Feedback
Guidelines on traditionals
Well I was thinking about the discussion about "Working on a Building" in this thread: http://www.secondhandsongs.com/forum/index.php?topic=3853.0.
To me, if a song is traditional, it means there is no original, usually because it dates back to before the beginning of the recording era and also because no-one knows who wrote it. Probably no-one wrote it, it evolved. This means in a sense every recorded version is a cover. Quite often, on a record, a traditional song is credited to "Trad. arr. the artist". Perhaps I'm misunderstanding, but in that thread you seem to be saying that one particular version "Trad. arr. The Carter Family" is an original. In the database (I'm Working on a Building this song is credited to The Carter Family, but in the thread you say it is traditional. So for the Cowboy Junkies' version to be a cover, do they have to be using The Carter Family's (or someone else's) arrangement? If they are using their own arrangement, does it still count as a cover? For me it does, but for SHS? Or are you saying something else? Are you saying the Carter Family's version is the first known recorded version and counts as an original by virtue of being first?
A contradictory case is "Scarborough Fair" (Scarborough Fair This is in the database without any claim of who recorded it first, it follows the "every recorded version is a cover" philosophy. I guess versions are added to the database as they are discovered, and in time the earliest version will gradually emerge. This is quite an easy approach, as I guess setting out to discover the first recorded version could in some cases take a lot of detective work...
I would like to research some other traditional songs, such as "The Great Silkie" and "She Moves Thro' The Fair", which are not in the database at the moment. "The Great Silkie" may have inspired the Byrds' "I Come and Stand at Every Door". "She Moves Thro' The Fair" was recorded by The Trees, Fairport Convention and Davey Graham and possibly others, and may have inspired Led Zeppelin's "Black Mountain Side". I have the Trees' version (they recorded quite a few traditional songs) and they credit it to "Trad. arr. Trees". Is it enough for me to post to "Add Songs To Database" saying Original: traditional; Cover 1: the Trees; Cover 2: Fairport Convention; and so on? Do these count as covers for SHS purposes, if they all use their own arrangements? (If not, I suppose there is little point in doing the research as far as SHS is concerned, although to me there is value in someone going to SHS and finding a Trees song was originally traditional.) Or do I need to find out how close the Trees' arrangement was to Fairport's (assuming theirs is credited to "Trad. arr. Fairport"), to discover if one of them is a cover of the other's arragement, and so on? Do I need to make an attempt at discovering the earliest recorded versions of these songs, or is the more relaxed approach of Scarborough Fair sufficient?
So, that's why I need a guideline... :-)
Jon
______
Really wild, General!
Jon, I am pleasantly surprised by your views of the matter. Yes, we have a problem: we haven't defined clear guidelines for traditionals. We simply didn't see it coming, and by consequence we dealed with it "ad hoc".
But you're right, we need a clear guideline, and your post will help us to define one.
1) Defining the original
SHS defines the original as "the first public version" of a song, "public" being the key of the definition. In this sense, traditionals have no original.
Example: Scarborough Fair
2) Cover hierarchy
One of the strengths of the db is in my opinion that we can work with cover hierarchies: if one artist altered a song in a significant way we can set him as "new original": all songs clearly based on the latter are linked to it.
Example: Scarborough Fair/Canticle
What are your thoughts about this first try?
Unfortunately, this is one of those issues that the more you talk about it, the more complicated it becomes...
"1) Defining the original
SHS defines the original as "the first public version" of a song, "public" being the key of the definition. In this sense, traditionals have no original."
Well, yes I agree that traditionals have no original, or at least that the original will never be known. But nevertheless, a song should at least be acknowledged in SHS as being traditional, as it is in "Scarborough Fair". So, in this sense "Working On A Building" is not correct. At the moment, it is:
* I'm Working on a Building by The Carter Family (May 8, 1934)
o I'm Working on a Building by The Jordanaires (October 1950)
o I'm Working on a Building by Bill Monroe and the Bluegrass Boys (December 5, 1954)
o I'm Working on a Building by Jimmie Davis (1962)
o Working on a Building by Old & In the Way (1997)
If it follows the "Scarborough Fair" example, this is wrong, it should be:
* I'm Working on a Building written by traditional () * I'm Working on a Building by The Carter Family (May 8, 1934)
o I'm Working on a Building by The Jordanaires (October 1950)
o I'm Working on a Building by Bill Monroe and the Bluegrass Boys (December 5, 1954)
o I'm Working on a Building by Jimmie Davis (1962)
o Working on a Building by Old & In the Way (1997)
This assumes "Working on a Building" actually is traditional, perhaps even that is difficult to establish sometimes...
