Lecture 1: Theory of Mind

theory of mind

-ability to ascribe mental states to one’s self and to others (Wellman)

-understanding that mental states that do not always accurately reflect reality

-understanding that mental states can be the causes of behaviour of others

understanding mental states

-information may be vague, incomplete or not relevant

-but we still make complex judgements

-interpreting mental states can depend on background knowledge

rock, paper, scissors

-if you figure out something about your opponent’s plan or mislead them, your odds of winning increase

-people tend to pick all three options equally but not at random

-helpful to know information about likelihood of choices, but if your opponent also knows it you need to make a judgement about how to use that information

precursors to mental-state understanding

-long before being fully-linguistic, children orient to other humans

-newborns orient towards human faces

-6 month olds follow other people’s gaze

-12 month olds shared mutual gaze and protodeclarative pointing

Woodward - method (humans as goal-oriented agents)

-9 month olds were habituated to an adult reaching for one of two toys

-measure how long the child looks at this

-the toys are then swapped

-infants aw either the same reach, to a new toy or a new reach to the same toy

Woodward - results

-9 month olds looked longer when the adult reached for the other toy

• despite the fact this was the same gesture they were habituated to, shows the infants expected the person to reach for the same object

-the children didn’t show this pattern for non-human reaches

humans as goal-oriented agents

-infants seem to understand that people act in intentional, goal-directed ways

-in other words, they can take account and have expectations of mental states when observing the actions of other

ToM and language learning

-from 18 months children use their mental state awareness to help them learn language

-vital for learning words and meaning

Tomasello - method

-show 18 month olds novel objects

-then have adults point to it and give it a name e.g., this is a modi

-then ask the child to pass them the modi whilst looking at the object

Tomasello - results

-if the child is not told which one is the modi, they will reach for one of the items at random

-show they use ToM to understand langauge

Repacholi & Gopnik - method

-goldfish crackers are high reward for children

-broccoli is low reward

-researcher would pick up goldfish crackers and model behaviours of disgust

-researcher would pick up broccoli and show they enjoy eating it

-researcher would then sit between the broccoli and crackers and ask the child to “give me some”, not specifying which one they want

Repacholi & Gopnik - results

-younger children would give the researcher crackers, because they like crackers

-15 month olds would give them broccoli because they understand that the researcher prefers broccoli, even though that is different to them

Bartsch & Wellman

-2 year olds can explicitly contrast desires

-it’s not until 3 years that children spontaneously use words like think and know

acquiring a ToM

-from infancy, children understand that other people may be driven by intentions

-this is unlikely to be explicit awareness, infants don’t have insight into mental states → just based of innate expectations of the world

-explicit mental state understanding emerges between 12 months and 6 years

mental states are difficult for young children to apprehend

-mental states are:

• insubstantial

• frequently non-obvious

• rapidly changing

• often dependent on real-world knowledge

extending own knowledge (acquiring ToM)

-often our thoughts tend to correspond closely with those of other people

-extending their own knowledge to other people is a useful heuristic for children

“does your friend like chocolate?”

-three year olds will probably answer yes

-however, their response may be based on their own mental state, not that of their friend

-genuine insight into their understanding of mental states would only become apparent when the mental states of others differ from their own

false belief tasks

-experimentally convenient as they avoid the problem that many beliefs match up with reality

-these tasks cannot be passed by substituting one’s own beliefs for those of the characters in the task

unexpected transfer task

-a character is introduced

-an event happens that induces a true belief, then the character leaves

-a second event, unseen by the character occurs, making their belief false

-the character returns, and a question is asked about their belief

passing the unexpected transfer task

-passing this task marks a step forward in terms of cognitive sophistication

-flexible understanding of mental states is important for empathy, planning, deceiving, learning and dealing with other people

unexpected contents task - method

-object looks like it has familiar contents, but when shown it has unfamiliar contents

-asked additional questions of:

• “what would your friend say was in it?”

• “when I first showed you this, what did you think was inside it?”

unexpected contents task - results

-if children are using their own knowledge + mental state they would say their friend would reply with the unfamiliar contents

-when asked what they thought was inside - three year olds reply incorrectly with the actual contents, despite the fact they previously responded otherwise

-children can’t predict other people’s mental state or acknowledge their own mistaken belief

Astington (unexpected contents task)

-shown juice carton and asked what is in there

-children shown the carton contains ribbons

-children as then asked what they thought was in the carton before it was opened

• children before ToM will say ribbons

-shows that even asking about their own mental state they can make mistakes

Gospnik & Astington (unexpected contents task)

