The mind-body problem

The relation between \(\mathbf{1}\) and \(\mathbf{2}\), once rid of all the theoretical apparatus we’ve just developed, ultimately outlines an ever-central question of philosophy, in every era and every civilization, that’s often called the mind-body problem. Are our mind and body distinct in nature, or are they the same? And if they are the same, what is that sole fundamental substance like?

\(\{\mathbf{1}\}\) : eliminative materialism

This position provides a simple answer to the mind-body conundrum: that mental states don’t exist This goes quite a bit further than traditional materialism, which we’ll consider next, and merely asserts that mental facts can be wholly derived from, or even identified with, physical facts.

What’s the difference? Proponents of eliminativism typically describe mental states as part of "folk psychology". Broadly speaking, this folk theory compiles the "laws" we all unwittingly use in our daily life to explain and predict behavior; for instance: "if Alice smiles and waves at me, then she knows who I am", or "if Carl yawns, then he feels tired or bored". In so doing, eliminativists deny any ontological import to mental states: they see belief, desire, fear or self-consciousness as akin to the bodily humours from ancient medicine, or to the demons that supposedly were the cause of mental disease: "common-sense" concepts (at least at some point in time) that are by no means devoid of explanatory power, but that also carry no ontological import. To use a popular image (popular among ontologists, I should add...), they argue that folk psychology does not carve reality at the joints. Taking another example, Newton’s concept of a gravitational field is extremely powerful from a methodological perspective, but since it is now assumed that Newtonian gravitation can be fully reduced to Einstein’s general relativity, there is no sense in which Newton’s gravitational field is an ontological brute fact. This is close to how eliminativists would look at mental states.

Historically, eliminative views have been carried by proponents of radical behaviorism. Folk psychology generally seeks to causally model how physical facts (sense-data, stimuli...) lead to certain mental states, which in turn lead to specific behaviors. Why then talk of an elusive mind and its properties when we can get rid of that black box and only focus on what’s observable, namely the stimuli/behavior complex? Pains, fears, beliefs, then become mere dispositions to behave in a certain way.

Beyond this Socratic undermining of our deeply held conceits, there are of course unsettling consequences. Descartes doubted everything, but at least he believed in his own mental states. Ultimately, it seems that denying beliefs as sensible properties of human minds would pose a serious challenge to the mere possibility of ethical or political discourse.

Furthermore, eliminativism appears to me as representing a very unstable pole in the philosophical state-space. As most radical positions, it has its uses as a way to shake up the status quo. But Feyerabend’s contention that folk psychology will prove to be completely false can hardly help but lead to a form of mental nihilism that runs against one’s own experience. Between this and physicalist approaches that identify mental states with physical states lies a spectrum of possibilities. Those would include views that folk psychology might only be partly false, or incomplete, or approximate, but that large swathes of it may end up being vindicated in some form. I called the radical version of eliminativism an unstable pole, because these weaker versions, sometimes called "revisionist materialism" seem bound to slide towards physicalism. They’re built on the promise that the "correct theory of mind" won’t include anything like our mental states, but don’t really make a case why the basic bricks of this future theory cannot be successor entities to those, thus eliminating eliminativism.

\(\{\mathbf{1}\mathbf{\succ}\mathbf{2}\}\) : physicalism

Simply put, physicalism is the idea that everything is physical. In the context of the mind-body problem, this means that mental states can be derived from physical states. Contra eliminativism, this involves accepting mental states as part and parcel of our ontology, even though they are not ontological primitives and can be fully explained by physical facts. In line with our discussion of inter-level metaphysics, we’ll conventionally assume that "real" physicalism shall at least imply the grounding of mental facts in physical facts. Supervenient physicalism ("no mental change without a physical change") without grounding is ambiguous, and compatible with many forms of dualism.