"2) Cover hierarchy
One of the strengths of the db is in my opinion that we can work with cover hierarchies: if one artist altered a song in a significant way we can set him as "new original": all songs clearly based on the latter are linked to it."
In the case of "Scarborough Fair" we know what Simon and Garfunkel added to/changed in the song, but do we know what The Carter Family did to "Working On A Building"? Does it mean we need to know an earlier version to compare (in this case it seems we do not have one)? Perhaps we can only compare with anecdote, because there is no earlier recorded version. For me, this is difficult, because the web is really my only research resource and if the anecdote is not on the web I won't know it. Is the anecdote believable/verifable? Maybe there is not one traditional version, but many. And, what is 'significant' (probably a whole discussion in itself...)?
I think it's best to keep things simple. If a song is known to be traditional, that should be acknowledged. Then, we can proceed as for any other song. So, if I start to research "She Moves Thro' The Fair", I can start as follows:
- She Moves Thro' The Fair written by traditional ()
o She Moves Thro' The Fair by The Trees (1971)
o She Moves Thro' The Fair by Fairport Convention (197x)
Even this simple information is valuable for the database. We can add more as we discover it. The Trees' version (the only one I own) is credited to "Trad. arr. Trees". Perhaps this actually means "Trad. arr. Trees based on a 1950s record from the Trees' childhoods", in which case we can add this when/if we discover who made that 1950s record (or add a comment if it's not a 1950s record but a 1950s performance in a pub). It does not make the initial list wrong, just incomplete. Another strength of the db is that new versions and hierarchies can be added at any time. In this sense, the only difference between a traditional song and a non-traditional one is in the first line.
It seems to me we have to keep it simple. Even in the recorded era, traditional songs can evolve in a way that non-traditional ones do not. But the path of that evolution will be different in every case, so we can't really make any rules. Maybe the "ad hoc" approach is best...
Jon
Last edit: 2006-06-18 13:53:31 UTC by baggish
______
Really wild, General!
Jon, I think we're moving into the right direction. Given your comments, I'll reformulate my draft guidelines for traditionals:
1) Defining the original
SHS defines the original as "the first public version" of a song, "public" being the key of the definition. In this sense, traditionals have no original.
In practice this means we will create an original, with credits to "traditional", and without any performer.
Example: Scarborough Fair
2) Cover hierarchy
If an artist altered a song in a significant way we can set him as "new original": all songs clearly based on the latter are linked to it. Significant means in this case or we know exactly what has been changed or added in the new original.
If it is unclear who changed what, it is better to keep it simple and link a cover to the "original original".
Example: Scarborough Fair/Canticle
And yes, given these guidelines we need to add a traditional original for "Working on a Building". As for later covers, I guess they often credited "Alvin carter", and we therefore linked these later versions to the Carter Family version... does this make sense to you?
Well, I've done some research into "Working On A Building", and I will add to the existing thread http://www.secondhandsongs.com/forum/index.php?topic=3853.0 in a few days' time. But the song is quite a good test for any guidelines about traditional songs (hopefully they won't all be as complicated as this...).
i) It's difficult to find information about such old songs on the net. It's easy to find half-truths and supposition. I found one comment on Amazon that I think gives a good picture of how Alvin Carter may have worked: http://www.amazon.com/gp/product/B00005TPB7/ref=dp_return_1/103-8183640-5017417?… Tony Thomas' comment. He collected songs, and also wrote his own. I don't know this for sure as I haven't seen a picture of the label, but it's possible the Carter Family release of "Working On A Building" is credited to Carter. If so, a question for the guideline is what to do when the first known release of a traditional song is credited to someone.
ii) I found four definite covers of "Working On A Building". Two of them, Bill Monroe (already in database) and Charlie Daniels (2005 cover not in database), credit Carter. Two of them, Old & In The Way (already in database) and Cowboy Junkies (not in database, origin of this discussion...), credit Traditional, but both of them acknowledge Carter on their websites. They all sing the same lyrics. Old & In The Way is recognisably the same song, Cowboy Junkies is a bit slower than the others (this is normal for them). So, a question for the guidelines is what to do about these different credits, how do they affect the cover hierarchy (if at all)?
