The focus on document lifecycles drove a number of important features that are missing from contemporary word processing platforms and relational database systems.

Markup language files are text files. That simple fact improves accessibility and lowers lifecycle costs by orders of magnitude. Nearly any computing device and system can interact with a markup-encoded text file. The Extensible Markup Language (XML) dramatically lowered SGML's cost profile, and there are a wealth of XML-based markup technologies.

Documents are naturally hierarchical. Books contain chapters. Chapters contain paragraphs... Markup is organized around addressable trees and nodes and nested containers. These explicit data structures lower level-of-effort for both human and machine processing.

Semantics enter the picture when naming the nodes in the tree. When creating custom data structures to solve real-world authoring and publishing issues, the names are often associated with real-world meanings and organizing concepts. These terms, their meanings, and various encoding decisions take us into the world of semantic technologies and formalized ontologies.

Semantic markup is the phrase given to this approach to document management, where XML is used to define an application-specific markup language, where the names and data structures reflect complex conceptual frameworks and models.

The challenge when using markup for single sources is to focus all authoring on the single, highly-refined source document, so that all derivatives flow from there.