Unknown performer, known recorder


There is a lengthy discussion in edit notes regarding the release Music from the Morning of the World and how it should be credited. The performers are unknown (except for one track). David Lewiston made the recordings and is credited as such on the cover (“Recorded in Bali by David Lewiston”). Lewiston also wrote the liner notes.

The two questions are:

  1. Should the release artist be “Various Artists” or “David Lewiston”?
  2. Should the individual tracks/recordings be credited to “[unknown]” or “David Lewiston”?

Note that @reosarevok has already expressed support for crediting Lewiston for the release.

The previous discussion is here, although it might be good if the participants in that thread were to summarize their positions here.

  1. Should the release artist be “Various Artists” or “David Lewiston”?

David Lewiston. Due to his role being heavily noted (written on cover, back and CD) on this release. Crediting compilers is not unusual; for example, Gilles Peterson, working only as a compiler, is credited as the release/release group artist on here.

  1. Should the individual tracks/recordings be credited to “[unknown]” or “David Lewiston”?

I believe that avoiding [unknown] in this case is the right thing to do; the guidelines point towards avoiding the use if at all possible. Even though he did not perform these recordings, he is still heavily involved them.

The case of where an unknown performer was credited in a more specific way was brought up. For example, a release might list “Cayapa harmonica player” (Lowland Tribes of Ecuador) as the performer for a track. I believe that it is still a good idea to avoid [unknown] with a more specific artist credit.

Relationships can still be used on recordings for unknown performance (with the credited alias), but if possible, nothing should be filed there.

In the case of performers being only partly unknown, for example, this credit on the same Folkways album as my previous example, emphasis on the unknown part: “Hernando, Luis & ayahuasca ritual participants”, the whole credit should be listed as is, without linking [unknown].

I think it is a good idea to avoid using special purpose artists if there is an alternative. Even though you might find it offensive to the artists who are being recorded, field recordings are seen as a kind of photography (“phonography”), and is a more intimate practice than the one of a recording engineer. If it is not applicable to credit field recordists for recordings of unknown music performed by humans, it should probably be noted that [unknown] is preferable on Style / Unknown and untitled / Special purpose artist.


I am personally quite confident that Lewiston is better than Various Artists or [unknown] as the release artist - that’s the only credited name in there, and his involvement with the release as a whole is significant as well.

I wouldn’t consider either of the track artist options incorrect, so I’m quite on the fence about question 2 :slight_smile:


If [unknown] is correct, the style guidelines are very cryptic about it. Correct me if I’m wrong, but the only other exception of it written down, other than a last resort artist credit, is when it is used for a classical recording with unknown performers (instead of the composer being recording artist as a placeholder).


I believe that

[unknown] - Used when a particular artist is unknown, but is potentially discoverable with further research.

applies to these recordings, even if the emphasis at this date has to be on “potentially” (in as much as Mr. Lewiston, and probably many of the other involved parties, are now dead).

I have always read the guidance on [unknown] as essentially “do your research before you use this” rather than “use absolutely anything else in preference”.

I find it disrespectful to the actual musicians (already marginalized by their complete anonymity in the liner notes) to attribute their performance to Mr. Lewiston.

Looking through the Alan Lomax discography, I don’t see a very clear precedent for crediting the recordist for unknown performers. While not completely consistent, those releases seem to lean more toward [unknown] - see here and here for a couple of examples.


The approach on that release seems to be overly specific, in as much as it implies there is only one “Cayapa harmonica player” who recorded both pieces. That may or may not be true, and certainly if we followed that precedent with “prison work gang” or “subway busker” we would end up conflating different groups/individuals.


Well, the conversation of what is respectful or not respectful might not be a priority if this information is impossible to obtain without time travel, and it is only used for when there is no information.

I find it to be a bit irrelevant for the purpose of MusicBrainz, but I get what you are coming from. One could argue that it would be more respectful to musicians if all of them would be listed alphabetically in the artist field for an album credited to one person, for example. Coltrane Time is an example of a ‘non’ respectful release, it was later changed to credit John Coltrane instead of The Cecil Taylor Quintet due to his commercial success after the album was recorded. I don’t think either of the artists gave this the OK sign. The music industry can be harsh, and a database should not carry a moral burden based on the actions of people in it.

