ISBN numbers are the unique code given to books that enable booksellers and bookbuyers to identify a particular publication without ambiguity. ISBNs are also used by library services when users search their catalogue, and for the monitoring of their stock. In the RADAR project at Swinburne and Melbourne, we’re using the ISBN as a cornerstone for connecting users of print books with electronic books at their institution. When you look closer, there’s a lot of things about ISBNs that you might not expect.
ISBNs come – as many will know – in either 10 or 13 digit forms. The extension in number format to 13 digits came as a result of the increasing number of ISBNs that were needed – and potentially running out of numbers in the 10-digit scheme. In both formats, the last digit doesn’t in fact tell us anything about the book – it is a check-digit that is used to validate the previous digits. One key use of ISBNs is when a bookseller scans the barcode a book at a point-of-sale terminal. A check digit is found in most of the retail formats of barcodes, and the ISBN is no exception.
The extension of the ISBN scheme from 10 to 13 digits is a problem created by its very success. However, there are other problems with the use of ISBNs and the scheme which reflect problems its inventors would not so readily have foreseen. Nowadays, books are available in multiple formats. In the 1980s, regular and large-print, paper-back and hard-back texts were also known. Many books were printed in only one format (hardback) and more popular texts in two or three (paper-back and hard-back, and perhaps large-print). In a few cases, a very popular text would appear in even more.
With the rise, first, of the audio-book, and that in multiple formats (CD, tape, etc.), then in the electronic book, again in multiple formats, a single “book” can now occupy many separate ISBN numbers. There is seldom a systematic order in which these different formats are numbered. For either a library- or a computer-scientist, this is a surprisingly ineffective approach, and leaves many opportunities for error.
In an idealised format, one might have an identifier for the book in any format, followed by a format-specific number. This would make it easy to surmise the right ISBN for the electronic partner of a print book – you simply swap out the format number accordingly. However, this is not done. There are in part good reasons for this – IBSNs are allocated to publishers, and different publishers may be responsible for the different formats. Furthermore, the same book may be published by different publishers in different countries under licence.
In practice, this means that finding the corresponding electronic book for a print book is a complex affair: there is no simple ‘trick’ to finding the eBook for a print book. If that were not enough of a problem, there are many nominal ebook ISBNs that have never been published, and ISBNs have been erroneously assigned to more than one ebook. While it is probable that the same has happened with print books, it seems much less common.
Publishing is a complex industry. ISBNs were primarily intended to help stock control and pricing in shops; not as a means to enable scholarship. Libraries have to add their own barcodes or RFID tags to identify individual copies of books, and record the relationships between different versions of the same book. However, in the digital age, should we be rethinking this paradigm, and arriving at a solution that is more systematic, more future-proof and more able to enable sophisticated library services?