From language to ontology

A lot of work in ontology – especially the ontology of the ordinary, spatiotemporal world – relies on linguistic analysis and intuitions. In order to establish what there is, one looks for the truth-makers of our true statements. Alas, this is no straightforward business. For one thing, the “surface grammar” of what we say can be misleading. But neither can we trust its “deep structure”, for there is no unique way of telling what it is. No analysis can reveal the deep structure of a given statement; at most we can fix it by dint of resisting alternative interpretations. Depending on what we think there is, we must attach a meaning to what we say. Going the other way around – I argue – is illegitimate; it amounts to attributing our own ontological views to the language we share with others.