Module 3 Flashcards
Infants’ responses suggestive of higher-level cognitive understandings.
But how much of that understanding is psychological understanding?
§ That infants understand the self and properties of
the self?
§ That infants understand that others act purposefully;
that agents take certain steps to achieve certain
goals, and context can gives clues to what actions
are relevant and irrelevant to those agents’ goals?
To answer this question we will look at …
To answer this question we will look at evidence that psychological understanding is available early in life, well before language gives it a boost.
Earlier findings we know of in the mobile task which demonstrates that infants can track causal contingencies in their environment. The link is still between the self-movement of their physical bodies and a visually rewarding object. Meaning in its current form it doesn’t tell us about whether infants discriminate what pertains to their own bodies, and what pertains to non-self-entities (i.e., that the self has unique characteristics).
Are there forms of evidence that infants have self-other awareness? At least 3.
(A) Own-Body Perception in Infancy
Bahrick & Watson (1985)
Steps: § 3 month and 5-month-old infants were seated on a
chair, and they were shown the camera output of two
videos; 1) showed a video of their feet moving in real
time(contingency/synchrony); 2) showed a video of
their feet moving with delayed reactions (imperfect
contingency). Which video will infants prefer to look
at?
Results: § Longer looking to non-contingent display of self-
amongst 5-month-olds (delayed unusual video
feedback), as compared to 3-month-olds (younger
prefer to watch contingent display, matches).
Implications: § Self-perception of body can be seen in the way
infants detect a link between proprioceptive
information for motion and visual display of that
motion § Younger infants prefer stability from cross-modal
information to perceive the world as a consistent and
stable place. Older infants slowly become more and
more interested in novelty. § Since both groups are able to discriminate between
the two visual images (show a preference) this tells us
they are able to make judgements of the self’s
properties which they can then relate to the visual
input in their environment.
- self-body perception is available within the first year of
life.
(B) Mirror Self-Recognition
Recognition of contingency between own action and
video feedback (self-image reactions) suggests that
infants should also be capable of mirror self-
recognition
Humans
Amsterdam (1972)
The Mark or Rouge Test
Q: Does the child, whilst looking in mirror, notices and
touches red mark on face/nose?
Steps:
§ Researchers put a secret red mark on the infants face
and then they’re placed in front of a mirror to see if
they will notice the mark, realize that it is their
reflection, and explicitly try to rub the mark off their
own face.
§ To pass the mirror test, infants need to touch their
nose where the red mark was placed to indicate that
they have EXPLICIT self-recognition (requires an
objective direct response).
Results:
§ There is a robust age-related pattern in how they
respond to the mirror and the mark on their face:
- 3-5-mths: look at mirror and treat image as playmate
(curious about what they see in the mirror, but they
still treat their reflection as another person; don’t
touch the mark).
- 18-20-mths: 42% touch mark on own face (less than
half touch their face)
- 21-24-mths: 63% touch mark on own face (more than
half will reliably and consistently touch their face;
explicit mirror self-recognition is in place by 2 years of
age).
- Has been replicated with all the great apes which all
seem to pass the test.
Non-Human Animals
Non-human animals in complex social groupings (apes, dolphins, certain birds like magpies, elephants) show some mirror recognition.
Plotnik et al. (2006) on Asian Elephants
- Elephants, happy, patty and Maxine.
Do they show:
(A) Physical inspection of the mirror? Yes, they all
touched it and checked behind it using their trunks.
(B) Repetitive mirror testing behaviour? Yes, all made
non-species typical trunk and body movements to
see if their reflection moved in the same way.
(C) Self-directed mirror behaviour? No, only happy
passed the mark test, tried to rub off the white x
painted on their face. She only did this when she was
in front of the mirror and not anywhere else.
Showing it is a very specific behaviour she is
demonstrating.
Implications:
§ Only 1/3 elephants passed the mark test indicating
there is some degree of self-awareness present.
§ It is challenging to interpret negative effects (2/3 not
passing the test) because absence of evidence
doesn’t mean evidence of absence (the use of
grooming behaviours of the trunk as the behavioural
approach to passing the test but in real life they really
do this behaviour, they’re more likely to throw dust or
debris on themselves with their trunk) this behavioural
measure may have underestimated elephants self-
awareness. Other authors which use a more naturalist
mode of behaviour for the specific non-human group
they work with found significantly more success,
cleaner fish are able to pass the rouge test. This
indicates that self-awareness is not zero sum all (have
or don’t have; not a discontinuous cognitive
accomplishment only in humans) it is more gradual
and continuous in nature where different groups of
species have different degrees of self-awareness.
