In lieu of an abstract, here is a brief excerpt of the content:

Reviewed by:
  • The structure of coordination: Conjunction and agreement phenomena in Spanish and other languages by José Camacho
  • Grant Goodall
The structure of coordination: Conjunction and agreement phenomena in Spanish and other languages. By José Camacho. (Studies in natural language and linguistic theory.) Dordrecht: Kluwer, 2003. Pp. xi, 184. ISBN 1402015100. $98 (Hb).

Over the last few decades, work on the phenomenon of coordination has produced a number of claims about the construction that have become widely accepted. One such claim is that each conjunct is licensed as if it were not conjoined with anything else. With conjoined subjects or [End Page 887] conjoined objects, for example, each conjunct typically carries the case that it would have if it alone were the subject or object. Possible exceptions to this generalization have been widely noted and analyzed (see e.g. Johannessen 1998), but still it is commonly assumed that the default situation is the one in which each conjunct is licensed symmetrically, as if each were the only one.

Another widely adopted claim about coordination is that the first conjunct asymmetrically c-commands the second. This allows us to explain contrasts such as 1.

(1)

a. Johni’s dog and hei went for a walk.

b. *Hei and Johni’s dog went for a walk.

This claim too has not gone unquestioned (see e.g. Progovac 1997), but the basic idea that conjuncts are structured asymmetrically has been highly influential and conforms to what we would expect if the conjuncts are both members of a single phrase in X′-format.

Despite the wide acceptance of both of the above claims, they would seem, on the face of it, to be incompatible with one another. If the conjuncts are asymmetrically embedded within a phrase, it is not immediately obvious how either one will be licensed as if the other were not there, let alone how they will be licensed symmetrically.

This incompatibility between the two properties is the point of departure of José Camacho’s book. Rather than trying to avoid the conflict by eliminating one or the other of the properties, C faces the problem head-on and proposes a novel structure for coordination in which both properties are straightforwardly satisfied. In his analysis, the reason that each conjunct is licensed as if it were alone is that each is the specifier of its own licensing head. Conjoined subjects, for instance, have the structure in 2.

(2)

The conjunction and here appears in an I position, and from there it copies the features of the lower I. Each I is thus able to license its own specifier, and the result is the sequence DP and DP, where each DP is licensed by virtue of its spec-head configuration with I. Conjoined objects are handled analogously, where each DP is the specifier of a licensing head. The structure in 2 automatically gives us the second property as well: since each DP has its own licensing head, the first clearly asymmetrically c-commands the second.

Structure 2 thus represents the heart of C’s proposal for the phenomenon of coordination, and it is a very clever solution to the seemingly intractable problem posed by the two conflicting properties mentioned above. There is no doubt, however, that it is also an extraordinarily counter-intuitive solution, in that the standardly assumed constituenthood of the DP and DP sequence is not represented here. It is in fact partly this counterintuitive quality that makes the book so interesting to read, because any evidence the author can find in favor of 2 comes across as something of a surprise. [End Page 888]

C does indeed find evidence for 2, and much of it is worth noting. Despite its originality, 2 does share certain features with claims that have already been made in the literature. The structure in 2 represents the conjunction as a head, for instance, as has been extensively argued for in Johannessen 1998. C’s new evidence for this position is suggestive, though not compelling. The structure in 2 also presupposes that coordination is in a sense propositional, in that each conjunct is the specifier of the head of a propositional projection...

pdf

Share