Also see the Consciousness Realist blog for additional FAQs and discussion about critiques of IIT. 

Coming soon. 

In the broadest sense, physicalism is the claim that all existing entities and their properties are physical. Let’s call this “ontological physicalism” because it is a doctrine (or assumption) about what exists. Standard notions of physicalism are usually synonymous with materialism and not in line with IIT (more details below) [1].

IIT has a “phenomenology-first” approach to both epistemology and ontology: experience exists, and it is not primarily physical but phenomenal. Even more, physical existence is inferred from within experience. Therefore, IIT does not endorse standard ontological physicalism but is rather committed to operational or methodological physicalism. The main idea is that we observe regularities within our experience, and these regularities make sense in terms of persistent and reliable causes and effects. Put otherwise, cause–effect power is a guide to figuring out what exists independently of our consciousness. 

Even though mainly methodological, the physicalism of IIT also has ontological import: experience itself is identical to a certain arrangement of cause–effect power (namely, a Φ-structure unfolded from a substrate that fulfills the five postulates). In this sense, IIT is an intrinsic powers ontology, which could be called ontologically physicalist, but only insofar as (intrinsic) physical existence is identical to—and always inferred from within—phenomenal existence (see From Phenomenology to Physics). 

As elaborated below, IIT’s physicalism is quite at odds with standard conceptions in science and philosophy of mind.

IIT & standard scientific physicalism/materialism

The concept of “physical” is not well defined [2]. Rather than being a fully articulated position, standard scientific physicalism, or materialism, is a collection of assumptions that scientists tend to endorse. It usually runs in the background of much empirical research, which is not particularly concerned with metaphysical issues. There are many varieties of physicalism, but to illustrate the contrast with IIT, we will outline two of its standard assumptions in broad strokes: 

1. Everything is physical down to the fundamental level. From quarks and quantum fields to atoms, molecules, cells, organisms, societies, and galaxies—all is physical through and through, consciousness included. This is an extrinsic view of physical existence: consciousness depends on a brain, whose neurons exist as well as the organelles and proteins that compose them, down to elementary particles and fields. These things are ontological substrates because they are considered to exist as such, having fundamental physical properties and obeying the laws of physics. 

2. The microphysical level exists in a stronger sense than higher levels. For example, higher-level constructs such as molecules, cells, or brains are usually said to “emerge from” (in a “weak” sense) or to be “reducible to” the microphysical level; the same holds for “functional states” (e.g., desires, beliefs) and even consciousness itself. In other words, the assumption is usually that higher-order constructs are necessary to understand and explain what complex systems do, so they are somehow real, but not “as real as” the microphysical level. Their existence is derivative rather than fundamental—they may be indispensable from an epistemological point of view, but they are reducible from an ontological one. Therefore, the ultimate ontological substrates reside at the bottom level of physical reality.

IIT is at odds with these two widespread core assumptions of physicalism [3]. Regarding the first, the physical in IIT is not fundamental but inferred from within consciousness, which is ontologically basic (see 0th axiom above). Thus, even if an experience is identical to a Φ-structure, it does not somehow “emerge” from it (see IIT and emergence FAQ). From the phenomenological point of view, experience is—period. IIT rather takes an intrinsic view of physical existence: It is certainly expedient to talk about “a brain” as an extrinsic object in the world, but from its intrinsic perspective, the brain as a whole may not actually “hang together” as an entity. The whole brain is rather an operational substrate for us as observers; we can systematically unfold its cause–effect power to determine specifically which parts of it indeed hang together—from its intrinsic perspective as a Φ-structure and from the observer’s extrinsic perspective as a substrate of consciousness (or a complex).

This is the sense in which IIT is operationally physicalist. The concepts of a Φ-structure and a substrate are indeed “physical” concepts in that they are based purely on assessing cause–effect power. Yet at no point does IIT argue that consciousness is “nothing but” a Φ-structure or its substrate (see FAQ: Is IIT reductionist?). These concepts are rather instrumental in describing an experience from the third-person perspective. Yet the experience itself is and always will remain first and foremost real. 