The conversation should be on how data can be filed when data is lacking, and I find that a (often highly-billed) field recordist is an acceptable alternative for filing unknown recordings to, which makes them easier to find. The context of the performance (unknown credits are often things like “male singer”, “Buddhist monks”, can be given with relationships linked to unknown).


I meant that those should still be credited to a recordist if no identifying profile can be made. I submitted edits to merge those to [unknown], because they’re essentially anonymous. This is how I would list that specific recording there:

Title: Harmonica “Song of Ourselves”
Artist: David Blair Stiffler
[unknown] [credited as “Cayapa harmonica player”] - harmonica
David Blair Stiffler - recording engineer

This keeps the information intact, and links the recordings and track artist to a ‘real’ entry.

The next example is only partially unknown:

Title: Spirit Manifestation (Wire Bow, Boots, Seed Pod Necklace)
Artist: {Hernando}[, ]{Luis]}[ & ayahuasca ritual participants] (curly brackets represent artist links, square bracket is the joining phrase)
Hernando and Luis are credited using their profiles, with [unknown] credited to “ayahuasca ritual participants”.
David Blair Stiffler - recording engineer

I hope this makes sense.


I see. In that case, I agree with the merge to [unknown] but not with the recrediting to David Blair Stiffler. (But you probably figured that already. :slight_smile: )

I just don’t see the value of avoiding [unknown] as outweighing what is, to me, a misleading attribution.

For the second example, I do like the idea of relegating “additional unknown participants” to the join phrase (as here with Ed Lewis & prisoners ).


When I look at this example, I think artist in the more abstract form. David Lewiston has painted a picture with music, sound, noise. He is the artist who has put this together. He created the album as a whole and chose what to record. No doubt he edited it too. And then chose what to release to the rest of the world. He should get the main credit and not Various Artists.

Same as when a musician like Peter Gabriel will have used musicians to create his music. His name goes on the track and the album, and the musicians are in the credits. Gabriel chose what to put together. He gets that main credit.

I find it more disrespectful to loose something like this into the mass pit of “Various Artists”. At the very least the actual release should be credited to David Lewiston as he created the release. He should not be demoted to a small bit part of “just a recording engineer”.

I very much agree that those [unknown] musicians should be getting credits for their works. But that should be the same way that Gabriel’s keyboard player does. They are the performers on the tracks. It is David Lewiston who has decided to put the performance together - even if he is “only holding a microphone”. So, without a better name to use, David Lewiston is better than [unknown] for the track credits.

Or I’d make up a new “artist” to be “Bali Musicians” or something more specific than a plain “unknown” as we know something about them.

To me, the “Various Artists” artist should be more about those releases with a list of disconnected bands on a single disk. Various recording dates. Generally brought together to sell as a theme. Whereas this example is a clear project with a focus of - record the music of Bali.

1 Like

Indeed, the artist field is often quite abstract. An instrumental version of a track might often be credited to a singer who contributed nothing to the instrumental part, for example.

I disagree with making new artists for really generic unknown artist credits like “Bali musicians”, though. However, it should be noted that Discogs does it like this, example Baka (3).


When I, admittedly briefly, looked at this one it seems like they are not that “generic”. It is a specific group of people on an island. So they are not as “unknown” as some performers.

I know I have added some obscure albums where an artist has had to be added for that one and only one release due to them playing the kazoo on one track. So it would not seem too weird to me to give this unknown collection of people a slightly less than anonymous credit.

Tricky though.

1 Like

Release, yes. Recordings, no.

While I don’t strongly object to creating a “Bali musicians” artists, I would think that in this case, using that as an artist credit for [unknown] serves that purpose sufficiently.


I would like elaboration on why exactly it is undesirable as a recording artist.

The artist field is an abstract value, which is fleshed out with relationships. It is really common to credit vocalists for instrumental versions of their songs which they did not perform anything on, for example. I can’t see how the popular music example is any better, in the case of field recordings, the recordist is directly involved in the production of the recording and documentation of the performance.

