6.1 An intrinsic property (Integrated Information Theory)

The Italian neuroscientist Guilio Tononi has considered what’s required for an experience to be conscious. One important facet of this theory is that different aspects of the experience are initially processed separately in the brain, then somehow ‘integrated’ together into a single conscious experience.

Take the experience of picking up an object, for example. The brain will process visual information about the colour, shape and position of the object, as well as other sensory information like touch. This information is shared between different regions of the brain to create the unified conscious experience. Imagine picking up a tomato. In doing so, we aren’t aware of each separate attribute – that it’s red, looks round and feels smooth – unless we’re focusing on them specifically. We just experience it as a tomato.

This is the basis of Tononi’s ‘Integrated Information Theory’. The more information sharing and integration that takes place, the higher the level of consciousness. This requires the neurons in the brain to be highly connected, as indeed many are.

Experiments using transcranial magnetic stimulation (TMS) suggest that Tononi may be on the right track. When a localised region of an awake person’s brain is stimulated, other regions far from the stimulated site respond in complex feedback patterns, which is suggestive of integrative processing and conscious experience. However, if a subject is under induced anaesthesia or in a state of non-dreaming sleep – and therefore without any awareness – then the brainwaves that are generated are more confined and have a much simpler form.

Tononi has derived a mathematical formula that can measure how integrated, and hence how conscious a system is. An intriguing aspect of this theory is that the connected components don’t have to be biological neurons for this intrinsic property of consciousness to emerge from the system. They could be artificial neurons, or silicon transistors. The question, then, is whether an artificial physical system with a high level of information integration could actually be conscious. Another implication of this theory is that a computer program can’t be conscious, however sophisticated its simulation might be – just as the Chinese Room idea argues.

This theory may give us a way of measuring consciousness, which is certainly highly significant. But what would remain unanswered is how that consciousness is actually created.