Past studies of the variable (ING) have demonstrated regular and stable social and stylistic conditions across English speech communities around the world, factors which shape the patterns of variation between the /n/ and /ŋ/ variants of (ING). This dissertation is an inquiry into the conditions, both linguistic and social, which gave rise to the evolution of modern (ING). The purpose is to demonstrate the existence of an observable continuity between the past morphological history of (ING) and its present-day status as a sociolinguistic variable, following Sturtevant's idea (1917) that the social evaluation of linguistic forms can be viewed as the result of a competition between forms originally not in variation with each other, but which are later brought together in a shared environment, becoming variants of a single form. This study is based on a quantitative analysis of 7950 tokens of -ing, including synchronic data based on 68 speakers (60 British and 8 American) and diachronic data based on textual materials (letters and diaries) from the 15th - 19th centuries, and a supplement of data from an earlier study (Irwin 1967) based on textual materials from the 9th - 15th century. A grammatical effect on modern (ING) is found, seen most clearly in the British data, showing that /ŋ/ is correlated with nominal categories, and /n/ with verbal ones. This alignment does not pattern according to discrete syntactic features, but aligns probabilistically along a linear continuum. This effect is interpreted as the reflex of a partial merger between Old English morphemes ing and ind . There is an observed correspondence between the synchronic British data and the Middle English isogloss c. 1450, which established the replacement of the participial suffix ind with the verbal noun suffix ing in the south of England (Moore, Meech and Whitehall 1935). The modern cities which show probability of /ŋ/ less than .5 fall outside the isogloss, those with probability of /ŋ/ greater than .5 lie within it. The difference in probabilities is consistent with the idea that the replacement of -ind with -ing c. 1450 occurred in southern England sooner than in northern England because of a difference in the pronunciation of the two suffixes in these regions, northern -and versus southern -ind. Evidence supports the view of a syncretism between verbal noun and present participle subsequent to the replacement of -ind with -ing. (Abstract shortened with permission of author.)