Consider a glass sphere sitting on a table. It has a certain mass, a certain chemical composition, a certain shape. It is also three feet from a coffee mug, owned by someone in Prague, and the only sphere in the room. Intuitively, these two clusters of properties feel different. The first group seems to belong to the sphere in itself. The second group depends on what else exists and how things are arranged around it.

This intuitive difference between what's intrinsic to an object and what's extrinsic—dependent on its surroundings—turns out to be one of the most fundamental and surprisingly elusive distinctions in metaphysics. It underwrites debates about duplication, supervenience, physicalism, and the very structure of fundamental physics.

Yet pinning down exactly what makes a property intrinsic has proved remarkably difficult. Every proposed analysis encounters stubborn counterexamples. And recent developments in physics raise the possibility that the distinction itself may not carve nature at its joints. Let's examine why this seemingly simple contrast resists simple treatment.

The pre-theoretical idea is straightforward: an intrinsic property is one an object has just in virtue of how it is, independent of anything external. Mass, shape, and charge seem intrinsic. Being two miles from a barn, being someone's favourite object, and being alone in a room seem extrinsic. But translating this intuition into a rigorous definition is another matter entirely.

One early approach defines intrinsicality through independence from accompaniment. A property is intrinsic if an object can have it whether or not other things exist. Being spherical passes this test—a sphere in an otherwise empty universe is still spherical. Being two miles from a barn fails, since it requires the barn's existence. But this analysis stumbles on properties like being such that a prime number exists. Plausibly, prime numbers exist necessarily, so every object has this property regardless of accompaniment. Yet it doesn't seem to characterize how the object itself is.

Another strategy ties intrinsicality to naturalness. On this view, intrinsic properties are those that correspond to genuine, non-gerrymandered features of reality—the joints nature actually has. This draws on David Lewis's distinction between natural and non-natural properties. But it risks circularity: we often identify the natural properties partly by identifying which properties are intrinsic. And it faces the question of whether naturalness itself admits a clean, non-circular analysis.

A third family of approaches uses duplication as the key concept. Two objects are duplicates if they share all their intrinsic properties; a property is intrinsic if it can never differ between duplicates. This is elegant but potentially circular unless we have an independent grip on duplication. Lewis proposed grounding duplication in shared perfectly natural properties, linking the analysis back to his broader metaphysical framework. Each account illuminates something, but none has achieved consensus—suggesting that intrinsicality may be a more complex or primitive notion than it first appears.

Takeaway

The concept of intrinsicality is clear enough to be indispensable but resistant enough to analysis that it may be a primitive feature of our metaphysical framework rather than something reducible to simpler notions.

David Lewis and Rae Langton developed one of the most influential analyses of intrinsicality by focusing on a deceptively simple idea: if a property is truly intrinsic, then whether an object has it shouldn't depend on whether the object is lonely (the only thing that exists) or accompanied (surrounded by other things). A property that is independent of loneliness and accompaniment in this way, and that is also non-disjunctive and not the negation of such a property, qualifies as intrinsic on their account.

The test works beautifully for paradigm cases. Having a mass of 5 kilograms is independent of whether anything else exists—a lonely duplicate of our glass sphere still has the same mass. Being two miles from a barn, however, fails immediately: a lonely object can't be two miles from anything. The framework also handles subtler cases. Being the only round thing is compatible with loneliness (a lonely sphere satisfies it) but not with all forms of accompaniment (add another sphere and it fails). So it correctly classifies this as extrinsic.

But edge cases push back. Consider the property being lonely itself—existing as the only object. This property is independent of accompaniment in a trivial sense and seems to pass certain formal tests, yet intuitively it is extrinsic: whether you're lonely depends on what else exists, not on your own intrinsic nature. Lewis and Langton handled this by excluding properties that are necessarily coextensive with loneliness or accompaniment, but this fix introduces its own complications.

More challenging still are properties involving necessary existents. If abstract objects like numbers exist necessarily, then coexisting with the number seven is a property every object has in every possible world. It is trivially independent of accompaniment, since the number seven always accompanies everything. Yet it seems profoundly extrinsic—or at best, metaphysically trivial. Such cases reveal that the loneliness test presupposes substantive metaphysical commitments about what kinds of entities count as "accompaniers." The analysis is not metaphysically neutral; it is embedded in a broader picture of what exists.

Takeaway

Lewis's loneliness test shows how metaphysical analysis works at its best—turning an intuitive distinction into a precise formal criterion—while also showing that formal precision always inherits the assumptions of the broader metaphysical framework.

Classical physics seemed friendly to intrinsic properties. A particle had mass, charge, and position as features it possessed in its own right. But developments in modern physics have put this picture under severe pressure. In quantum mechanics, entangled particles exhibit correlations that cannot be explained by assigning independent intrinsic states to each particle individually. The state of the composite system is not reducible to the states of its parts. This phenomenon—quantum holism—suggests that at the fundamental level, the world may be characterized by irreducible relations rather than intrinsic properties of individuals.

Structural realism in the philosophy of science pushes further. On its most radical version—ontic structural realism—the fundamental level of reality consists entirely of structures and relations, with no intrinsic properties grounding them. Objects, on this view, are merely nodes in a relational web. They have no hidden inner nature; what they are is exhausted by how they relate. If this is right, the intrinsic/extrinsic distinction loses its grip at the most basic level of physical description.

This has striking consequences for debates about reductionism and emergence. Traditional reductionism assumes that complex systems are explained by the intrinsic properties of their fundamental parts plus the relations between them. But if fundamental physics is purely relational, there is no bedrock of intrinsic properties to serve as the reduction base. Emergence might then be reconceived: higher-level intrinsic properties could be genuinely novel, not derivable from lower-level intrinsic features, because there are no lower-level intrinsic features.

Not everyone accepts these conclusions. Some philosophers argue that even in quantum mechanics, individual systems retain intrinsic dispositional properties—propensities to produce certain measurement outcomes—even if these properties are not always determinate. Others maintain that apparent relationalism can be reinterpreted: the wave function of the universe, taken as a fundamental entity in a high-dimensional configuration space, may have its own intrinsic properties. The debate remains open, but the pressure from physics forces metaphysicians to consider that intrinsicality—far from being a fixed presupposition—may itself be a feature that the world could lack at its deepest level.

Takeaway

If fundamental physics turns out to be irreducibly relational, then the intrinsic/extrinsic distinction may be a feature of our conceptual framework rather than a feature of ultimate reality—a humbling possibility for any metaphysician.

The distinction between intrinsic and extrinsic properties seems as obvious as the difference between what something is and where it happens to be. Yet sustained analysis reveals a concept that resists clean formalization, generates persistent edge cases, and faces a genuine challenge from fundamental physics.

This is not a failure of metaphysics—it is metaphysics working as it should. The difficulty of the analysis tells us something real about the structure of the concepts we bring to reality. Whether intrinsicality is primitive, analysable, or ultimately dispensable remains one of the field's sharpest open questions.

What we take from the investigation is a clearer picture of the landscape: formal tools that partially capture the distinction, deep connections to broader metaphysical commitments, and the live possibility that nature's fundamental level may not respect a boundary we find intuitive.