Most staunch physicalists defend some form of reductive approach, which in itself is sufficient for grounding, as we’ve seen. Sure enough, the concept of reduction looms large, since the explicit goal of modern cognitive science is to reduce psychology to neurobiology. Philosophers’ ideas have at times evolved in step with the latest developments in that endeavour, and I’d like to describe some of the main strategies they’ve outlined.

The identity theory

holds that mental states and processes are identical to physical states and processes. Since most discourse about the mind is centered around human beings, or more broadly animals, these physical states are generally brain states (hence the concept is often referred to as mind-brain identity). Naturally, this general conceit has taken hold as knowledge about the brain has increased through the twentieth century. I try not to harp on about terminology too much, but in this case it seems important to acknowledge how poor of a fit the prevalent term of "identity" is. Mental states cannot be strictly identical to physical states: my leg pain may be dull and diffuse, but the corresponding brain state is neither. Having different properties is precisely what allows mental states the ontological difference that is a rebuke of eliminativism, even as they could be fully derivative of physical states.

A weak version of this thesis is often called token identity: it merely states that a particular mental state can be matched with a particular brain state. Taken literally, this can be seen as an utter truism that’s even compatible with full-blown dualism: the state of one Cartesian soul at a moment in time can be "identified" with the brain state of the soul’s owner at that moment. At least, token identity seems to exclude the possibility of disembodied spirits and other brain-less mental states, which is already something, I suppose.

Conversely, the strongest version of the thesis is type identity: the idea that types of mental states correspond to types of brain states. For instance: all mental states that share the property of being a leg pain would correspond to brain states that share a specific physical property (such as: a specific region of the brain is activated in a certain way). A complete account of type identity theory seems unlikely to materialise, but should it ever be attained, it would likely have achieved a full bridge reduction of psychology to neurobiology: observing a brain state would then allow to fully deduce via bridge law the corresponding mental state.

Functionalism

is the doctrine that mental states are not characterised by what they are ("identical" to physical states), but in terms of what they do, of their function in the cognitive system. This is of course reminiscent of functional reduction, and for good reason: that concept was developed as a generalization of mental functionalism, which came first. It will not be a surprise that this idea emerged in the 1950s, just as modern computer science found its footing, and with it the first serious considerations of artificial intelligence. The notion that future computers or hypothetical androids may share at least some types of mental states with humans seemed ruled out by identity theory, which relied on physico-chemical explanation, but were a perfect fit for functionalism. This multiple realizability of mental states acts as a double-edged sword of course: if we could gather as many humans as there are neurons in a brain, and each human node was sending some electrical signals to his neighbors, following the exact pattern of a brain in a certain mental state, would that collective be experiencing a mental state in any meaningful sense? It would be functionally equivalent to a human brain, but many would deny there is any form of phenomenal experience there.

I believe it was useful to contrast these two doctrines, given their omnipresence in the literature, however I’m not convinced the distinction has much use in terms of sharply contrasting physicalist views. Isn’t a functional role a type of physical state? The "type" in type identity is often left vague enough to accomodate many approaches, so it would seem peculiar to answer by the negative in this case. It is often argued that type identity is first-order, while functionalism is second-order (functions being "types of types"). In that case it still seems the two are sub-cases of a reductive physicalism rather than antithetical positions. Depending on the granularity of types, even multiple realizability could end up being compatible with forms of identity theory.

As is evident from the continuum we described from reduction to grounding, strong emergence and supervenience, there are versions of physicalism, real or ambiguous, that do not seek to nomologically explicate the links between mental states, brain states and behavior.

On a last note about physicalism, it is as close to a received view as can be in the year 2021. As a result, the standard of proof is often lower for proponents of physicalism than for its opponents. In addition, confusion often arises from all the ancillary theses that physicalism carries with it as part of a more holistic world-picture: methodological naturalism, scientific unitarism, empiricism, nominalism, atheism, may often find themselves a part of the same "package deal", but none of those is necessary nor sufficient for physicalism.

