If the use of computers to write texts was a logical step forward from the classical typewriter, maybe using these machines to read and study them isn’t so strange after all.
The point of XML is to provide encoding rules to computers, so there is no discrepancy between items of the same type, for instance, enlarging the first letter of a new chapter.
The newer trend is to use XML to give the computer clues about what it is in fact processing. So it can give information when queried and perform the typesetting automatically.

This is possible thanks to the new frameworks which allow writers to specify more than just encoding options, but actually help define style. These would allow a computer to understand if the given passage was spoken, thought or written and in which representation it was shown for instance. But such a system being too limited, it was further extended to add more information about the style. For instance, whether what is said is a metaphor, a fact or a hypothesis. All these could potentially make studies of texts easier, as they would allow the computer to exhibit styles more in depth.

One of the applications could be the study of the differences between editions of the same text. To see and understand why some metaphors were removed, or if the style remained constant. Yet another use would be to feed a novel inside a system that would showcase the evolution of the characters or the plot, since those would be mapped using XML. This way to view information would prove itself quickly to the weary-eyed digital humanist trying to figure out tendencies in a 50-novel collection across 15 different editions.

But such a system wouldn’t be of much use if the result was only for those who use computers to read. Since most of the publications today need to be printed out, and XML can’t be used on a typewriter, there exists a need for some editing software which processes it and still produces high quality prints. XML-Print was invented with precisely that goal in mind, as it allows authors to work with XML while not worrying about how the end document is going to look.

In the last few decades, computers have turned from advanced typewriters to the ultimate study tool for large collections of books. There seems to be a trend going to make computers see digital books as more than just a collection of letters and words. By making them understand the structure and content of a book, even on a basic level, we can use queries to see changes, tendencies, and since a computer has more memory and processing power than any human being, they would make for very handy bookworms.

References :

-An XML Annotation Schema For Speech, Thought and Writing Representation.
By Annelen Brunner.

-Leaves Of Grass : Data Animation and XML Technologies.
By Brett Barney, Brian Pytlik Zillig.

-XML-Print : Typesetting Arbitrary XML Documents in High Quality.
By Lucas Georgieff, Marc Wilhelm Küster, Thomas Selig, Martin Sievers.