Conditional Realism

A minimal ontology and the case for reality

Donald Hoffman has a flair for the dramatic. In The Case Against Reality he tells us that everything we see is a lie: the world of colors and objects, of space and time, is nothing more than a user interface designed by evolution. The metaphor is irresistible — just as your desktop icons hide the messy circuits of silicon, your visual field hides the quantum froth beneath.

Readers typically respond in one of two ways. Some nod in awe, as though Hoffman has revealed a cosmic conspiracy. Others scoff, dismissing him as another idealist in scientific drag. Both reactions miss the point. The right question is not “Is Hoffman right?” but under what conditions would Hoffman be right?

This is where Conditionalism shows its power. Every truth claim presupposes background conditions; interpretation is never unconditional — it always runs on hidden rails. So to test a sweeping theory like Hoffman’s, we do not choose between True and False. We expose the conditions under which his thesis holds, and the conditions under which it collapses. This chapter runs that method in full — and then asks what kind of realism is left standing when the method has done its work. The answer, it turns out, has no foundations at all, and is stronger for it.

The Conditional Matrix

Here is the skeleton of Hoffman’s case, laid out as conditions:

Every row has the same shape. There is a condition under which Hoffman’s claim goes through, a countercondition under which it fails, and a verdict that flips with the switch. His book never states the conditions; it needs them left implicit, because stated in the open they are exactly the ones most readers would refuse. Two rows deserve to be worked in detail.

The Interface Metaphor

Hoffman’s most effective rhetorical move is the desktop metaphor. Just as your computer icons are convenient fictions — tiny blue folders and trash cans that conceal circuits and voltages — so too, he says, your perceptions are convenient fictions that hide the machinery of physics.

Conditionalism cracks this open.

Condition A: if interfaces only hide, the metaphor works. Your icons tell you nothing about transistors, just as your perception of an apple tells you nothing about quantum fields. The apple is an icon, not a thing.

Condition B: but if interfaces also preserve structural invariants, the metaphor is fatally misleading. A subway map is an interface, but the topology it encodes is real enough that you will get lost if you treat it as fiction — the lesson of maps, models, and understanding. Your perception of an apple may not mirror physics, but it still preserves causal constraints: solid versus liquid, edible versus toxic. It is not pure fiction; it is compressed structure.

Hoffman’s metaphor is powerful precisely because he sneaks in Condition A while most readers tacitly assume Condition B. The argument never confronts the substitution; it depends on the reader not noticing it. Conditionalism’s whole job is to notice it.

Emergence Is Not Illusion

Another of Hoffman’s recurring moves is to cite cutting-edge physics. He likes to point to Nima Arkani-Hamed’s amplituhedron, or the growing consensus that spacetime is emergent in quantum gravity. His conclusion: if spacetime itself is emergent, then what we perceive — tables, trees, stars — isn’t really there. It’s just our interface.

Again, the hidden condition does all the work.

Condition A: if “emergent” means “illusory,” the claim follows. Spacetime is a trick of the eye, not part of the furniture of reality, and our perceptions of objects in space are doubly fictional: fictions within a fiction.

Condition B: but if “emergent” means what physicists actually mean — higher-level structures that are real even though they derive from deeper laws — the leap is illegitimate. Temperature emerges from molecular motion, but that does not make temperature unreal. It makes it a macro-level regularity with predictive and explanatory power. By the same token, spacetime can be emergent without being fictional.

This is an equivocation, plain and simple. Hoffman swaps the physicist’s technical sense of “emergence” for the layperson’s sense of “illusion,” and the swap carries the entire argument. Once it is unmasked, the physics citations stop being evidence for his thesis and become evidence for the ordinary scientific picture: layered structure, real at every layer.

Navigating the Branches

There is a grain of truth in Hoffman’s picture, and the Quantum Branching Universe (QBU) — the Everettian picture introduced in Measure and Credence — locates it exactly. In Everettian physics, perception really is a heuristic: a way of compressing the branching welter of possible worlds into a coherent agent-relative experience. We do not see the universal wavefunction. We see a navigational slice of it.

But unlike Hoffman, I do not need to claim that perception is nothing but fiction. Under Everett, perceptions are partial but structurally anchored: they track enough of reality’s constraints to keep an agent alive, and science extends those anchors deeper into the branching structure. Theory and instruments reach layers the senses never will, and each extension is validated against the constraints the previous layer already tracked. The interface is not a wall between us and the world. It is the first rung of a ladder into it.

