Music Publishers - What Do We Want to Capture

Hello all, I’m probably opening up a can of worms here but that seems to be something I’m good at.

We’re going to be talking about the confusing hellscape that is Music Publishing :smiley:

I’ve been trying to add Publishing metadata more frequently these days, initially going by what is printed on the release packaging, adding this information to the work entry and then going a step further and performing a lookup of that work on the various music publisher search engines available, most notably ISWCNet, BMI Songview and ASCAP Ace.

When you search on these services (mostly Songview and Ace) you can reveal information that might be missing in MusicBrainz that could be useful. The most important element I find are IPI’s (Interested Party Information) which when we are talking about an individual (assuming CISAC are doing their jobs properly) should be unique to that person.

In regards to MusicBrainz this can be super useful as it can quickly confirm who the real individual is when there might be variances in the way their name is printed. For example ReleaseABC may state that “Super Song” was written by “Joe Jones”, and ReleaseXYZ may state that “Another Love Song” was written by “Joseph Michael Jones”, if one of these CISAC enabled tools returns a person IPI of “00012345678” for both that is the same person, and thus the entities should be merged.

That part is fundamentally agreed as “working” in MusicBrainz currently, as Modbot will alert if two IPI’s for a person exist.

Where it gets hairy is when you get to IPI’s for “companies”/“labels”. When looking at results on say BMI SongView, if a label is involved with the publishing of a work it should have a related IPI#. Now I would believe much like when talking about people this IPI should be unique to the company. Now I’m yet to find an example where that isn’t the case (where an IPI has been used against what would appear to be two distinctly different companies), but there is I guess the possiblitiy of this happening when one company absorbs the other.

Now if the world was all pretty roses and daises we’d see the proper use of the terms “Administered by” or “Care of” (sometimes just printed “c/o”). This way if a smaller publishing company was having their publishing tasks managed by a larger entity it would be clear as to whom that would be. An example I have found is below:

This screenshot (from Songview) states that the publisher Trumpet Blast Music has its publishing tasks undertaken by BMG Bumblebee which is part of BMG Rights Management US LLC.

As for MusicBrainz relationships we should then see something like

When You Find Out (work) was published by Trumpet Blast Music (label) who is being “Administered by” BMG Bumblebee who’s parent is “BMG Rights Management US LLC”.

We can do most of that chain of relationships except accurately link Trumpet Blast Music to BMG Bumblebee, there was an old request on JIRA (STYLE-800) to make this a Work>Label relationship but my query is would it not be better for this to be a Label>Label relationship.

Now I’m going to be very clear here, I am not a music publishing or rights lawyer - this is definitely getting into the weeds of that entire sector; and one could argue that to the average music enjoyer and even the slightly-more-than-casual MusicBrainz enjoyer this information is not of any interest but we are already capturing a lot of this supposed extraneous information (things like PRO specific ID’s and ISWC’s [although again I would fight that these do have a valuable purpose] ).

So with that I have begun to educate myself on this a little further, yes that does mean I am badgering ChatGPT as I don’t have any contacts who work in this world… and I think if I came knocking on the door of BMG Rights Management to explain all of this they’d swiftly eject me from the premises (or ignore my email pleas).

The problem I have with tracking music publishers, and then their relationships is that it seems that when a comapny is involved in music publishing their involvement can be ephemeral. What I mean by that is from my readings, a company’s involvement can sometimes only be for a set period. Example would be that companies involvement in the publishing of that work lasts 3 years, after which it hits the market for someone else to pick up. Another example would be that the work was originally on a “perpetual” agreement (i.e. no end date) with a company, who then go bust and those rights are transferred to another company.

What this means is that when the MusicBrainz editor is adding publishing information regarding companies they should be specifying the years in which that publisher has control; the problem is that it is nearly impossible (unless someone can educate me) to identify the terms in which the company printed within the packaging has rights.

So here are some questions, and a poll.

The questions are here to generate further research within the community, and maybe if we can get the attention of someone who works in the business to explain to us peons what this all means.

Q1 - How do I, a MusicBrainz contributor, identify the terms of a companies involvement in regards to publishing?
Q2 - When we see “care of” or “administered by” is that a relationship for ALL works published by that company or a relationship for that SPECIFIC work?
Q3 - When a publishing company is acquired by another, does the IPI assinged by CISAC get reassigned to the new owner or does it stay with the original company
Q4 - When a publishing company does not want to be involved in day-to-day tasks of music publishing (royalties, licensing, registration etc.) and requests another company undertake this responsiblity, what are the terms of that agreement (is it perpetual, for a few years etc.)

POLL

The following poll is single choice, anonymous - however if you choose Maybe please comment in this thread what information you think is valuable to the project.

Do you think we should aim to capture ALL metadata regarding a companies involvement in the publishing of a work?

