Article 7 traced Tomas's changed availability to a deadline conflict. Giving a tool the same job requires explicit descriptions of the choices, facts, and requirements involved.

The relationship maps have already given us a vocabulary: people, tasks, reports, and time windows. A phrase pattern is a sentence template with named slots and a defined meaning. Filling those slots names specific things and connects them. Every use of C must identify the same report.

Illustrative phrase patterns and filled examples for the review problem
Illustrative phrase patternFilled example
<report> is planned ready at <time>.C is planned ready at 11:00.
<person> is available from <start> to <end>.Tomas is available from 11:30 to 12:30.
Review of <report> takes <minutes> minutes.Review of C takes 60 minutes.
Review of <report> must finish by <time>.Review of C must finish by 12:00.

These patterns are illustrative. A tool needs a defined translation for each one. The 24-hour times share a workday and time zone. Draft readiness is assumed, availability is confirmed, and the deadline is required.

Every report needs one qualified reviewer who did not prepare it. A review runs uninterrupted for its duration, starts at or after the draft's ready time, fits the reviewer's availability, and finishes by its deadline. A person cannot overlap a review with another assigned task, including a meeting.

A candidate supplies choices: assign Tomas to review C starting at 11:30. The assignment becomes a labelled connection; its start and 60-minute duration place the bar in the schedule. Both views refer to the same proposed review:

Relationships and schedule for the same rejected candidate The relationship map assigns Tomas to Review C, which examines Report C and lasts 60 minutes. Tomas is available during the 11:30 am to 12:30 pm time window, and the review is scheduled during that window. Review C must finish by the separate noon deadline. The schedule shows the same proposed review from 11:30 am to 12:30 pm: it fits Tomas’s availability but crosses noon and finishes 30 minutes late. This is a rejected candidate, not evidence that work happened. People are ovals, tasks are rounded rectangles, and reports, time windows, and deadlines are square-cornered artifacts. Relationships in the candidate being checked Tomas Review CDuration: 60 minutes Report C Time window11:30–12:30 DeadlineNoon assigned to examines available during scheduled during must finish by Schedule 11 am11:30 amNoon12:30 pm Tomas Review C 30 minutes late

A relationship map assigns Tomas to a 60-minute Review C, connects that review to Report C, and links Tomas's availability and the proposed review to the same 11:30–12:30 time window. A separate requirement links the review to a noon deadline. The matching schedule runs from 11 am to 12:30 pm and places Tomas's Review C from 11:30 to 12:30, fitting his availability while crossing the dashed noon deadline by 30 minutes. Both views describe the same rejected candidate and do not establish that work occurred.

Two views of the same rejected candidate. Review C fits Tomas's availability and crosses its noon deadline. The map names both relationships; the schedule makes the timing conflict visible.

Define the boundaries: by 12:00 includes 12:00. Consecutive tasks may share an endpoint without overlapping. These meanings determine which arrangements pass.

The bar fits the availability window and crosses the deadline: two relationships expressed in time. A full model also needs the other reports, reviewers, permissions, and approved meeting changes. Searching requires eligible choices and allowed times. Missing information needs a declared policy: does it block a decision or leave a candidate awaiting confirmation?

Keep sources and confirmation times with inputs, including Tomas's 10:30 update. A passing result applies to the declared model. Actual completion still needs evidence. Start with examples you can check by hand, then compare the tool's results with your reasoning.

Next: Which Observations Can Fit the Same Explanation? We will apply arrangement reasoning to data fusion.