Why are Certain Sentence Constructions Mnemonically Robust for Modifiers?.
24 October 1988
The work reported here extended earlier findings (Rothkopf, et al., 1986) that certain sentence construction improved recall of key modifiers following rapid reading of expository text. For example, the geographical origins of Risdulph, the physician, were recalled more accurately after reading the phrase Risdulph, a physician from London ... (here called robust), then after reading Risdulph, a London physician (here called fragile). Fragile constructions differ from robust in that they remain well formed English sentences when the modifier is deleted, while robust constructions do not. In the work reported here, we confirmed the construction effect with new samples of expository text and with a new recall measure. We also tested the syntax- as-error-signal hypothesis for the mnemonic superiority of robust constructions. The hypothesis states that a) for robust constructions, the inadvertent omission of the modifier during rapid reading results in an syntax violation; and b) this violation can act as an error signal and prompt rereading of the phrase in which the error occurred. We rejected this hypothesis in a computer-controlled reading experiment in which rereading was prevented. The results from another experiment supported the conjecture that robust phrasings stress the importance of the modifier.