The manifest image is the picture of the world as it shows up in everyday human experience: a world of solid tables, vivid colors, free choices, and people who deserve praise or blame. The philosopher Wilfrid Sellars introduced the term in 1962 to distinguish this familiar picture from what he called the “scientific image,” the account of reality offered by physics, chemistry, and neuroscience. The gap between the two has never been fully closed, and the attempt to reconcile them touches everything from whether your kitchen table really exists to whether artificial intelligence deserves something like trust.
Where the Concept Comes From
Sellars laid out the distinction in his essay “Philosophy and the Scientific Image of Man.” His claim was that human beings operate with a refined, centuries-old framework for making sense of the world. This framework includes the assumption that objects are more or less as they appear, that other people have inner mental lives, and that actions can be genuinely free. Sellars called this the manifest image not because it is naive or primitive but because it is the image that manifests itself to us in ordinary life. It has its own internal logic, its own standards of evidence, and its own sophistication. It is the world as described by careful common sense and the humanistic disciplines.
The scientific image, by contrast, describes reality in terms of entities that no one directly perceives: subatomic particles, electromagnetic fields, neural firing patterns. Sellars argued that the central task of philosophy is to figure out how these two pictures relate. Can they be fused into a single coherent view? Does one of them ultimately replace the other? Or do they describe different aspects of the same reality without conflicting? These questions remain unresolved, but the terminology Sellars introduced has become a standard way of framing them, and the debate continues to generate new work across several fields.1Bloomsbury Academic. Wilfrid Sellars’s Metaphilosophy
The Two Tables Problem
One of the most vivid ways to feel the tension between the manifest and scientific images is to think about a table. In the manifest image, a table is a solid, colored, bounded object you can lean on. In the scientific image, that same region of space is overwhelmingly empty, with sparse particles interacting through forces at distances that would make the table seem more like a cloud than a block. The physicist Arthur Eddington described this as having “two tables” and the puzzle has been debated ever since.
The obvious philosophical move is to say the scientific table is the real one and the manifest table is an illusion, or at least an approximation. But a recent analysis pushes back hard on this idea. The argument is that the supposed “scientific table,” built from swarms or clouds or aggregates of microphysical particles, does not actually hold together as a coherent object. Attempts to define it run into a basic problem: there is no principled way to draw boundaries around which particles belong to the table-swarm and which do not. If you try to define it as a fixed collection of specific particles, you get a rigid fusion that changes identity every time a single atom is exchanged, which does not match how any real table behaves. If you allow the membership to change over time, as the philosopher Kit Fine’s framework of “variable embodiments” would, you end up describing something that is functionally identical to the ordinary table of the manifest image. The scientific table, in this view, is not a discovery but a metaphysical fiction.2Topoi. Turning the Tables on the Table Deniers: Metaphysical Mythmaking About Ordinary Objects
This matters because the “two tables” picture is often treated as a settled demonstration that science has shown everyday objects to be unreal. If the argument above is right, the demonstration fails: contemporary solid-state and condensed-matter physics do not supply a determinate set of particles “arranged tablewise” that could serve as a rival to the manifest table. The manifest image’s objects may be more defensible than the textbook version of this debate suggests.
Consciousness and the Limits of Structure
The manifest image includes the richness of conscious experience: the redness of red, the sharp sting of a stubbed toe, the feeling of recognition when you see a friend’s face. These qualitative features of experience, often called qualia, have been a flashpoint in the debate because they seem to resist translation into the scientific image. You can describe every neural event involved in seeing red without apparently capturing what it feels like to see it. This is the so-called “hard problem” of consciousness, and it is sometimes used to argue that the manifest image contains something the scientific image can never fully explain.
Some neuroscientists and philosophers have pushed back on the claim that qualia are inherently beyond the reach of science. One line of argument holds that the apparently non-structural aspects of consciousness, the raw “what it’s like” quality of experience, can in fact be analyzed in structural terms once you adopt the right theoretical framework. The idea is that the relationships between different experiences, the way red relates to orange, or the way a loud sound relates to a soft one, are themselves the qualia, not something hiding behind the structure. If this approach works, it would mean the hard problem is not a permanent barrier but a sign that the scientific image needs to become more sophisticated in how it models structure.3PubMed Central. Structural qualia: a solution to the hard problem of consciousness
Whether this solution satisfies everyone is another matter. Critics argue that even the most complete structural description of experience leaves something out, that knowing every relational fact about red does not tell you what red looks like to someone who has never seen it. The debate is genuinely open, and it is one of the places where the manifest and scientific images seem furthest from reconciliation.
Free Will and the Readiness Potential
In the manifest image, you are an agent who makes decisions. You weigh options, choose, and act, and your choices are yours in a way that makes praise and blame appropriate. The scientific image complicates this picture, because it describes the brain as a physical system whose states unfold according to causal laws, leaving unclear where voluntary choice enters.
