The meta-problem of consciousness, according to David Chalmers, is (roughly) the problem of explaining why we think there is a problem of consciousness. In his paper, titled ‘The Meta-Problem of Consciousness’, published last year in this journal, David Chalmers did a great service to the field of consciousness studies by laying down a framework for dealing with the meta-problem and with its consequences in a way that cuts across the traditional theoretical divisions on the topic of consciousness. This collection — of which this volume constitutes the first part, the publication of the second part being planned for May 2020 — is devoted to discussing Chalmers’ approach to the meta-problem. It features a series of response pieces targeting Chalmers’ article written by philosophers and scientists, as well as a response to commentators by Chalmers (which will feature in the second of the special issues).
I am very happy to be editing this collection, which I think is likely to bring together new ideas (or new versions of old ideas) in order to help us make progress in the study of consciousness. In this editorial introduction, I will provide an overview of Chalmers’ meta-problem paper as well as the contributions featuring in this first volume.
1. Chalmers’ Meta-Problem
In 1995, in an influential paper published in this journal and titled ‘Facing Up to the Problem of Consciousness’, Chalmers coined the expression ‘hard problem of consciousness’ to refer (roughly) to the problem of explaining how material processes in the brain give rise to conscious experience. He distinguished it from the ‘easy problems’ — problems concerning the explanation of the fulfilment of various cognitive functions by brain processes. Twenty-three years later, in 2018, in the same journal, Chalmers published ‘The Meta-Problem of Consciousness’. The meta-problem of consciousness is the problem of explaining why we (or at least many of us) say and think that there is a hard problem of consciousness: why we say and think that consciousness is particularly hard to explain, and puzzling in various ways. This meta-problem is in principle one of the easy problems: explaining why we say and think certain things about consciousness does not necessarily require explaining consciousness itself and seems to consist entirely in explaining the fulfilment of certain functions.
However, it bears a special relation to the hard problem, which suggests that solving the meta-problem might shed light on the hard problem itself. As is the case with the hard problem, reflections on the meta-problem predate its naming by Chalmers. For decades — if not centuries — philosophers have been trying to explain why we say and think certain things about consciousness (or the mind in general) in a way which does not necessarily require explaining consciousness itself (or the mind) and all its putative peculiarities. The strength of Chalmers’ article is that it brings together ideas from various fields and approaches and discusses them in a unified framework, which I think is likely to prove useful to all. I think and hope it will be instrumental in bringing about the kind of multidisciplinary research work that Chalmers believes is likely to solve the meta-problem — and thus help us make substantive progress in our understanding of consciousness.
In his article, Chalmers first defines the meta-problem. He notably addresses two questions: (a) what is the explanandum corresponding to the meta-problem — what should we explain, in order to solve the meta-problem? (b) What would count as a satisfying explanans in order to solve the meta-problem — how should we explain what we have to explain, in order to solve the meta-problem? Chalmers’ answer to the first question is that the explanandum of the meta-problem consists in problem intuitions. Problem intuitions are defined as dispositions to make certain judgments and certain reports regarding consciousness and its puzzling character. Conceiving of the explanandum of the meta-problem as consisting in behavioural, verbal, and cognitive dispositions is what makes it an easy problem. Chalmers then categorizes problem intuitions (what kind of problem intuitions are there? His list includes explanatory intuitions, metaphysical intuitions, knowledge intuitions, etc.) and discusses their strength (how robust are they?) as well as their distribution (who exactly has these intuitions?). He presents and discusses some of the empirical work that has been done on these topics and ends up embracing the working assumption that these intuitions are robust and widely shared (across individuals, cultures, etc.). As for Chalmers’ answer to the second question, it notably requires the explanans of these intuitions to be topic-neutral — that is, to exclude terms mentioning consciousness. Solving the meta-problem of consciousness thus means explaining our (supposedly widely shared) problem intuitions about consciousness in a way that does not mention consciousness. It might be that there can be no such solution: in that case, one would have to explain why exactly there can be no such solution.
