Merker et al. suggest that empirical measures of the joint presence of differentiation and integration (presented as stand-ins for Φ) are not indices of consciousness per se but reflect the extent to which “performance requirements for cortical operations” are preserved during wakefulness rather than relaxed during other states “of less than full wakefulness” (sect. 4, para. 11). Furthermore, the authors advocate that, in order to provide evidence for their reliability in indexing consciousness, experimental dissociation “of the consciousness variable itself from the operations that are dependent on efficient cortical information transfer and crowned by systems consolidation,” would be needed and that this fundamental requirement is “unlikely to be achieved in the foreseeable future” (sect. 4, para. 12).

Here, we provide a more informed, and positive, view of current empirical efforts in the science of consciousness; specifically, we highlight recent experiments – ignored in the target article – that dissociate experience from performance requirement and system consolidation, suggesting that theory-inspired measures of integration and differentiation are valid indices of consciousness.

A practical way to estimate the joint presence of differentiation and integration in the human brain is quantifying the complexity of electroencephalographic (EEG) responses to direct cortical perturbations with transcranial magnetic stimulation (TMS). This causal, information-related measure – called the perturbational complexity index, PCI (Casali et al., Reference Casali, Gosseries, Rosanova, Boly, Sarasso, Casali and Massimini2013) – was explicitly designed to approximate the integration and information postulates of integrated information theory (IIT) and has been systematically applied to a variety of conditions ranging from wakefulness to unresponsive states such as sleep, anesthesia, and disorders of consciousness. In the context of the present commentary, we would like to note that a first dissociation between consciousness and the performance requirements typical of wakefulness is provided by measurements performed during dreaming sleep (Casali et al., Reference Casali, Gosseries, Rosanova, Boly, Sarasso, Casali and Massimini2013; Casarotto et al., Reference Casarotto, Comanducci, Rosanova, Sarasso, Fecchio, Napolitani and Massimini2016; Massimini et al., Reference Massimini, Ferrarelli, Murphy, Huber, Riedner, Casarotto and Tononi2010). Indeed, dreaming is a state of disconnection from the external environment, in which such requirements are significantly relaxed, yet it is characterized by intense subjective experiences. A second, more extreme case is represented by the dissociative anesthetic ketamine. At the appropriate dosage, this N-methyl-d-aspartate (NMDA)-antagonist renders subjects behaviorally unresponsive, fully disconnected from the external environment (they can undergo surgery) and unable to retain memory of external events occurring during sedation. At the same time, subjects typically undergo very vivid hallucinatory-like experiences. Crucially, PCI values measured both in dreaming and in ketamine-anesthetized individuals were found to be high and similar to those obtained during wakefulness (Casarotto et al., Reference Casarotto, Comanducci, Rosanova, Sarasso, Fecchio, Napolitani and Massimini2016; Sarasso et al., Reference Sarasso, Boly, Napolitani, Gosseries, Charland-Verville, Casarotto and Massimini2015), consciousness being present in all cases but “performance requirements for cortical operations” being clearly different across the three conditions. Dissociations have also been demonstrated in a large cohort of severely brain-injured individuals including vegetative (VS) and minimally conscious state (MCS) patients. The latter show minimal but definite signs of awareness (Giacino et al., Reference Giacino, Ashwal, Childs, Cranford, Jennett, Katz and Zasler2002) typically in the context of severe cortical damage and impaired cognitive functions including memory consolidation (Aubinet et al., Reference Aubinet, Murphy, Bahri, Larroque, Cassol, Annen and Gosseries2018; Bodien, Martens, Ostrow, Sheau, & Giacino, Reference Bodien, Martens, Ostrow, Sheau and Giacino2020). Crucially, MCS patients (including the so called “MCS minus,” who are unable to follow commands) attain PCI values in the range of awake healthy subjects (Casarotto et al., Reference Casarotto, Comanducci, Rosanova, Sarasso, Fecchio, Napolitani and Massimini2016; Sinitsyn et al., Reference Sinitsyn, Poydasheva, Bakulin, Legostaeva, Iazeva, Sergeev and Piradov2020). Thus, measures of integration and differentiation are extremely accurate in detecting consciousness, irrespective of levels of arousal, connectedness to the external environment, behavioral performance, cognitive, memory abilities, and the severity of lesions. This large body of evidence provides empirical validation of the theoretical assumptions motivating the development of indices of integration and differentiation and contradicts the thesis of the target article, yet it is not even considered.

