Consciousness

The Phenomenal and the Physical

My approach to consciousness is based on a distinction between the phenomenal world of our experiences and the physical world described by science. The physical world cannot be experienced by us and conceptual problems with consciousness often arise when people mistakenly imagine that the physical world has phenomenal characteristics - for example, in a classic formulation of Chalmer's hard problem of consciousness McGinn tries to imagine how the colour red arises from soggy grey neurons, when in fact, as Metzinger points out, physical neurons are completely invisible and not grey or soggy at all. A detailed discussion of the phenomenal and physical distinction can be found in Chapter 2 of my recent thesis and in What We Can Never Know, pp. 71-87.

Science of Consciousness

According to my interpretation, the science of consciousness starts by looking for correlations between the phenomenal and physical, which could include the physical, neural, functional and/or cognitive aspects of a system. Once a combination of brain experiments and phenomenology have identified systematic patterns, a theory of consciousness is developed that links states of the phenomenal and physical worlds. This theory is tested by making predictions about the phenomenal states associated with human subjects. Figure 1 is a preliminary illustration of how experiments on the correlates of consciousness could be carried out. The most significant parts of Figure 1 are shown in Figure 2.

Figure 1. Preliminary diagram illustrating scientific experiments on the correlates of consciousness. The scientist measures the brain of the subject within the scientist's phenomenal world. Representational relationships in the physical world connect these phenomenal experiments with the physical brain of the subject. The subject's descriptions of his phenomenal experiences appear as experienced sounds in the phenomenal world of the scientist.

Figure 2. The science of consciousness formulates precise descriptions of the physical brain and the phenomenal experiences associated with the physical brain and uses these to identify correlations between the physical and phenomenal worlds. The predictions that different theories of consciousness make about these correlations are then experimentally tested.

Type I and Type II Potential Correlates of Consciousness

The search for the correlates of consciousness is limited by the fact that it relies on the external behaviour of a system to measure the presence of consciousness. In some cases parts of the system can be changed that have no effect on behaviour, but could in principle affect consciousness. For example, it is possible that haemoglobin could be a necessary but not sufficient correlate of consciousness. However, this is impossible to test experimentally because the substitution of artificial blood for haemoglobin-based blood would have no effect on a subject's behaviour.

To handle situations of this kind I have developed a distinction between two different types of potential correlates of consciousness.

• Type I potential correlates of consciousness are either behaviour-neutral or they cannot be separated from the behaviour that is used to measure consciousness in a system. Their key characteristic is that no experimental measure of their connection with consciousness can be devised or suggested. The temperature of the brain and the presence of haemoglobin are type I potential correlates of consciousness because experiments that changed the brain's temperature or substituted artificial for real blood (without affecting anything else about the brain) would leave a person's behaviour untouched, and so we would have no way of measuring whether the experiments had any effect on consciousness.
• Type II potential correlates of consciousness can be separated out using behaviour and there is no overlap with the parts of the system that are used for measuring or reporting consciousness. When a type II potential correlate of consciousness is removed or altered, the system's reports of conscious states can change. Activity in particular brain areas is a type II correlate because we can vary this activity through transcranial magnetic stimulation or observe brain damaged patients and measure the change in consciousness through verbal or other behaviour. Functional correlates also fall into this category because it is conceivable that we could disable a person's capacity for imagination or emotion, for example, and probe their conscious states.

This distinction between type I and type II potential correlates of consciousness is described in detail in Chapter 2, Section 2.5, of my recent thesis, and it has important consequences for my approach to synthetic phenomenology.