Second, Chalmers focuses on potential solutions to the meta-problem. He presents a number of candidate solutions that have been put forward in the pre-existing literature (quite often in combination) by various philosophers and scientists, and points to the strengths and weaknesses of each of them. He starts with seven approaches he finds particularly promising: (1) introspective models, (2) phenomenal concepts, (3) independent roles, (4) introspective opacity, (5) immediate knowledge, (6) primitive quality attribution, and (7) primitive relation attribution. He then describes six other kinds of approaches that he takes to be somewhat less promising: (8) introjection and the phenomenological fallacy, (9) the user illusion, (10) the use-mention fallacy, (11) underestimating the physical, (12) historical and cultural explanations (other approaches are grouped under a last ‘other proposals’ category). Chalmers also presents the solution he favours, a combination of the seven first candidate solutions, which gives pre-eminent roles to approaches (5) and (7).
In the rest of the article, Chalmers examines the way in which dealing with the meta-problem could shed light on the hard problem, and more generally on consciousness itself. He starts by explaining how a solution to the meta-problem should constrain realist theories of consciousness. He explains that we should reasonably expect a correct realist theory of consciousness — one that takes at least most of our central intuitions about consciousness to be correct — to show how the mechanisms that are responsible for consciousness itself also play a central role in explaining our intuitions about consciousness. Chalmers shows that this is no trivial task: the meta-problem challenge thus arises for major theories of consciousness (integrated information theory, global workspace theories, first-order representational theories, higher-order thought theories, quantum theories, panpsychist theories, etc.).
Chalmers then goes on to describe six stances one can take regarding the meta-problem and consciousness itself. The first one (‘meta-problem nihilism’) states that there is no solution to the meta-problem of consciousness: there is no topic-neutral explanation of our intuitions about consciousness. The five other reactions consist in admitting that there is a solution to the meta-problem, but each of them takes a different stance on the relation between consciousness itself and the meta-problem processes (the topic-neutrally characterized processes which supposedly explain our intuitions about consciousness).
The second and third positions are compatible with a robust form of realism regarding consciousness (i.e. with the correctness of most, if not all, of our central intuitions about consciousness): one states that consciousness correlates with the meta-problem processes (‘correlationism’), the other that consciousness realizes meta-problem processes (‘realizationism’).
The last three positions all require the rejection of a number of central intuitions about consciousness. This means, according to Chalmers, that they require an element of illusionism regarding consciousness (as they state that consciousness is not as we intuit it to be). The fourth option bluntly denies the existence of consciousness (‘strong illusionism’), the fifth equates consciousness with lower-order meta-problem states (the states that are the usual target of meta-problem processes), and the sixth equates consciousness with higher-order meta-problem states (states that consist of the activation of meta-problem processes, themselves targeting lower-order states). Chalmers takes options five and six as leading very naturally to a form of weak illusionism. According to weak illusionism, consciousness exists, but does not have many of the important features we intuitively think it has: for example, it is not a primitive, non-physical, non-functional process, it is not true that it is fully knowable only in the first-person, etc.
The paper then focuses on illusionism regarding consciousness and on the opposition between illusionism (weak or strong) and realism. Chalmers first shows that the thesis stating that there is a solution to the meta-problem can lead to the formulation of debunking arguments for illusionism. His idea is that if we have a correct explanation of our intuitions regarding consciousness, and if this explanation is independent of consciousness itself, this undermines the justification of our intuitions regarding consciousness (alternatively: this makes it so that their truth would be a mere coincidence). Chalmers examines ways for a realist to answer this argument, and concludes that none of them is fully satisfying at this stage — although he notes that the most promising ways might lie with a form of realizationism, in which consciousness is crucial for realizing meta-problem processes, so that the truth of our intuitions regarding consciousness is not a mere coincidence. The debunking argument nevertheless puts some serious pressure on realists.
Chalmers then compares weak and strong illusionism. He argues that only strong illusionism can help us in making progress with the hard problem of consciousness — by dissolving it. Such dissolution, on the other hand, is not a genuine possibility for weak illusionists. Weak illusionists might be right, but their weak illusionism in itself does not give them a way to avoid the hard problem — they face it just as much as realists do. If one is concerned with avoiding the hard problem, then, strong illusionism fares better than weak illusionism. Only strong illusionists can claim that solving the meta-problem of consciousness eliminates the need to solve the hard problem.
