When and Why Do Scholars Use Secondary Scholarly Sources?
1) To establish, at the start of the paper, what has been said in general about their topic. (E.g., Hawthorne's treatment of women in his novels; Hawthorne's use of "nature vs. civilization" images; the early nineteenth-century public norms for females' affectionate behavior in public places) This allows you to set up your best readers' understanding of what other scholars have said, and you can launch your own thesis as an addition to, modification of, or negation of one or more of those opinions.
2) At the point at which the question arises, to prove that an author could have known about or even cared deeply about some event, work of literature, idea, person, place, or any other kind of thing which you believe the text you're interpreting refers to or which should be part of our context for interpreting that text. (E.g., Hawthorne's reference to Swift in the Lost Notebooks strongly suggests familiarity with Gulliver's Travels but did he ever publicly comment on Swift's satire?; Robin ("MK,MM") looks into a tavern and sees people smoking and drinking, activities the narrator describes as sinful, but does NH ever use them as representative sinful activities in his other works?; Giovanni ("R's D") takes rooms in Padua to attend its university--does NH have any personal experience of the Paduan city-scape or does the town figure in any literary works he would have been aware of?) This allows you to develop intuitive hunches about the emphasis you believe you see the author adding to an apparently trivial detail so that its meaning can help your thesis explain it as part of a related pattern of significance in other details.
3) In the introduction if essential, or at the place where it's needed if more minor support, to establish the date of a text or event that your author may refer to or to identify an unusual thing or person in the text which may be an important piece of period or thematic detail. (E.g., the innkeeper ("MK,MM") is a "French Protestant," so when and under what circumstances did the Huguenot refugees come to the American colonies and how might it affect the innkeeper's attitude toward what the town plans to do to the Major; at Goodman Brown's burial, the narrator tells us that "they carved no hopeful verse on his tombstone," but what might they typically have carved and what does its omission suggest?; Baglioni calls Rappaccini a "vile empiric . . . not to be tolerated by those who respect the good old rules of the medical profession"--so what does it mean to be an "empiricist" in Hawthorne's day and what were the "good old rules" that empiricists violated?) This allows your thesis to become more accurate in its description of the text's "mental world" by infusing it with the facts of the author's own world as they are relevant to the text. (Remember that the O.E.D. can give you quickly very precise, dated definitions of words in their meanings at the time the stories were written.)
4) In the introduction if essential to the thesis, or at the place where it's needed if more minor support, to supply statistical information that might put the paper's topic or thesis in perspective. (E.g., how many of Hawthorne's short stories involve male protagonists?; in the year in which Hawthorne wrote the story you're working on (ca. 1830) or in the year in which it was set (ca. 1730), how many violent crimes were committed in his region, especially crimes involving mob action?; what percentage of the plant world is actively poisonous to humans in its native, unenhanced state?) This information allows your reader to gauge the "unusualness" or "frequency" or "enormity" of the issue you are exploring, perhaps to see whether NH is inflating or deflating it for literary effect, or whether he takes unusual interest in it for some literary-structural (rather than "historical") reason.