This one acknowledges Carter's role in collecting and preserving the song and putting his name to it:
* I'm Working on a Building written by traditional () * I'm Working on a Building by The Carter Family (May 8, 1934)
o I'm Working on a Building by Bill Monroe and the Bluegrass Boys (December 5, 1954)
o Working on a Building by Cowboy Junkies (1988)
o Working on a Building by Old & In the Way (1997)
o I'm Working on a Building by Charlie Daniels (2005)
This one is based strictly on the credits of each version:
* I'm Working on a Building written by traditional () * I'm Working on a Building by The Carter Family (May 8, 1934)
o I'm Working on a Building by Bill Monroe and the Bluegrass Boys (December 5, 1954)
o I'm Working on a Building by Charlie Daniels (2005)
* Working on a Building by Cowboy Junkies (1988) * Working on a Building by Old & In the Way (1997)This one simply states that everyone covered a traditional song, but does not leave any room for significant variations:
* I'm Working on a Building written by traditional () * I'm Working on a Building by The Carter Family (May 8, 1934) * I'm Working on a Building by Bill Monroe and the Bluegrass Boys (December 5, 1954) * Working on a Building by Cowboy Junkies (1988) * Working on a Building by Old & In the Way (1997) * I'm Working on a Building by Charlie Daniels (2005)
iii) This is not really about traditional songs, but about defining the original as "the first public version" of a song. For 'Alabama Song' (Alabama Song for example, "the first public version" seems to be a live performance. On the other hand, for 'Echoes' (Echoes by Pink Floyd, "the first public version" is on the Meddle album. However Pink Floyd performed embryonic versions of this song live many times before that album was released, revising and perfecting it and testing audience reaction. Even the final version of 'Echoes' was performed live before it was available on an album, because the Meddle tour started in the US a month before the album was released. The Beatles playing their songs in Hamburg would be another example. So, my question is, when is it acceptable for a live performance to be the "the first public version"? For older songs? For show tunes? or something else (perhaps lack of information)?
Jon
______
Really wild, General!
Jon,
1) Working on a building
You have formulated some interesting insights in this traditional. My suggestion would be to post all of them to the "Report errors" board and continue the discussion over there. Otherwise it's becoming very messy here. Also the editor responsible for the entry will join the discussion.
2) "The First Public Version"
You're question is indeed very relevant: sometimes the first public version is a live peformance, sometimes a recording, sometimes a release. Normally you should take the first release as original. However, sometimes we pick the first live performance or first recording to keep things intuital.
The "Alabama Song" case: If we picked the first release here, you would attribute the original to the artist who by accident happened to release it first, although it was already very famous because of the musical. We thus pick the musical version.
The "Echoes" case: We keep it simple and set the first release as original. As nobody released it "by accident" before Pink Floyd, nothing "forces" us to pick an older Pink Floyd recording or performance.
Do you get the intuition? Now read again:
http://www.secondhandsongs.com/pmwiki/pmwiki.php/Guidelines/Original
Member
Posts: 15
Guys, this is a complicated (and interesting) question, but I expected you'd have a different answer.
The term "traditional" refers to a song that is known to have been around a long time, but for which the author/composer of the song iis not known. However, there was still at some point a first time that it was released on a recording medium. And wouldn't that be the "original," by your own guidelines? True if it is traditional, many people (maybe generations) would have sung and performed it previously, but as they were not recorded and released on a medium, wouldn't SHS exclude those versions from consideration?
as they were not recorded and released on a medium, wouldn't SHS exclude those versions from consideration?
That guideline only applies for covers, not for originals. An original doesn't need to be released.
Retired Editor
Posts: 1015
What qualifies as someone 'basing' their version of a traditional on a specific previous version - does it have to be 'credited'? or are we trying to listen to musical arrangements?
I believe it must be credited in some way to count as being based on it...is that what you guys are all thinking?
Well, if someone covers workin' on a building and credits it to A.P. Carter, this means he was somehow inspired by the Carter Family, but I think also a less explict evidence would be good to clarify the "inspiration degree" 
Some weeks ago I added this:
Song of the Shrimp ; it's not a traditional, btw.
As you may read in the comment, Frank Black never listened to Elvis' version, so it might be correct to say he didn't cover Elvis, but Townes Van Zandt. I'm not 100% sure this is the right approach, but it does make sense to me. Of course the problem comes out when you don't have such explicit "confessions"; in those cases you have to guess, and if one has to guess, he better make it easy and credit the song to the "original original", instead of taking a degree in musicology 
Maybe it's easier to compare the lyrics (at least it is to me).
I remember the time when Guns'n'Roses covered Dylan's "Knockin' on heaven's door": all the bands in my town did their version of the song, and they were all "inspired" by G'n'R: the singers all went: "Knock-Knock-Knockin' on heaven's door... ah, ah, ah-ah yeah!" They were quite sad, actually... But if I had a band and had to cover "Wanted Man", I'd use this version as a source ---> Wanted Man instead of the original...