Interestingly, dissociations between stand-ins for Φ and the kind of cortical processes considered by Merker et al. were also found in the opposite direction. For example, non-rapid eye movement (NREM) sleep has been directly implicated in the consolidation of declarative and visuospatial memories (Marshall, Helgadóttir, Mölle, & Born, Reference Marshall, Helgadóttir, Mölle and Born2006; Rasch, Büchel, Gais, & Born, Reference Rasch, Büchel, Gais and Born2007) as well as in that of motor skill (Cousins, El-Deredy, Parkes, Hennies, & Lewis, Reference Cousins, El-Deredy, Parkes, Hennies and Lewis2014; Laventure et al., Reference Laventure, Fogel, Lungu, Albouy, Sévigny-Dupont, Vien and Doyon2016; Rasch, Pommer, Diekelmann, & Born, Reference Rasch, Pommer, Diekelmann and Born2009; Schönauer, Geisler, & Gais, Reference Schönauer, Geisler and Gais2014) in humans. These processes, also involving a dialog between the hippocampus and neocortical areas (Merker, Reference Merker2004), occur during NREM rather than REM sleep (Rasch & Born, Reference Rasch and Born2013). Thus, systems consolidation (Frankland & Bontempi, Reference Frankland and Bontempi2005), used as a central argument for the mistaken identity of Φ in the target article, can happen when PCI is low (Casarotto et al., Reference Casarotto, Comanducci, Rosanova, Sarasso, Fecchio, Napolitani and Massimini2016; Comolatti et al., Reference Comolatti, Pigorini, Casarotto, Fecchio, Faria, Sarasso and Casali2019; Rosanova et al., Reference Rosanova, Fecchio, Casarotto, Sarasso, Casali, Pigorini and Massimini2018). That PCI can dissociate processes subserving system consolidation (present during NREM sleep) from experience (often absent) should, thus, be seen as an encouraging sign rather than as a problem.

More generally, we would like to highlight a remarkable disconnect between the rather pessimistic view depicted in the target article and the recent surge of original studies (almost 200, reviewed in Sarasso et al., Reference Sarasso, Casali, Casarotto, Rosanova, Sinigaglia and Massimini2021) converging on measures of integration and differentiation as reliable markers of consciousness. Notably, such empirical convergence involved many independent researchers endorsing disparate frameworks and employing different techniques across various conditions such as sleep, anesthesia, epilepsy, hallucinatory states, and coma. Although these measures still lack the optimal spatiotemporal scale, do not incorporate all relevant theoretical constraints, and suffer from other important limitations (we agree with Merker and colleagues that there is still much work to be done), they definitely represent a step forward compared to previous indices. For example, they show a better association with consciousness compared to delta power (Frohlich, Toker, & Monti, Reference Frohlich, Toker and Monti2021), alpha power (Esposito, Nielsen, & Paquette, Reference Esposito, Nielsen and Paquette2004; Westmoreland, Klass, Sharbrough, & Reagan, Reference Westmoreland, Klass, Sharbrough and Reagan1975), or the P3b (Sergent et al., Reference Sergent, Corazzol, Labouret, Stockart, Wexler, King and Pressnitzer2021). For this reason, measures of integration and differentiation, and their broad experimental validation as reliable indices of consciousness, are recognized as important advancements from a clinical (Bai, Lin, & Ziemann, Reference Bai, Lin and Ziemann2020; Comanducci et al., Reference Comanducci, Boly, Claassen, De Lucia, Gibson, Juan and Massimini2020) and ethical (Bayne, Seth, & Massimini, Reference Bayne, Seth and Massimini2020) standpoint; we think that they may also offer empirical ground for a more positive appraisal of their theoretical foundations.