That is the position I call conditional realism: perception and theory alike are agent-relative, condition-laden interpretations — and they are interpretations of something, anchored to structure that pushes back. Drop either half and you get a familiar failure. Drop the conditionality and you get naive realism, the pretense that we see the world as it is. Drop the anchoring and you get Hoffman.

The Demand for Bedrock

At this point a realist of the traditional stripe will object that anchoring is not enough — that realism worth the name needs a foundation, a bottom layer of reality on which everything else rests and from which all interpretation ultimately borrows its authority. Hoffman himself, having demolished the everyday world to his own satisfaction, promptly supplies a new bedrock: a network of “conscious agents” as the ontological primitives beneath the interface. Tear down one foundation, install another.

The demand itself is the mistake. Ontology — the investigation into the categories and relationships of existence — typically seeks solid ground: linear hierarchies, unchallengeable axioms, stable groundings. But every candidate foundation is stated in some vocabulary, and a vocabulary must be interpreted, and interpretation always relies on existing interpretations. There are no context-free starting points. Foundationalism does not escape this; it just refuses to look at the interpretive conditions its axioms presuppose. That refusal should be familiar by now — it is absolutism’s blindness transposed into metaphysics.

So what happens if we deliberately reject linearity in favor of a carefully constructed circle?

A Minimal Ontology

Here is a concise, deliberately circular ontology:

• Property: an observable interpreted by an agent.
• Object: a set of related properties.
• System: a set of related objects.
• Process: a system that evolves over time.
• Agent: a process that interprets observables as properties.

At first glance this is self-referential — and indeed it is, explicitly so. Follow the chain: properties presuppose agents, and agents are defined, four steps later, in terms of the properties they interpret. Rather than a defect, the circularity embodies the Conditionalist insight directly: all interpretation demands context, and contexts depend on interpretations. By openly embracing the circle, the ontology acknowledges the mutual dependency that foundationalist schemes are forced to hide.

Two design choices deserve defense.

First, defining “agent” as a type of “process” makes agency dynamic, temporal, and interpretive rather than static or foundational. An agent is not a substance that has interpretations; it is an ongoing activity of interpreting, continually reconstituting meaning. This aligns agency with systems theory, active inference, and the branching-universe framework, where an agent just is a self-locating process threading through the branch structure.

Second, “observable” is purposefully left undefined. Is observability physical, quantum, computational, perceptual? The ontology does not say — not as an oversight but as deliberate openness to context-specific clarification. What counts as observable is itself subject to agent interpretation, which returns us, again, to the central claim: meaning always hinges on conditions. A definition that pretended to fix observability once and for all would smuggle a foundation back in through the side door.

The same openness lets the ontology scale. Because agency is framed as interpretive process rather than sophisticated cognition, the circle covers minimal computational systems, thermostat-grade regulators, human minds, and potential artificial intelligences without requiring separate frameworks. What varies across that range is the richness of the interpretation, not the ontological category.

Reality Without Foundations

Now the two halves of this chapter close into one argument.

Hoffman’s case against reality fails because it equivocates on its conditions — but the deeper reason it ever looked compelling is the foundationalist picture both he and his realist critics share: the assumption that either perception bottoms out in the really real, or it is fiction all the way down. Reject the picture and the dilemma dissolves. Interpretation does not need a bottom. It needs anchoring — structural constraints that survive translation from layer to layer, from icon to circuitry, from subway map to tunnels, from perceived apple to quantum froth. The minimal ontology says what a world of such anchoring looks like: not a pyramid resting on bedrock but a circle of mutually defining categories, closed under interpretation, with agents inside the circle rather than beneath it.

He is right under Condition A: if evolution always punishes truth, if interfaces are pure icons, if emergence equals illusion, if consciousness is primitive. He is wrong under Condition B: if survival requires approximate veridicality, if interfaces preserve structure, if emergent levels are real, if consciousness is derivative. The trick is not to choose sides but to recognize the hidden switch: change the background conditions, and the verdict flips.

That is Conditionalism working as a method rather than a slogan. It does not merely arbitrate the dispute; it dissolves it into a map of assumptions, showing precisely when the theory is compelling and when it is empty. Hoffman is useful not because he overturned realism but because he demonstrates how easily sweeping claims collapse into conditional ones. And that, in the end, is the real case against reality: not that the world is illusory, but that every theory of the world — his, mine, and the naive realist’s alike — is hostage to its conditions. Conditional realism is simply the position that owns up to this and keeps the world anyway.