Discussion
The ten pairs read here sit at the point in NADIR's frame where a relationship is wearing away and nothing has yet broken. They are a model's conditional interpretations of transposed, redacted scenes, unreviewed, and every one has the status of a hypothesis. Read as a record set, not as reports, they pose a decision that is easy to misstate. Their openings pose it as a verdict: the trio is already dead, power is already a corpse, the late salary is the heart failing. Their structure poses something else: a later sign to watch for, evidence that would undo the reading, an unknown about other people, and a choice between speaking up and stepping back, each priced.
The decision before the break, on this reading, is a decision about what to wait for. It asks what would have to be seen and what would have to be known before a small sign could be treated as the end. Of the series' two questions, what to decide now and what would change the decision, the pairs answer the second with care, in their conditions, refutations and missing facts, and the first with force, in their openings; the two answers do not fit.
Three features of the records carry this reading, and none is a slot the format guarantees. First, the judgment's own words are a small part of each judgment, between 18 and 35% of its English, and those words are the ones that read the sign as the end; the rest repeats the scene, the condition and the refutation. Second, undoing a judgment usually takes evidence, objective proof or dated receipts, that forming it did not. Third, the missing fact concerns other people's knowledge, dealings or intentions in nearly every pair, and it is in each case the fact that would decide what the sign means.
Contributions
To the debate on weak signals, the paper offers a small published case in which a format built for conditional reading carries a headline voice that reads the faintest information most strongly. Where Ansoff (1975) describes matching a response to the quality of its information, the NADIR pairs hold the materials for that scaling: the second sign, the missing fact, the graduated options. Their openings set them aside. The point is not that the openings are wrong about invented scenes. It is that a record can hold both a weak-signal discipline and its opposite in the same few lines, and that a reader taking the first sentences as the judgment would take the opposite.
To the debate on how relationships come apart, the paper offers a reading in which the records compress a process into a moment and lay out the process anyway. Baxter (1984) rejected a single sequence of stages; Lee (1984) described stages from discovery through exposure and negotiation; Andersson and Pearson (1999) located a spiral at a starting point and a tipping point; Sias and colleagues (2004) reported mostly indirect disengagement among workplace friends. Each pair, taken whole, has its own short course with a starting sign, a tipping sign and an outcome that waits on facts. Its opening then reads the course as finished at its first step.
To the debate on exclusion in groups of three and at work, the paper offers the observation that the records seat the reader in every chair of Caplow's triad at once, and that their overturning conditions often fall to the one excluded or to the one with authority, not to the pair whose coalition is in question. Read with Lewicki, McAllister and Bies (1998), the options assume, in places, that trust and distrust stand together in one relationship; the openings assume that a single sign of distrust has ended trust.
Limits
The evidence is ten records of one kind, from one studio, made by one process: transposed scenes with a model's interpretation, the source reviewed and the interpretation not. They are synthetic. Nothing in this paper is a claim about real friendships, firms, clubs or audits, and the paper does not say whether any opening, condition or option is right about the scenes it reads. The English is this study's, made from the Korean originals; a different translation could shade the strength of an opening. The counts of wording were made by fixed word lists on the Korean, which catch the words listed and nothing else. The frame was measured only among the ten, not across the whole list, and the paper did not read the blackbox pairs in other layers or phases, so it cannot say whether the gap between opening and structure is particular to relationships, to erosion, or to the blackbox pairs as a whole. The works are read only through their published abstracts, and two of them were written about romantic couples, not friends or colleagues.
Further research
Three lines of work follow. The first would set these ten beside the relationship pairs the studio files after the break, to see whether the openings change once the frame says the end has come. The second would read the 46 blackbox pairs across their four layers for the same asymmetry between the standard of a reading and the standard of its overturning. The third would set the openings aside and ask readers to judge each scene from the condition, refutation and missing fact alone, to see what the structure, without its headline, asks a person to decide. Each would keep to the same limits: the records are hypotheses, the scenes are synthetic, and the judgment, in the end, belongs to the person reading.