last changed in 2002
Both books on WG (Hudson 1984 and 1990) contain substantial chapters on semantics, and semantics takes up the whole of Hudson 1995, an elementary textbook on lexical semantics; but very few other WG publications discuss semantics, so WG is less well-known for its semantics than for its syntax.
One of the main characteristics of WG is its clear separation of syntax and semantics: syntax is about words whereas semantics is about their meanings, and since the difference between words and their meanings is very obvious, so is the difference between syntax and semantics. A word is connected to its meaning by the relationship `meaning', which is subdivided into two types: `sense' and `referent'; so these links always cross the boundary between syntax and semantics.
By convention this boundary is often shown as a horizontal dotted line in WG diagrams, and one of the most general conclusions to which WG analysis points is that a sentence's semantic structure is far richer than its syntactic structure. The easiest way to demonstrate this is to give the complete structure of the simple sentence Help me to make a cake! which contrasts a tiny syntactic structure (which is complete) with a vast semantic structure which could still be expanded; see the figure below.
Where does all the semantic detail come from?
In short, whereas words are virtually the only units recognised in syntax, they have no special status in semantics. The units of semantics are concepts which are not usually in a one-one relationship with words:
Why doesn't WG use a more conventional semantic analysis based more directly on the predicate calculus of logic?
The WG analysis is more similar to other approaches which are based on semantic relations and an ontology of concept types:
However it now seems that it may be possible to combine some aspects of the predicate calculus with the relation-based approach by recognising propositions. The other important part of the predicate calculus is the treatment of quantifiers, but at least some of these can already be analysed in WG structures that relate the 'quantity' of one set to that of another - see scope.
This is an article from the Encyclopedia of Word Grammar and English grammar. If you refer to it, please give the url as "http://tinyurl.com/wg-encyc".