Finally, Chalmers presents a ‘Moorean’ argument against strong illusionism, which uses as its crucial (and ‘obviously true’, according to Chalmers) premise that people sometimes feel pain. He then describes the dialectics between the illusionist and the realist in the form of a schematic dialogue leading to what seems to be a stalemate. Chalmers concludes that, even though the illusionist position is dialectically more interesting, the realist position is the right one.
Chalmers finishes with a description of the task at hand for illusionists and realists. Illusionists, he says, should find a way to explain why, even though we are not phenomenally conscious, things are like this (where ‘this’ ideally captures something about our mental lives which is more than mere dispositions to judge and to speak, but is not phenomenal either). Realists, he says, should manage to explain how consciousness and meta-problem processes are intertwined, in a way that makes it clear that the truth of our intuitions regarding consciousness is not merely coincidental. Research towards a solution to the meta-problem of consciousness, however, cuts across the distinction between illusionists and realists (or between physicalists and anti-physicalists, as well as, incidentally, between scientists and philosophers). As Chalmers puts it, ‘the meta-problem is a potentially tractable research project for everyone’.
2. The Meta-Problem of Consciousness
Special Issue: First Part
This first part of the special issue features 22 response pieces by various authors — philosophers and scientists. These contributions constitute a rich and fascinating set of responses, objections, remarks, and suggestions regarding the meta-problem of consciousness. I will now attempt to give a brief presentation (together with a rough categorization) of these pieces — even though it is naturally impossible to do full justice to these contributions in a few pages.
2.1. The explanandum of the meta-problem
A number of pieces focus on the definition of the meta-problem, and more particularly on the characterization of the explanandum corresponding to the meta-problem: problem intuitions. Two pieces criticize Chalmers’ view of problem intuitions from the point of view of empirical research. Justin Sytsma and Eyuphan Ozdemir use the methods of experimental philosophy and present new data, which suggest (in agreement with some previous studies) that most laypeople do not have a unified concept of phenomenal consciousness (a concept that would notably encompass perceptual experiences and pain experiences). They argue that this shows that problem intuitions cannot be really widespread amongst lay people (pace Chalmers). Anna Wierzbicka tackles a closely related question from the point of view of linguistics. She casts doubts on the widespread character of problem intuitions — and more particularly on their cross-cultural nature — by showing that Chalmers articulates these intuitions using terms such as ‘consciousness’, which, according to cross-linguistic data, do not find equivalents in all human languages. She argues that, if we purport to express potentially universally shared intuitions about our minds, we should use, instead of parochially English terms such as ‘consciousness’ or ‘experience’, universally cross-translatable terms such as ‘think’, ‘know’, and ‘feel’.
The characterization of the explanandum corresponding to the meta-problem — problem intuitions — is also the focus of three conceptual pieces. Elizabeth Irvine examines the consequences of potential variations in problem intuitions. She argues that the (plausible) existence of variations in the intuitions people have regarding consciousness (and in the psychological weight they give to such intuitions) generates important constraints on solutions to the meta-problem, that Chalmers tends to ignore — and which might create a problem for the views Chalmers deems promising. David Papineau sees a tension in Chalmers’ approach — between the view that we have a number of irreducible ‘problem intuitions’ about consciousness, and the view that these intuitions reduce in the end to a derivation gap between the physical and the phenomenal. He also argues that, once this tension is recognized, ‘orthodox a posteriori physicalism’ (as opposed both to strong illusionism and to anti-reductionist realism) becomes more attractive than what Chalmers claims. David Rosenthal casts doubt on Chalmers’ idea that problem intuitions are widespread and pre-theoretical. He argues against the view that we should require theories of consciousness to provide (even indirectly) some kind of explanation for the arising of these intuitions. He also rejects the idea that realism about consciousness should be defined by the endorsement of these problem intuitions — unless we provide reasons to think that these intuitions are precisely true of consciousness itself.
