Consider the cat on your mat. You take yourself to be referring to a single, well-defined creature. But now consider the countless microphysical configurations that differ from your cat only by a whisker or a stray hair. Each seems an equally good candidate for being the cat.

This is the Problem of the Many, articulated forcefully by Peter Unger and David Lewis. It threatens something we ordinarily take for granted: that singular reference picks out one thing, and that counting yields determinate answers.

The puzzle is not merely academic. If the boundaries of ordinary objects are metaphysically indeterminate, then either our ontology contains vastly more objects than we suppose, or our practices of individuation rest on some subtler metaphysical relation. Either horn demands systematic analysis.

Unger's original example concerns a cloud in the sky. Peripheral water droplets stand in a continuum of degrees of integration with the central mass. Any principled criterion for cloud-membership—density, proximity, dynamical coupling—yields not a unique cloud but a vast plurality of equally eligible aggregates, each differing from its neighbours by a single droplet.

Lewis extends the case to cats. Tibbles the cat has hairs that are variably attached, cells that are variably integrated, and molecules exchanging with the surrounding air. For each borderline particle, there is a maximal aggregate including it and one excluding it, both apparently possessing all the intrinsic properties requisite for being a cat.

The pressure this generates is severe. If each of these aggregates satisfies the qualitative conditions for cathood, then there are billions of cats on the mat. Our assertion that there is exactly one is either false or requires substantial metaphysical rehabilitation.

Note that the puzzle is not resolvable by appeal to precisification of predicates alone. Even granting a sharp criterion for being a cat, the underlying microphysical situation supplies too many equally good candidates satisfying that criterion. The problem is one of ontic abundance, not merely semantic vagueness.

Takeaway

Vagueness is not merely a feature of language reaching toward the world—it appears to reflect a genuine multiplicity in the world itself, one that ordinary reference and counting must somehow navigate.

The supervaluationist proposes that vague singular terms like 'Tibbles' are indeterminate between multiple sharpenings. On each admissible precisification, 'Tibbles' refers to exactly one of the candidate aggregates. A sentence is super-true if true on every admissible precisification.

This strategy preserves classical logic and, crucially, our ordinary counting. 'There is exactly one cat on the mat' comes out super-true because on every admissible precisification of 'cat' and every admissible assignment of reference to the singular terms, exactly one cat is on the mat—albeit a different one on each sharpening.

The elegance is considerable, but the costs are real. Supervaluationism requires that we accept the existence of the candidate aggregates as legitimate referential targets. It also generates higher-order vagueness: which precisifications count as admissible is itself a vague matter, threatening regress.

Moreover, the supervaluationist owes us an account of why, if the candidates are all equally eligible, exactly one is selected on each precisification. The response secures determinate counting at the semantic level while leaving the metaphysical proliferation untouched.

Takeaway

Preserving ordinary discourse sometimes requires accepting a richer semantic apparatus rather than a thinner ontology—but the metaphysical problem may persist beneath the linguistic solution.

Lewis offers a strikingly different diagnosis. The candidate aggregates, he suggests, are not many cats but almost one. They share almost all their parts, differ in only negligible respects, and stand in a relation of near-coincidence that approaches but does not quite reach strict identity.

This proposal introduces a non-standard relation into our metaphysical toolkit. Almost-identity is reflexive and symmetric but not transitive, and it comes in degrees. Two aggregates differing by a single molecule are almost-identical to a very high degree; aggregates differing by a whisker somewhat less so.

The payoff is that counting-by-almost-identity yields the ordinary answer. When we count cats, we count equivalence classes of the almost-identity relation, or we count by identity-enough-for-present-purposes. There is one cat because the candidates collapse together for counting purposes.

The account is not without difficulty. Almost-identity is a strange relation, and its non-transitivity generates familiar sorites-style pressures. Yet Lewis's proposal has the virtue of taking the metaphysical situation at face value while explaining, rather than explaining away, our ordinary practices.

Takeaway

Sometimes philosophical progress consists not in eliminating apparent multiplicity but in identifying the coarser-grained relations by which we legitimately treat the many as one.

The Problem of the Many reveals that even the most mundane acts of reference and counting rest on substantive metaphysical presuppositions. Any adequate response must engage with the underlying ontology, not merely the semantics.

Supervaluationism and almost-identity represent two systematic responses—one preserving classical semantics through precisification, the other introducing graded metaphysical relations. Each illuminates a different feature of the puzzle.

What remains clear is that our ordinary talk of ordinary objects is metaphysically less innocent than it appears. The cat on the mat is either many, or almost one, or determinately singular only relative to a chosen sharpening.