When Two No’s Nearly Make a Yes: A Study of Conditional Imperatives

Abstract

Performance was monitored in various tasks, in all of which people carried out instructions to respond appropriately to specified events. The tasks varied in presentation modality and in the temporal interval between the instruction and the event requiring a response. Types of instruction were compared, having either zero, one, or two negative elements. When there was only one negative in an instruction, it occurred either as the element “not” in the main clause or as the connective “unless” introducing the subordinate clause. The instructions with double negatives included both “not” and “unless.” It was found that, in some tasks, performance with instructions having double negatives was better than with instructions having only one negative. However, the instructions showed no consistent rank order of difficulty across tasks. It is suggested that psycholinguistic models of comprehension need to incorporate situational factors explicitly. Generalis-ability across task and contextual variables can be as important as generalisability across subjects and specific sentence contents. Negative imperatives are commonplace (Do not touch, No Parking, No admittance). At times, the importance of correctly understanding such instructions can be critical (Do not remove cover unless power is disconnected). Nevertheless, relatively little of the empirical work on negation has been done with imperatives. Clark and Lucy (1975) introduced negative elements into conversational imperatives (Will you close the door?). They found that performance was determined more by the negativity of the conveyed meaning than by the negativity of the surface structure. Consequently a sentence having two negative elements (“sad” and “unless”) and a positive conveyed meaning (ril be very sad unless you make the circle blue) showed a pattern of response latencies that was typical of affirmative sentences (“True” responses were always faster than “False” responses). Clark and Lucy interpreted this as suggesting that subjects recoded the double negative into some affirmative internal representation prior to responding. The binary stimulus set they used may have facilitated such recoding. In the test, subjects could be sure that “not blue” meant “pink.” Such recoding may not occur, indeed may not even be possible, in other situations (No smoking).