(C) Detecting adult contingent behaviours
Infants as young as three months of age are sensitive to
and can discriminate when adults are making
contingent actions upon, rather than ignorant of, infants
themselves. Infants seem to realise when adults
treating them as social partners and prefer it when
adults do this.
Striano et al. (2005)
3-month-olds
Steps
(A) Normal interaction (visit 1 control)
§ Mother and infants arrived at the lab and the
researcher’s recorded audio of their normal
interactions.
§ One-week later infants and their mothers came back
and were randomly assigned to the non-contingent or
imitation contingent group.
(B) Non-contingent (visit 2 interaction)
§ Parents wore a pair of headphones and were
instructed to press play on the audio recorder and to
repeat the audio of what they had said to their infant
in visit 1. The researcher insisted that the parent
repeats what they said in the first interaction back to
the infant (interaction between parent and infant non-
contingent).
(C) Imitation contingent (visit 2 interaction)
§ Parents were instructed to perfectly mirror infants’
facial gesture, arm movements and vocalisations (if
the infant smiled, cooed etc.). Would 3 months olds
show differential reactions between the two groups
(discriminate)?
Results:
§ 3-month-olds gazed more intently (sustained gazing;
more dilation) in imitation interaction condition. Infants
are able to pick up in conversations if they’re being
treated as social agents, being responded to and
mimicked and prefer this (self-other awareness;
harmony in what they do and what is reflected back to
them = contingent, my action causes your action).
These three stands of evidence encourage us to consider what mechanisms might support social psychological understanding of others?
(A) & (B)
(A) Contingency detection
§ Being sensitive to situations in which one’s
behaviour is followed in time by another stimulus
event (timing of events; immediate vs delayed) or
being sensitive to how much energy someone puts
into an instrumental act and the consequences
obtained could be important.
§ Preference for or sensitivity to certain contingencies
can allow infants to realise that their own behaviours
can exert control over others’ behaviours (harmony
in what they are doing and what they see is
reflected back to them; by changing what they do it
influences what is reflected back to them; my action
causes your action, your action causes my action).
(B) Imitation
§ Mapping actions of other people onto actions of
one’s own bodies (like me hypothesis that my
behaviour is goal orientated therefore other actions
must also be purposeful; self-other psychology
awareness) is also important.
§ Imitation can give insight into what it might internally
be like to those actions, and in so doing give a
glimpse into the minds of others.
*both mechanisms ensure the maintenance of proximity
with caregivers and facilitate attachment bonds (infant-
parent; ensure they learn the required skills to reach
sexual maturity and be functional social beings able to
be harmonious with other beings in their
social or cultural group).
Q: when we (infants and adults) observe others, we assume that other actions are directed to outcomes (that the actor is acting to achieve CERTAIN goals). But so far, the evidence is suggestive.
- Yes, infants can imitate another person’s action upon
an object even when the action is not seen (i.e.,
completed) but how can we be sure that infant’s re-
enactment to ‘pull apart toy with hands’ because it
done because it is viewed as the most effective way
to achieve the goal? - We only have evidence of the first part of
teleological framework; purposeful-action not
evidence of efficiency.
Are there imitation studies that can provide more direct evidence that infants only re-enact an action if they perceive it to be the most effective means to achieve a goal?
Direct evidence (or stronger evidence) using imitation task that infants view actions of others as goal-directed (infants regard other actions in a more careful way as being goal oriented).
Light-box study: Gergely et al. (2002)
Light-box study: Gergely et al. (2002)
14-month-olds
Steps
§ Infants were randomly assigned to one of two
conditions: 1) Hands Free condition: There is a light
box and lamp that can be activates on the table in
front of the infant. The experimenter comes in and
sits in front of them with a cloth around their
shoulder. They place their hands on the table and
leans forward to turn the light box on with their
forehead (unusual action being modelled). 2) hands
occupied: the experimenter comes in complaining
that they’re cold, wraps themselves in a blanket with
the hands occupied and turns the lightbox on with
their forehead.
§ To what extent will 14-month-old infants imitate this
unusual behaviour? Will they regard the action
(turning light panel on with forehead) as having a
different meaning in the two conditions?
Results:
§ In hands occupied condition, majority of infants in
this condition turn the light on with their hand
instead of imitating the behaviour (using their
forehead).
§ In the hands-free condition, majority of the infants
imitated that unusual behaviour (turn light on with
forehead).
Implications:
§ Fit with teleological reasoning:
§ Infants assume agents choose the most efficient
means to ends.
