The claim that “MARC Must Die” because its labeling system is too elaborate and abstract to facilitate consistent use naturally resonates with anyone learning about MARC for the first time, especially given that it uses a format that was invented at a much earlier stage in the development of computing technology. At the same time, the author’s seemingly total confidence in his hypothesis may obscure certain practical advantages of MARC format that justify continuing to use it. For instance, while the article points out that XML would allow for more clarity and range in the relationships between (if I understood the table of contents example correctly) subfield and field, it doesn’t explain whether XML would offer the same opportunities to identify relationships between fields that MARC 21 does with its tagging system. The use of three-digit numbers allows a given pair of tags to indicate multiple kinds of similarities at once and also to distinguish between content that has to follow a special set of conventions and content that can be entered verbatim from the information source. Having no knowledge of XML whatsoever, I’m reluctant to subscribe to the author’s argument without a more concrete comparison of MARC with other possible practices for digital cataloging.