Three Issues for Theories of ConsciousnessAt present, ToCs are generally used as ‘narrative structures’ within the science of consciousness. Although they inform the interpretation of neural and behavioural data, it is still rare for a study to be designed with questions of theory validation in mind. Although there is nothing wrong with employing theories in this manner, future progress will depend on experiments that enable ToCs to be tested and disambiguated. We conclude our review by identifying three issues that need to be addressed for a mature regimen of theory-testing to flourish in consciousness science.Of course, the above challenges are already being addressed, to varying extents, by consciousness research­ ers. These efforts are now complemented by initiatives such as the adversarial collaboration model, which is encouraging proponents of ToCs to devise experiments with the specific goal of differentiating between alterna­tive ToCs. Consciousness remains scientifically contro­versial, yet there is every reason to think that the iterative development, testing and comparison of ToCs will lead to a much deeper understanding of this most profound of mysteries.NatureNTR_4133ebd6-cfb0-11ec-834a-a6f41283ea00Anil K. SethCo-AuthorTim BayneCo-AuthorConsciousness science flourishes_4133ed3e-cfb0-11ec-834a-a6f41283ea00To identify issues to be addressed in consciousness science_4133ee42-cfb0-11ec-834a-a6f41283ea00TheoryConsciousnessSciencePrecisionDevelop theories with precision_4133ef0a-cfb0-11ec-834a-a6f41283ea001First, ToCs need to be developed with precision, for theories that appeal only to vague and imprecise con-structs can generate only vague and imprecise predictions. For example, HOTs and predictive processing and re- entry theories need to specify the kinds of meta-representations and re-entrant or predictive processes that are distinctive of (specific aspects of) conscious-ness; IIT needs to make precise its implications for the functional profile of consciousness and the impact of the environment and embodiment on consciousness; and GWTs need to provide a principled account of which workspaces qualify as ‘global’ in the relevant sense.Computational ModelsUse computational models to bring mechanistic specificity to ToCs_4133efaa-cfb0-11ec-834a-a6f41283ea001.1A promising approach here is to use computational models to bring mechanistic specificity to ToCs that may have been formulated in relatively abstract or conceptual terms. In addition to grounding the generation of fine-grained predictions, such models might also pro-vide a shared language in which the relative merits of rival ToCs can be compared, which can be especially useful for comparing ToCs originating from different starting points.Top-Down SignallingReveal shared principles of top-down signalling_4133f036-cfb0-11ec-834a-a6f41283ea001.1.1For example, computational models could reveal shared principles of top-down signalling among HOTs and re-entry and predictive processing theories, while clarifying the distinctions between meta-representation (for example, see REF.35) and global broad-cast (for example, see REFS127,128) that separate HOTs from GWTs129.ProcessesAllow contrasts between ToCs to be reframed in terms of processes rather than regions_4133f0cc-cfb0-11ec-834a-a6f41283ea001.1.2The development of computational models might also allow contrasts between ToCs to be reframed in terms of (potentially distributed) processes rather than, as is currently popular, in terms of broad neuroanatomical regions (for example, as in the debate between ‘front-of-the-brain’ and ‘back-of-the-brain’ theorists).Phenomenological PropertiesAccount not merely for the functional features of consciousness but also phenomenological properties_4133f162-cfb0-11ec-834a-a6f41283ea001.1.3A key challenge for the computational approach is to develop models that do not merely account for the functional features of consciousness but also account for its phenomenological properties — a challenge that can be described by the general label of ‘computational (neuro)phenomenology’ ...Subjective ReportsCollect subjective reports at the appropriate levels of phenomenological granularity_4133f202-cfb0-11ec-834a-a6f41283ea001.1.3.1This brings the additional challenge of how to validate, or disambiguate between, computational models using phenomenological data ... — a challenge that can be met, at least in part, by collecting subjective