Zenan Ruan1,2*- 1Zhejiang University—University of Luxembourg Joint Laboratory on AIs, Robotics and Reasoning (ZLAIRE), School of Philosophy, Zhejiang University, Hangzhou, China
- 2Department of Philosophy of Science and Technology, School of Philosophy, Zhejiang University, Hangzhou, China
Introduction
In the wide field of cognitive science, there are many works on consciousness. These studies can be roughly divided into two categories, i.e., the philosophical (especially the philosophy of mind) and the scientific (especially cognitive neuroscience). Neuroscientists have been searching for “neural correlates of consciousness” (NCCs, cf. Koch, 2012; Tononi and Koch, 2015), an approach that prioritizes the study of the correlation between specific brain activity and aspects of consciousness, while a philosophical approach always focuses on identifying explanatory links between neural mechanisms and consciousness.
The philosophical approach to consciousness was at one point difficult due to the “hard problem” argued by Chalmers (1995), that is, there seems to be an unbridgeable gap between the first-person perspective and the third-person perspective. Nonetheless, the current empirical Theories of Consciousness (ToCs) lay in an unoptimistic situation, where they “talk” past each other (Seth and Bayne, 2022; Yaron et al., 2022). As the number of theories keeps increasing (Signorelli et al., 2021), certain projects like adversarial collaboration (Melloni et al., 2021) are proposed and valued. A theory [for a preliminary case, see Safron (2020)] is expected to eventually surpass such situations in the future (Koch, 2018), yet there remain challenges to be met.
Criteria for the competition of various theories
Since Crick and Edelman, some neuroscientists have tended to support setting aside the “hard problem” and starting with the easy ones. Ideally, testing, contrasting, and iterating those ToCs might ultimately lead to a deeper understanding of the mystery of consciousness. Nevertheless, all theories need not be directly adopted to empirical tests without effective constraints.
As a beginning, Doerig et al. (2021) proposed four challenges that constrain current theories, the so-called “hard criteria,” which are “hard” on their emphasis on rigor and articulation and different from some evaluations that are vague and difficult to compare effectively. Overall, their initiations and contributions seeking empirical standards of ToCs are inspiring, but there are certain criticisms of their specific proposals, especially those that treat consciousness only as a function (Del Pin et al., 2021). As for discussions on conscious experience (Edelman and Tononi, 2001) in terms of its states and contents are concerned, they appear to be overlapping—an experience (as identified by its state) is always created along with at least one of the specific aspects of contents (cf. Damasio, 2021). Thus, functionalists tend to take much count of the contents of consciousness (Fleming, 2021) that are typically revealed in experiments through verbal or behavioral reports—several kinds of measurements. Even in the no-report paradigm (Tsuchiya et al., 2015; Koch et al., 2016), we cannot say certainly that we captured the experiential markers of the subjects, which can exclude the cognitive processes that are the consequences of consciousness (Northoff and Huang, 2017).
Recently, Seth and Bayne (2022) explicitly surveyed four typical ToCs based on their theoretical origination, the basis for prediction, and the specific problems they target. From these analyses, an avenue from the various theories toward a comprehensive and satisfactory theory should be to supply more comparable explanatory objects and discussion bases (such as the preference for the phenomenological or functional characteristics of consciousness and the global or local states of consciousness). Furthermore, they proposed three conditions for effective consideration of any feasible future theory referring to explicitness, comprehensiveness, and measurement, respectively.
Constraints for the study of consciousness are a form of meta-theoretical study that certainly makes sense, and in principle, we need to address them all. Moreover, such criteria may be as great in number as the diversity of theories of consciousness itself. We should consider the fundamental criterion underlying these challenges and why it matters. How, or at least to what extent, might the empirical rivalries between the various ToCs ultimately iterate to a “true” theory?
Criterion necessary and sufficient for the models of consciousness
Specific models in ToCs describe experiences from a particular point of view, and some of them are less rigorous, especially when they usually originate only from a single empirical finding and reflection. For an early example, inspired by the regular discharge in certain conscious activities, Crick and Koch concluded that the cortical mechanism responsible for gamma oscillations could be the NCC (Koch, 2019, p. 98–100). In addition, the NMDA theory was proposed because the NMDA synapses were found to work along with conscious experience (Flohr, 1992; Doerig et al., 2021).