______
坐低飲啖茶，食個包
Well, I just bought a 5CD set with all the Carter Family recordings 1927-1934. And I can't stop listening to it... I remembered of this thread, and I have some ideas about traditionals:
1) A traditional (unknown artist) must be added in the database only if it had been composed (by an unkown artist), published or recorded (by an unkown artist) before the first cover was recorded/released.
Examples:
A traditional (original) of this kind must have a (composition, publishing, recording) date. If the date is uncertain ("composed around 1550", "published ca. 1770"), we could use those dates and explain they're uncertain in the comment.
2) If there is no trace of a composition, publishing or recording date before the first "cover", then the first cover must be considered as if it was the original, and its words and lyrics credited to "traditional".
Examples:
I'll Take You Home Again, Kathleen
This condition applies when the artist doesn't copyright the traditional (=doesn't say he is the writer of the song).
3) "stolen" traditionals, just like a lot of A.P. Carter's songs.
If there is no trace of a composition, publishing or recording date before the first "cover", but it has been copyrighted by the first one who recorded/published it, then it should be credited to the first one who copyrighted it (eg A.P. Carter). This may (or may not, I'm not sure) be the case of Working on a Building. The reason is simple: a song, in itself, it's nonmaterial, so to speak. Sometimes it just floats in the air. The "owner" of a song could be considered the one who gives it a material form, for instance putting his signature on a music paper or recoding it on a medium.
Case 3 is very similar to case 2, the only difference is that, in case 3, the artist has been less "honest" and credited to himself a song he didn't write. But we're not here to judge, are we?
In any case, I think, a comment about the origin of the song and why we chose case 1, 2, or 3 is mandatory.
Hierarchy
:This is a though one, probably it'e easier to "feel" it than to know it for certain...
1) Does the covering artist give credits to another covering artist?
2) Does the covering artist quote another covering artist as a source?
3) Does the cover share some {+unequivocal+} similarities with another cover? A particual variant of the title, some particuar lyrics, a particual way to paly the song that {+clearly+} comes from another covering artist? Of course, what's {+unequivocal+} and {+clear+} may be subjective, and could cause a little trouble... So this option should be used carefully.
If the answer to any of these questions is yes, then the cover should be put under the "mother" cover, which becomes the new original.
Self-objection:
The Cowboys Junkies say that Working on a Building is a tradtional, instead in the database is credited to A.P. Carter!
Self-answer:
It may be that they're clever people and do know that A.P. Carter didn't actually write the song; it may also be that A.P. Carter's copyright had expired and wasn't renewed; it may also be that they just didn't know the song was copyrighted by A.P. Carter and thought it was a traditional. I've been an editor here for four months, now, and I must say that at least 20% of what's written on a record sleeve is simply wrong...
The point is that Working on a Building reached the fame thanks to the Carter Family, and it's unlikely that the Cowboys Junkies were inspired by a version of the song recorded prior to the Carter Family's, or that they found by a book from the 1910's publishing the song and decided to give their rendition...
______
坐低飲啖茶，食個包
I don't think you can make such a clear distinction between (1) and (2). The problem is that quite often it will be impossible to establish definitively what is the first cover. As you will have seen, I'm currently working on the 18th century traditional "Fair Margaret and Sweet William". The earliest recording I've found so far is from the late 1950's. There must be earlier recordings, but I probably won't find them. I think I would be inclined to go for (1) in almost every case. Except in some circumstances where the first cover 'discovered' the song (Lomax recordings, Carter Family, etc), the first cover is not that important.
Yes, I think there are many examples from the 1920's and 1930's of 'stolen' traditionals, as you describe for the Carter family. Probably you have read Tony Thomas' comment at http://www.amazon.com/gp/product/B00005TPB7/ref=dp_return_1/103-8183640-5017417?… which suggests Carter was not allowed to record copyrighted songs and so he had to find and claim traditionals. I'm sure he wasn't the only one with such a contract. I think we should still have a traditional original in these cases, if we know it is a traditional, something like this:
* I'm Working on a Building written by traditional () * I'm Working on a Building by The Carter Family (May 8, 1934)
o I'm Working on a Building by everyone else
Of course, supported by appropriate comments. This acknowledges that the Carter Family "own" the song and that they probably inspired every other cover, and also acknowledges that it existed before they claimed it, althugh perhaps not in recorded form.
So, I guess I'm an advocate for putting every traditional into group (1). Keeps things simple...