This brings us to the second assumption of standard physicalism, that the microphysical level is somehow more “real” than higher levels. IIT does not prioritize the microphysical level. It argues, rather, that our view of what actually exists should be guided by what we know to exist—immediately and irrefutably—namely, experience itself. Thus, in a physical sense, a complex unfolded as a Φ-structure is what actually exists because it is identical to an experience. In IIT’s framework, therefore, the Φ-structure unfolded from a complex is “more real” than its atomic constituents and cannot be reduced to them. Moreover, IIT follows the principle of maximal existence in arguing that over a given operational substrate, the only entity that exists is the one with the highest quantity of integrated existence, defined by system integrated information (φs). This means that atomic constituents of a complex do not “also exist”; they exist for us as observers to manipulate and observe, but their intrinsic existence as entities is ruled out since they form part of a greater intrinsic entity [4].

IIT & materialism in consciousness studies 

Various materialist (or physicalist) theories of consciousness have been proposed in philosophy of mind. Following Chalmers (2003), we can classify materialism about consciousness into types A, B, and C. These types are identified by the way they deal with the so-called “explanatory gap” between physical and phenomenal truths.

Type-A theories deny there is an explanatory gap between the two. Examples of type-A are eliminativism, analytic functionalism, or logical behaviorism, which hold either that consciousness does not exist, or that it is fully reducible to functions, behaviors, or a set of physical properties. 

Type-B theories accept there is an unclosable explanatory gap but deny there is an ontological one. Classical identity theories (e.g., Smart and Armstrong) are examples of type-B materialism, which claim that experience can be identified with certain physical or functional states; we can discover this identity empirically, but we can never know why it obtains—it just does.  

Type-C theories accept there is an explanatory gap in our current state of knowledge but hold it is closeable in principle by acquiring more knowledge about the physical world. Chalmers (2003) claims such theories prove to be either untenable or to collapse into other views, such as type-A or -B materialism (but also non-physicalist options).

Since its physicalism is operational, IIT rejects consciousness materialism and defies Chalmers’s categorization. Let’s start with IIT’s stance on the explanatory gap. First, because the physical is postulated from within the phenomenal and understood operationally, there is no gap. The difference between the intrinsic and extrinsic perspectives does not lead to an unbridgeable chasm between the phenomenal and the physical. On the contrary, with proper operational assumptions, the cause–effect power of a substrate can be fully characterized, and an intrinsic physical account of experience can be given. Second, there is also no ontological gap because IIT’s operational physicalism leads to its fundamental identity between an experience and a Φ-structure. They are one and the same, although the experience will always remain primary, both epistemologically and ontologically.

Regarding Chalmers’s categories, IIT is certainly not a type-A theory because it says consciousness is real and irreducible to functions or behaviors. It is tempting to call IIT a type-B theory since it is an identity theory. However, unlike classical identity theories, IIT does not lead to a “brute identity”—that is, an empirical identity with no logical sense as to why it should hold. Instead, IIT claims that every property of experience corresponds one-to-one with a property of a Φ-structure. (To get a sense for how IIT does this, see the contents of experience page.) In this way, IIT’s identity “scratches our itch” of explanation in a way type-B theories do not. Hence, IIT is closest in spirit to a type-C position by agreeing that the gap between the intrinsic and extrinsic perspectives is closeable. Yet it’s far cleaner to say that IIT defies Chalmers’s categorization altogether because the theory demands that we reconceive of the “physical” in purely operational terms from within experience—which remains both epistemologically and ontologically primary. 

So, is IIT physicalist?

It depends. If physical means that the particles and forces in the standard model of physics are the “most real” ontological substrate from which everything else emerges or to which everything else is reducible, then no, IIT is not physicalist. If physical implies one of Chalmers’s categories of materialism, then, again, no, IIT does not conform to this. But if physical is understood in an operational sense—as the cause–effect power we can unfold by perturbing and observing—then yes, IIT is physicalist. And if physical is meant in an intrinsic sense—as required to account for experience—then, again, yes, IIT’s intrinsic powers ontology can also be called a physicalist ontology. 

Coming soon. 

Is IIT an emergentist theory? More precisely, does IIT claim that consciousness “emerges” from a neural or physical basis? IIT’s unequivocal answer is no—neither in a “strong” nor “weak” sense [1]. 