\(\{\mathbf{1}\mathbf{\parallel}\mathbf{2}\}\) : mind-body dualism

Our starting point in this chapter was to identify realms of seemingly distinct ontological nature. If physical entities and mental states seem different on the surface, the position that they are different, that none can be fully derived from the other, should be the default.

Predicate dualism

, the weakest dualist thesis, simply states that mental events cannot be described using only physical terms. In other words, even if we successfully reduced mental states to brain states, language about the mental (beliefs, desires, pains,...) could not be "reduced" in the same way, and that idiosyncrasy would stay essential to psychology. Clearly, predicate dualism is not at all incompatible with physicalism. However, it is incompatible with eliminativism; as a matter of fact, the key idea of eliminativism — that the concepts of folk psychology will ultimately be dispensed with — is precisely the negation of predicate dualism.

Property dualism

goes much further by saying that mental states and physical properties are ontologically distinct, even though they both apply to the same (physical) entities. The fact that I am happy or sad can then not be fully grounded in the physical attributes of my brain or of my body in general. Aristotle’s view of the soul as the form of a human body (the same way its shape is a form that makes a bronze object into a statue) exemplifies this idea. We’ve defined the outer border of "real" physicalism at the level of a grounding relationship between the mental and the physical. Now I’ll go further and say where real physicalism ends, property dualism starts. It is quite clear that reduction is incompatible with property dualism, since it either identifies mental states with physical brain states outright, or with functions thereof, which can then be expressed in terms of physical properties. Grounding alone is weaker than reduction, as we’ve seen, but its metaphysical nature should also suggest that the derivation at hand implies a form of monism. Conversely, property dualism is compatible with many forms of non-reductive, non-grounded, "ambiguous" physicalism. For instance, it is possible to be a property dualist, yet believe that the mental supervenes on the physical (obviously, it is not necessary).

On that note, it is worth mentioning that the contemporary prevalence of physicalism/property monism about the mind has dulled the edge of that position. Only one century ago, property dualism about life was still standard: it had been established long before that living matter was not different in its nature from inorganic bodies, yet "vital force" or some other biological attributes seemed indispensable to explain the unique properties observed in living beings. Today, this has mostly devolved into a predicate dualism about life: even though it may never be described via a neat physico-chemical criterion, it is believed that life is susceptible to a full physical explanation. This is one of the last steps in the progressive gathering of properties under the wide umbrella of today’s unitarian scientific project. The fact that biological and chemical properties can all be uncontroversially placed within the same realm as physical ones was far from obvious for a long time.

Substance dualism

is in turns an even stronger position: it refuses the concept of mental states as properties of living bodies (or other physical realizations), nor does it consider "the mind" as an amorphous bundles of these mental states. Instead it posits the existence of a specific substance that is ontologically distinct from material entities, and of which mental states are properties. Plato calls it psuchē and Descartes âme — the soul —, which brings a host of spiritual connotations that are not integral to the concept. One feature of substance dualism that was certainly attractive to both these thinkers is the allowance for disembodied mental entities: a soul and its mental properties do not a priori need a physical host. (It’s interesting to note that more unusual versions of substance dualism exist, such as E.J. Lowe’s distinction between body and person). Generally speaking, substance dualism in its most general guise is not compatible with supervenient physicalism, nor with other types of ambiguous physicalism.

The problem of interaction

Now, it looks like dualism should be the default position: physicalism is a thesis that relies on the promise of mounting scientific evidence, again following the trail that was blazed by biology — a daunting endeavor that advances in neurobiology and cognitive sciences have only partially proved equal to. Dualism is often on the defensive though, and it’s facing one problem in particular: that of interaction. It is indeed asking for trouble to insist on the ontological distinctness of two categories, whilst all of our day-to-day experiences make it look like they’re constantly interacting: surely, my feelings are borne out of what I see and hear, and those feelings lead to specific behaviors and bodily reactions. How can mind and body influence each other thus if they are so fundamentally different?

