Francik and Clark (1985): How to Make Requests That Overcome Obstacles to Compliance #

This file formalizes [FC85]'s obstacle model of requests for information. The speaker estimates the greatest potential obstacle to getting the information, a preparatory condition of [Sea69] that may fail, and overcomes it with a request conditional on that condition, as Do you remember what time the concert begins? means "if you remember, tell me" and is void if the answer is no. Conditions are ordered by specificity (PreparatoryCondition): having read the announcement is a way of knowing the time, which is a way of being able to tell it, so Did you happen to read in the newspaper when the lecture is?, Do you know when the lecture is? and Can you tell me when the lecture is? overcome the same obstacle with decreasing specificity. A request Overcomes an obstacle, the set of doubted conditions, when each lies in the scope of its Query: empty for a direct request, the queried condition and every way of satisfying it for a conditional request, and everything for the general Can you tell me? and Could you tell me?, whose excuse is that the hearer is somehow unable or unwilling. A direct request overcomes no obstacle (overcomes_direct_iff), and a request whose scope misses the obstacle presupposes what is in doubt: Do you happen to know your middle name? presupposes willingness, Do you want to tell me what time the lecture is? knowledge. The speaker is as specific as the situation allows: a request Pinpoints the obstacle when it is the least that overcomes it, which for a conditional request means the queried condition is the least upper bound of the doubted ones (pinpoints_condition_iff) and for the general request that no single condition bounds them (pinpoints_general_iff), as when a stranger asked for directions may be unable or unwilling.

Three experiments support the model. In Experiment 1 thirty speakers formulated requests for twenty-four scenarios in a high- and a low-obstacle version, twelve on the hearer's ability, eight on willingness and four on the speaker's memory of having asked already; direct requests fell from 49% to 35% with the obstacle (Table 1), and the indirect forms named the obstacle of the scenario, Do you know? where knowledge was in doubt, Do you remember? where memory was, Did you see? where the source was and Have I asked you? where the speaker's memory was, while in the permission and willingness scenarios speakers preferred the general Could you? or sidestepped the obstacle by asking for related information. Experiment 2 ranked the forms for directness; Table 2 gives the nine forms for asking the time of a student with or without a watch, the direct What time is it? produced only with the watch and every form produced without it overcoming the doubt (time_produced_overcomes). In Experiment 3 eighteen raters judged five forms in every scenario: each conditional form was rated higher, at either obstacle level, in the scenario type whose obstacle it overcomes than in those whose obstacle it does not (ratings_crossover), and the general forms, preferred overall, were rated lower at the high level than at the low (ratings_general_lt), as forms that pinpoint no single condition. The requests the paper discusses are predicted by Overcomes (rows_predicted), and its gradient of specificity descends the order on queries (gradient_antitone).

Implementation notes #

Can you tell me? and Could you tell me? are the general query rather than the ability condition, following the paper's gloss of their excuse as "somehow unable or unwilling" and its finding that they stay appropriate in the speaker-memory scenarios. No form in the data queries the bare ability condition, which describes the obstacle of a speaker unsure whether the hearer could find out the information, remember it or be allowed to tell it.
Table 3 pools the knowledge, memory, source and permission scenarios as the ability type; its rows carry knowledge as the type's obstacle. Ratings are recorded in hundredths of a point and directness scores in tenths; sidestepping and politeness stay in the prose.
The examples are Data.Examples.FrancikClark1985.

Francik and Clark (1985): How to Make Requests That Overcome Obstacles to Compliance #

Implementation notes #

References #

The paper's data #