fise:TextAnnotation may define the fise:selected-text property with a literal containing the selected text as value. When selecting bigger sections of the text (e.g. a sentence or even a chapter ...) this becomes inefficient so the model allows for using fise:selection-head and fise:selection-tail instead.
The TextAnnotationNewModel engine is expected to add those properties in case a fise:TextAnnotation does not define them. However this is currently not implemented.