2.2. Solutions to the meta-problem
Many pieces in this special issue focus on presenting or defending a particular solution to the meta-problem of consciousness (sometimes conjoining this solution with a recommended reaction to the meta-problem). Andy Clark, Karl Friston, and Sam Wilkinson present a tentative solution to the meta-problem, using the predictive processing framework. They see phenomenal states as kinds of inferred states (playing a certain role in the prediction of future sensory input). These phenomenal states correspond to ‘mid-level’ inferences (in the hierarchy of inferential processes that constitutes the mind). This mid-level location, when considered together with adaptive pressure, might explain, according to the authors, why phenomenal states are represented as states about which one has quasi-certain knowledge, and why they end up being seen as irreducibly special and puzzling. This solution leads the authors to a position regarding consciousness itself which is intermediate between illusionism and realism. Gary Drescher identifies a series of confusions that we very naturally make regarding consciousness, because of the way in which we usually represent our own conscious states (and the conscious states of others). These confusions make the hard problem seem hard. Resolving these confusions, on the other hand, allows us to solve the hard problem, while vindicating at the same time the Type-B materialist view of consciousness that Drescher favours. Brian Fiala and Shaun Nichols defend a solution to the meta-problem, according to which an intuitive (but merely apparent) explanatory gap regarding consciousness emerges because of a discrepancy between our third-personal descriptions of brain states and our first-personal intuitive ascriptions of consciousness. These first-personal intuitive ascriptions consist in applications of primitive representations and are triggered only by a restricted set of inputs. They claim that this kind of explanation solves the meta-problem of consciousness, and might be generally applicable to other apparent explanatory gaps arising about other entities (intentionality, normativity, causation, etc.). Michael Graziano presents his attention schema theory, and argues that it solves the meta-problem of consciousness. In his view, the brain constructs a schematic model of attention, which discards as irrelevant the physical and mechanistic aspects of real attentional processes. This leads us to think that we enter into ‘ethereal’ and puzzling subjective states devoid of physical properties (with the important exception of location). Nicholas Humphrey presents his view of consciousness: according to him, consciousness should be understood as an intentional object: consciousness is how privatized ‘sentition’ (reflex motor responses to stimulation) is represented. Making consciousness an intentional object allows us to understand how it can have all kinds of properties that nothing physical could have, which is why we face the hard problem when we try (hopelessly) to locate consciousness in the physical world. Consciousness does not exist at the level of physical reality. However, says Humphrey, this does not mean that it is an illusion. Moreover, the representation of consciousness corresponds to our normal cognitive functioning, which has been fostered by natural selection. François Kammerer (the author of these lines) proposes an evidential approach to the meta-problem of consciousness. He argues that, in order to solve the meta-problem of consciousness, we need to understand how our introspective representations of phenomenal states and our naïve epistemological representations of evidential situations — or data — are deeply intertwined. He presents a speculative view of introspection and naïve epistemology (according to which phenomenal introspection relies on an ‘evidence-by-resemblance’ mechanism) which he thinks solves the meta-problem, and suggests conjoining it with an illusionist view of consciousness. Hakwan Lau and Matthias Michel put forth a socio-historical take on the meta-problem. They suggest that the persistence of the intuition that consciousness is hard — if not impossible — to explain might stem from the scientific weakness of the field of consciousness studies. They claim that the field tends to allocate too much attention to ‘gurus’ who claim that consciousness is mysterious. This prevents good scientific research from being carried out concerning the real explanans of consciousness; this in turn reinforces the intuition that consciousness is mysterious. Derk Pereboom draws on some of his past work, in which he developed an illusionist account of phenomenality (the ‘qualitative inaccuracy hypothesis’). He proposes a two-tier explanation of why we have a persistent and robust intuitive resistance to illusionism (even though illusionism is, to a certain extent, true), which allows him to defend his conception against the objections mentioned by Chalmers in the target article. Bradford Saad develops a form of interactionist dualism on which consciousness realizes meta-problem processes. In his view, experiences cause problem judgments, and such causation relies on fundamentally teleological laws, which relate normative features of experiences and the content of problem judgments. He thinks that this allows the realist to say that problem intuitions are generally correct and that this correctness is not a mere coincidence: we judge that consciousness has a peculiar nature because it does. Wolfgang Schwarz draws on previous work and defends the view that phenomenal states are states belonging to an imaginary dimension, which some cognitive systems (such as ours) represent — as it helps them update their world model in an efficient way. These systems thus have certainty about certain aspects of these imaginary states. Propositions about these states, in turn, will never logically imply anything about the real world. Schwarz claims that this explains the persistently puzzling character of phenomenal states. He combines this account with a view of consciousness which is neither illusionist nor realist: in his view, there is a strict philosophical sense in which ascriptions of phenomenal states are neither true nor false. Galen Strawson, drawing on previous work, defends a solution to the meta-problem according to which consciousness seems uniquely puzzling and hard to explain only to those who have a (wrong and unjustified) conception of the physical according to which consciousness cannot be physical. Once one recognizes that nothing (and notably nothing in physics) forces us to embrace this conception, one can accept the view that consciousness is in fact fully identical with the physical — a thesis which is not uniquely puzzling (not more puzzling, say, than wave–particle duality).
