In the context of the scientific method a hypothesis is a testable explanation for an observation or set of observations. A statement which appears to be a hypothesis but is constructed in such a way that it cannot be "dis-proven" (because of ambiguity in its syntax, for example) is non-falsifiable.