There is no consensus on the meaning of “emergence” in consciousness science, though there are proposals (e.g., [3]). Broadly speaking, however, the literature converges on two core features of emergence [3, 4, 5, 6]:

• the emergent phenomenon has some sort of “base” as a precursor, from which it emerges, 

• and the emergent phenomenon has some novel properties that its components or precursors lack.

Proposed examples of emergent phenomena are traffic jams, bird flocks, weather systems, biological life, entangled quantum states, covalent bonding, or—some argue—even consciousness.

To understand IIT’s approach to emergence, it is essential to begin with the key distinction in IIT between intrinsic and extrinsic existence. Anything that we can observe and perturb can be said to exist extrinsically. Hence, we may speak of “causal emergence” when discussing extrinsic entities or processes, whether bird flocks, cells, organisms, or organizations. In fact, previous work has used the tools of IIT to develop causal analysis (which is extrinsic, Albantakis et al 2019), demonstrating not only that “emergent” macro levels can be causally just as good as micro levels, but that the “macro can beat the micro” (Hoel et al. 2013). 

When it comes to intrinsic existence—consciousness, that is—IIT is decidedly non-emergentist [2]. It does not argue that experience “emerges from” or “is generated by” the substrate of consciousness. IIT rather posits an explanatory identity: the phenomenal properties of an experience—its quality or character—correspond one-to-one to the physical properties of the Φ-structure unfolded from the complex. In other words, all the contents of an experience correspond to Φ-folds in the Φ-structure: a patch of blue color, a whiff of lavender, a pang of nostalgia—any content whatsoever. And the quantity of experience is measured by the Φ value of the Φ-structure, capturing the structural information (or richness) of its countless distinctions and relations.

How can we be sure that a relationship is of identity and not emergence? A simple test is to ask whether you can have the base without the emergent property. Can you have cars without a traffic jam? Yes. Can you have birds without a flock of them? Yes [9]. Can you have a Φ-structure without an experience? No. In IIT, these are one and the same “thing,” merely expressed in phenomenal terms vs. physical (causal) terms. And this identity is “explanatory” in the sense that we employ the physical language and analysis to explain the fact of experience. 

Moreover, given the phenomenology-first approach of IIT, it would be logically absurd to assert that experience emerges from a Φ-structure: we know experience to exist immediately and irrefutably, while we posit that Φ-structures exist using inference and operational procedures. Why would one assert that irrefutable existence emerges from inferred existence? 

Beyond the broad question of emergence, there could be at least five additional, more nuanced ways in which one might ask whether emergentism is at work within the IIT framework:

1. Does a Φ-structure emerge from a complex?

2. Does a complex emerge from its mechanisms?

3. Do higher-order mechanisms emerge from first-order mechanisms?

4. Do mechanisms emerge from their units?

5. Do macro units emerge from micro units?

1. Does a Φ-structure emerge from a complex?

No, a Φ-structure is unfolded from and specified by a complex (or a substrate of consciousness). 

IIT claims that the physical (i.e., causal) properties of the substrate account for the phenomenal properties of experience. To qualify as a complex, a candidate must satisfy the postulates, and we use the term unfolding to describe the operational steps in determining the causal structure of a complex—its Φ-structure, comprising distinctions and relations. Ontologically speaking, the Φ-structure does not exist separately from the complex; it is rather specified by the complex—it is the fully unfolded cause–effect power of the complex. A Φ-structure thus doesn’t emerge from a complex; rather it is the complex seen from the intrinsic perspective.

Like in the previous sub-question, we can apply the same test of emergence by asking whether we can have the purported base without the purported emergent property: Can we have the complex without the Φ-structure? No. A complex only exists if it fulfills all postulates and, thus, specifies a Φ-structure. (In the same vein, a Φ-structure only exists with the complex that specifies it.)

2. Does a complex emerge from its units?

No, a complex is constituted by units but does not emerge from them.

Just like complexes, the units constituting it must also fulfill the postulates (save composition). This means that a complex must have irreducible units as constituents. However, the complex is neither novel nor does it appear in time; it is either there, with all its constituents, or not. 

Again, we can test for emergence by asking whether the purported base can exist without the purported emergent property: Can the units exist without the complex? No. The units only exist as constituents of a complex; in the limit case, they are complexes themselves (see monad).

3. Do higher-order mechanisms emerge from first-order mechanisms?

No, higher-order mechanisms co-exist with first-order mechanisms. 

For example, for system ABCD, unit A and unit B may each be mechanisms, and units AB together may also be a separate mechanism. This simply means that mechanism AB has cause–effect power in the system that cannot be reduced to that of unit A and unit B (see FAQ: Why do we assess the causal power of all orders of mechanisms?). To assess if a mechanism—of any order—exists, we check whether it satisfies the postulates. Regardless of order, each mechanism exists separately from the others. This is also shown by the fact that mechanism AB can exist even if mechanism A does not. 

The emergence test we’ve done so far doesn’t quite work in this case. In other words, if we ask whether the first-order mechanisms (the “base”) can exist without higher-order mechanisms (the “emergent property”), the answer is trivially yes. Instead, to see that this is not a case of emergence, we can ask whether the purported emergent entity can exist without the emergence base. Higher-order mechanisms do not emerge from first-order ones because, at least in principle, higher-order mechanisms can exist even if no first-order one exists.

For more, see 

• Composition FAQ: Why do we assess the causal power of all orders of mechanisms?

• Grasso et al. 2021, “Causal reductionism and causal structures.”
  
  

4. Do mechanisms emerge from their units?

No, mechanisms are constituted by their units. Units are not the base from which new mechanisms with new powers emerge. 

Units must themselves satisfy the postulates (with some caveats; see units), and to exist there must be at least one mechanism that includes them. For instance, unit A exists intrinsically only if it is a constituent of a first- or higher-order mechanism (say, AB), and therefore contributes to the cause–effect structure specified by the system (say, ABC). Of course, in an operational sense, we have to start with a guess about what might be units—say, units A, B, and C. But in the end, something exists intrinsically (including units) only if it is a constituent of an irreducible entity or it is an irreducible entity itself.

Applying the usual test for emergence shows that mechanisms don’t emerge from units because units cannot exist if they are not constituents of mechanisms.

5. Do macro units emerge from micro units?

No, macro units are constituted by micro units, they do not emerge from them. 

In IIT, we aim to dissect the cause–effect power of a system down to the smallest grain that we can possibly manipulate and observe. This operational reductionism—or “atomism”—is necessary for a complete assessment of a system’s cause–effect power. But this does not mean that the bottom level somehow exists more fundamentally. The reason is that the smallest constituent and update grain may not be the one at which the system has maximal integrated information (φs), which measures the irreducibility of a system in IIT [7]. It may turn out that φs is maximized by macro units (i.e., sets of micro units), which should thus be considered the relevant constituent grain for that system [8]. 

As a simple example, say that neurons are the relevant constituent grain for maximal φs in a brain; if so, the neurons are constituted by their molecules (meso-units). We don’t define the neurons’ existence in terms of being aggregates of molecules (emergence); we define them rather by virtue of maximizing φs. The molecules, in turn, are not considered “more fundamental”; rather they only exist by virtue of being constituents of the macro units. (In IIT literature, the same idea is sometimes expressed by saying that the macro units “subsume” the micro units.) 

For more, see 

• FAQ: If IIT assumes "atomism," does this make it a reductionist theory of consciousness?

• Marshall, W., Findlay, G., Albantakis, L., and Tononi, G.. (Forthcoming). Intrinsic units: a mathematical framework for identifying the causal grain of a system from its intrinsic perspective.
  
  

_________

In conclusion, let us return to the two standard assumptions of emergence—that there is a “base” level of reality from which other phenomena emerge, and that the emergent phenomena have some novel properties. 

The base reality in IIT is that experience exists—your and my experience, right here, right now, and presumably the experience of countless other entities. The IIT method aims to explain the fact of experience in operational terms—with units, mechanisms, systems, substrates, Φ-structures, and so on. This lets us obtain a third-person explanation of experience (i.e., in physical terms), but this does not replace the ground truth of first-person existence. 

