The EpiDoc DTD permits only the appearance of character data and the g/ element within unclear/unclear. When an editor or encoder encounters a continuous string of unclear characters in a text, two encoding options are available: each character may be encapsulated within its own unclear/unclear or the entire string may be encapsulated within a single unclear/unclear. No other markup may appear inside unclear/unclear, so if other aspects of a long string of unclear characters warrant additional tagging internal to the string, the per-character solution may be preferable or even required. For background on this issue, see a discussion on the markup list in July and August of 2005, subject: EpiDoc DTD issues [1: unclear].

TEI defines a reason attribute for the unclear element. In the EpiDoc DTD, this attribute has three possible values: "damage", "execution" and "undefined." "Undefined" is designated as the default/automatic value. See further the markup list discussion referenced above. The default stylesheets render all instances of unclear with the standard subscript dot, regardless of the value of the reason attribute.

In cases where a letter trace is compatible with a range of characters but the context does not offer an unambiguous choice but the editor wishes to offer a default value, the alternative interpretations may be offered in a series of rdg elements within an app; in which case the particular reading to be rendered must be labelled with an attribute n='default' (see e.g. 6 above).

It is also envisaged that the possibility will exist to record the general shape of the surviving traces using embedded SVG or similar elements which could then be rendered as a drawing.

Where the editor is not willing to commit to a default reading, see vestiges.