5. The notation of valency frames and its semantics

In the following text we describe the way of writing valency frames of the PDT Valency Lexicon. The notation applies to the manual, some papers, and other written materials (Section 5.1, "The notation of valency frames"). The second part of the text provides a brief description of the semantics accompanying the way valency frames are recorded in the PDT Valency Lexicon. (Section 5.2, "Semantics of a valency frame record").

For improved clarity, we use a formal grammar productions to describe the notation of a valency frame. Note that some other constraints on the annotation (e.g. a rule acc. to which a functor name may not occur more than once in a frame notation) are not recorded in the formal grammar but dealt with only in the text. In the grammar productions, non-terminals are marked with lower case; terminals are either literals described in quotes, or special terminals FORM representing a token consisiting of letters and numbers, CHAR representing any character, and LEMMA representing a string which is the main part of a morphological lemma of PDT. A more detailed structure of the terminalLEMMA is not dealt with in this text. All characters except for letters, numbers, and a hyphen which match the occurences of the terminals CHAR, LEMMA, and FORM in the productions must be in the notation of a frame introduced by a backslash character.

The notation of valency frames described herein can be further simplified by the use of abbreviations (see Section 5.5, "Abbreviated forms of realization records"). The notation of valency frames which contains abbreviations is valid if it conforms to the grammar discussed below after all abbreviations have been expanded.