Some multi-word lexical units are represented as consisting of several separate nodes and their complex character is captured by assigning them a special structure. This applies mainly to:
non-inflected titles.
Individual parts of a title are represented by individual nodes. The complex character of the whole expression is indicated by assigning the effective root node of the relevant subtree the ID
functor. In some cases, this analysis is supported also by adding a node with the t-lemma substitute #Idph
to the structure. This node is on top of the whole subtree representing the title (for the analysis of titles, see Section 8, "Identifying expressions").