reports at the appropriate levels of phenomenological granularity ...ComprehensivenessMake ToCs more comprehensive_4133f2fc-cfb0-11ec-834a-a6f41283ea002In addition to being made more specific, ToCs also need to be made more comprehensive. For the most part, ToCs have tended to focus on particular kinds of local states (perceptual experiences, with an emphasis on vision), on particular kinds of global states (ordinary waking awareness) and on particular kinds of conscious creatures (adult human beings). Although there are good reasons why theorists have tended to focus on a restricted class of conscious states and creatures — experimental accessibility being an important factor — a fully comprehensive ToC must do justice to the rich diversity of consciousness...Although there is nothing wrong with ToCs that have a restricted focus, theories that provide a more comprehensive account of consciousness have obvious advantages over those that do not, especially if they can identify explanatory connections between different aspects of consciousness.Affect, Temporality, Volition & ThoughtAccount for affect, temporality, volition and thought_4133f3f6-cfb0-11ec-834a-a6f41283ea002.1With respect to local states, ToCs must go beyond perception and account also, for example, for affect, temporality, volition and thought.Dreaming, Meditation & DisordersAccount for the distinctive modes of consciousness associated with dreaming, meditation, disorders of consciousness and the psychedelic state_4133f4dc-cfb0-11ec-834a-a6f41283ea002.2With respect to global states, ToCs must go beyond ordinary wakefulness and account also for the distinctive modes of consciousness associated with, for example, dreaming, meditation, disorders of consciousness and the psychedelic state.Infants, Animals & AIAddress questions regarding consciousness in human infants, non-human animals and even artificial systems_4133f5c2-cfb0-11ec-834a-a6f41283ea002.3With respect to conscious creatures, ToCs must go beyond adult experience and address questions regarding consciousness in human infants, non-human animals and even artificial systems.MetricsIdentify trustworthy measures of consciousness_4133f66c-cfb0-11ec-834a-a6f41283ea003The third issue to be addressed is the measurement problem: the problem of identifying trustworthy measures of consciousness. Solving this problem is crucial, for detailed and comprehensive ToCs are unlikely to be of much use unless we also have the capacity to verify their predictions. It is useful to distinguish two (closely related) versions of the measurement problem.UnconsciousnessIdentify ways of distinguishing conscious from unconscious mental states_4133f720-cfb0-11ec-834a-a6f41283ea003.1The first concerns the detection of conscious contents. Here, the primary challenge is to identify ways of distinguishing conscious from unconscious mental states that do not make controversial assumptions about the functional profile of consciousness (such as that conscious contents must be reportable or otherwise available for high-level cognitive control).Animal KingdomDetermine the distribution of consciousness in the animal kingdom_4133f7d4-cfb0-11ec-834a-a6f41283ea003.2The other version of the measurement problem focuses not on contents but on creatures. The questions here include how we might determine the distribution of consciousness in the animal kingdom; whether certain classes of cerebral organoids or artificial intelligence systems are conscious; when consciousness first emerges in ontogenesis; and when consciousness is retained in the context of traumatic brain injury. Here, too, the challenge is to develop ways of measuring consciousness that avoid controversial assumptions about its functional profile.Organoids & AIDetermine whether certain classes of cerebral organoids or artificial intelligence systems are conscious_4133f888-cfb0-11ec-834a-a6f41283ea003.3OntogenesisDetermine when consciousness first emerges in ontogenesis_4133f93c-cfb0-11ec-834a-a6f41283ea003.4Traumatic Brain InjuryDetermine when consciousness is retained in the context of traumatic brain injury_4133f9fa-cfb0-11ec-834a-a6f41283ea003.52022-05-032022-05-09https://www.nature.com/articles/s41583-022-00587-4.epdf?sharing_token=YcY6bzXl0iqFYKrqtykdLNRgN0jAjWel9jnR3ZoTv0OlRlPtg3bVLf-Jc8wcElS4cYy8AzDVCWBxQOzhq6tjCaPtzaUOCVNudwUX_DHiGRbrwwYvSfYcJ-WgeYee3uFDjHJggIjwukEF0eyKzcSGFjW47xrxnt_yGTuxSkm_API%3DOwenAmburOwen.Ambur@verizon.net