Comma separated list within citation...

jdchess, to answer your spcific question, often it is a matter of juggling words to find the clearest expression of our intent. The two basic grammar rules that apply are these:

1. Items in a series are separated by commas, with an "and" before the last item (as you've noted).
2. When one item in the series carries internal commas, then we substitute semicolons for the major divisions.  (Example: Flags were on vivid display: the red, white, and blue of the United States; the red and yellow of Spain; the red, white, and green of Italy …)

In citations, things get more complicated. A complex citation may already have semicolons to separate the major divisions—as with layered citations. In those cases, we can sometimes clarify matters by adding bridge words rather than a punctuation mark. To use your example:

To better clarify divisions between items, we might handle that string of commas this way:

"South Carolina, Wills and Probate Records, 1670–1980," digital images, Ancestry (https://www.ancestry.com : 2015), will, inventory, and administrative papers for Shirley Whatley, 5 July 1852, Edgefield District; citing...

However ....

In the case at hand, we might reconsider the wisdom of featuring the database as the first element of our citation and then clumping all the significant records in to the field where we normally cite a specific database item. Especially should we consider this when we are not citing the database entry at all—but, rather, citing an image of the actual record.  The approach above raises a number of issues.

In the citation above, you have three separate documents but just one date. For certain, all those documents don’t carry the date 5 July 1852. Which document does the date represent? What are the dates for the other documents? Is this a probate case that was settled in the same year the will was written or does it string across several years?

We also need book: page identification—or collection: file identification—for each record cited. Which one of those documents supports the assertion to which you are attaching this citation?

When I call up this database at Ancestry and type “Shirley Whatley” into the name field of the search form, with “1852” as the year and “Edgefield” as the county, I get no results. When I choose “browse” for the database and then select Edgefield, I’m given two separate options. Within those options, I’m given many other options that break down by dates (with overlapping dates between the options) and by record type (bound registers vs. packets of loose papers).  How do we go about finding the specific document(s) that support whatever assertion you are making in your research files?

A record such as this really begs for a different approach. Layer 1 needs to cite the original record. Layer 2, then, would say it was imaged in “Database Title,” Website Title (URL : date), path > ____ > _____ > [down to the actual image that supports your assertion]

QuickLesson 25 goes into more instruction on paths (aka waypoints) at https://www.evidenceexplained.com/content/quicklesson-25-arks-pals-paths-waypoints-citing-online-providers-digital-images. EE has numerous other examples at 7.18, 8.18, 8.24, 9.9, 10.6, 10.39.

jdchess, your last draft works—with the exception of the date, which I'll discuss below. Your reasoning in messages 4 and 5 is on target.

As for citing a file versus a document: There are some instances in which we might cite an entire file; but almost always our citations will be to the document, not the file. As an analogy: when we cite a book, we don't usually just cite the book without saying where in the book one looks to find what we've asserted. We identify the book, and then we cite the specific page, plate, or part. Document files work the same way. We don't take a specific piece of information from the file as a whole. We take that piece of information from a specific document; so we cite that document, the creator of that specific document, and the date of that specific document.  Sometimes, a piece of information  might appear in multiple documents, in which case we'd cite them both or all of them.

As for the date: within a file, different documents involve different people and different sets of circumstances. When we take a piece of information from a specific document, we need to know the date that information was valid, or the date that event occurred. or the date that Persons A, B, and C participated in that event. This date affects our analysis of the data in endless ways. In those relatively rare occasions when it is appropriate to cite a whole probate case, whatever date we cite for the case needs to be explained as to what that date represents. In this case, would it be the date that probate was opened? Or the date probate was finalized?

As for a layer 3 (citing the source of your source), most of the data is already cited in the first layer. In layer 3, you might add where Ancestry says that record set is located.