We outline an argument that a single-particle universe (a universe containing precisely one pointlike particle) can be described mathematically, in which observation can be considered meaningful despite the a priori impossibility of distinguishing between an observer and the observed. Moreover, we argue, such a universe can be observationally similar to the world we see around us. It is arguably impossible, therefore, to determine by experimental observation of the physical world whether the universe we inhabit contains one particle or many—modern scientific theories cannot, therefore, be regarded as descriptions of ‘reality’, but are at best human artefacts. Our argument uses a formal model of spacetime that can be considered either relational or substantivalist depending on one’s preferred level of abstraction, and therefore suggests that this long-held distinction is also to some extent illusory.