> It's a familiar conceptual thing, not a familiar real thing.

Not exactly. Unlike unicorns, 1954-style songs *do* exist.

The point is that there's little true consensus about how many of their identifying features must be present in a song to mark it as "folk." And for transcultural studies, Gibb raises the very pertinent point that the 1954 def. may not make any sense.   

And the waters have been muddied further by the transatlantic, intentional, and intellectually unscrupulous (surprise!) commercial marketing of other songs as "folksongs." And the popular perception - furthered in America by the very influential schoolbook "The Fireside Book of Folk Songs" around 1950 - that national songs like "The Star Spangled Banner," "God Save the Queen," and the Soviet version of "Meadowlands" are *also* "folksongs.

So besides a manageable complex of expert interpretations of and adjustments to the '54 specialist def, we also have a competing and universally familar *second* definition that's only tangentially related to the original.

Copmpare the popular use of "virus" to mean "any disease-causing micro-organism, including a bacterium."

It drives microbiologists crazy! Viruses and bacteria are far more different biologically than cats and dogs. Or cats and snakes.

"Virus" (like "proton") designates something whose misidentification by experts could lead to serious real-world consequences. But if an expert "mislabels" a "folksong," nothing much happens. What's more, as I mentioned yesterday, it's to the advantage of some scholars (or "scholars," Gibb might put it) to stretch the category even further in the direction of their choice, because it gives them something new to write about. (It also makes them look like bold innovators to
other "scholars" ("So 'Yes, We Have No Bananas!!' is really a folksong! I never thought of that! Care for tenure?").

That encourages semantic change even at the top. That's what we see in the various definitions, interpretations, and perceived permutations of the word "folksong."

And that, as they say, is life in the swamp. The label, as I cannot repeat too often, is far less important than what we have to say about the song.

If I can show that "Yes, We Have No Bananas!!" has some unnoticed relationship to "Barbara Allen" (joke) how much does it matter how I label them? That would *not* be true if I were a specialist talking about protons or viruses.