This phenomenal–physical inversion is not just epistemological but also ontological. In other words, consciousness is not a new “thing” or property that emerges from the phi-structure unfolded from a substrate; rather it is this structure. 

Just as IIT has no ontological reduction, it also has no ontological emergence. The IIT method aims to “carve nature at its joints” to assess what exists—right here, right now. Since entities are (they exist), they cannot emerge from entities that are not. In IIT, therefore, nothing emerges; everything is. 

In philosophy, intrinsicality is defined in different ways. One widely cited definition is that an object has a property intrinsically if and only if it has it independently of the way the rest of the world is [1,2,3]. Another way to put it is that a property is intrinsic if it can be instantiated in a “lonely” object—that is, an object that doesn’t coexist with any contingent objects wholly distinct from it. Typical examples of extrinsic properties are “being bigger than” or “being the brother of,” which require reference to something else. Examples of intrinsic properties, in contrast, are “being an electron” or “being a triangle”—which do not require reference to something else. An electron is defined by its mass and negative charge, and a triangle by being a shape with three angles summing to 180°. 

The notion of intrinsicality in IIT is different. Phenomenally, intrinsicality means that experience always exists for itself, from the point of view of the subject of that experience. Intrinsicality is translated into physical terms as the fact that a system has cause–effect power within itself from its own intrinsic perspective. Note that there is no mention here of the system’s relationship with the rest of the world—with what’s outside of the system. 

One important difference between these senses of intrinsicality is this: in philosophy, the word intrinsic can be a label for many different properties, whereas in IIT, it can’t. For example, some philosophers use intrinsic to describe the shape of an object, or the charge of an electron, or the mass of an object, etc.—and at least some of these properties are accidental, not essential (e.g., an object can change its shape). In IIT, however, intrinsic is only ever an essential property of existence—first of phenomenal existence (experience) and, by inference, of physical existence (the cause–effect power of its substrate). 

IIT aims to account for a system’s intrinsic cause–effect power—the way a system causally constrains itself. But to say that a system has intrinsic cause–effect power does not mean it is a “lonely” system—that it is not influenced by its environment. A system’s intrinsic cause–effect power can change if the system’s background conditions change,  and this doesn’t make it less intrinsic (see FAQ: Does the intrinsicality postulate mean that the substrate of consciousness is not influenced by its environment?). 

This FAQ is under construction. For the time being, please see the paper Only what exists can cause: An intrinsic view of free will:

"If IIT is right, we do have free will in the fundamental sense: we have true alternatives, we make true decisions, and we—not our neurons or atoms—are the true cause of our willed actions and bear true responsibility for them. IIT's argument for true free will hinges on the proper understanding of consciousness as true existence, as captured by its intrinsic powers ontology: what truly exists, in physical terms, are intrinsic entities, and only what truly exists can cause."

This FAQ is in development. For the time being, see

Findlay, G., Marshall, W., Albantakis, L., David, I., Mayner, W. G., Koch, C., & Tononi, G. (2024). Dissociating Artificial Intelligence from Artificial Consciousness. arXiv preprint arXiv:2412.04571.

"Developments in machine learning and computing power suggest that artificial general intelligence is within reach. This raises the question of artificial consciousness: if a computer were to be functionally equivalent to a human, being able to do all we do, would it experience sights, sounds, and thoughts, as we do when we are conscious? Answering this question in a principled manner can only be done on the basis of a theory of consciousness that is grounded in phenomenology and that states the necessary and sufficient conditions for any system, evolved or engineered, to support subjective experience. Here we employ Integrated Information Theory (IIT), which provides principled tools to determine whether a system is conscious, to what degree, and the content of its experience. We consider pairs of systems constituted of simple Boolean units, one of which—a basic stored-program computer—simulates the other with full functional equivalence. By applying the principles of IIT, we demonstrate that (i) two systems can be functionally equivalent without being phenomenally equivalent, and (ii) that this conclusion is not dependent on the simulated system’s function. We further demonstrate that, according to IIT, it is possible for a digital computer to simulate our behavior, possibly even by simulating the neurons in our brain, without replicating our experience. This contrasts sharply with computational functionalism, the thesis that performing computations of the right kind is necessary and sufficient for consciousness."