Interactionism accepts those intuitions from folk psychology and recognizes that our minds and the physical world are influencing each other. There is no contradiction a priori, but a host of difficulties, to do with principles like the causal closure of the physical universe or locality, or with laws such as the conservation of energy, or the second law of thermodynamics.

Epiphenomenalism holds that physical states can influence mental states, but that the reverse isn’t true. As Huxley had it, this makes mental events similar to the steam that accompanies a locomotive, but doesn’t contribute to its motion. One common way to support epiphenomenalism is to distinguish between aspects of the mind: one can be a reductionist about intentional states, but see phenomenality as a distinct byproduct of neural processes. This class of views is often conflated with forms of non-reductive, supervenient physicalism. Simply put: assuming that the pain I feel from stubbing my toe is caused by some specific neurons firing, then my reaction (wincing, yelping...) is not caused by that feeling, it is a mechanical consequence of the initial neuronal activity. Epiphenomenalism has the advantage of being less disruptive to the laws of physics since it can be seen as an "add-on". Nevertheless it doesn’t avoid the same fundamental problem as interactionism, since the mode under which physical systems can influence ontologically remote mental states still seems quite mysterious. In addition, epiphenomenalism has severe problems of its own, like the apparent redundancy of the mental: if consciousness cannot influence our behavior and if philosophical zombies (mindless replicas of ourselves) could exist, then what is the point of consciousness? In evolutionary terms, pain is often considered as an advantage for species whose members will then avoid injury. If the same behavior could occur anyway as a form of painless reflex, what’s the point of pain? Not that it needs to have a point, but if causal roles are out of the picture, as is evolutionary emergence, then only sui generis teleological explanation remains (e.g. that God created pain for a specific purpose). Without that, epiphenomenalism could be labelled an ontologically profligate doctrine.

Parallelism avoids the problem of interaction entirely by saying there is none. Since it is nevertheless a fact that mental and physical events are correlated, some sort of justification must be found. Leibniz had it that God had created a pre-existing harmony between the mental and physical worlds that ensured both stream of events would run parallel. Malebranche held that God was the cause of all events anyway. Even if we were to admit God in our ontology, that seems like a bit of a cop-out. If a deity is acting as a bridge between the two realms of our dualist universe, then isn’t it just disguised monism?

One last problem, in the same vein, is specific to substance dualism: if mental states are properties of a mental substance which could, in principle, exist independently of any material entity, how should we explain that souls (at the very least, the ones we’re aware of) are tethered to specific material beings. This is similar in nature to the problem of interaction, and can be seen as a special case thereof, whils adding another layer of puzzlement.

\(\{\mathbf{1}\mathbf{\prec}\mathbf{2}\}\) : idealism

I’ll call idealism the converse of physicalism: the doctrine that physical facts are grounded in mental facts. In other words, it is the thesis that our world is mental at its core.

The term itself is historically charged, to say the least. Such prominent philosophers have been labeled, by historians or by themselves, as idealists, that the concept itself is now quasi-synonymous for their doctrine, with some confusing results: some of them were not idealists in the sense I’ve defined above. Here I’ll try to ignore these historical contingencies and to start from the beginning, in a treatment that owes a lot to David Chalmers’s brilliant work.

First of all, it’s important to clarify what we’re talking about: the motivation for idealism is often semantic — that any statement about the physical world can be analyzed into a statement about the mental —, or epistemic — that our knowledge of the physical world only exists in our mind. However those theses do not entail the sort of ontological doctrine which I’ve just defined. Kant’s transcendental idealism qualifies as an epistemic form of idealism, but not as an ontological one, since he introduced the thing in itself, which is not knowable, but is existent.

