In Head-driven Phrase Structure Grammar a linguistic sign, as usually, consists of two main components, a signifier and a signified. The signifier is defined as a phoneme string (or a sequence of such strings). Taking into account the correspondence between acoustic and written forms of language, such signifier can be identified as conceptually coinciding with elements of Text in the MTT.
FIGURE IV.15. Syntactics of a linguistic sign.
As to the signified, an object of special type SYNSEM is introduced for it. An object of this type is a structure (namely, a labeled tree called feature structure) with arcs representing features of various linguistic levels: morphologic, syntactic, and semantic, in a mixture. For a minimal element of the text, i.e., for a wordform, these features show:
· How to combine this wordform with the other wordforms in the context when forming syntactic structures?
· What logical predicate this word can be a part of?
· What role this word can play in this predicate?
Simple considerations show that SYNSEM in HPSG unites the properties of Meaning and syntactics of a sign as defined in the framework of the MTT, i.e., SYNSEM covers syntactics plus semantics. Hence, if all relevant linguistic facts are taken into consideration equally and properly by the two approaches, both definitions of the linguistic sign, in HPSG and MTT, should lead to the same results.