Goff, Philip
▶ Watch on YouTubeCore ideas
Goff defends panpsychism — the view that consciousness is a fundamental and ubiquitous feature of physical reality — against the two dominant alternatives, materialism and dualism. Against materialism, he argues that physical science describes the world in purely quantitative, structural terms (what particles do, their causal relations) but cannot in principle capture the qualitative "what it's like" of experience (the redness of red, the smell of coffee), so no amount of neuroscientific detail about brain activity closes this explanatory gap. Against dualism, he argues (while noting he has softened this view somewhat since writing his book, given how little is still understood about brain mechanics) that positing consciousness as a wholly separate, non-physical substance is needlessly unparsimonious. Panpsychism, in his account, offers a simpler alternative: rather than starting with matter and asking how consciousness could arise from it, start with simple forms of consciousness at the level of fundamental particles (electrons, quarks) and show how physics — the mathematical description of these entities' causal behavior — can be derived from that base, comparing this to Bertrand Russell's idea that physics only ever describes the "causal skeleton" of reality, not its intrinsic nature. He stresses this is not a dualist add-on (particles don't have separate physical and consciousness properties); rather mass, spin, and charge just are, in their intrinsic nature, simple forms of experience — physics tells us what matter does, panpsychism proposes what matter is.
Goff distinguishes his view from philosopher Daniel Dennett's "multiple drafts" theory of consciousness (which Goff associates with "heterophenomenology" — treating only third-person-reportable behavior as legitimate scientific data) and from "illusionism" (the view, held by his Mind Chat co-host Keith Frankish, that phenomenal consciousness as philosophers describe it doesn't exist at all). He discusses the "philosophical zombie" thought experiment (a being physically identical to a human but with no inner experience) as clarifying why he thinks consciousness — not mere behavioral sophistication — is the basis of moral status: he states he would not consider it wrong to "turn off" a genuine philosophical zombie, but stresses the deep epistemological problem of ever knowing, from the outside, whether a sufficiently sophisticated AI or robot is a zombie or a genuine experiencer, given that consciousness (unlike the data of physics) is not publicly observable — only privately, directly accessible to the subject having it. He extends this to animal ethics, arguing panpsychism (if trees and plants are conscious in some very minimal sense) removes any clean line between "conscious, morally relevant" and "non-conscious, fair game" organisms, and describes his own practical ethical compromise as avoiding factory-farmed animal products specifically, rather than full veganism, as a more broadly adoptable harm-reduction standard.
On measuring degrees of consciousness, Goff discusses Integrated Information Theory (IIT) sympathetically — its proposal that consciousness correlates with how much a system's information is dependent on its internal interconnection, rather than being decomposable into independent parts — while expressing philosophical unease with the idea that "experience itself" admits of degrees (as opposed to the complexity or structure of an experience varying). He discusses suffering as itself a form of conscious experience, arguing humans likely have access to categories of suffering (existential dread, heartbreak, fear of death) unavailable to non-human animals with simpler cognitive lives, while remaining agnostic about whether shared experiences like physical pain are "deeper" for humans than for other animals. Discussing free will, Goff argues for treating it as a more genuinely open empirical/philosophical question than consciousness — he sees no decisive argument against libertarian free will (responsiveness to reasons, as philosopher E.J. Lowe argued, distinguishes a free, reason-guided choice from mere randomness) and is skeptical that Libet-style experiments (measuring brain activity preceding simple, arbitrary button-presses) meaningfully bear on the kind of deliberative, reasons-responsive choices philosophers are most interested in.
Goff also discusses his intellectual disagreements with physicist Sean Carroll over whether panpsychism can account for genuine "mental causation" (Goff's analogy: consciousness is like hardware and physics like software — the fact that the same software could in principle run on different hardware doesn't mean the hardware isn't doing real causal work), and recounts a personal anecdote from a philosophy cruise organized by Russian philosopher/entrepreneur Dmitry Volkov, where he engaged Daniel Dennett directly on a technical point about dualism and conservation of energy — a moment he cites as a small, hard-won concession from Dennett. He explains why he does not believe consciousness could be preserved by mind uploading, since (unlike substrate-independent computational/software views of mind, which he attributes to philosophers like David Chalmers) panpsychism holds that consciousness is tied to the intrinsic nature of the physical "stuff" itself, not merely its organizational or computational pattern — meaning replicating a brain's software elsewhere would not, in his view, preserve the original subject's experience. He closes by connecting panpsychism to broader questions of meaning: arguing the widespread cultural sense of "disenchantment" and alienation (Max Weber's term) partly stems from a scientific worldview many people intuitively feel is incompatible with the reality of their own felt experience, and by describing his own stance toward cosmic purpose as a matter of chosen hope rather than confident belief — comparing it to choosing to have faith in a gravely ill friend's recovery even without strong evidence that recovery is likely.
Excerpts
On why he thinks consciousness poses a uniquely difficult scientific problem compared to everything else science explains:
“In the whole of science... we postulate unobservables in order to explain what we can observe. In the case of consciousness, the unique case of consciousness, the thing we are trying to explain is not publicly observable, and that is utterly unique. If we want to fully bring science into consciousness, we need a more expansive conception of the scientific method.”
On the moral stakes of not knowing whether a sufficiently advanced AI is conscious or a "philosophical zombie":
“I'm inclined to think, you know, if it doesn't have experience, it doesn't really suffer, it doesn't really have moral rights at all... but there's a difficult epistemological question there — because consciousness is not publicly observable, it is a very difficult question how we decide which things are and are not conscious.”
On why mind uploading would not, in his view, preserve consciousness:
“I think consciousness is the stuff of the brain, it's the stuff of matter — so I think just taking the organizational properties, the software in my brain, and uploading them, you wouldn't get the stuff of my brain... it could be we're just committing suicide, we're just getting rid of our consciousness.”