The passive construction
As we have seen throughout this book, the passive construction has received a great deal of attention in the literature. This interest in the passive construction is not restricted to cognitive approaches. Indeed, in the transformational generative model, the passive construction is described in terms of what is probably one of the best-known syntactic transformations (the other being the ‘wh-movement’ transformation that derives wh-questions in languages like English).
As we saw in Chapters 17 and 18, the cognitive account emphasises the meaningful aspects of the construction, in particular the fact that it effects a TR-LM (figure-ground) reversal. Furthermore, the Cognitive Grammar account holds that the passive participle is synchronically related to other functions of the same form (PERF). In contrast, the formal account holds that examples (12a) and (12b) have distinct structures and that the category of the verbs is different in each case, as is the category of broken.
According to this view, example (12a), as an active clause, is not derived by transformation but has the standard structure of the canonical English active declarative clause. The copular verb is treated as a lexical verb because it is the only verb in the clause, but it is viewed as a ‘defective’ lexical verb because it lacks semantic content and because it displays the same ‘operator’ features as auxiliary verbs. It can invert with the subject to form an interrogative clause (Was her heart broken?), and it can carry negation (Her heart wasn’t broken). In this example, broken is a predicative adjective. Example (12b), on the other hand, is derived by transformation from an active clause. In this example, was is an auxiliary verb and broken is the past participle form of the lexical verb.
In the transformational model, the fact that an active clause like George betrayed her paraphrases a passive clause like She was betrayed by George (and vice versa) motivates the view that the passive clause is related to the active clause by derivation. This relatedness is captured by a lexical process that passivises the verb by altering its morphology as well as its argument structure: the passive verb only requires a single argument, which is interpreted as PATIENT but surfaces in subject position. To capture this pattern, the transformational analysis holds that the PATIENT argument originates as the complement of the passive verb, because PATIENT arguments are canonically linked to complement positions of transitive verbs. This accounts for the fact that the single obligatory argument in a passive clause is interpreted as a PATIENT despite the fact that it appears in the position that is typically associated with the AGENT. The PATIENT raises to subject position by transformation. As we mentioned earlier, this trans formation relies upon case as a mechanism. The fact that the PATIENT of a passive verb is not licensed to occur in its canonical post-verbal position is interpreted as an indication that passivisation affects the case-marking qualities of a verb. This claim is supported by the ungrammaticality of example (13a), in which it is a ‘dummy subject’ that is inserted into the example to discount the possibility that the sentence is ungrammatical because it has an empty subject position. According to this analysis, the PATIENT NP raises to subject position in order to be case-licensed and surfaces with nominative case (13b).
The by-phrase is viewed as ‘underlying’ subject in the sense that the subject of the active form of the verb is ‘absorbed’ by the passive morphology. This means that if the optional by-phrase occurs in a passive clause, it has the status of a modifier. Finally, the active and passive counterparts are seen as truth conditionally synonymous (if one is true, the other is true, and vice versa). Any meaning-related distinction (that is, the fact that the speaker chooses to make one referent rather than another discourse prominent) falls within the domain of pragmatics and thus beyond the immediate ‘responsibility’ of the grammar. All the grammar has to do is to generate a well-formed output. Hence, the emphasis in the formal approach is upon accounting for the syntax of the construction rather than accounting for its discourse function.
According to Langacker, the intuition behind the transformational account is valid in the sense that broken in (12a) describes a STATE, whereas in (12b) it describes a PROCESS. Furthermore, Langacker’s claim that the copula verb be designates a schematic imperfective PROCESS but contains no other information means that the Cognitive Grammar analysis is also rather similar to the formal account in this respect, as we saw in Chapter 17. A final similarity relates to the modifier status of the by-phrase in both accounts. Despite these similarities, the cognitive account differs from the formal account in emphasising the different semantic properties of the active and passive constructions: from this perspective, their grammatical properties are only superficial ‘symptoms’ of the cognitive representation that these constructions evoke.