In most cases, these synapses are all acceptable. Experiments within a theory can still be kept going as long as other diverse theories are set aside. After all, thought experiments of Chalmers could not have “jumped out” as a threat to their independent empirical experiment. Given the comprehensive considerations, N–S (necessary and sufficient), a fundamental criterion for a qualified model of consciousness that is necessary and sufficient. If the description of a ToC meets the N–S criterion, then its model would be entirely identical to consciousness itself. In the empirical test, its predictions of a subject would not be inconsistent with what it indeed is.
However, this intuitive insight would not yet bring us much in practice. In a sense, it is an empirical version of the “hard problem.” Furthermore, we must note that there would be a triple identity of the expected equation “theoretical model = consciousness” from an empirical perspective, corresponding the model to the hypothetical consciousness (T), the data of measurements (M), and the experience itself of the subject (E), respectively (see Figure 1). The most straightforward solution is the traditional philosophical approach from T to E (T~E), which may have been mired in the “hard problem.” The empirical approach to consciousness consequently has to go through the “proxy” of measurement (T–M–E) to the extent that such a complex situation emerges.
Figure 1. The N–S criterion of ToCs. (1) The general: A “proper” ToC specifies a model identical to consciousness, whose “necessity” is to get a fairly “wide” scope covering all the mental activities and whose “sufficiency” is to contain a “narrow” enough scope of those mental activities (see bottom left, ToC, denoted as T). Both the “wide” and the “narrow” are shown by dashed circles, while the filled portion within the solid circles is the necessary and sufficient region. (2) The specific: Similarly, the experience itself (top, denoted as E) and the measurement (bottom right, denoted as M) also get in the proper scope of where the consciousness is located throughout the mental activity. E is considered proper a priori and the identity between T and E (T~E, ~ for a transcendental relation) is the goal of the traditional philosophical approaches. Theoretically, only one (gray-filled) of the four underlying relations between them can meet the N–S criterion. For empirical tests, three levels of the N–S criterion (T–M, M–E, and then T–M–E) are a must and each of level would constrain the corresponding procedure of the ToC tests successively.
Three levels of the N–S criterion
Considering the practical details, the N–S criterion needs to delineate three progressive levels due to the triple identity of consciousness in empirical tests.
1. The candidate theories
A theory of consciousness is completely self-consistent inherently and eligible for experiments. Because of the inappropriate thinking of the proponents or the limited resource or data, the proposals of some theories may have obvious or obscure logical weaknesses or may be contrary to the recognized laws of nature. In such situations, the theories would be first eliminated or need to be improved further. Then, rather than being mysterious, their models must be experimentally measurable. They need to be also precise enough in their description for they must be able to address most aspects (Doerig et al., 2021) of the discussion on consciousness. Through such self-examination, a theory would meet the first level of the N–S, referring to the relation T–M (see Figure 1, bottom). Thus, they qualify to be the candidates and can undergo valid empirical tests with their own assumptions independently.
The first level is usually met relatively easily. Theories that tentatively meet the first level of the N–S criterion may retain their place in the study of consciousness. For example, the Integrated Information Theory (IIT), once with the so called “coarse-graining problem” (Mørch, 2018), has been initially applied to measure the consciousness of patients clinically.
2. Intra-paradigm tests
When a theory meeting the first level has undergone adequate experiments independently, it could be ready for the second level. Here, those candidate theories with the same assumptions have to be tested together as competitors for their experiments with the same paradigms. The relation M–E (Figure 1, right) is referred to by this level in the empirical tests.
There is clearly no generally accepted recognition of what consciousness exactly is. Thus, diverse theories could prefer to have their own choices. For example, the Global Workspace Theory (GWT, cf. Mashour et al., 2020) has fixed it to “conscious access”, while IIT has fixed it to phenomenal consciousness. Whatever their preferences, they must be rigorously held accountable for their choices and make corrections when the empirical findings deviate from the fixed goals. In this case, the paradigm of IIT should try to avoid the interference of typical cognitive functions such as language and working memory. Instead, IIT and the Recurrent Processing Theory (RPT, cf. Lamme, 2006) may be considered a valid adversarial collaboration under the specific paradigm, as both of them are causal structure theories and have asserted no necessity for frontal regions.
