Probable genetic relationships among Old and Middle English manuscripts have been established according to affinities in their variant readings. Most early twentieth-century scholars in this field have collated manuscripts for differences and similarities in word order, lexical choice, and lacunae, and have based their stemma upon evidence of constant, consistent, and non-contradictory groupings of variant readings. In these cases, the variant forms have served to support inferences about descent. However, the manuscript forms are interesting for another purpose; they can provide significant linguistic data for studies in the phonological history of the language. Those manuscripts which share a common ancestor may show in their variant readings spelling differences attributable to uniform diachronic or synchronic changes in sound patterns. Old English wills and charters which can be dated either internally or externally and can be extra-linguistically identified for geographic origin can be used to establish whether those of the same geographic area and same period consistently represent the same words or sets of words with the same graphic configurations, whether those of different geographic origin and the same period show a regularity of correspondences among them to be properly called dialect differences, and whether those from the same geographic area but of different periods reflect consistent differences in form. Single manuscripts transmitted by more than one scribe may contain orthographies or spelling practices peculiar to each. In addition to manuscripts, the term "text variant" can be extended to include strictly linguistic material: idiolectal or dialectal phonetic transcriptions, utterances of a single informant in differing extra-linguistic contexts, alternative phonemic analyses or lexical representations of a single body of data, mispronunciations or misspellings of a group of non-native English speakers of same or similar first language origin. Common to all of these examples is a text or a set of texts which contain enough similarities and enough differences of vocabulary and spelling to present a problem which computer processing can best handle. The programs described here can aid in accumulating details of similarity or difference in terms of word-form and word-graph by (1) building a matrix, columns representing texts and rows representing words, to show what words appear or do not appear in the texts, (2) providing a list of forms unique to any given set of texts, (3) supplying a list of text-coded