1. Foundational issues
Mind is the constitutive problem of Psychology. Psychology’s disunity is
contrastively interpreted. For some it is inherent (e.g., Koch 1981) for others it is sign
of a “would-be-science” in need of unification to achieve the status of a mature
science (e.g., Staats 1999). I assume that full unification is not necessarily impossible.
At this point the following caveat should be made. If practically possible, the working
hypothesis objective is both far away in the future and not achievable along the lines
of most current research programs. Therefore, my main objective in this article is to
briefly argue for the need of the cognitive science community to take a radically new
approach to the study of mind and subsequently illustrate it by systematizing the
following three core research areas of the nature and workings of mind: 'thinking',
'representation' and 'communication'. That is to squarely tackle the foundations of
cognitive science.
The first explicit statement of the need to discuss the foundations of psychology
was Staats (1981, p. 253): “it is important to consider the nature of unified theory and
the methods involved in unified theory construction. These are topics our science has
thus far not addressed.” To date the most comprehensive and theoretically coherent
description of the current foundations of cognitive science is contained in Newell’s
(1990) Unified Theories of Cognition.1 It constitutes the undisputed basis of the still
dominant computational paradigm and is squarely based on Turing’s (1936, 1950)
work and the physical symbol systems hypothesis (Newell & Simon 1976). Recent,
more or less, promising alternatives are still much less developed in terms of their
foundational descriptions of the nature of mind.2
The rest of this section, except the last paragraph, proposes modifications with
respect to the received scope, methodology, and evaluation criteria of a unified theory
of mind (UTM) in accordance with the requirements of evolutionary compatibility
and of a mature science.
Newell (1990 p. 16) was fully aware of the enormity of the task. “Clearly, I can’t
mean a theory of all that! A unified theory of cognition is just a fantasy.”
Consequently, he proposed the unification of a subset of behaviour along the lines of
the following priority list: Problem solving, decision-making, routine action.
Memory, learning, skill. Perception, motor behavior. Language. Motivation, emotion.
Imagining, dreaming, daydreaming, .... He was crystal clear. A unified theory “is a
cognitive theory that gets significantly further down the list cumulatively than we
have ever done before. If someone keeps asking how the unified theory explains
dreaming or how it describes personality, I will simply point back to this list. It is
right to ask, but wrong to insist.” (ibid. p. 15). Although, Newell was right on the
cumulative aspect of a unified theory he was mistaken about his priority list. A theory
of mind that does not address a minimum core of mental phenomena like meaning,
thinking, emotion, and communication should not count as a UTM. As a result,
Newell’s foundational vocabulary cannot account for phenomena like language and
consciousness.
Newell’s methodology is probably the most important characteristic of his work
and it is widely followed. He argued for unified theories of cognition to be
formulated as architectures.3 The choice of architecture as the theoretical tool for
developing a unified theory is conceived to be particularly important because it
provides the interface between structure and content (Newell 1990, p. 82) or the
abstraction that gets at the essence of mind (Anderson 2007a, p. 7). This position can
be based on the familiar philosophical distinction of the personal sub-personal level