Briefly, the constraint imposed by the N–S criterion at this level can be dedicated to making those similar theories compete effectively together. Given a fixed description of consciousness, the predictions of an ideal theory under the paradigm should be kept consistent with all the experimental results.
3. Tests between paradigms
Constraints at this level require the consciousness fixed in its experiments by a theory is indeed the subjects' experience. As illustrated, the total identity T–M–E (Figure 1, top) is referred to by this level. In a sense, it requires transforming the “hard problem” to be a series of hard but promising work (Baars, 2021).
A certain set of theories may remain logical and tested well within their own paradigm, without considering the metaphysics of consciousness or their meta-level. Therefore, these ToCs with diverse paradigms would completely be eligible to start this one and compete with each other after they have met the constraints of the first two levels of the N–S criterion. This is where the rivalry between IIT and GWT could finally come in if they would have both succeeded in their own paradigm. The competitions are indeed some of the preferences to model the consciousness in each theory. In other words, which choice is indeed the most precise—or closer to consciousness itself?
Conclusion
Chalmers' “hard problem” does point out a central problem in the study of consciousness, while his thought experiments are always ignored in empirical experiments. Thus, the N–S criterion would be its concrete, empirically enhanced version and highlights especially the assumptions and paradigms in the experiments. In fact, the criterion and all its specific levels are concluded from various arguments on the constraints of the study of consciousness.
The final level partly reflects the requirement of the paradigm shift toward first-person science (Ascoli, 2013; Ellia et al., 2021; Pinto and Stein, 2021). Studying subjective phenomena can be possible, where we have internal measurements like “first-hand experience” (Kleiner, 2020), which allows scientists to first be used to introspection or meditation. Perhaps, some inspiration may come from the “interhemispheric communication” (Watanabe, 2022) between the referee and the measured variables.
The science of consciousness may transcend the current one, but it is also fine now. Our opinion is that the tests of the various theories of consciousness should not be a one-step kind test. As the level of the fundamental criteria increases in tests, a further theory would be more accurate, at least in the sense of probability and chance. The opinion aims at an initiative for further discussion of constraints in the empirical tests of contemporary ToCs.
Author contributions
ZR conceived the opinion and wrote the entirety of the manuscript.
Funding
This study was supported by the Chinese National Social Science Foundation (20BZX045 and 18ZDA029).
Acknowledgments
I have to thank Hengwei Li and Shiling Cai for their helpful suggestions on my work, and Li's steadfast desire to the problem of consciousness and its theories drove me further into the field.
Conflict of interest
The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
Ascoli, G. A. (2013). The mind-brain relationship as a mathematical problem. ISRN Neurosci. 2013, 1–13. doi: 10.1155/2013/261364
Baars, B. J. (2021). The hard problem is mainly hard work. Cognit. Neurosci. 12, 63–64. doi: 10.1080/17588928.2020.1838474
Damasio, A. (2021). Feeling and Knowing Making Minds Conscious. New York: Pantheon Books. doi: 10.1080/17588928.2020.1846027
Del Pin, S. H., Skóra, Z., Sandberg, K., and Overgaard, M. (2021). Comparing theories of consciousness: why it matters and how to do it. Neurosci. Consciousness 2021. doi: 10.1093/nc/niab019
Doerig, A., Schurger, A., and Herzog, M. H. (2021). Hard criteria for empirical theories of consciousness. Cognit. Neurosci. 12, 41–62. doi: 10.1080/17588928.2020.1772214
Edelman, G., and Tononi, G. (2001). A Universe Of Consciousness: How Matter Becomes Imagination. New York: Basic Books.
