Every expression is either a number-tree, a fraction-tree, a text-tree, a table-tree, a function-tree, a native-tree, a variable-tree,etc... . Each tree is identified by a tag (such as NBR, FRC, TXT, TAB, FUN, etc.). This tag is e.g. used by the evaluator and the printer to dispatch over the different kinds of trees. Of course, it is the reader that constructs the trees with the right kind of tag. The trees are further elaborated on in the other productions.