2.3. Reactions to the meta-problem
Other pieces focus mainly on reactions to the meta-problem. Justin Clarke-Doane examines the debunking argument for illusionism discussed by Chalmers. He compares it to similar debunking arguments regarding mathematical knowledge and moral knowledge, and he claims that a solution to the meta-problem would not undermine our belief in consciousness (unless we accept very strong premises, that lead to general scepticism). Not all consciousness realists will rejoice in this, however, as he also points out that the same line of reasoning implies that modal arguments against physicalism have little force. Daniel Dennett explains how Chalmers’ perspective should now lead him to embrace strong illusionism, a position Dennett himself has defended for many years. He also indicates which steps researchers (including Chalmers himself) should take to move forward on solving the meta-problem of consciousness, and why these steps would most naturally lead to strong illusionism about consciousness. Keith Frankish argues for strong illusionism, by showing that maintaining our belief in consciousness in the face of a solution to the meta-problem raises the hard meta-problem (the problem of explaining how consciousness realists such as Chalmers could have the kind of robust direct access to consciousness required to ground their Moorean confidence in the existence of consciousness). Strong illusionists, on the other hand, can avoid this hard meta-problem: for them, the meta-problem is the only problem of consciousness. Joseph Levine recapitulates the difficulties that arise for materialists and non-materialists when they want to unify their views of what consciousness is (or is not) and their views of how we know of it. He ends up defending a form of non-reductionist strong emergentism. He then wonders how, within such a view, it is possible to give a crucial role to consciousness in the generation of our judgments about it. He sketches speculative solutions to this problem, which rely on abandoning the thesis that the non-identity of two contingent things entails their possible decorrelation, and giving a central role to cognitive phenomenology. Lisa Miracchi argues that the distinction between the hard problem and the easy problems is misguided, as it rests on an inaccurate view of scientific explanation. Once we recognize this, we can also understand that solving the meta-problem (supposedly, an easy problem able to shed light on the hard problem) is unlikely to have the kind of consequences described by Chalmers. Helen Yetter-Chappell argues that some solutions to the meta-problem of consciousness (the ones that make problem intuitions relevantly contingent) might undercut the dialectical force of arguments against physicalism regarding consciousness. Indeed, these solutions to the meta-problem give support to the idea that there is no theory-neutral conception of the notion of ‘idealization’ (while such a conception is required in order for standard anti-physicalist arguments to take off).
I want to thank Valerie Hardcastle (Editor-in-Chief of the Journal of Consciousness Studies) and Graham Horswell (Managing Editor of the Journal of Consciousness Studies) for their invitation to edit this special issue, as well as for their support and their help. I want to thank David Chalmers for his incredibly valuable help in editing this volume. I received very useful input and advice regarding editorial work from Keith Frankish, as well as useful comments on this introduction from Ismahan Wayah and Sonia Paz-Higgins. I want to thank them all here. Finally, I am very thankful to all the authors participating in the special issue, as well as to the reviewers who helped the authors improve their papers.