The article provides an example of the application of the techniques and results of historical linguistics to traditional problems in the philosophy of language.$<$span class='Hi'$>$$<$/span$>$ It takes as its starting point the dispute about the nature of facts that arose from the 1950 Aristotelian Society debate between J.$<$span class='Hi'$>$$<$/span$>$ L.$<$span class='Hi'$>$$<$/span$>$ Austin and P.$<$span class='Hi'$>$$<$/span$>$ F.$<$span class='Hi'$>$$<$/span$>$ Strawson.$<$span class='Hi'$>$$<$/span$>$ It is shown that,$<$span class='Hi'$>$$<$/span$>$ in some cases,$<$span class='Hi'$>$$<$/span$>$ expressions containing the noun fact refer to actions and events;$<$span class='Hi'$>$$<$/span$>$ while in other cases,$<$span class='Hi'$>$$<$/span$>$ such expressions do not have a referring function at all.$<$span class='Hi'$>$$<$/span$>$ Thus,$<$span class='Hi'$>$$<$/span$>$ nothing corresponding to Strawson's $<$span class='Hi'$>$$<$/span$>$'pseudomaterial correlate'$<$span class='Hi'$>$$<$/span$>$ need be postulated in order to account for the reference of the noun fact $<$span class='Hi'$>$$<$/span$>$. It is suggested that many philosophically problematic expressions may be better understood by tracing their historical evolution in natural language.