Okay, who can help me out with this.
My ISBN-13 assumptions:
1) ISBN-13s that begin with 978 can be “converted” to a 10-digit ISBN that still uniquely identifies the same ‘thing’ (roughly, a manifestation).
2) However, ISBN-13s that begin with 979 can not be.
3) ISBN-13s that begin with 979 are already allowed by the standard, and probably already exist somewhere, if not in the US, or if not now very soon. (It’s up to the national ISBN registrar to decide when/how to begin using 979s).
Okay, so far so good. But these assumptions were challenged when on a listserv recently, someone provided examples of cataloging copy that seemed to present a 979 ISBN-13 with an equivelent ‘converted’ ISBN-10.
Were these in error, they blindly applied the “conversion” routine appropriate to a 978 to a 979? Or was I wrong? (Or, it occured to me only later, as you’ll see, was it a typo, and those really were 978 not 979 ISBNs?)
Here is an example:
Leadership in post compulsory education : inspiring leaders of the future / Jill Jameson.
Published London : David Fulton, c2006.
So I figure, let’s look these up in Books In Print, see what it says. Look up the 10-digit, 1843123398, yep it corresponds to that book. But that book is reported as having a 13-digit ISBN identical to the one given in the example but beginning with 978. 978-1-84312-339-2. Identical except for the different trailing checksum digit, which was computed based on the 978. Well, sure, if it’s a 978 ISBN, it can be converted to a 10-digit losslessly.
So was it a typo in the cataloging? Doesn’t sound like it, a typo wouldn’t result in the different ‘correct’ checksum digit too.
Well, let’s look up that 979 ISBN-13 in BiP, and see what happens. When I look up 9791843123391, BiP still gives me the same result list, containing Leadership in post compulsory… . But the actual records in BiP don’t contain that 979 ISBN, they just contain the 978 ISBN.
Huh? What’s going on? Has this item been issued both a 978 and a 979 13-digit ISBN? That doesn’t seem right.