  • Yes - I think we should track all metadata related to Music Publishing companies (IPI’s, locations, contract terms, who takes care of, ownership etc.)
  • No - I think this information is superflous to the project, bear little to no relevance on documenting music and brings little to no value to the average user of the data we hold
  • Maybe - I think we should capture some information, but not all of it (explain what below)
0 voters
2 Likes

The more I read about this (links below) the more I think “I actually don’t really care”. For that reason I have voted Maybe, I think it is interesting to track companies who are involved in publishing but I do think it would be helpful if there is a way to find out the contractual obligations behind that and then we can specify the Start/End date attribute for the relationship. IPI’s for companies are only going to be useful if they are unique and exclusive, at the moment I am unsure if this is the case. The “Care of” part which is where I started this journey now seems to only really matter if you were a party who is wanting to make contact to say get a license to use that song in a film production or such; as I have a feeling in my bones that this likely changes at whim, random contributors trying to capture this information and then have it publicly presented might cause more (reptuation) problems down the line to the quality of data we produce, and instead we should simply hold the mentality that if you do need to find that information you should seek it via a Professional Rights Organization (such as BMI, ASCAP, PRS, SACEM etc.).

1 Like

Recordings on compilations created in countries other than the source recording country may have publishing organizations of the compilation country labeled on the release notes instead of the actual publisher. I found this on a number of Roy Orbison compilations from Europe. I also found on two of the compilations what I believe to be the wrong work creator credited. Also some of those European publishing organizations are listed as publishers on some of the works.

Real care needs to be used when getting information from BMI, maybe some of it has been fixed but I found where a lawyer managing the interests of a number of artists had put their name(IPI) on all the artists works showing a financial interest in the work but they were not the creators.

As said before in this form BMI, and others are about collecting royalties. As times change the publisher information in BMI (and others) may change from the original liner notes, to include new managers, publishing mergers and/or spelling changes, financial interests, other. Not sure if that helps.

1 Like

I personally might like to see both, since (at least in theory) a publisher could be bought by another parent, but it will still have been administered by the old parent label in the past

as to the rest, I don’t know that I have any input, as I’m not familiar with how music publishing works, I just put the data in from my booklets, lol

edit: that said, I have been adding relationships between publishers and their administrators (as parent/subsidiary) on occasion

So a few hours on and it seems that the community want to try and capture this information, which is good but my concerns are still valid in the sense of how we can accurately enter and present this information; this has to be accomplished to allow for suitable documentation and guidance to be written (likely by @reosarevok ).

I’m going to respond to the two contributors who have commented so far.

I wonder if there is anyway we could quickly identify on a work page which publisher relates to which country/region, or do we instead just accept that setting the location on the Publisher label entity is enough?

I assume you are talking about an individual here, and not a company?
If the former then yes I have also seen this, actually it happens a lot with traditional music (folk, jazz, blues etc.) where lazy record labels might simply credit “Traditional” or “Anonymous” yet there is an established writer; I have also seen where they credit maybe only half of the individuals who were involved in creating the work entity.

Indeed, as mentioned above we’re hoping that these commercial entities (which is what they are) are doing their job and presenting accurate and correct information; the question pertains however how do you challenge information to ensure it is valid and accurate?

Indeed, I believe the idea is that we want to try and provide a historical record of who might have been involved over time; however no-one has yet given a method on how we would establish time-frames (i.e. the terms of the contract) in regards to Publisher; what we could end up with is a list of 20-odd publishers attached to a work with no idea where one’s responsiblity ended and the next one began!

The problem we have here is potentially doubling up on metadata, which is confusing and difficult to then document. I think we should choose one or the other, and not both unless we can accurately establish that both scenarios could exist.

None of us really seem to, which is why this thread exists as some kind of fishing net to try and find someone who may know the final answers to some of these questions; there is a clear drive from the community to track this information so we need to work this out.

Adding the content from booklets is fine, but again we are assuming that what is printed is accurate. As someone who dabbles in the messy world of budget releases it isn’t uncommon for those record companies to do the worst job possible in listing publishers, and many could simply state factually incorrect information.

Indeed what I have done in some cases, but I don’t believe this is the correct way forward as it seems that it may not always be the case that the administrator is the owner of the publishing company.

3 Likes

In my experience, local publishers are marked as sub-publishers in JASRAC and SACEM databases, alongside publishers.

So I also add them as sub-publishers in MB, it’s enough difference with the “genuine” publishers.

2 Likes

IANAL, but from my understanding, in the US at least “publishing” occurs when a music composition, soundrecording, or audiovisual work is first distributed to the public. There’s 3 different things, though a soundrecording being published could also include the music composition being simultaneously published. So the entities that “published” the works could be captured. In the case of soundrecordings and audiovisual works, the music composition is generally not also published but is used under a license. I suppose there is some value in understanding who licensed the composition used in a soundrecording but that seems pretty limited interest. Then there is also the question of other works besides the music composition included when a soundrecording is published, such as liner notes, artist NIL, etc. Seems to me more of interest to whomever wants to be paid (for eg public preformance).