§ In hands-occupied condition, agent’s hands hold
blanket and the goal is to turn on light, with
whatever means available to her ( infants pick up on
the goal and perceive the actor to be rational under
constraint and when imitating see no need to copy
turning light on specifically with head because
they’re not constrained and can complete the action
in the most efficient way possible- with their hand).
§ In hands-free condition, agent’s hands are free so
goal must be to turn light on specifically using head
– so copy by turning on light with head (if there
hands are free and the actor is choosing to use the
inefficient method so they assume there must be a
reason for why they use their head so they imitate
it).
§ Infants think about someone’s goal transforms the
way infants perceive their bodily motions as being
purposeful behaviour.
The Debate: Lightbox Study
§ Interpretation 1: perceptual saliency and not
teleological perspective
- The blanket completely covered upper body, and is
eye-catching, and this feature may be like a
perceptual distractor which competes with the
head-touch action, drawing the 14-mth-olds’ infant’s
attention away from it (there is a perceptual
difference between the two conditions; the hands
occupied condition having the blanket cover the
upper part of the body is very eye-catching, is
salient enough to cause infants to be distracted and
are not focusing on their forehead movement
relative to when their hands were on the table in the
hands free condition; blanket less salient means
they attend more to the unusual action and thus
more likely to imitate the behaviour).
§ Interpretation 2:
- Differential imitation of same action because of
changes depending on the context which provides
strong evidence that infants understand that others’
actions are goal-directed (teleological view of other
people’s social actions).
Solution
Solutions to the Debate: Lightbox Study
Solution
To rule out the alternative explanation that the effects in the light box study are due to perceptual saliency.
Consider Bisert et al.’s (2012) test of perceptual distraction as shown in picture C (their ‘Hands-free with extra distraction’ condition). They added an additional condition, a “hands free with extra distraction condition” where at the beginning of the experimental phase, the experimenter put a blanket loosely around her shoulders, and then put two smiling faces to left and right of table.
Results:
• Hands occupied: most infants used their hands to
turn on lamp.
• Hands free: most infants imitated the unusual action
to turn the lamp on with their head.
• Hands free with distraction: They found that 50%
imitate and 50% do not imitate the head touch action
(chance levels). Being able to demolish the effect by
adding a distractor in the hands-free condition,
according to Bisert, supports that the original
findings of the lightbox study were due to perceptual
salience being a distractor from imitating social
behaviour (its relatively dumb and not a
sophisticated mental inference about teleological
goal and efficiency).
Alternative Explanation:
§ Does the light-box study show a cognitive effect or a
motor effect?
Remember:
- Hands occupied condition: agent’s arms folded
across chest
- Hands free condition: agent puts hands on table
when performing head touch
Consider this:
- Notice that when infants perform a head touch
themselves, they put their hands on the table next to
the light-box – to maintain a stable position. This
action motorically matches closely the movement of
adult in the hands free but not in the hands occupied
condition (infants imitate the action sequence in a
motor sense rather than being evidence of
sophisticated cognitive reasoning where actors are
goal orientated and efficient, they are actually
focusing on the motor components. Monkey see
monkey do. I will imitate the behaviour as long as it
is within my physical capabilities).
- = motor driving the effect matching visual of others
with own motor capabilities (hands-occupied is too
difficult, being able to lean forward and turn light on
head is not possible in their current motor stage
without hands on the table to stabilize them).
Given challenges with light-box method, how else can we study that infants show differential imitation of same action when context changes (how infants decide what to imitate and what not to imitate)?
- Yes, use method by Carpenter, Call, & Tomasello
(2005) : Mouse in the House
Carpenter, Call, & Tomasello (2005):
Steps:
1. 12–18-month-old infants were assigned to one of two
conditions.
2. In House condition, Infants were placed in front of a
table with two toy houses and a mouse on the table.
The goal is to put the mouse in the house but the
experimenter made movements (hopping or sliding
motion with sound affects) into the house.
3. In No-House condition, there was no house on the
table. The experimenter would make the movements
and sound affects as the mouse moved around the
table but there was no house at the final location.
4. Did infants (12-18-mths) copy the hopping or sliding
action between the conditions? Is it relevant in some
conditions and irrelevant in others?
Results:
In House condition, infants imitate by copying the
final goal (putting straight into house) and did not
copy the hopping or sliding.
In No-House condition, infants copied the motion
and movements (and even sound effects); these
actions themselves treated as goal.
