Because they seem interested in the experience of space, the authors should have noticed that integrated information theory (IIT) has set a specific project in explaining the spatial structure of experience (Balduzzi & Tononi, Reference Balduzzi and Tononi2009; Haun & Tononi, Reference Haun and Tononi2019; Tononi, Reference Tononi2014). The most recent study shows how IIT can account, in a highly detailed manner, for spatial phenomenology, demonstrating that topological features of spatial experience are identical to the features of the cause–effect structures specified by the kinds of networks that compose early sensory cortices. In this study, we showed that the extendedness of a spatial experience – its connectedness, the locatedness of its points, and so on – is all to be found in the intrinsic cause–effect structure of a lattice network. This study is just one example, but it would seem to give the lie to any notion that IIT is unconcerned with detailed perceptual phenomenology.

It should be pointed out as well that the authors seem to conflate the concepts of the “just noticeable difference (JND)” and the “difference that makes a difference,” taking the latter to refer to the former. The JND is a concept from classical psychophysics that traces back to Weber and Fechner, identifying a difference in some physical stimulus intensity, like the luminance of a light source, that is needed to make a difference to some perceptual experience, like the perceived brightness of the light. It is a relationship between stimulus and percept. (Strictly speaking, the Fechnerian goal in measuring JNDs is to infer the relation between a physical stimulus and an absolute perceptual quality like brightness; not between the stimulus and the JND.) Furthermore, the just noticeable difference is a statistical relationship – the JND does not figure in any single experience, rather it is derived by observing many experiences as the stimulus is varied.

Meanwhile, the phrase “difference that makes a difference” is sometimes used in science to explain the meaning of “information” in a domain-neutral sense, and traces back to Gregory Bateson (Bateson, Reference Bateson1972) – but it does not carry any special currency in psychophysics. In the context of IIT, the phrase is always used to point out the importance of neural (i.e., physical) state to the notion of causal constraint. That is, some neural mechanism might vary in its state in many ways, but it is those differences in state that matter to the system that are important to the system's cause–effect structure: those are the differences that make a difference. This is a relationship between the system and itself. Furthermore, the difference that makes a difference is a feature of a system now: it is actually in one state or another, and which state it is in is exactly what matters to integrated information. Therefore, as used in the IIT literature, the “difference that makes a difference” is a completely distinct concept from the “just noticeable difference” of psychophysics. The fact that these could be so confused in the target article suggests a lack of depth in the authors’ investigation into the relevant literature.

Of course, one could recontextualize “the difference that makes a difference” to mean something like the JND – the JND is the difference in some stimulus that makes a difference to perception. Again, this is a matter of a relation between consciousness and the outside world, not of consciousness per se, and IIT is first and foremost a theory of consciousness. But certainly we could draw some connection between JNDs and integrated information. Changing a stimulus along certain property dimensions would be expected to change the neural state of the observer, and if the state is changed in a way that matters to the system – if it's a difference that makes a difference – then according to IIT, then phenomenal changes will ensue. As a very basic example, it was shown in Haun & Tononi (Reference Haun and Tononi2019) how changing the state of nodes in a grid network changes the intrinsic causal structure of that network in a way that is akin to changes at locations or regions in a spatial experience.

For further questions regarding perceptual experiences – how does IIT explain brightness, or color, or timber, or pain – we need explanations for those phenomenal qualities first, before we can try to explain the connection between the phenomenal qualities and an external stimulus. The link between a psychophysical function (which arrays a series of statistical measures of percepts, measured over many moments, along a variable stimulus axis) and a cause–effect structure (which is a single experience at one moment in time) is ultimately an empirical project, and one domain in which IIT can be tested: The theory should predict what kind of experience an observer has, depending on the states of the neurons constituting their brain. But this is not a problem of “level of abstraction” or “misconstrual” as posed by the authors; rather it is a problem of what we now know and what we are now able to do – at present, we do not know enough about the brain or about the fine structure of such experiences, and we are not able to apply the IIT formalism to such complex systems. But no other theory of consciousness, or of perception for that matter, would seem to be in a better position to explain why experiences feel the way that they do – the problem is at least well-defined for IIT. From Haun and Tononi (Reference Haun and Tononi2019) alone, it should be clear that a course for this project has already been set.