I did quite a lot of research into "Working On A Building" and had the opposite problem to you and "Rye Whisky": I found at least three different songs with the same title (the only variation being A or THE). I have no idea if they are completely separate songs or three developments of the same traditional original. The differences in the lyrics can support both views. However I enter these songs to SHS (three separate songs or three variations, or some combination) I will be making a statement of fact that I can't support. I will write a thread about it one day... I certainly feel that both songs (and presumably many other traditionals as well) reveal the limitations of being an internet researcher (speaking for myself...) and maybe there is a need for some more expert input from a musicologist. I certainly can't fully get to the roots of an American traditional song from the south coast of England... But it seems every traditional has developed in its own way, so it's almost impossible to formulate detailed guidelines because there will always be exceptions.
As for the Cowboy Junkies, they know "Working On A Building" is a traditional, and they know it is because of the Carter Family that they are aware of the song: http://www.cowboyjunkies.com/albums/trinitysession/index4.html.
Jon
______
Really wild, General!
[I should have said before that all my previuos consideration about so called traditionals start from SHS definition of an original as "the first public version" of a song]
If you say that "Fair Margaret and Sweet William" is an 18th century traditional, it means that there are traces of this song that go back to 18th century (a book, a music paper, some indirect attestation, or a recording 
). I agree that searching for them could be a pain, but I'm afraid there's nothing we can do about it... For this reason, I'd say that "Fair Margaret and Sweet William" is definitely a case (1): you don't necessarily need the first recording, IMO, because we know the date (although not the exact date) of its first public version.
When I wrote about case (2) I was thinking about Big Rock Candy Mountain (one of my favourite songs!). I remember I read the story of it: apparently Harry McClintock learned the song from a hobo, and recorded it some time later. In this situation, we don't know:
- who wrote the song
- when the song was written
- how different the "original hobo version" was from the Harry McClintock "cover".
Probably, instead of "traditional", we should talk about words + music: unknown or public domain. In this case, clearly the first "public version" is the Harry McClintock version, just like the first public version of Penny Lane is the Penny Lane single, not the McCartney demo, if there's any.
Another example: Suzanne's original version is Judy Collins'. Is it? I know that Leonard Cohen sang the song to Judy Collins on the telephone (!), she liked it and recorded it. Was the "telephone version" the first "public version"? I don't think so. Do we know for certain that Leonard Cohen never performed the song in front of an audience before Judy Collins recorded it? Likely, not sure.
Case (3) is just like case (2). The only difference is that now the first performer claims to be the author. As with the beatles' songs, everybody knows that Lennon and McCartney seldom wrote together: Eleanor Rigby is McCartney's, Tomorrow Nevers Knows is Lennon's. But if it's ok with them as authors, it must be ok with me as editor to say that those songs are Lennon-McCartney.
Alvin Carter didn't write Working on a Building, and this is a known fact. But if there is nothing in the world to prove that someone else wrote Working on a Building, the only thing we can do is credit it to A. P. Carter.
Hope you know what I mean, I have a feeling I was a bit confusing.
PS: AFAIK, Carter family songs were of three kinds:
- songs A.P. Carter wrote (yes, he did write some songs )
- public domain songs (or "traditionals"), which weren't copyrighted until A.P. Carter put his hands on them
- copyrighted songs, that A.P. Carter copyrighted again "forgetting" he didn't have the rights do it (example: Will the circle be unbroken).
The story as I know it is that he earned (more) money recording songs copyrighted by himself. But Tony Thomas may be correct when he says he simply wasn't allowed recording copyrighted songs.
______
坐低飲啖茶，食個包
Another example: Auld lang syne. Robert Burns composed it in 1788, but he actually just wrote someone else's song out.
http://www.chivalry.com/cantaria/lyrics/auldlang.html
Robert Burns sent a copy of the original song to the British Museum with this comment: "The following song, an old song, of the olden times, and which has never been in print, nor even in manuscript until I took it down from an old man's singing, is enough to recommend any air."
http://incompetech.com/authors/burns/auld.html
How to handle this?
One way would be:
auld lang syne written by traditional ()
* auld lang syne written by Robert Burns (1788) * auld lang syne by others...But this would be against our guidelines, that say a cover (in this case, Burns covering the old man singer) must be released and (this is implied) performed by someone...
The other way would be:
auld lang syne written by traditional, Robert Burns (1788)
* alud lang syne by others...Meaning that Robert Burns added some music + lyrics to a piece of music that already existed. This would be acceptable only if we are sure that song existed in some form prior to Robert Burns adaptation. But if we are sure about it, then we should enter the "old man singer" version as the original, and quote robert burns only as the one wrote it out, not as the composer. [Case 1]
Finally, another way, the one I personally prefer, would be:
auld lang syne written by Robert Burns (1788)
* auld lang syne by others... [Case 3]with a comment, saying exactly what Burns said to the British Museum. I'm sure the old man singer won't be very offended by that!
______
坐低飲啖茶，食個包