Infants’ action processing is flexible. They notice the
context behaviour is modelled in and use contextual
cues to determine what aspects of the action stream
are relevant to imitate and what is irrelevant to the
actor’s goal. For example, in the no house condition
infants assume since there is no end goal location
the experimenter has purposefully done the motion
and sound affects for a reason or relevant to the
task. In the other goal, to get into the house is the
end goal and the motion doesn’t matter so they do
not imitate it = flexible.
Studying infants’ responses to accidental actions is also a useful way of investigating flexible understanding of goal-directed actions (how flexible are infants ability to pick up on goal directed actions)
Carpenter et al. (1998):
Carpenter et al. (1998):
Do infants (14-18-mths) discriminate between Purposeful (relevant to goal) and Accidental (irrelevant to goal) actions when reproducing others’ actions.
Steps
Watched adults produce 2-step actions that produce
a result; sometimes an action is intentional “There!”,
and some vocally marked as accidental “Whoops!”
Infants watched adults interact with a blue wooden
box with a handle to get a toy. The trick is to pull the
handle to get the toy and the wheel is irrelevant to
their goal. They watched adults complete a 2-step
sequence which was either rational (pull handle
“there” spoken) and irrational (touch wheel
“whoops” spoken to indicate its irrelevant to goal).
Results:
Infants more likely to copy purposeful rather than
accidental actions. Shows infants screen out
meaningless actions (discard information which is
irrelevant to the current goal).
Summary:
1.
Propensity for social interaction may be partly very
early-developing (part innate basis).
Infants, for example, notice contingencies between
their own bodies and the environment, and will soon
go on to explicitly pass the mirror test of self-
recognition. If infants understand that their actions
are directed to outcomes/goals, and if they treat
others as “like me” and worthwhile imitating, then –
by analogy – should also understand that others’
actions are directed to outcomes/goals.
Evidence for goal-understanding is diverse but also
challenging to interpret. Nonetheless, scientists are
super interested because infants’ capacity to
understanding goals is a significant marker of infants’
insight into the psychologies, the minds, of human
agents.
Recap:
14–18-month-old infants are able to exploit context to identify which action are most relevant to them in an action stream to imitate. These tasks rely on some degree of motor skills to be able to complete the task. Is there a technique which doesn’t rely on infants having certain motor skills and will allow us to see if the effect applies to younger infants (teleologically; goal directed).
Technique:
Woodward (1998)
Technique:
Visual Habituation as a measure of young infants’ perception of actions as goal directed.
Woodward (1998)
- Habituate
- A ball and bear are placed on the table in front of
the infant. There are then habituated to a repeated
action sequence (e.g., a hand reaching for the teddy
bear on the left side of the screen).
- Infants are either shown the new goal trial or the
new side trail. - New goal trial:
- We disrupt agent-object contingent relationship
(new goal), but preserve motion. For example, the
agent still reaches for an object on the left side of
the table which in now the ball (e.g., ball and bear
were swapped). - New side trial: in this condition the relationship
between the agent and object is preserve
preserved, but we disrupt motion e.g., agent grasps
bear (same) on right side (new). Which of these trails
will infants pay more attention to? - Results:
- 6-9-month-olds show selective recovery to new goal
trials (dishabituation to new goal/end-state; reach for
the ball instead of the bear).
*the benefit of this paradigm is that we can introduce
control conditions to be really clear that infants are able
to distinguish between inanimate (machine) or animate
objects (humans). Primarily that humans are more likely
have goal-directed action relative to inanimate objects.
Woodward (1998) Control condition (human hand vs machine claw)
- Habituation:
- 6-month-old infants were shown the hand reach for
the object (hand or claw) on the left side of the
screen till they’re bored (habituate). - Test:
(A) New Side Trial
- In the new side trial infants are shown the agent
reach for the same bear but on the other side of the
table (same goal; different action).
(B) New “Goal” Trial
- In the new gal trial infants are shown the agent
reaching for the ball (new goal) on the same side of
the table (same action). - Results:
- Hand condition: infants look longer (dishabituate) to
the new goal trial (same action new goal).
- Claw condition: infants look longer (dishabituate) to
the new side trial (same action but different goal;
directional properties temporal spatiotemporal
relationships; where in space is the claw moving to).
- Infants treat hand and claws different. They’re more
likely to attribute goals to animate action rather than
inanimate action (i.e., perceive actions as goal
orientated).
Note: In Woodward (1998), the ending is revealed to infants (visual information
If infants are tracking and generating predictions about the goal of an impending action then in the absence of visual information (conclusion or outcome shown) infants should be able to anticipate or predict what the agents action will be (before it does it; via eye tracking).
