In his article "South Mountain: Three Gaps, One Battle," Steven R. Stotelmyer opens his essay with a note of finality: "The purpose of this paper is to set the historical record straight regarding the nature or character of the Battle of South Mountain."
The historical record has a way of changing with each new discovery in primary research; taking the hard line of "case closed" (or as Sears once named an essay, "Last Words on the Lost Order") is ahistorical. You might fairly try to square the current historical record with "the nature or character of the Battle of South Mountain," but you are making an argument and we, the jury of readers, will decide whether we have been convinced or not.
The author sets his problem forth with a question: "Do the events of September 14, 1862, occur as separate battles on South Mountain, or are they actually part of a single battle of South Mountain."
This question has two components. "To set the historical record straight" can try to answer whether the historical record currently supports separate battles or not and to what extent if so. But there is also the philosophical issue of what constitutes a separate battle. This philosophical part has an historical piece (what convention was used at that time for determining the question) and an ahistorical piece (what is in fact, in all times and cases, a separate battle).
The core failure of this essay is that its author ultimately will develop an ahistorical solution - the modern doctrinal definition of battle - to solve what the author has earlier defined as an historical problem. After taking a tour of the historical evidence, he travels off into Army doctrine to close his case.
The historic evidence handling is not good and the Army doctrine is erroneous by my experience and sources, however these are smaller problems than having defined the problem one way and then solving it through other means.
More on this to come...