Ellia, F., Hendren, J., Grasso, M., Kozma, C., Mindt, G., Lang, P., et al. (2021). Consciousness and the fallacy of misplaced objectivity. Neurosci. Consciousness 2, niab032. doi: 10.1093/nc/niab032
Fleming, S. M. (2021). Theories of consciousness are solutions in need of problems. Cogn Neurosci 12, 86–88. doi: 10.1080/17588928.2020.1841744
Flohr, H. (1992). “Qualia and brain processes,” in A. Beckermann, H. Flohr, and J. Kim, eds Emergence or Reduction. p. 220–238. doi: 10.1515/9783110870084.220
Kleiner, J. (2020). Brain states matter. A reply to the unfolding argument. Consciousness Cognit. 85, 102981. doi: 10.1016/j.concog.2020.102981
Koch, C. (2018). What is consciousness? scientists are beginning to unravel a mystery that has long vexed philosophers. Nature 557, 9–12. doi: 10.1038/d41586-018-05097-x
Koch, C. (2019). The Feeling of Life Itself: Why Consciousness Is Widespread but Can't Be Computed. Cambridge, Massachusetts: MIT Press. doi: 10.7551/mitpress/11705.001.0001
Koch, C., Massimini, M., and Boly, M. (2016). Neural correlates of consciousness: Progress and problems. Nat. Rev. Neurosci. 17, 307–321. doi: 10.1038/nrn.2016.22
Koch, C. (2012). Consciousness: Confession of a Romantic Reductionist. Cambridge, Massachusetts: The MIT Press. doi: 10.7551/mitpress/9367.001.0001
Lamme, V. A. (2006). Towards a true neural stance on consciousness. Trends Cognit. Sci. 10, 494–501. doi: 10.1016/j.tics.2006.09.001
Mashour, G. A., Roelfsema, P., and Changeux, J. (2020). Conscious processing and the global neuronal workspace hypothesis. Neuron 105, 776–798. doi: 10.1016/j.neuron.2020.01.026
Melloni, L., Mudrik, L., and Pitts, M. (2021). Making the hard problem of consciousness easier. Science 372, 911–912. doi: 10.1126/science.abj3259
Mørch, H. H. (2018). Is the integrated information theory of consciousness compatible with russellian panpsychism? Erkenntnis 84, 1065–1085. doi: 10.1007/s10670-018-9995-6
Northoff, G., and Huang, Z. (2017). How do the brain's time and space mediate consciousness and its different dimensions? Temporo-spatial theory of consciousness (TTC). Neurosci. Biobehav. Rev. 80, 630–645. doi: 10.1016/j.neubiorev.2017.07.013
Pinto, Y., and Stein, T. (2021). The hard problem makes the easy problems hard—a reply to Doerig et al. Cognit. Neurosci. 12, 97–98. doi: 10.1080/17588928.2020.1838469
Safron, A. (2020). An integrated world modeling theory (IWMT) of consciousness: Combining integrated information and global neuronal workspace theories with the free energy principle and active inference framework; toward solving the hard problem and characterizing agentic causation. Front. Artificial Intell. 3, 30–30. doi: 10.3389/frai.2020.00030
Seth, A. K., and Bayne, T. (2022). Theories of consciousness. Nat. Rev. Neurosci. 23, 439–452. doi: 10.1038/s41583-022-00587-4
Signorelli, C. M., Szczotka, J., and Prentner, R. (2021). Explanatory profiles of models of consciousness—towards a systematic classification. Neurosci. Consciousness 2, niab021. doi: 10.1093/nc/niab021
Tononi, G., and Koch, C. (2015). Consciousness: here, there and everywhere? Philos Trans R Soc Lond B Biol Sci. 370, 117–134. doi: 10.1098/rstb.2014.0167
Tsuchiya, N., Wilke, M., and Fraessle, S. (2015). No-report paradigms: exracting the true neural correlates of consciousness. Trends Cognit. Sci. 19, 757–770. doi: 10.1016/j.tics.2015.10.002
Watanabe, M. (2022). From Biological to Artificial Consciousness: Neuroscientific Insights and Progress. Cham: Springer International Publishing. doi: 10.1007/978-3-030-91138-6
Keywords: consciousness, models of consciousness, theories of consciousness, IIT, GNWT
Citation: Ruan Z (2023) The fundamental challenge of a future theory of consciousness. Front. Psychol. 13:1029105. doi: 10.3389/fpsyg.2022.1029105
Received: 26 August 2022; Accepted: 14 December 2022;
Published: 12 January 2023.
Edited by:
Frank W. Stahnisch, University of Calgary, CanadaReviewed by:
Adam Safron, Johns Hopkins Medicine, United StatesFlorin Gaiseanu, Independent Researcher, Bucharest, Romania
Copyright © 2023 Ruan. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Zenan Ruan, 
bestfantafr@zju.edu.cn