The Stylistics of Syntactic Complements: Grammar and Seeing in Flannery O'Connor's Fiction - Critical Essay

Style, Spring, 2000 by Donald E. Hardy, David Durian

(30) a. You know that Bill is sick. [much greater than] Bill is sick.

b. You don't know that Bill is sick. [much greater than] Bill is sick.

(31) a. John knows that Bill is sick. [much greater than] Bill was sick.

b. John doesn't know that Bill is sick. [much greater than] Bill is sick.

(32) a. I know that Bill is sick. [much greater than] Bill is sick.

b. I don't know that Bill is sick. [sim][much greater than] Bill is sick.

When the subject of a factive verb is second or third person, if no other special circumstances occur, the finite complement is presupposed, as is shown in (30) and (31). But if the subject of the factive verb is first person, the finite complement is presupposed only in the positive, as in (32a). If the factive verb is negated, the presupposition is canceled, as in (32b), since speakers cannot presuppose that which they currently deny any knowledge of.

The subtlety of the pragmatics of presupposition is partially revealed when we notice that a simple shift to past tense restores the presupposition to (32b):

(33) a. I knew that Bill was sick. [much greater than] Bill was sick.

b. I didn't know that Bill was sick. [much greater than] Bill was sick.

Thus, while (32) shows that a speaker cannot presuppose what s/he currently denies any knowledge of, (33b) shows that s/he may presuppose current knowledge of what s/he at one time did not know.

It may have appeared from our discussion above that implication is purely semantic, i.e., not dependent on contextual factors. Thus, consider the following:

(34) a. John saw Mary leave. [greater than] Mary left.

b. John saw Mary leaving. [greater than] Mary was leaving.

c. John saw the house painted. [greater than] The house was painted.

It would seem incontrovertible that if someone sees an event occur, it necessarily, or implicationally, occurred. It is perhaps a sign, however, of an overzealous quest for rigorous semantic rules such as implication in linguistics that Kirsner and Thompson must point out what sounds, once it is pointed out, like a truism: "that people can perceive things which do not exist and events which do not take place" (212). Consider the following examples from Kirsner and Thompson (212):

(35) The delirious patient saw the room spinning around him, but we know it wasn't spinning.

(36) When the neurologist stimulated that particular area of her brain, Susan saw the light turn red even though it really did not.

Example (35) does not implicate "the room was spinning around him," nor does (36) implicate "the light turned red" for the simple reason that the factivity of the complements is defeated, or canceled, within the sentences in which they occur. Thus, just as presuppositions may be canceled on the basis of contextual factors such as person of the matrix verb subject and tense, implications may be canceled by means of contextual indications that the perceptions are not real. We may note, however, that it takes more linguistic effort to defease implications than presuppositions. Examples (37) and (38) illustrate this fact: