A computational cognitive modeling approach to the development of second-order theory of mind

Burcu Arslan

    Research output: ThesisThesis fully internal (DIV)

    1777 Downloads (Pure)

    Abstract

    Most 3-year-olds fail on the hide and seek game. For example, they close their eyes as if: “Nobody knows where I am if I do not see them”. When they turn 4, they understand that others can have different beliefs and knowledge, and this theory of mind makes them successful at hide and seek.

    Theory of mind is also crucial in more socially complex situations, such as catching others’ lies and keeping a secret. Imagine, a father and his son want to surprise the mother and want to keep secret what they bought for her until her birthday. To keep this as a secret, the son should reason: “Mom should not know that I know what her birthday present will be.” If the son is younger than 6, it is not likely that he will apply this more complex theory of mind reasoning, and probably will reveal information before the mother’s birthday.

    Why does this development take another two years? This dissertation combines computational cognitive modeling with empirical research to provide an answer to this question. Simulating children’s minds with a computer program allowed us to generate a new prediction: 5-year-olds’ development can be accelerated with exposure to complex theory of mind. We trained 5-year-olds with many examples of complex theory of mind. Confirming the prediction, 5-year-olds were able to use complex theory of mind after two weeks of training but also after four months from the training. Coming back to the question, the answer is: “`Just lack of experience!”
    Original languageEnglish
    QualificationDoctor of Philosophy
    Awarding Institution
    • University of Groningen
    Supervisors/Advisors
    • Verbrugge, Rineke, Supervisor
    • Taatgen, Niels, Supervisor
    Award date21-Apr-2017
    Place of Publication[Groningen]
    Publisher
    Print ISBNs978-90-367-9693-4
    Electronic ISBNs978-90-367-9692-7
    Publication statusPublished - 2017

    Cite this