specifies that the file is a token-based TAN alignment file. Root element. TAN-A-tok files take exactly two sources all declarations must have ids TAN-A-tok sources must be named identifies how the text on one text-bearing object relates to that on another by specifying a relationship, e.g., B is a direct copy of A, B and A descent from a commont parent, etc. See main.xml#tan-a-tok for theoretical background. In most cases, there will be need for only one of these elements. But multiple values may be helpful for cases where a bitext has a complex history, for example, a textual object that was created over time, and in different phases. This element should not be used to describe any activities (e.g., translation, copying), reserved for <reuse-type>. For examples see main.xml# keywords-bitext-relation identifies a category of text reuse (e.g., translation, mistranslation, paraphrase, ridicule). See main.xml#tan-a-tok for theoretical background. For examples see main.xml# keywords-reuse-type points to one or more <bitext-relation> @xml:id values that qualify the type of relation that holds. Multiple values assume inclusive or (A or B or A and B) This attribute is inheritable. See main.xml#inheritable_attributes points to one or more <reuse-type> @xml:id values that qualify the type of textual reuse that holds. Multiple values assume inclusive or (A or B or A and B) This attribute is inheritable. See main.xml#inheritable_attributes declares one or more groups of words that should be aligned with each other. <align> specifies that all the tokens invoked for one source collectively align with the tokens in the other. alignments may claim bitext relations and reuse type alignments must contain one or more toks