In this article we propose a bottom-up approach to higher-level mental states,

such as emotions, attention, intention, volition, or consciousness. The idea behind

this bottom-up approach is that higher-level properties may arise as emergent

properties, i.e., occur without requiring explicit implementation of the phenomenon

under examination. Using a neural architecture that shows the abilities of

autonomous agents, we want to come up with quantitative hypotheses concerning

cognitive mechanisms, i.e., to come up with testable predictions concerning the

underlying structure and functioning of an autonomous system that can be tested

in a robot-control system.

We do not want to build an artificial system that is, for example, conscious

in the first place. On the contrary, we want to construct a system able to control

behavior. Only then will this system be used as a tool to test to what extent descriptions

of mental phenomena used in psychology or philosophy of mind may be

applied to such an artificial system. Originally these phenomena are necessarily

defined using verbal formulations that allow for interpreting them differently. A

functional definition, in contrast, does not suffer from being ambiguous, because it

can be expressed explicitly using mathematical formulations that can be tested,

for example, in a quantitative simulation. It is important to note that we are not

concerned with the “hard” problem of consciousness, i.e., the subjective aspect of

mental phenomena. This approach is possible because, adopting a monist view, we

assume that we can circumvent the “hard” problem without losing information concerning

the possible function of these phenomena. In other words, we assume that

phenomenality is an inherent property of both access consciousness and metacognition

(or reflexive consciousness). Following these arguments, we claim that our

network does not only show emergent properties on the reactive level; it also

shows that mental states, such as emotions, attention, intention, volition, or consciousness

can be observed, too. Concerning consciousness, we argue that properties

assumed to partially constitute access consciousness are present in our network,

including the property of global availability, which means that elements of

the procedural memory can be addressed even if they do not belong to the current

context. Further expansions are discussed that may allow for the recognition of

properties attributed to metacognition or reflexive consciousness.