Following military service (1952–54), he took up a MacKinnon scholarship at Magdalen College, Oxford, with a place to study Law.
His PhD was awarded in 1964 for a study of environmental, genetic and hormonal influences on emotional behaviour in animals.
He retired from the chair of psychology in 1999, but continued his experimental research as an emeritus professor, and spent a productive year at the Center for Advanced Study in the Behavioral Sciences at Stanford University, California.
He suggested that intentionality is another aspect of the "binding problem", as to how the different modalities, such as sight and hearing, are bound together into a single conscious experience.
Without such unifying binding, he argued that objects would be just meaningless shapes, edges, colours, and tastes.
However, Gray stressed that these perceptual signals arise in the brain, and do not have any kind of external existence.
Thus, for example, in his view, visual perception is a good guide to the reflectance of surfaces, which in turn often have survival value for the organism.
Late-error detection permits more successful adaption, if a similar situation emerges in the future.
Gray accepted that there are many unconscious systems that detect errors, so this on its own does not establish a survival value for consciousness.
Gray argued that the brain takes account of plans as to what to do next, plus memories of past regularities, in assessing what is likely to be the next stage of a particular process.
In this way, the sketch is causal in the sense that it performs the function of recalling or assisting memories, but it is not directly active in the brain.
Consciousness is causal, in the sense that downstream unconscious systems respond to it, mainly in the area of error correction.
In discussing this question further, Gray looked at synaesthesia, where he described modalities as becoming mixed, for example when sounds are experienced with colour.
Brain scanning studies have shown that when words are spoken, in addition to the normal activity in the auditory cortex, the V4 colour vision area in the visual cortex became active, in a way which does not occur in people without synaesthesia.
He also questioned the temporal aspect of Hameroff's model, where the proposed 25 milliseconds to wave function collapse is only a tenth of the time considered necessary for a conscious perception to form.