It should be noted that libraries with huge and well organized collections of ethnic music recordings credit the recordist in absence of a noted performer. Example: Ragas on bamboo flute [Enregistrement sonore] / Deben Bhattacharya, collecteur (Bibliothèque nationale de France).

In my view, listing something under [unknown] is highly undesirable, equivalent of an unsorted stack of books in a library.

If the recordist is credited, one can click on the recording artist link and get a list of credited and uncredited field recordings by that recordist. In the other case, one would be wading through thousand of things like “CD Track 01” from “Instrumental Relax Music Vol 3”, imported from FreeDB 10 years ago, if they were only going by track artist.


In this example, though, Deben Bhattacharya is clearly identified as the “collecteur”; one would not confuse him with the actual flautist. In that regard it’s a bit more like an MB relationship.

The same is true when the recordist is credited via a ‘recording engineer’ relationship, isn’t it? For someone like Alan Lomax, who was both a field recorder and an occasional performer, this has the benefit of not mixing the two roles indiscriminately.


One can see in what capacity Alan Lomax was involved with a certain recording by looking at his relationships tab, and find the ones where he’s an actual performer if you want. It doesn’t mix the roles.

Yes, one could also find the recordist by looking at relationships. However, doing that would also make any argument about potential conflation invalid, since in my examples, the performer would be linked to [unknown] with a relationship.

In the case of just having the artist and title, like in tagged music files, one still has more context about the recording than [unknown] provides. “Ragas on Bamboo Flute” by Deben Bhattacharya is much more identifiable than “Ragas on Bamboo Flute” by unknown or “Ragas on Bamboo Flute” by “Indian flautist”, even though it might be confusing what the actual role is. If someone played this recording on the radio, and said one of the three variants here, which one would give you the most helpful information to locate the recording played?

There is no denying that the system of crediting a sound recording to one person is often very flawed and can lead to confusion. However, the given whole with relationships I suggested would give people an usable name of a directly involved person to attribute something to, with the full context provided by relationships. Is no usable name of a directly involved person to attribute something to really a better alternative?


For another example of western releases of anonymously presented foreign music where I think [unknown] is justifiable: Folk and Pop Sounds of Sumatra, Volume 1

This release is compiled by Alan Bishop, and consists of recordings he has compiled from old cassette tapes. There are some tracks here which are not credited, nor is there anyone who you can attribute the recording itself to. I would absolutely oppose crediting Alan Bishop as a track artist (and as a release artist, since his name isn’t prominently promoted) on this since his role is limited to putting together the compilation.


I have no idea why you decided to use Peter Gabriel as an analogy here. He composed music and wrote lyrics for his songs. He sang them. It is so much more than recording a performance or even making a decision to record it.

I don’t think this matters at all, but out of curiosity – how do you know that he put the performance together?


@insolite I picked a name off the top of my head. I just meant someone leads a project. Someone puts the artwork together. It was supposed to be a vague example of how one name appears on the front of a product.

The only name on this project is Lewiston. I know he is “only the man with the microphone”. I did read off a number of links. IMHO that is better that loosing it to the never ending pit that is “Various Artists”.

I’ll drop out of this discussion. Don’t want to distract with more misunderstandings as you guys know more than me on this stuff :frowning:

1 Like

I still haven’t gotten a response on why filing field recordings of any other kind under the recordist is okay, but not in the case of performances where the musicians are impossible to track down.

I don’t want to repeat myself, but this still seems ignored: The artist field is not based on performance or the role of the artist. The guidelines do not explicitly prevent any role to be credited for an recording (other than composer for classical recordings by unknown performers). Nearly every instrumental version of pop and hip hop tracks would have to be changed if that were the case. Tadanori Yokoo, a visual artist who did not contribute anything to the actual recording, is credited as a release/recording artist here because the album itself is co-credited to him.

There are countless examples outside of field recordings one could look to where the credited artist is objectively “irrelevant”; they are not involved in any way. But here, the artist traveled to a foreign country, recorded performances, edited the performances for release as a commercial recording and is clearly credited on the front cover. I cannot help but read this as a double standard.