Allerton vs. Radford Contrasting Valency and the English Verb with Transformational Grammar

Transformational Grammar by Andrew Radford and Valency and the English verb by D.J. Allerton are both books about linguistics, or to be more precise about phrasal syntax. This being their only common feature, both writers approach the topic with their own distinct, individual conceptions. These ideas are outlined and explained in the 4th and 5th Chapters of Transformational Grammar (entitled Noun Phrases and Other Phrases) and the 1st Chapter of Valency and the English Verb (entitled Syntactic Structure and Functional Relations).

Transformational Grammar is a textbook, therefore it is very elaborate, very detailed, which can not be said about Allerton's book. This one is rather essay-like, not much detailed, less explicit. Radford wrote for students of English linguistics, so he had to make sure that everybody understands his conceptions. To succeed in doing so, he overexplains his statements. For instance his 'endocentricity constraint' rule (5.6, pp282--262) is explained through four pages and is revised four times. This is of course necessary for a textbook. Radford uses many examples to show the validity of his ideas, but always the same ones, i.e. "The King of England" or "a student of Physics". However this type of illustration suits the needs of first-year students, it does not show us the 'global' validity of each statement. I agree with the view that examples should be as clear as possible but not as consistent as this. Allerton on the other hand wrote his essay for an audience of scholars, assuming that the readers know about his previous works and those that he refers to.

Each of the two linguists has his different idea that he sets as the basic phenomenon for his description. Allerton calls this phenomenon 'verb valency'. While Radford puts the emphasis on categories, setting two types, i.e. the word-level categories and the phrase-level categories, his most important concern is to establish a third type of category that he calls small-nominal phrases, Allerton in the meantime is more concerned with 'valency'.

Function, unlike in Radford's typology, plays a very important role in 'valency'. Allerton supplements Haas' suggestion (1966:126-7) of the description of grammatical patterns by a third one (i.e. (syntagmatic, paradigmatic, functional). He claims that "The importance of the functional aspect to a description is that it makes us view the constituents of a construction as more than just dismembered pieces." Function in his point seems to be a part-whole relationship (as in subject of a sentence), so he also sets another link, he calls relational (as in subject of a verb). He suggests that "function is partly derivable from class". Since Radford on the other hand disregards function and is only concerned with classes, not taking into consideration the constituent's relationship between its function and class, the class of its neighbors and the class of the construction of which it is a part, he does not arrive to the Bloomfieldian division of constructions.

While Radford claims that rules should only be constructed in a way that they create endocentric (i.e. properly headed) structures, Allerton divides up the most common patterns for two-member constructions of all kinds according to the following:

Subordinative:

(1) lexical head + lexical modifier

(2) lexical head + grammatical modifier

(3) grammatical head + grammatical marker

Semi-subordinative:

(4) lexical core + lexical specifier

(5) lexical core + grammatical specifier

Exocentric:

(6) lexical base + grammatical convertor

(7) lexical base + pure marker convertor

(8) grammatical base + pure marker convertor

Irreducible:

(9) lexical cornerstone + lexical cornerstone

(10) lexical cornerstone + grammatical (pure marker) cs.

Coordinative: (not used in Radford's book as a separate type, but only as a possible test.)

(11) lexical coordinate (+co'or) + lexical coordinate

(12) grammatical coordinate (+co'or) + grammatical coordinate

This procedure of dividing up the pattern types is much more sophisticated than what we find in Transformational Grammar.

Radford on the other hand sets 'Constraining Categorial Rules', one of which, i.e. the 'endocentricity constraint' rule (which was already mentioned above) claims, deduced also from Bloomfield, that

Xn ... Xm ... (n£m),

where m¹n if ... is null (i.e. if Xn is non-branching),

this way achieving the "endocentricity requirement".

In Radford's typology phrases can be of any size, they may contain any number of constituents, but Allerton sets this number a minimum of two. Radford in this matter is unique with his idea of one word being able to make up a complete phrase.

Another difference between the two writers is that Radford only writes about the English language, Allerton (although he mentions the "English verb" in his title) picks examples from other languages as well, this way wishing to make his typology universal. Unfortunately the Russian examples sited contain some misprints this way not proving his point exhaustively and leaving doubt about the matter.

Finally one more difference has to be pointed out when contrasting these two works, namely the divergence between the two types of drawings used to illustrate the syntax. Allerton breaks from the well-tried tree structure type of drawing and indicates alternative ways of illustration. He uses a "dependency type of diagram" for illustrating one kind of multi-member construction. Also, he makes up his own diagrams for his own pattern division. Since the division itself is very sophisticated, the diagrams also have to be rather colorful to be able to meet the requirements of the pattern division. Unfortunately, Allerton does not justify the validity of these diagrams, but they are somehow forced on the reader that by adopting his pattern division, the diagrammatic conventions also have to be adopted.