides a way to organize the natural
world in an organized way, linear.
> Man was naturally placed on the top and
the following went in order of how similar
they were to man: mammals, cetaceans,
(reptiles, birds, amphibians and fish),
cephalopods, crustaceans, other
arthropods, other molasses, ascidians,
holothurians, sponge, plants and inanimate
matter.
> This idea that man was more complex and
above all other life is an ideal adopted by
the church; the only thing above man is
God and the angles.
When thinking about animal intelligence from an evolutionary lens:
> Darwin was the first to propose that “the
difference in mind between man and the
higher animals, great as it is, certainly is
one of degree and not of kind”
The mind is subject to the forces of natural
selection. We should share fundamental
cognitive traits with many species to whom
we are closely related to.
What is intelligence?
> How it is defined and measured
influences what our findings are and what
animals will appear intelligent.
> As a construct it is not useful to us if it is
not clearly defined.
(A) Clark’s Nutcracker:
Using a measure of the ability to recall
specific locations, which
is the more “intelligent” animal?
○ Bird species that hoard seeds for the
winter and has very accurate recall of
where they are located/hidden.
○ If we compared it to humans to test our
recall for 10’s of thousands of animals we
would find that birds were more intelligent
than humans.
(B) Complex Behaviours are not always
“Intelligent”
○ Looking at intelligence by looking for
complex looking behaviour doesn’t always
mean intelligence (i.e., they could be fixed
action patterns; behaviour pattern
triggered by a releasing stimuli which
continues even if the trigger is removed).
(C) Clever Hans
o You may assume that animals behave
based on human-like intelligence when it is
not (e.g., Hans the clever hors who
travelled around completing mathematical
questions but the trick was he was
stamping and reading the actors body
language to know the right answer; he did
not actually have conceptual
understanding of mathematics).
“Intelligence”
Isn’t particularly useful as a concept for investigating the mental lives of non-human animals
Cognition
Cognition
A more operationalized definition of intelligence is “cognition”
Cognition:
o The mechanisms by which an animal
acquires, stores, processes and acts on
Information from their environment.
The Challenge:
o The biggest challenge to cognition is how
do we test something we cannot observe
directly? Mechanisms in animal minds
which are not directly observable.
Non-Human Cognition:
How can we test cognition in non-human animals? How can animals’ observable behaviors be used to test hypotheses for cognition?
To answer the question: Do animals share ‘uniquely human’ cognitive abilities? We need to build falsifiable hypotheses based on animal behavior.
Uniquely human cognition?
What does uniquely human cognition refer to?
What does uniquely human cognition refer to? Characteristics of human cognition that is argued to only be found In humans such as:
o Tool manufacture and tool use, (animals
crows, octopus etc. all use tools)
o Mental Time Travel, (future and past
thinking)
o Syntactical-grammatical language
o Consciousness,
o self-awareness,
o Imitation (observe and imitate behaviour
in other conspecifics; imitate goal and
actions to achieve goal)
o deception,
o Theory of Mind
Do animals have TOM when you can only use non-verbal based tests?
What is Theory of Mind?
Two Tasks
What is Theory of Mind?
> The ability to perceive others’ actions as
governed by internal states (desires,
beliefs, intentions, knowledge).
> Agents’ behaviour is goal directed and
driven by internal states. Its an
unconscious process we do in humans. Are
animals able to do the same?
False Belief Task
Sally Anne
● People watch the doll being placed in the
box when she leaves the room the doll is
moved (location change) then re-enters
the room. Where will she look for the
object?
o Pass: where she last saw it in the wrong
location
o Fail: where we know where it is
● Do we predict others behaviour based on
our own knowledge or can we appreciate
that others have own belief that can be
false (i.e., not reflect reality).
● Passed by 4-year-olds
● Failed by younger children (under 4)
● Failed by autistic children
● This task is language-based, and some
people argue the task is too demanding
for young children to answer it correctly.
Non-verbal versions have been developed
to assess this.
Non-verbal False Belief task
Anticipatory eye-tracking tasks
● Watch a video and we are testing where
they anticipate the actor to search for the
object before they do it.
● Tested 2 y.o. children
● Puppet moved a ball around then removed
it when actor is not looking
● 17/20 children look first at the correct
location
● Children look longer at correct locations
*This indicates this paradigm could be
applied to non-human animals
Is ToM uniquely human?
● Can TOM be attributed to apes?
● It’s still a current debate
● Penn and Povinelli are adamant that
they do not.
● Carli and Tomaselli argue they do.
● The first animal that people have tried to
attribute TOM to and is most common in
chimpanzees.ToM in chimpanzees
Tells us when TOM developed is it specific to humans or from a common ancestor.
Early chimpanzee failures…
Early chimpanzee failures…
Originally, they found that chimpanzees fail explicit FB tasks. Can humans see them when their vision is obstructed in ways that are not actually ecologically valid to what they experience in the real world. Therefore, cannot understand which keeper is able to see them and can give them food.
Some argue that this visual perception TOM test is not ecologically valid and too far from their behaviour repertoire to be able to reliably assess TOM in chimps.