Macro-idealism

So, if everything is grounded in mental facts, we need to be clear about what those are again. So far, we’ve described mental states as attributes of human minds. This can be extended to at least some animal minds, and to the minds of hypothetical extra-terrestrial beings. The resulting version of idealism holds that everything physical is derived from the mentality of macroscopic beings. Following Chalmers, we’ll call this macro-idealism. Which macro-mental states ground physical entities? Those are generally assumed to be perceptions, or sense-data. (It is not strictly inconceivable that the physical could derive from non-perceptive states, but nobody ever mounted a credible defense of that position, to my knowledge).

This general idea has been around since ancient times, in particular within Indian philosophy, but has taken its modern form in the West through the work of Berkeley. One could argue that Berkeley’s theory, which is often labelled as subjective idealism and which he himself called immaterialism is a natural progression from dualism. Parallelists and occasionists had sidestepped the problem of interaction between matter and soul by positing the mental and the physical were on parallel tracks, mirroring each other but never crossing. Berkeley’s instinct was that once interaction is ruled out, one is best advised to do away with the physical world as a separate entity altogether.

Simply put, the fact that I see a book on the table is ontologically prior to the book existing as a physical entity in the universe (whilst a materialist — and truly almost anyone in the world — would say the reverse is true). It is natural to feel incredulous: Samuel Johnson, to expound his own incredulity towards this theory, famously kicked a big stone with mighty force, claiming: "I refute it thus!". Two questions immediately come to mind. First, what to make of imaginings, dreams or hallucinations? They can be mentally identical to perceptions, yet I want to draw a clear line between the book in front of me now and a book I’ve seen in my dreams. Idealists will typically invoke some form of coherence between perceptions, which often extends to a form of intersubjective worldview, to distinguish physical objects from illusions. A second question is about unperceived parts of the universe: even the staunchest idealists wouldn’t deny that objects can exist and persist even if no mind can perceive them. The Berkeleyan way to answer that is to appeal to God, who can of course perceive everything, and in that way fill the gaps from human experience. This actually shows Berkeley was not what we defined as macro-idealist, since he accepts God’s perceptions as ontologically fundamental — and a supreme being is clearly not "macro". The alternative emerged in the nineteenth century under the name of phenomenalism, as epitomized by Ernst Mach, or later by A.J. Ayer. The phenomenalist answer to the problem of gaps is that the existence of physical entities is grounded in the continued possibility of experience. A materialist would say that if I open the bedroom door, I do see the bedroom because it bedroom exists independently of me anyway. A phenomenalist would say the flux of sense-data characteristic of my opening that door is always followed by the sense-data associated with perceiving that room.

Of course, it is easy to see how this view can be unpalatable to many. One problem is to what extent it is solipsistic: could there be a world if there were no minds? What is the ontological status of objects and events ten billion years in the past, before there were any macro-minds (at least to our knowledge)? Is the mere possibility that macro-mental beings may one day evolve and that if they had been present they could have perceived these events enough to make them "existent"? Or is it solely our indirect perception of them today — via complex instruments and theories — that grants this existence? The question might not even be sensible in the context. Another significant issue is that of explanation. Looking at a table, I perceive a certain image. Walking to the other side of the room and looking again, I see a different image, which I still recognize as a table. To a non-idealist, the relation between those appearances is no mystery: the table is a single three-dimensional object, as are my eyes, and to predict and explain the resulting appearance is a simple matter of projective geometry. To the phenomenalist, each possible experience is a fundamental fact, and the relation between them is dispositional rather than categorical: this prevents the recourse to strict scientific laws in the form we are used to, and makes for an enormously complex worldview.

Pan-idealism

An alternate view of idealism can arise from the doctrine of panpsychism: the idea that everything has a mind. Largely through misunderstanding, we would argue, panpsychism is often considered a somewhat fanciful doctrine. It is true that an assertion like the one I’ve just made can lend itself to confusion. Panpsychism does not suggest that nothing is truly inanimate (that’s hylozoism), nor that inanimate objects have feelings, perceptions, pains, like humans have (that’s animism). It’s also different from pantheism, the religious doctrine that identifies the physical universe with God. Rather than saying that everything has a mind, it would then be more evocative to say that everything has mental qualities, or perhaps, a mentality. As such, that is of course vague and unsatisfactory — we will see why that route is still enticing and what can be done to make it a more concrete theory.

