‘The ideal procedure would have been to make a slip for each word, with the provenance of each clearly stated (i.e. the year, journal, day of week and month, … page and column); to have kept these slips to the end of the war; sorted them out alphabetically; and sufficiently annotated them’, wrote Andrew Clark, reflecting on his attempts to collect and collate the language of WW1. Like James Murray (editor of the OED 1878- 1915), he, too, was preoccupied by the image of the ‘Ideal Dictionary’. Clark’s self-criticism, however, fails to do justice to what he achieved. Archived in the Bodleian Library are almost 100 notebooks and files, many of which are indeed replete with carefully dated slips, provenance, annotation, and alphabetic organisation. In reality, in what remains an almost entirely neglected work, Clark appropriated OED methodology (with some strategic innovations), but chose to focus on the documentation of ordinary discourse, excluding the canonical writers who increasingly featured in the then on-going first edition of the OED. His prime sources are located in newspapers, advertising, and ephemera of all kinds.  In tracking language on the move in WW1, anything might, he argued, be a text, and capable of exhibiting the ways in which language mediated a period of unprecedented historical change.  The geological metaphors of OED1 are discarded. Clark’s historical principles are reoriented to real-time history. This paper will examine Clark’s lexicographic contributions, and his attempt to create a strikingly contemporary narrative of words in time.