The most famous experimental challenge to the manifest image’s account of free will came from the neuroscientist Benjamin Libet in the 1980s. Libet asked people to flex their wrists whenever they felt like it and to note the moment they became aware of their intention to move. He found that a brain signal called the readiness potential, a ramp-up of electrical activity in motor areas, began several hundred milliseconds before participants reported being aware of any intention. The implication seemed stark: the brain had already “decided” to move before the person felt like they were deciding. Subsequent work using brain imaging pushed the timeline even further, identifying patterns of brain activity that could predict a person’s choice several seconds before they reported making it.4PubMed Central. Free Will and Neuroscience: From Explaining Freedom Away to New Ways of Operationalizing and Measuring It
Taken at face value, these results suggest the manifest image’s picture of conscious, deliberate choice is an after-the-fact narrative the brain constructs rather than the actual driver of behavior. But the story has gotten more complicated over time. Critics have questioned whether the readiness potential really reflects a “decision” at all; it may simply reflect the buildup of neural noise that eventually crosses a threshold. Others have pointed out that the wrist-flexion task is about as trivial as a decision can be, and that extrapolating from it to the kind of deliberation involved in, say, choosing a career is a stretch. Neuroscience has not settled the question of whether free will is real, but it has made it impossible to discuss free will without also discussing what the brain is doing underneath the experience of choosing.
Moral Responsibility in a World of Brain Scans
If free will is one pillar of the manifest image’s social world, moral responsibility is another. We hold people accountable for their actions, assign praise and blame, and structure legal systems around the idea that individuals are responsible agents. Neuroscience has begun to press on this pillar too, particularly through the use of brain imaging.
Neuroimaging studies have raised pointed questions about how responsibility should be understood. When brain scans reveal structural or functional abnormalities in people who have committed violent acts, it becomes tempting to conclude that the person’s brain, not the person, is at fault. Courts have already wrestled with this: defense teams sometimes introduce brain scans as evidence that a defendant lacked the capacity for the kind of control that responsibility requires.5PubMed. Seeing responsibility: can neuroimaging teach us anything about moral and legal responsibility?
The philosophical question is whether this kind of evidence actually changes the concept of responsibility or merely adds detail to it. One view is that if every action has a neural cause, then no one is ever truly responsible in the deep sense the manifest image assumes, and our practices of blame and punishment are a kind of useful fiction. Another view is that responsibility was never meant to require freedom from physical causation. To be responsible, on this account, is to have the right kind of brain: one that can respond to reasons, learn from consequences, and adjust behavior. Showing that a brain does these things through neural mechanisms does not undermine responsibility any more than showing that a car moves through combustion undermines the claim that the driver is steering.
The legal and ethical stakes are real. If neuroscience gradually erodes the manifest image’s account of responsible agency, the downstream effects on criminal justice, education, and everyday social life could be enormous. If it instead enriches the account by showing how responsibility is implemented in the brain, the manifest image absorbs the science and emerges more detailed but fundamentally intact.
How Classical Reality Emerges from Quantum Physics
Physics offers its own version of the manifest-versus-scientific tension, and it may be the most disorienting one. Quantum mechanics describes a world of superpositions, entanglement, and probabilities that looks nothing like the stable, definite world of everyday experience. A quantum particle can be in multiple states at once; a cat, famously, can be described as both dead and alive until someone checks. Yet nobody has ever seen a dead-and-alive cat. Why does the manifest image’s solid, definite world emerge from an underlying reality that is so radically different?
Two closely related ideas from physics address this. Decoherence describes how quantum superpositions effectively vanish when a system interacts with its environment. A single isolated particle can maintain a superposition, but a table, which interacts with trillions of air molecules and photons every second, loses its quantum fuzziness almost instantly. The environment acts like a constant observer, collapsing superpositions before they can become visible at the scale of everyday objects. Quantum Darwinism extends this idea by explaining why different observers agree on what they see. When a quantum system interacts with the environment, certain states are copied into the environment many times over, like information being broadcast to multiple channels. Only the most robust states survive this process, and those are the classical, definite properties that populate the manifest image.6arXiv. Decoherence, quantum Darwinism, and the generic emergence of our objective classical reality
The upshot is that the manifest image’s picture of a stable, objective, classical world is not a convenient illusion. It is the image that physics predicts will emerge for creatures of our size, embedded in environments of our complexity. The quantum world is not more real in any straightforward sense; it is the deeper description from which the manifest image’s solidity is a reliable and inevitable consequence. This is one area where the two images fit together more neatly than you might expect, though debates about interpretation continue.
When We Treat Machines as If They Have Minds
The manifest image is not just a description of the physical world; it is also a way of relating to other agents. You assume the people around you have beliefs, desires, and intentions, and you predict their behavior by reasoning about what they want and what they think. Philosophers sometimes call this the “intentional stance.” It belongs squarely to the manifest image, because beliefs and desires are not things you can find under a microscope. The scientific image describes brains, circuits, and firing patterns, not hopes or plans.
