Theory of Mind: A Worked Example of the Scientific Process

What is Theory of Mind?

Definition: The ability to attribute mental states (beliefs, desires, intentions, knowledge) to oneself and others, and to understand that others' mental states can differ from one's own and from reality.

In other words: Recognizing that other minds exist and have their own perspectives.

Typically developing kids generally start showing theory of mind around ages 4-5. 

The Classic Paradigm: False Belief Tasks

The Sally-Anne Task:

  1. Sally puts a marble in basket A and leaves
  2. Anne moves the marble to basket B
  3. Sally returns - where will she look for her marble?

Typical responses:

  • 3-year-olds: "Basket B" (where it really is)
  • 5-year-olds: "Basket A" (where Sally believes it is)

What are the 5-year-olds getting that the 3-year-olds aren't?

Origins of the Concept

1978: Premack & Woodruff introduce "theory of mind" studying chimpanzees

  • Question: Does the chimpanzee have a theory of mind?

1983: Wimmer & Perner - first false belief study with children

1985: Baron-Cohen, Leslie & Frith - Sally-Anne task

  • Also examined difference in autism

This process -- making an ancient philosophical question (the problem of other minds) -- into a testable psychological question is generally referred to as operationalization.

How Does Falsification Work Here?

Think of all the claims we might make about ToM:

  • There is a thing, defined basically as the idea that other people have states of mind, that is a trait typically developing adult humans possess.
  • Typically developing kids start to show it around age 4 or 5.
  • Nonhuman primates have it only in very rudimentary form.

How might we operationalize these as a falsifiable prediction?

Falsifiable prediction: "If children under 4 lack representational theory of mind, they should fail false belief tasks even when memory, language, and executive function demands are minimized"

How Could This Be Falsified?

Possible falsifications:

  • Show that 3-year-olds consistently pass when task demands are reduced
  • Demonstrate that "failure" reflects other factors (language, executive function, pragmatics)
  • Show the shift isn't about conceptual understanding but performance factors

Fun fact: All of these ostensibly falsifying events have actually happened.

Challenge 1: Task Demands as Confounds

The critique: Standard false belief tasks require:

  • Complex language comprehension
  • Memory for story sequence
  • Inhibitory control (to override "where it really is")

Scientific response: Create simplified versions

  • Nonverbal anticipatory looking tasks
  • Implicit measures without verbal responses
  • Reduced executive function demands

Result: More nuanced understanding of competence vs. performance

Challenge 2: Infants Pass Implicit Measures!

Onishi & Baillargeon (2005) and others:

  • 15-month-olds look longer when agents search in the "wrong" place
  • Infants seem sensitive to false beliefs much earlier

This potentially falsified: "Theory of mind emerges at 4-5 years"

Competing interpretations:

  • Early implicit forms exist; explicit use develops later
  • Infants track behavioral rules, not mental states
  • Standard tasks have pragmatic/communicative demands

Challenge 3: Cross-Cultural Variation

Finding: Cultural differences in timing of false belief understanding

Possible interpretations:

  • Universal cognitive development with cultural variation in expression?
  • Different socialization of mental state talk?
  • Pragmatic norms affecting task interpretation?
  • Testing context effects?

Scientific response: Examine what drives variation

OK, so are we just not able to falsify claims in psychology?

Unlike most standard examples of problems to which we apply the scientific method, psychological constructs are:

  1. Theoretically defined - Can't directly observe "theory of mind," only behaviors we interpret as reflecting it

  2. Multiply realizable - Different mechanisms might produce similar behaviors

  3. Context-dependent - Performance varies with task demands

Thus: Falsification works at multiple levels, not just accept/reject

How the Field Progresses

Not through simple falsification, but through:

  • Triangulating across multiple measures
  • Refining what we mean by the construct
  • Specifying boundary conditions
  • Developing process models

After decades of research:

  • Not "When does theory of mind emerge?" with a single answer
  • But richer understanding of components, development, and measurement

The Messy Reality = Good Science

What we now understand:

  • Theory of mind is not all-or-nothing
  • Implicit and explicit understanding may diverge
  • Task demands matter enormously
  • Multiple components develop at different rates

This represents scientific progress:

  • Refined constructs
  • Improved methods
  • More nuanced theories

In psychology, falsification often means progressively constraining and refining theories rather than simple rejection

Discussion

  • How does this example fit within your understanding of "the scientific method"?
  • What makes falsification different in psychology versus other sciences?
  • How do we decide when a construct needs refinement vs. rejection?
  • What does "progress" look like in this kind of research program?
  • Have you encountered similar challenges to falsification?

As you do your own research...

Consider:

  • How will you operationalize your constructs?
  • What alternative explanations should you rule out?
  • How might task demands affect your results?
  • What would it mean to "falsify" your hypothesis?
  • How does your work fit into an ongoing research conversation?