[mb-style] RFC-327: Featured Artists (attempt 2)

[Andii Hughes]

On 22 July 2011 08:20, Ryan Torchia <anarchyriot at gmail.com> wrote:
> On Thu, Jul 21, 2011 at 3:26 PM, Andii Hughes <gnu_andrew at member.fsf.org>
> wrote:
>>
>> On 21 July 2011 18:07, Kuno Woudt <kuno at frob.nl> wrote:
>>
>> > On most of the physical releases I've looked at the featuring artist is
>> > credited in the artist column if there is an artist column on the
>> > release.
>> >
>> > So, considering featuring artists are commonly credited as
>> > "track title - artist Z feat. artist Y" on back cover tracklistings,
>> > there
>> > shouldn't be any problem with us standardizing on that at the recording
>> > level.
>>
>> +1
>>
>> This is my experience too.  I can't think of any releases where the
>> existing MB
>> behaviour of using the track title occurs, other than studio albums where
>> the
>> primary artist is unlisted altogether (because it's their album).
>
>
> That's a rather large "except" though, isn't it?  That essentially narrows
> down the scope of your sample to title tracks of singles.

No, I was primarily thinking of compilations, mainly various artists
ones.  I actually forgot about singles, which pretty much go with the
studio albums in that they have a dominant primary artist.  It's not
so much an exception as you wouldn't expect the primary artist to be
repeated on every track of a release when their name is also on the
front cover e.g. you wouldn't expect to see:

1. A by X
2. B by X
3. C by X
4. D by X feat. Y
5. E by X

on the back of X's album.  It's implied.  So instead the feat. is part
of the title because there is no artist credit:

1. A
2. B
3. C
4. D feat. Y
5. E

but if track 4 then appears on a various artists compilation, it's
likely to be credited as D by X feat. Y.

The examples of singles you posted pretty much go along with what I
was thinking of with regard to studio albums.  You have the primary
artist dominating the cover, and the featured artist next to the track
because they don't appear on other tracks by the artist (including
B-sides on that single), their appearance being a feature of that
track.

What I had in mind was something like:

http://www.discogs.com/viewimages?release=2238114

where the feat. is clearly part of the artist credit (in bold).

  That also seems
> very anecdotal,

Well yes, I did say 'in my experience', offering things up for counter
examples.  They may well be compilations that credit feat. in the
track title but I can't remember seeing any.

 and hasn't been my experience -- but that comment isn't any
> less anecdotal.  So I started looking around, searching for "feat." in
> recording titles on singles and EPs, and honestly, I found a lot more
> releases like this:
> http://www.discogs.com/viewimages?release=1958183
> http://www.musicbrainz.org/release/7df28f89-7039-40a0-929d-3e7e296dee0a
> http://www.discogs.com/viewimages?release=335941
> http://www.discogs.com/viewimages?release=69180 (<-- this one's interesting
> because it shows a clear distinction between a "feat." spot and a
> collaboration.)
> http://www.discogs.com/viewimages?release=1904322
> http://www.discogs.com/viewimages?release=1884794
> ...where the featuring artist is linked with the song title, or this:
> http://www.musicbrainz.org/release/150bbdaf-95ca-4016-abcb-9ac3ce11d287
> http://www.amazon.com/gp/product/B00009VGGT
> ...where the guest isn't listed on the title, or this:
> http://www.discogs.com/viewimages?release=41461
> http://www.discogs.com/viewimages?release=8506
> ...where the guest name placement is ambiguous, or this:
> http://www.musicbrainz.org/release/b83f146d-0374-4cbc-8d65-7e7de7f71fca
> http://www.musicbrainz.org/release/d75937fa-1ea8-4799-8389-700c575f9ca0
> ...where MB has misrepresented the join phrase.
>
> (I tried to keep my search unbiased and invite people to try for themselves:
> advanced search for Recordings, "feat AND type:single", click open a few
> random entries from random pages, try to find cover art and take a look.)
>
> Actually, it was remarkably difficult to find releases where the guest was
> credited "Artist feat. Guest", even when limiting the search range to
> singles.  The one genre where it was very common was for single-titled
> releases (i.e. versions of one track, no B-sides) from electronica acts with
> guest vocalists; those almost always were "Artist feat. Guest" format,
> though frequently the "feat. Guest" was in a smaller typeface and on a
> different line, not billed equally:
> http://www.musicbrainz.org/release/84604323-a490-4d66-998e-c0f565315d08
> http://www.discogs.com/viewimages?release=301847
> http://www.discogs.com/viewimages?release=238949
> http://www.musicbrainz.org/release/9be39dcb-7110-4de1-baa5-0e1a9bb985b3
> http://www.discogs.com/viewimages?release=339973
> http://www.discogs.com/viewimages?release=112659
>
> (These turned up more frequently when searching for "feat" in Releases
> instead of Recording titles.  Searching for it in Artists often resulted in
> bands where the "featured" artist was one of the band members.)
>
> What we can take from this is that there definitely isn't any consistency on
> how "feat." appears on singles, and nothing that indicates consensus of use
> or meaning.  There are some cultural biases and tendencies, but that's not
> something we should base a universal standard upon.

I didn't look at them all in detail, but it pretty much reinforces
what I'd assumed.  The featured artist is credited in the title when
there is a dominant primary artist for the whole release, and the
exceptions are where the featured track is the only track, so the
credit for that becomes the credit for the whole release.  That's not
so much how the credit should be as a simplification to avoid
repeating the primary artist attribution ad nauseum.  You need to
consider how the track is credited in isolation and the (not ideal)
examples of that come from compilations.

My primary motivation for moving the featured artist to an artist
credit is that it stops being merely text and becomes a reference to
the actual artist, giving more structured data which can be parsed
reliably by a machine.  I think that's pretty valuable.

>
> --Torc.
>
> _______________________________________________
> MusicBrainz-style mailing list
> MusicBrainz-style at lists.musicbrainz.org
> http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
>



-- 
Andii :-)