The Four Dimensions of Artifacts

The aim of this paper is to build an upper-level ontology of artifacts. To this end, two formal tools are employed: theory of consequence operation and ontology of states of affairs. It is argued that the adequate representation of an artifact consists of the representation of the artifact’s purposes, the representation of its design, the representation of the background knowledge according to which it has been designed, and the representation of instructions of use relevant for that artifact. I submit a series of constraints to be satisfied if our production and use of artifacts are to be rational activities. Since mereology is claimed to be an important part of every theory of artifacts, I sketch a few preliminary remarks on the prospects of formulating the mereology for the functional parthood. In particular, I explain why the principle of transitivity, the axiom of general sum, and the supplementation principles are false for that relation. Finally, the ontology of artifacts is evaluated against the desiderata formulated by P. Vermaas and W. Houkes for theories of artifact functions. Imagine that the knowledge modeller complains to the philosopher about her headaches with artifacts. The latter persuades her that philosophical ontology provides universal and unique tools for representing knowledge on all conceivable kinds of objects. The knowledge modeller decides to corroborate her new conviction. She adopts mereology, one of the formal tools recommended by philosophers, in order to model some domain of artifacts. To her surprise, she finds out that given a collection of artifacts she may combine them into only one artifact because of the extensionality of the mereological parthood. To her even greater surprise, she discovers that the object called ‘the mereological sum of all diodes in the world’ is an artifact and the description ‘the mereological sum of all diodes and all bicycles’ denotes also a single artifact. The knowledge modeller lodges a complaint against these absurdities with her friend. What will the philosopher say? I would say that it is well known that the standard system of mereology (devised by Stanisław Leśniewski as a substitute for set theory) is not directly applicable to all domains of objects. If you do not use it in accordance with Leśniewski’s prescription, you may apply it to abstract objects such as regions of space or periods but not to artifacts.1 Artifacts have functional parts that are 1 Applying mereology to regions of space Casati and Varzi (Casati and Varzi 1999) claim that (i) the standard conceptually different from mereological parts. In order to give a sound definition of functional part I construct a formal ontology of artifacts. The resulting ontology is a tentative proposal of a reference ontology for the domain of artifacts (cf. Guarino 1998). My proposal is both ontological and representational as it answers both the question ‘What is an artifact?’ and the question ‘How to speak about artifacts?’