But first, we need to bring the focus back on our topic of idealism. That everything is mental does not preclude that everything may be physical as well, so panpsychism is entirely compatible with dualism. Pan-idealism would stipulate that mental facts ground physical facts, but unlike macro-idealism, it wouldn’t assume those mental facts are solely about the mentality of humans, animals, and other macro-beings.

Often, one special class of mental facts grounds everythings else (that is similar to the current doxa of materialism: assuming everything is physical, not every physical thing is equally fundamental — facts about the entities described in the Standard Model and in General Relativity ground all others). The prevalent form of pan-idealism, which we’ll call, following Chalmers, micro-idealism, states that everything is grounded in the mentality of micro-entities, often taken to be the elementary particles of fundamental physics. To put it bluntly, this means that an electron or a gluon shares with me some form of subjective mental quality. The first response to this unverifiable, unfalsifiable hypothesis is usually an incredulous stare.

Yet, if we can move past our initial puzzlement, it’s arguable that micro-idealism could combine theoretical virtues from competing doctrines, and avoid some major drawbacks. First, micro-idealism benefits from the main dividend of idealism in general: that of avoiding the problem of consciousness: reductive physicalists aim to explain how phenomenal properties are grounded in the physical, but in so doing disregard some of its characteristic features. The concept of emergence has been built to provide a robust response to such concerns, but to some, they’re quite insurmountable. For instance, emergence provides no satisfactory answer to the question of continuity: where does phenomenal experience start? Can we draw a clear line? This synchronic puzzle also arises in the context of evolutionary theory: what’s the one mutation that "makes" phenomenality? All this exposes phenomenal experience to a sorites paradox — idealists, at least, would argue that consciousness is not vague. Micro-idealism does not fall prey to such issues: even if humans’ phenomenal experiences are grounded in physical facts, those are themselves grounded in micro-mental facts. There is no rupture of continuity since mental facts are present at all levels. Secondly, micro-idealism is a monism: on top of providing a neatly unified view of the world, it does not suffer from the same problems of interaction as dualism.

But those would be true of any form of idealism; let’s turn to what’s more characteristic of micro-idealism. Like physicalism, it is supportive of a worldview that sees the laws of science as key to the very fabric of our world. The impressive success of science in past centuries makes doctrines whose ontology espouses its methodological tenets particularly attractive. Like any idealists, micro-idealists see mental facts as fundamental, but since they are intrinsic to the entities described by fundamental physics, the spectre of solipsism is kept at bay: the laws of physics can simply be re-cast as a consequence of some hypothetical laws of micro-mentality. I’ve called mental facts intrinsic and that’s actually in reference to an idea that’s often referred to as Russellian monism but has roots back to Leibniz and beyond. It stems from the idea that the physical worldview is entirely relational: we know what a quark does, not what it is. It is characterized by various quantum charges and other properties, all of which are a way to describe its propensity to react in certain ways in certain situations. That may seem banal, but can lead, and has led to a continuous regress: as Russell noted, "there are many possible ways of turning some things hitherto regarded as “real” into mere laws concerning the other things. Obviously there must be a limit to this process, or else all the things in the world will merely be each other’s washing". The gravitational field was considered a real object in Newtonian physics; in general relativity, it is but a mere property of spacetime (its curvature). Schrödinger wavefunctions seemed like fundamental objects in early quantum mechanics; in quantum field theory, they are eigenstates of a quantum field. Russellian versions of micropsychism, and thus of micro-idealism, hypothetize that the functional role played by physical properties could be realized by micro-experiential properties. The intrinsic nature of microphysical entities would then be found in their mental qualities, with no impact on their extrinsic, relational, physical properties.

