Sunday, 23 May 2021

The Absence Of The Unit 'Word' In The Cardiff Grammar

Fawcett (2010: 228-9, 229n):
There is one final point to be made about the representation of morphology — at least in English and other languages with little or no inflectional morphology. In English (and probably in all such languages) it is in fact possible to represent all of the structures that occur in such a way that there is no need to introduce the 'word' as a syntactic unit — i.e., as a unit with an internal structure of elements such as 'prefix', 'base' and 'suffix', each of which is in turn expounded by an item. And yet such a grammar is still able to show that, in a 'word-form' such as eating, the items eat and ing are the realisations of different choices in the system network, and that in the word boys the items boy and s are similarly the realisations of different meanings.
How is this achieved? The grammar's rules for generation simply show that each of the lowest elements in the tree diagram (e.g., the head of a nominal group) is expounded by an item (a 'free morpheme'). Then, whenever it is needed, another item (a 'bound morpheme') is simply added to the head as a further exponent of it. In this way we achieve the same effect that we would if we first generated an abstract unit (e.g., the word class 'noun'), and then gave it two elements (which we might call 'base' and 'suffix'), and then expounded each element by the items boy and s. The first approach has the great advantage of avoiding adding a whole new layer of structure to the tree.
To summarise: by allowing an element to be expounded by two items (one as its main exponent and one as an affix), we avoid the need to have an additional layer in the tree diagram, thus considerably simplifying the description. The full analysis of an English sentence using the Cardiff Grammar therefore has one layer of structure less than there would be in an analysis based on "Categories", since the latter's 'rank scale' model predicts that each element of a group is filled by a word and each element of a word by a morpheme. Thus in Berry (1975:85), the example When the bus broke down the boys walked to school is analysed into thirteen elements at the 'rank' of the morpheme.


Blogger Comments:

[1] To be clear, this views the word 'from below' on the rank scale: in terms of the lower ranked units, morphemes, that constitute it. Taking the SFL perspective, 'from above', the word is a constituent of the group, and realises functions in the structure of a group, as when an adjective serves as the Epithet of a nominal group.

[2] To be clear, here Fawcett is presenting his adaptation of theory to the limitations of computers as if this adaptation were a theory of language, as spoken, signed and written by human beings.

[3] To be clear, the "advantage" of Fawcett's model is that it omits the word as a distinct unit from its model of lexicogrammatical structure.

No comments:

Post a Comment