Conscious perception of brief presentations rules out the complex time-consuming interactions that are essential for integrated information. There is ample evidence showing that images are accurately perceived when presented for very short times. Relatively simple objects displayed for intervals ranging from microseconds (Greene & Visani, Reference Greene and Visani2015; Nordberg, Hautus, & Greene, Reference Nordberg, Hautus and Greene2018) to tens of milliseconds (Keysers, Xiao, Foldiak, & Perrett, Reference Keysers, Xiao, Foldiak and Perrett2001; Rolls, Tovee, Purcell, Stewart, & Azzopardi, Reference Rolls, Tovee, Purcell, Stewart and Azzopardi1994), were accurately perceived. Using small objects that included high-spatial frequencies, Gur (Reference Gur2018) showed that subjects can discriminate between <15′ wide faces displayed for a mere 33 ms, and Gur (Reference Gur2021) has recently demonstrated that it has been possible to recognize briefly flashed small doodles, contours, and faces that were followed by a masking pattern, thus ensuring that the neural activity generated by the displays lasted much less than a 100 ms. We note that short stimuli generate short neural responses so that all the above experiments show that we perceive briefly displayed objects in a strict feed-forward fashion. There is simply no time for any lateral or feed-back interactions.

Shifting from the laboratory to normal behavior, it is well known that humans continuously sample the world by quickly shifting their eyes from target to target (“saccades”) and lingering on a target for 200–250 ms (“saccadic pauses”). Because it takes >100 ms to plan for the next saccade, and given ~50 ms before neural responses are generated in primary visual cortex, there are <100 ms that the visual system uses to process information during saccadic pauses. Such a time limit is very consistent with flashed stimuli data.

The most important observation of the integrated information theory (IIT) is that consciousness is characterized by the experience of a holistic undivided view. Such an experience is achieved, according to the IIT, by a complex, elaborate interaction between constituent elements that results in an integrated information. Brain processes leading to integration consist, inter alia, of back and forth local lateral interactions, exchanges between numerous cortical areas, and re-entrees from “higher” to “lower” cortical areas. However, such an integrative process takes hundreds of milliseconds which is quite at odds with our ability to consciously perceive images presented for times ranging tens of microseconds to tens of milliseconds.

Integration which is fundamental to the IIT erases individual characteristics of constituent elements. This contradicts conscious perception where both holistic view and an exquisite details of individual elements co-exist. Conscious perception of space is manifested first and foremost in the detailed, high-resolution, topographically precise depiction of spatial elements. When we see an object, we perceive it as an integrated whole but at the same time we are consciously aware of all the minute spatial elements that make up that object. Indeed, it is our perception of details that enables us to discriminate between objects. This dual perception of the forest and the trees cannot be emulated by any computational system – the IIT included. Any mode of integration, biological or artificial, necessarily discards all information properties defining the integrated-over elements. To take a biological example, in primary visual cortex, cells representing small, discrete spatial features converge onto “simple cells” which represents some collective spatial properties such as preferred size and orientation (Hubel & Wiesel, Reference Hubel and Wiesel1962; Fig. 1). Clearly, the output of a simple cell that integrates over many input cells lacks any information regarding individual, discrete spatial elements (cf. Gur, Reference Gur2015; Fig. 1).

Fig. 1. Collection of elements that generate an integrated view – a triangle, but at the same time all individual characteristics are discerned.

The IIT posits that consciousness is a manifestation of the level of information integrated by a system. If a system consists of a very large number of elements and if by a process of elaborate interactions these elements are integrated over, then the system is conscious. However, as detailed above, such an integration renders the system blind to the characteristics of its individual elements – which is in a stark contrast to our conscious perception where all the details making up the whole are preserved.

What is obvious about our conscious perception is that it presents two characteristics that are contradictory to any computational/integrative/interactive mechanism: Our perception is holistic yet we see all details, and, furthermore, we can compare all characteristics of a given region in space to another, a spatially remote one. For example, we can see in a flash that at the right corner of the triangle (our holistic percept; Fig. 1) there is a small blue pentagon whereas at the left corner there is a larger yellow ellipse.

Because the unique properties of conscious perception cannot be explained by lateral, feed-forward, or feed-back axonal transmissions, it may be useful to think in terms of a “wireless” mechanism – a field. An electric field, for example, is shaped by the number and strength of its electrical charges; changes in the charges’ characteristics instantly (at the speed of light) affect the shape of the force distribution of the whole field, yet the properties of individual charges (location and strength) are preserved. We do not know what the nature of this presumed “neural field” is but we do know which essential characteristics it must have – and they are not what the IIT suggests.