Chimps TOM
Ecological validity
● Hare et al. (2001)
● Criticism of Hare et al. (2001)
● Hare et al. (2001)
● Study with two chimps competed with a
dominant conspecific. Can they
understand what another agent can see
● Subordinates could always see where
the experimenter hid the food and
whether the dominant chimp could see
it.
● Dominant chimp either new the location
of the food, didn’t know or was
misinformed.
● Subordinate always had a head start, so
shouldn’t have responded to dominant’s
choice
● Chimps less likely to approach food
seen by the dominant. They preferred
to approach food when the dominant
was unaware of its location.
● They argued this is evidence of chimps
being able to understand other chimps’
knowledge of the food location based
on their visual field and use this
information to guide their own
behaviour.
Criticism of Hare et al. (2001)
● A behavioural rule would produce the
same result:
● If another animal can see then they
must have knowledge of where it is
hidden then “If a dominant individual
orients to a piece of food in a particular
location, then that food must be
avoided”
More recent work: False belief in apes
● Krupenye et al. (2016)
More recent work: False belief in apes
● Krupenye et al. (2016)
● Anticipatory eye tracking study in apes
(non-verbal FB task).
● Their results indicate that the most first
look locations were to the target
location (FB) and there were no species
differences in their performance =
implicit TOM rather than explicit TOM.
Q: Are you convinced that apes attribute
false beliefs?
● How do you test whether they
understood the content of the videos?
● What tests would provide us the
information to confirm this?
● Is anticipatory eye looking a reliable
measure?
● Yes:
● Explanation
● In the previous videos, the subjects watched an actor interacting with a ‘naughty ape’. In the scenarios, the actor had to guess the location that the naughty ape was hiding, or the location that the naughty ape hid the stone. You can see the locations that subjects watching the videos were looking based on the red circles in the clips. These scenarios were designed so that at the end of the clip there was always a ‘correct location’, which was where the actor should look (based on a false belief), as well as a ‘distractor location’. In watching the videos without any explanation, how hard did you find it to tell what was happening in them? Do you think that you would have looked at the ‘correct location’ in this task?
A spanner in the works…
● Rachel (lecturer) said she was on the
fence and says there is one important
flaw to consider in non-verbal FB tasks.
● The study is consistent with infants’
performance on non-verbal FB tasks
but there is a replication crisis where
they are not able to replicate the
original infant study. Is it measuring
what we think it does? unlikely.
● We do not yet have the tools or the
paradigm to answer the question
chimps have TOM robustly.
False belief in apes…?
Non-verbal FB task in apes suggests they have implicit theory of mind and can anticipate actors search behaviours but is still a struggle to replicate
Can we look at other aspects or maybe less sophisticated components of TOM in non-human animals
A good place to start is to look at a lower-level TOM traits by looking at desire state attribution.
Children initially use desire-states to understand behaviour in others
Desires and beliefs are integrated later in development
Do other animals also use this precursor to adult ToM?
Desire state attribution in Eurasian jays
Desire state attribution in Eurasian jays
● We haven’t shared a common ancestor
with for millions of years.
● Their cooperative food sharing
behaviours provides a good platform
for examining TOM in avian species.
● Can males’ jays attribute desire to their
female mates?
● In the image the male passes food to
the females bills during pair bonding.
Do they give females food they like or what the females like using satiation techniques?
Manipulating desires:
Ostojić, Shaw, Cheke, Clayton 2013 PNAS
● Specific satiation: satiation for a specific
food and not all.
● i.e., always have room for desert after a
big meal but are too full for more
savoury food.
● In animals, we can satiate females via
refeeding them wax worms they would
prefer meal worms and vice versa. If
males and females are prefeed different
food they develop different desires. Will
males feed their female food they want
our based on their own desires.
Results:
• Indicate males reduce the proportion of
wax worms he shares with females to
baseline.
• After being prefeed on meal worms he
increases the proportion of wax worm
shared with her relative to baseline. They
are responsive to the food preference of
females (i.e., specific satiation).
How are they doing this?
● Is able to understand the preference of
females food desires. Reasoning via
analogy of their own specific satiation
on what food would be prefeed.
● The male could merely be reading the
behaviour of their mate to pick up on
what food they want. External behaviour
cues rather than an understanding
internal states.
How do we test this?
● We reran the same test but changed
one condition where the male couldn’t
see her prefeeding and had no
information on what she ate earlier. He
only has information about her response
now to food sharing. We see the effects
are not present to suggest they were
responding to the food preference of
the females (i.e., no pattern in males
food sharing choices; it’s much more
random).
What does this tell us?
● It intriguing that they only adjust their
sharing behaviours when they have
their own knowledge on the females
prefeeding experience.
● Males do not rely on the behaviour of
the sharing phase to figure out what to
feed her.
● They are inferring based on their
previous food eating on what she is
most likely to currently want. If this is the
case, this is consistent with desire state
attribution.
*understanding desires guide others actions
and they may differ from our own was
believed to be uniquely human but may be
seen in jays. First evidence of this in non- human animals.
