Hi John,

I have been wondering the same. Perhaps the only really "safe" way to do a literary survey is to go chronologically by date of publication...and yet my own interest is in something other than a history of *writings on* chanties, so the attempt has been to follow the chanties, themselves, chronologically. Considerably less certain, but up until the 1870s I don't think it was so bad. There weren't *too many* instances of the publication date being far later than the time described. And even those later publication dates might be said to have been in the "not yet too contaminated" era which is worth something.

As I said, 1870s got more tough. And I felt unsure, for example, about using THE CLIPPER SHIP SHEILA, 1912, to get statements reported to be about 1849. I'm a little worried I've mixed in bad data with good. Having included it speaks to the multiple purposes I have in mind. One of those is simply recording and extracting info from the texts that are not focused on chanties.

Even though the chronological (by date of publication) survey is not my main goal (it is certainly of interest to me, and it's sort of hovering in the back of my head at the same time)...nonetheless I think the best method at this point would be to stay grounded by following that technique. I am going to continue looking for and discussing references published in the earliest times possible -- now that basically means 1880s. I'll deal with, say, Whall's chanties (possibly all learned in 1860s-70s) once we get to the 1910s, and retroactively add them to the various lists.

I expect to go at least into the 1910s before stopping; I have lots of bookmarks up through then!

In the end, the data will still all be there, to be rearranged however one wants to. But I think if we move along this way -- keeping the date of attribution as close as possible to the date of 'publication' (/recording) -- then it will remain most coherent.

Gibb