-an object was placed inside a doll’s house

-it was then replaced with a second object

-two year olds could successfully report what was in the doll’s house first, so they do not have a problem with working memory

-so the wrong responses in UCT are not due to bad memory

second-order theory of mind

-the ability to think about other people thinking about other people’s mental states

Naito & Seki

-mental state understanding continues to develop beyond five years

-second order theory of mind develops around 6 years

emotions - age of onset

-understood in others by 12 months

-gradual developmental progression

-understand the most simple mental states first

desires - age of onset

-can understand that other people have desires

belief - age of onset

-4 years

-understand that ourselves and others can have false beliefs

explaining ToM development

-children’s ToM understanding is said to be theory-like

-unobservable constructs are used to explain human behaviour

-during childhood, children construct an implicit theory about the way minds work

-early, simple views of ToM are revised in the light of new experiences and observations

“theory theory”

-locates the source of improvement in children’s knowledge

-assumes that children’s improved ToM performance is caused by a better understanding of mental states

-fits with the finding that the first mental states to be understood are ones that are easiest to reason about

• can explain findings from non-mental state tasks

the false photograph task - method (Zaitchik)

-teddy (object of interest) is placed on the table

-then the teddy will be moved to the chair

-memory question → “where was the teddy when the photo was taken?”

-test question → “where is teddy in the photo?”

the false photograph task - result (Zaitchik)

-children will pass the first question quite easily

-3 year olds will say the teddy is on the chair for the second question

-make the same errors as in the false belief task → but questions involve no mental state components

the false photograph task - interpretation (Zaitchik)

-children fail to answer questions about representations correctly

-it’s not that they don’t understand mental states per se, but that they find it difficult to reason well about any kind of representation

the false belief task (Zaitchik)

-teddy is on the table

-Sally leave the room

-Teddy is moved to the chair

-memory question → “where was teddy when Sally left the room?”

-test question → “where does Sally think teddy is?”

representation

-describes how we mentally represent information about the world

-refers to a mental image, symbol or concept that a child creates to represent an object, event or idea in the real world

representations and ideas

-simple representations are fine if we want to think simple thoughts about reality

-less helpful for more subtle ideas

-just thinking in terms of representations will not give the ability to answer ToM tasks

metarepresentation

-describes how we mentally represent information about thoughts, beliefs, desires

-reflecting on our own thoughts

-better at this at 5-6 years of age

false photograph task and representations

-suggests that mental states are hard for children to think about because they find it difficult to reason well about any kind of representation

-find it hard to think about their own thoughts

family

-children who have siblings are more likely to pass false belief tasks than children without siblings

-suggests that social interaction can influence children’[s conception of minds

-helps to develop reasoning → links to ZPD

Goswami (family)

-families that talk about mental states tend to have children that are more successful at false belief tasks

-language used by families and their values links to children’s understanding of mental states

-families that explicitly model talking about mental states leads to earlier understanding of ToM → helps children reflect and articulate these concepts

culture

-many measures of ToM are language dependent

-the age at which ToM is reached on tasks varies across cultures between 4 and 7 years

-specific ages vary but the order in which competences are reached remains the same

executive function

-children with stronger inhibitory control tend to perform better on measures of false belief

-EFs and ToM emerge at similar ages but direction of causality is unclear

infants and false beliefs

-use VOE paradigm

-gaze following to study ToM

Onishi & Baillargeon - method (infants and false beliefs)

-infants were habituated to an actor reaching into one of two boxes to retrieve an object

-then either:

• the object remained in the same location

• object is moved to a new location

-infants watched where the actor reached to get the object

Onishi & Baillargeon - test phase (infants and false beliefs)

-the actor either:

• reaches to where the object was last seen

• reaches to where the object is

Onishi & Baillargeon - results (infants and false beliefs)

-infants looked longer when the actor reached to where the object is

-suggests infants found this suprising

-shows the infants expected the person to reach for where the object was last seen, not where it actually is

Onishi & Baillargeon - interpretation (infants and false beliefs)

-suggest that infants are surprised because the actor’s belief is different to their action

-this view suggests infants have some understanding of beliefs in other people

Onishi & Baillargeon - vocal task (3 year olds and false beliefs)

-same method but asked “where will Sally look for the pizza?”

-where she last saw it → infants look less long at this scene, suggesting it’s what they expected

-where we know it is → 3 year olds give this incorrect verbal answer

explanation for contradictory findings

-we can’t claim both that children understand false beliefs at 15 months and that they don’t do so until 4 years

-suggestion that we have two distinct systems for dealing with mental states

• one system is fast but simple, implicit and automatic → infants

• other system is slow-but-accurate, and explicit → 4 years