Re: ComicSeries publisher/imprint

I've not got a huge amount to add to this, but from the academic journal publishing side I think the picture is pretty similar. Imprint is definitely separate from 'publisher' in this case. I'd probably broadly boil your four definitions of Imprint down to two in this case:

* Brand (encompasses your 1 and 4)
* Subsidiaries (your 2 and 3)

Anyway - I think the thrust is very similar.

As you indicate the area is fraught with problems, but for me the biggest issue is that the data publishers and consumers are unlikely to cleanly differentiate (or often care?) about the differences - and as sometimes the same label may apply as first a publisher (independent publisher) then as a Subsidiary (publisher bought out by another company) and finally a Brand (purchasing company absorb business but keep absorbed publisher name as an brand type imprint).

The complexity of publisher/imprint relationships (especially once taken over time) and the fact that many data publishers and consumers are unlikely to make the distinction clearly makes me feel that any 'publisher' field needs to be read as 'publisher or imprint' from both sides. Where a string is used it can be taken as either publisher or imprint. Where a URL is used potentially there can be more information available behind it (and I can imagine this is an area where schema.org isn't going to be what is needed to describe the underlying complexity).

Just to through in one other complication in academic publishing sometimes titles are 'published on behalf of' - especially with "Learned Societies" (http://en.wikipedia.org/wiki/Learned_society) who often commission publishers to publish titles on their behalf. This isn't an imprint, it is an additional relationship but one that can be confused with the 'publisher'. In cases where such an arrangement is in place the user is often interested in the learned society responsible for the journal rather than the publisher (especially as Learned Socieities may decide to commission an alternative publisher at any point)

Owen

Owen Stephens
Owen Stephens Consulting
Web: http://www.ostephens.com
Email: owen@ostephens.com
Telephone: 0121 288 6936

On 9 Dec 2013, at 07:41, Henry Andrews <hha1@cornell.edu> wrote:

> Hi folks,
>   Here is the promised separate thread on publishers and imprints.  This is an area of particular interest for me in terms of historical research, so I'm probably going overboard for these purposes.  But just in case it's useful, here are some thoughts.
> 
> Most of this boils down to whether you're trying to define clear guidelines for uniformly managing the data (which the GCD is trying to do) or providing generally useful fields and allowing different users to tailor the exact usage to their own needs (which perhaps is closer to what y'all are doing).
> 
> Imprint appears to be the only publisher data being stored, although I see publisher fields in the examples at the PeriodicalVolume/ComicSeries level.  If there isn't a publisher field, is the imprint intended to link to some sort of publisher object that contains various imprints? (which can be a complex many-to-many relationship, or involve otherwise identical imprints in multiple places.
> 
> Assuming its separate from "publisher", what definition of "imprint" is intended?  Roughly speaking, there are four:
> 
> * "real" imprint, meaning some business entity has said "this thing here is an imprint"
> 
> * subsidiary company, often because one publisher bought another, but the bought company continues doing business as itself.  Sometimes treated as a full publisher.
> 
> * shell companies, i.e. the 80 or so companies Martin Goodman shuffled around from his pulp days through when he sold Marvel: http://www.comics.org/publisher/78/indicia_publishers/  In cases where we don't know as much as we do about Goodman's companies, shell companies may be treated as full publishers because we have no idea to whom they belonged.
> 
> * "branding" for lack of a better term- something buyers identify due to emblems of some sort (logo or tagline) but that may or may not correlate with any of the above.  The GCD just implemented a more useful way of organizing this info and the DC data is in pretty good shape: http://www.comics.org/publisher/54/brands/ (click on a brand to see the associated emblems).
> 
> Now I'm certain you aren't trying to track details like brand emblems, but it's a tricky question as to what you expect people to put in that field, or in the publisher field if you have one.  Here's an example:  Captain America Comics #1 (March 1941): http://www.comics.org/issue/1313/
> 
> Most people, who probably know Cap from the recent movies, would expect this to be published by Marvel.  If imprint is the only field available, they'd probably expect it to say Marvel.  However, it doesn't say "Marvel" *anywhere*.   it has absolutely no brand emblem whatsoever (Goodman used them only rarely, and with different names, until about 1949).  The indicia lists the publisher as Timely Publications, which is at least a name commonly associated with pre-Marvel comics in the 40s by comics historians.  But that's all you've got in terms of objective data on the comic.
> 
> And that's just issue #1.  Next issue is Timely Comics, Inc. for a while, then Complete Photo Story Corp. for most of the run, with the last few published by Marjean Magazine Corp.  Most have no branding, but variations on Timely, Atlas and Marvel appear sporadically.  None of those companies or logos are "imprints" in the modern sense (like, for instance, Vertigo is an imprint of DC, although the publishing company is usually whatever DC is calling itself at the time).
> 
> Anyway, I can go on about this sort of thing at length, with a broad range of examples, and that's just in the US.  The GCD currently uses four fields for publication data and its still not really sufficient, but it gets down to precision and data quality.  And also use cases.
> 
> The GCD is concerned about authoritatively recording data in a central location.  If the schema is just so various people can use a standard format for a range of purposes, it probably doesn't matter too much whether these fields match "correctly".  People dealing with modern comics will use pretty intuitive notions of "publisher" and "imprint" which probably have more to do with logos and common usage than the publishing company.
> 
> Historians trying to use this will probably go nuts making the data fit, but that's probably inevitable no matter the fields available (speaking from experience :-)
> 
> cheers,
> -henry