Kim & Song (2015)
Kim & Song (2015)
Steps:
1. Habituation:
- 6-month-olds sit behind a table with a red cylinder
(right) and a blue triangle/conical object (left), the
agent reaches repeatedly the blue triangular object,
2. Test Trial:
- In the test trail, the objects positions (blue left; red
right). The question is will infants, before the agent
reaches for an object, eyes follow the blue object
and remain there (predict that the goal is the same
and anticipate agents will reach for the blue object).
3. Results:
- 6-month-old infants selectively place their attention
to the goal outcome object before the agent moves
to reach for it.
*findings compliment Woodward’s (1998) study
Note:
All we can say from Woodwards study is that infants are able to track the agents’ actions and identify the end-state or goal from the sequence (get the bear; action is related to a certain object). The Woodward design is “old-skool” classic, but strictly speaking, we cannot precisely use it tell us about means to ends, that infants understand the hand’s actions (e.g., reaching and grasping) are performed IN ORDER TO acquire a certain object.
Do infants understand actions are performed to achieve goals?
- Yes, there are complementary studies that tell us that in the first year of life, infants understand that certain actions are performed IN ORDER TO acquire a certain object.
- teleology; with inanimate objects; two balls and a wall; jumping rational and irrational conditions; look longer at irrational/same behaviour in new condition.
We should be careful when choosing our interpretation of the discovery that infants track others’ goal-directed actions.
how do we test which interpretation is correct?
A. Rich (mentalistic) interpretation:
- Infants understand that agents act to achieve goals
because they DO understand and consider that
agents have mental states (e.g., beliefs, desires,
fondness, intentions, and emotions) that cause them
to act to achieve certain goals (inconsistent with a
teleological perspective and more consistent with
theory of mind).
B. Lean (non-mentalistic) interpretation:
- Infants understand that agents act to achieve goals
only because they are able to understand end-
states (outcomes), actions as means to end-states,
and that relevant aspects of reality (e.g., a visible
block on road) can constrain efficiency of actions.
The mental states of agents are NOT considered.
*They only understand actions as goal oriented because
they can understand outcomes and how physical
properties may constrain the actions efficiency to
achieve their goal (teleological; no mental state of the
agents’ motivation, belief, intention, feeling etc.).
How do we test which interpretation is correct?
- Using false belief tasks.
- If infants can genuinely understand the concept of
false-belief (representing that people can do things
based on false thoughts, thoughts can be incorrect
and not based on reality yet still guide actors
behaviour), only then might we choose a rich
interpretation that infants have a theory of mind.
There is some evidence that hints at mentalistic interpretations of infants’ understanding towards goal-directed actions.
What other evidence, other than false belief tasks, can we use to suggest that infants might be mentalizing (TOM) goal directing actions? Yes.
Kuhlmeier et al. (2003)
Kuhlmeier et al. (2003)
12-month-olds
Steps:
1. Habituation:
- Infants were randomly assigned to the square
helps movie or the triangle hinders movie
(counterbalanced; triangle is helper and square is
hinderer for half of the infants). In the square helps
movie we see the yellow square help the red circle
struggling to get up the hill. In the triangle hinders
movie the triangle comes down and blocks the red
circles path as they are struggling to get up the hill
and pushes it back down the hill.
2. Test Trial:
- Infants are then either shown two possible
outcomes; 1) the circle approaches the triangle; 2)
the circle approaches the square. Which outcome
will infants look longer at? Can they identify the
appropriate ending?
3. Results/Implications:
- Infants look longer at the approach-helper ending.
The authors conclude that these results reflect
infants ability to reason that the object is more likely
to approach the helper because they will be fonder
of the object that helped them and want to interact
more with them in the future. Infants thus are able to
socially evaluate agents they see around them which
is mentalistic in nature (fondness and distaste
judgments).
Is there convergence evidence that infants will prefer to interact with objects or agents which they socially evaluate as being good relative to objects they consider bad?
Hamlin et al. (2007)
Steps:
1. Infants (6- to 10-mths) morally evaluate, show
distaste for, the hinderer after watching the same
helping/hindering movies (hinderer pushing agent
down the hill, helper pushes the agent up the hill).
2. After being familiarised with both events; which
object will infants prefer to interact with when given
the chance? Kind-helpful-square or mean-hinderer-
triangle?
3. Results:
6-moths and10-month-oldd infants both show a
preference for interacting with the helpful agent
(square). This preferential selection indicates that
infants have an innate moral compass which guides
their social evaluations (TOM) which guides their
actions (interacting with moral agents).
We need to be careful about our interpretation of the Hamlin et al. study. The findings could be explained not by a rich explanation or a lean explanation.