The rise of artificial intelligence has made this distinction newly urgent. When a chatbot generates a plausible excuse for missing a deadline, or a medical AI recommends a treatment plan, people naturally slip into the intentional stance. They treat the machine as if it has goals, understands context, and makes choices. Research on how people interact with AI suggests this is not just a lazy habit but may be the only practical strategy for non-experts to understand, predict, and learn from artificial agents’ behavior. A pragmatic defense of this approach argues that as long as the agents are transparent about how they work, attributing intentions to their actions is not only useful but ethical.7Minds and Machines. A Pragmatic Approach to the Intentional Stance Semantic, Empirical and Ethical Considerations for the Design of Artificial Agents
This gets more concrete in medical settings. When an AI system recommends a diagnosis or a course of treatment, clinicians and patients face a trust problem. The system has no intentions in any manifest-image sense, but its outputs carry what one analysis describes as “technically mediated intentions,” meaning the goals and values of the designers are embedded in how the system processes information and arrives at recommendations. On this view, trusting a medical AI is conceptually coherent, because the trust is directed at the human purposes woven into the system’s design, not at a fictional inner life.8PubMed. Intentional machines: A defence of trust in medical artificial intelligence
Neuroscience adds another layer. When people are told that an AI failure happened because the system “intended” a certain approach versus being told the failure was a design limitation, their brains process the information differently. Brain activity in the theta frequency band, associated with cognitive control and conflict monitoring, is significantly higher when people are given intentional-stance explanations for emotional task failures compared to design-stance explanations. In other words, framing a machine’s behavior in manifest-image terms (as if it meant to do something) changes how deeply and effortfully the brain engages with the failure. The manifest image is not just a philosophical lens; it shapes neural processing in measurable ways.9PubMed Central. Intentional or Designed? The Impact of Stance Attribution on Cognitive Processing of Generative AI Service Failures
Why the Manifest Image Persists
Given how much the sciences have revealed about the hidden machinery behind everyday experience, it might seem surprising that the manifest image has not been quietly retired. But it keeps proving resilient, for reasons that go beyond nostalgia or stubbornness.
One reason is practical indispensability. Try navigating a single day without treating tables as solid, people as intentional, and your own choices as genuine. You cannot function using only the vocabulary of particle physics and neural firing rates. The manifest image is the operating system for social life, planning, morality, and law. Even researchers who study its inadequacies at work revert to it the moment they leave the lab and ask a colleague to pass the coffee.
Another reason is that the scientific image has not actually produced a complete replacement. Physics describes particles and fields but has no consensus on how to get from those descriptions to the experienced world of colors and sounds without invoking something that looks a lot like the manifest image along the way. Neuroscience can correlate brain states with reported experiences but has not closed the explanatory gap between neural activity and the felt quality of consciousness. The scientific image is powerful and expanding, but it is not yet a self-contained worldview that can do everything the manifest image does.
A third reason, and perhaps the most philosophically interesting, is that some features of the manifest image may not be reducible at all. If the argument about tables is right, that microphysics does not actually supply a coherent rival object to the ordinary table, then the manifest image is not a simplified version of the scientific image waiting to be replaced. It describes genuinely real features of the world that cannot be reconstructed from the bottom up. The same might be true of consciousness, agency, or moral facts. These are open questions, not settled ones, but the persistence of the manifest image across centuries of scientific progress is itself a data point worth taking seriously.
Common Misconceptions About the Debate
People encountering the manifest-versus-scientific distinction for the first time often make a few predictable mistakes. The most common is assuming that the scientific image is simply the “true” one and the manifest image is a collection of errors held by people who have not studied enough physics. Sellars himself did not hold this view. He thought the manifest image was a sophisticated framework with its own internal coherence, and that dismissing it was as much a philosophical error as clinging to it uncritically.
A related misconception is that the debate is about whether science is “right.” It is not. No one involved disputes that atoms exist or that brains produce behavior. The question is about the status of the everyday descriptions, whether tables, colors, choices, and moral responsibilities are real features of the world that science must accommodate, or useful fictions that science will eventually explain away. Reasonable, well-informed people land on different sides of this.
A third misunderstanding involves thinking the manifest image is fixed. It is not. Sellars described it as the product of long refinement. Concepts like “person,” “intention,” and “cause” have evolved over centuries of philosophical and cultural development. The manifest image of a twenty-first-century adult is not the manifest image of a medieval peasant. It absorbs some scientific findings over time: most people now accept that the Earth orbits the Sun and that diseases are caused by microorganisms, even though both claims contradicted earlier versions of the manifest image. The interesting cases are the claims that resist absorption, like the claim that your conscious experience is nothing more than a pattern of neural activity, or that your sense of choosing freely is an illusion generated by deterministic brain processes. Those are the frontier, and the manifest image’s response to them will determine what it looks like a century from now.