It is the goal of JATS to enable users to encode the information that will be needed to create a variety of interchange and presentation formats from the markup in the JATS-encoded document. It is not the intent for JATS-encoded articles to be directly ingested by typesetting, voice-synthesis, or web displays. So, for example, JATS provides tagging for both short/unstructured and longer/structured annotations (to a graphic, for example) to enhance accessibility. It does not position this information in the locations that are appropriate for end user display; the assumption is that the application creating the display will put the information where it is needed for display. (For example, some guidelines suggest that HTML's “Long Description” should be on a separate HTML page from the content. That content is embedded in the JATS-encoded article but can be separated when creating a web version of the document.)

Similarly, many page layout tools require heading levels be explicit in their input source; first level heads must have a different tag from second- and third-level heads. JATS uses a containment hierarchy to store this information: the “title” of a “section” that is inside only one other “section” would be displayed as a second-level head. It is the assumption behind the design of JATS that the application providing the JATS-encoded content to that typesetting system will convert the tagging to the style required by that system.

The JATS multi-language tagging includes some of the information described in the W3C's “Internationalization Tag Set (ITS)”. For example, JATS enables users to encode information that can be converted to ITS':

JATS does not provide tagging for information that is specific to display tools or that is unlikely to appear in published journal articles, such as: