A pattern of light falls on your retina, and the information is partly processed there before being transmitted to the visual cortex at the back of the brain, passing through hierarchical stages of processing, each generating a successively more complex representation than the last.

But at which stage of this process do you become conscious of the scene you are viewing, and exactly when do you become aware of the objects within it?

According to one popular hypothesis, called the 'global workspace' model, consciousness is an all-or-nothing phenomenon, with awareness of an object emerging only when information has been processed extensively, and has been made available to the brain's working memory system.

New research, however, suggests we can be fully conscious of some features of an object, such as its color or shape, while being completely unaware of others. The study, just published in the journal Consciousness and Cognition, provides some evidence of what neuroscientists call 'partial awareness'.

James C. Elliot of the University of California, Santa Barbara and his colleagues used a phenomenon called attentional blink to try to determine whether or not visual consciousness is all-or-nothing. First described in 1992, attentional blink occurs when two visual stimuli are shown in quick succession, and describes the fact that we consistently fail to detect the second stimulus when it is shown within about half a second after the first one.

In one experiment, 16 undergraduate volunteers were asked to sit at a computer screen, which displayed rapid sequences of letters. The participants fixated on a cross in the middle of the screen, and then saw a white capital letter, followed in quick succession by a number of black, and then colored, letters. Across 300 trials, each participant was instructed to identify the white letter, and also the color of the first colored letter they saw afterwards.

The second experiment involved 26 more participants, and was designed in exactly the same way, except that this time, some were asked to identify the white letter in each sequence, as well as name the next one, and also report its color, while the rest were told to ignore the first one and identify each second letter and its color.

The participants experienced attentional blink that led them to being aware of one feature of the letters presented to them but not the other.

In the first experiment, the attentional blink significantly decreased the accuracy with which they could name the colour of the second 'target' letter in each sequence, regardless of whether or not they could correctly identify the first letter. The only factor that seemed to affect the participants' performance was the time lag between the two visual stimuli in each trial, such that their accuracy increased with a longer delay.

This was confirmed by the second experiment. In some cases, the participants could name the target letter without correctly reporting its color, and in others they could correctly report its color without knowing what letter it was that they had seen. Again, attentional blink made it difficult for the participants to both identify the second letter and correctly name what color it was, and their accuracy decreased with the time lag between the letters.

The findings thus provide evidence for partial visual awareness, and suggest that consciousness does not, after all, occur in an all-or-nothing fashion. Even when it seems your brain has limited resources to absorb everything in a fraction of a second, it's still processing that information and some of what your eyes have seen may enter into your conscious mind.

"This is an interesting and novel study which challenges the assumption that consciousness is all-or-nothing, [and that] you either see something or you don't," says Steve Fleming, a cognitive neuroscientist at UCL.

He adds, however, that the findings are not conclusive. "The participants may have had similar awareness of both aspects [of the target letters] but failed to report one because of noise in the reporting process, so convergent evidence from other measures of awareness – such as visibility or confidence ratings – would be a useful next step."