Discussion
The ten scenes hear words judged by their surface as pointers. A hard word points at an order, a blank or a range; a short word points at what a grade, a sheet or a cell leaves out. The scenes follow the pointer and make their changes where it leads, in times, places and authority. They do not rule on the words.
The two kinds are not treated alike, and the difference is the most telling result. The scenes let a short word be information in its own right: a new nickname is a signal of a stoppage not yet classified, "ambiguous" a signal of a boundary seen first. They do not extend that credit to hardness. A hard word is allowed to point, but the scenes lean to no reading that calls its hardness the signal. One reading of this asymmetry is that the scenes keep apart two things the research on incivility and the research on occupational talk each study alone: the force of a word and its content. Another is simpler: a scene written for comparison may simply avoid crediting a raised voice. The scenes do not choose between these readings, and this study leaves the choice open.
Contributions
The study makes three contributions, each narrower than the works it reads. First, it reads two kinds of GAEKJU talk that no earlier GAEKJU paper had read as synthetic scenes, under one stated rule. Second, it shows that the scenes treat the surface of talk as a pointer: they move the hard word's target from a person to a thing, set the short word beside the account it leaves out, and credit short words but not hard ones as signals. Third, it shows that their remedies are of time and place rather than of speech: none of the ten proposals asks anyone to change how they talk. Read beside Andersson and Pearson (1999) and Cortina, Magley, Williams and Langhout (2001) on one side, and Orr (1996), Bechky (2003) and Bowker and Star (1999) on the other, the contribution is a position between them: words judged by their surface, followed to what they point at and answered with a place where two accounts can be seen together.
Limits
The scenes are synthetic. They show how GAEKJU frames a question, never what happens in an emergency room, a pharmacy, a factory or any other trade, and nothing here should be read as a claim about those settings or as a view on how anyone should speak, check or decide. The selection is small: four rough scenes and six code scenes. A single scene moves a count by a large share, and the rule that keeps only evidence without a digit sets aside 107 scenes of the two kinds whose framing may differ. One scene, the month-end push, differs in shape from the rest. The counts rest on patterns in the Korean that were checked by reading; a different reader might draw some lines differently, though the scenes each count rests on are listed. The reading in the grade-code scene that putting in all the information can blur the trade could be counted among those that take the short form to be enough; this study does not. The translation is this study's, and a word such as "hard" for the Korean of the rough scenes carries choices of its own. The readings through the works are this study's interpretation and go beyond the works where marked.
Further research
Three questions follow. The first is whether the asymmetry holds in the scenes the rule set aside, the rough and code scenes whose evidence brings in counts or years. The second is whether GAEKJU's record readings of the same two kinds frame hard and short words in the same way as its synthetic scenes do, a comparison this study did not make. The third is whether second lines in which a speaker takes up the next piece of work, which all four rough scenes and four of the six code scenes have, recur across other kinds of talk.