Barry Smith and Achille C. Varzi (2000), Philosophy and Phenomenological Research 60(2), 401–420 — DOI-10.2307/2653492
Key takeaways
Read in full (NOTE-721).
- A boundary is bona fide when it follows a spatial discontinuity or a qualitative heterogeneity of material, and fiat when it follows none and exists by demarcation: perception, language, law or a mathematical definition. The distinction is “absolute”: building a fence along a border, or cutting along a fiat line, makes a new bona fide boundary; it does not transform the fiat one.
- Bona fide boundaries obey ordinary closure topology: a boundary belongs to one side, so two closed bodies cannot touch. Fiat boundaries are parts of what they bound; each side has its own, and the two coincide (the equator, the Mason–Dixon line).
- Distinct bodies are never in genuine contact, so any object made of separate bodies is a fiat object, possibly scattered. Fiat objects are still not arbitrary: real properties of the material enable and constrain where fiat borders can be drawn.
- Self-connectedness has to be defined by fiat connection, so the bona fide theory presupposes the fiat one. The formal part is a sketch, with no models or metatheory, and the theory is spatial and synchronic.
Standing in the record
Filed for CLAIM-194, the counter that a gap or coupling criterion cannot tell a boundary a system produces from one imposed on it (a cell membrane and a beam joining two pendulums pass alike), and for QUESTION-035, where a combined form’s boundary lies. The survey of the Mathematics of Meaning corpus pointed to it (MOM-LIT-016).
Fiat/bona fide does not map onto produced/imposed. Bona fide asks whether a discontinuity is there; produced asks who makes and keeps the boundary. The membrane is bona fide and produced, the beam bona fide and imposed, the equator fiat and imposed, and the paper has no category for a boundary the bounded thing draws for itself. It sorts the membrane and the beam together, as the gap criterion does, which is indirect support for CLAIM-194’s point and no answer to it. The paper’s “imposition” of boundaries is an observer’s demarcation, not CLAIM-194’s externally maintained constraint.
The coupled pair’s boundary is fiat on the paper’s terms, and not only because a target is chosen: separate bodies are never in bona fide contact, so the pendulums with their beam, and the dancing couple (CASE-046), are fiat objects. The distinction therefore classifies every answer to QUESTION-035 as fiat and cannot choose among them. What it offers the record is the vocabulary of a fiat boundary that real coupling strengths enable and constrain, which fits a target-relative block (CLAIM-157).
Where the resemblance fails. The theory is spatial and has no time or process; its heterogeneity is of material, not of interaction strength. Reading a near-decomposability gap as bona fide would extend the paper. Its dependence of a boundary on a host is ontological, not the causal maintenance at issue in self-production (Weber & Varela, LIT-1018, filed beside this).
Varzi’s SEP entry on mereology (LIT-124) is the neighbour already held.