Mainstream contract theory often makes a distinction between things that are observable by the contracting parties but not verifiable by a third party. This distinction has, e.g. been used to explain why contracts are incomplete in some dimensions and claimed to provide the foundations of incomplete contract theory.

This article sets out to scrutinize the assumption and argues that if something is observable by the parties, the parties can ex ante, without prohibitive costs, make these features verifiable by a third party. This implies that the distinction between observable and verifiable cannot be a fundamental explanation for incomplete contracts. Ex post it can be difficult to verify statements by a third party, but then it must be explained why the parties had not ex ante taken measures to make it possible to verify ex post. A number of such explanations are discussed in the paper. The arguments put forward are based on two detailed examples and also supported by important theories from the philosophy of language, which underline that concepts must be related to something observable in order to be learnt.