Whether we lend credence to Russellian interpretations or not, the fact remains that for all its strengths, micro-idealism suffers from its apparent inverifiability. Another major issue is called the problem of combination: "pure" micro-idealism has it that mental facts at the macro level (such as our own beliefs, feelings, etc.) are grounded in micro-mental facts. One who doesn’t accept that is a pan-idealist of another form, which we could call micro/macro, whereby macro-mentality strongly emerges from micro-mentality somehow. That is a defensible position, but it re-ignites lots of the same issues we saw in a mind-body context for strong emergence and dualism. On the other hand, if one accepts that micro- grounds macro-mentality, there needs to be some explanation of how that works. Micro-idealism is not an animism, so micro-phenomenal states such as those of quarks or photons cannot be akin to humans’. Yet, once combined, they do give rise to phenomenal consciousness. Chalmers suggested that we need to be a bit more prescriptive, lest one could argue that reductive physicalism itself is a form of pan-idealism by identifying micro-mental properties with physical properties. Based on the irreflexivity of grounding however that is not possible, since the micro-mental grounds the physical. Two main options remain: the macro-mental could be fully grounded in the physical — that’s just re-creating physicalism with an additional fundamental layer of micro-mentality; or the macro-mental could be more directly grounded in the micro-mental (which Chalmers call "transparently"). In any case, to resolve the problem of combination is to wrestle with thorny issues such as subject-summing (how can several subjects come together to form a new subject), the palette problem (how can a finite number of micro-entities with presumably basic phenomenal experiences assemble to form the seemingly infinite palette of macro-phenomenality, and the grain problem (the related puzzle that experiences appear to be smooth and homogeneous, rather than a jagged array of micro-components, akin to physical micro-structure).

Another hurdle on the route to micro-idealism lies in the "micro". In the context of modern physics, does it even make sense to speak of micro-entities as discrete units? If the universe is best represented as a cosmic wavefunction, and its particle content as an eigenstate — or alternately if general relativity is ultimately about one object: spacetime, can the hypothetical mentality of these particles be truly fundamental? One way out of this possible conundrum it to reverse the micro-idealist approach by positing that every fact is grounded in mental facts about the universe. We’ll call this holo-idealism. It entails that mental facts about humans are grounded in the mentality of the universe, as are physical facts. This can be fleshed out with a similar Russellian approach to the one described above, whereby cosmo-experiential properties realize the structures and functions described by our physical theories.

The problem of combination in micro-idealism has an analog for holo-idealism: we’ll call it the splintering problem: how can the mentality of one entity ground the private experiences of macro-subjects. The same hurdles appear, and the same general options: one is to the identity of all subjects with the holo-entity, of which they are then modes or aspects; conversely, one could deflate the notion of subject, as some Buddhist traditions have, suggesting that there are experiences but no experiencer; lastly, the intermediary route is to recognize the macro-subjects’ distinct existence from universal mentality, in which case we still need to explain the grounding relationship at work there.

What to make of pan-idealism? Its "best-of-all-worlds" quality certainly is intellectually seducive. Out of all the mind-body doctrines considered here, it seems to have the most manageable issues. Yet it is also one that raises many unanswered questions. Like some scientific theories with much elegance but too many axioms and constants, it will need more refinement, and deeper connections with the natural sciences, to gain significant support.

\(\{\mathbf{2}\}\) : radical solipsism

Is there a symmetrical doctrine to eliminative materialism: one that recognizes mental states as the ontological basis of our world, but furthermore even denies the existence and usefulness of physical concepts?

This cannot apply to the mental features of elementary particles or of the universe, since those have been precisely hypothetised in relation to those hitherto archetypally physical entites. Therefore, "eliminative idealists" would need to only accept mental facts about macro-subjects in their ontology. They would hold a form of extreme phenomenalism that would hardly distinguish between our "physical reality" and dream-like states. For that reason we propose to call such views "radical solipsism".