Enter ‘Vegetius’ on Amazon Books, and one of the items that appears is Epitoma rei militaris (Oxford Classical Texts) by Vegetius and M. D. Reeve. Treating an editor as a co-author is an amusingly literal version of a simple fact: editors do create the authors they edit, in that the text of author X in editor Y's edition is made up of the sum total of editor Y's textual decisions. In most cases the variation between the authors created by different editors does not fundamentally affect the impression given of the text: Cicero will look Ciceronian whatever edition of his work is used, and while my text of the Metamorphoses differs from W. S. Anderson's in more than 600 places, the two texts are recognizable as versions of the same poem. But there are instances where the overall impression of an author can be shifted by differing editorial decisions. Most such cases turn on the issue of anomaly versus analogy, that is, the extent to which departures from standard usage in the transmitted text of an author reflect genuine idiosyncrasies of style as opposed to scribal deformations. Tacitus (especially the Annals) is one example, Valerius Flaccus another; but nowhere is the question posed so starkly as in the case of Propertius. Here contrasting views of what kind of poet he was, and corresponding differences in dealing with his text as transmitted, have in recent years produced a divergence so great that J. S. Phillimore's often-cited phrase Quot editores tot Propertii (‘as many Propertiuses as there are editors’), originally a warning of the anarchy that would ensue if editors transposed large numbers of couplets, now seems like a fairly accurate description of the present situation.

Disputes about the text of Propertius involve many of the issues discussed in earlier chapters, for example, recension (specifically, disagreement over the shape of the stemma), the proper scope of conjecture, weighing of an author's habits of expression as they can be elicited from a controversial transmission, the place to be given to interpolation.