Atlas insight · Legal record method
All insightsHow to build a source linked legal chronology.
A legal chronology is often treated as a long table of dates. That is the least interesting thing it can be. In a serious matter, the chronology should expose what happened, what the record merely says happened, which source supports each proposition, where accounts conflict, and what remains unknown. Built well, it becomes a working instrument for interviews, motion practice, investigation, and strategy. Built poorly, it turns a large record into a polished source of false confidence.
Begin with the propositions that matter.
A chronology built by reading from the first document to the last will usually reproduce the disorder of the production. Begin instead with the questions the timeline must help answer. When did notice arise? Which version of an agreement governed? What changed after a particular communication? Which acts were contemporaneous and which accounts were created later? Those propositions determine the fields, level of detail, and source standard.
This does not mean deciding the answer before examining the record. It means defining why a date deserves a row. A billing event, an email, a policy revision, and a later recollection may all share a day while serving different evidentiary functions. A chronology that knows its purpose can preserve that distinction.
Separate event, assertion, and inference.
The cleanest working model distinguishes at least three layers. An event is something the source directly records, such as a contract execution or filed order. An assertion is what a person or document says occurred. An inference is the analytical relationship drawn from one or more sources. Those categories may eventually appear in one narrative, but they should not be collapsed in the research table beneath it.
Consider a later declaration stating that a person learned a fact in March. The declaration proves that the person later made that statement. It may support the March date, but it is not the same kind of support as a March email written by that person describing the fact. The chronology should make that difference visible without forcing the reader to reopen both documents.
Give every source a stable identity.
Document names drift. Attachments detach. The same record arrives through several custodians. A stable source identifier prevents the chronology from depending on whatever filename happens to survive. The source ledger should record the identifier, title, author or custodian, date, version, location, relevant page, and any authenticity or completeness concern.
A row should point to that identity and to the exact portion that supports the proposition. Page, paragraph, exhibit, message identifier, or spreadsheet cell is more useful than a general reference to a file. The objective is simple: another careful reader should be able to reproduce the support without guessing.
Use more than one date field.
Complex records contain several kinds of time. A document may be created, signed, sent, received, filed, entered, effective, amended, discovered, or described later. Compressing these into one date column creates silent errors. The relevant date type should be explicit, and the system should preserve both the recorded value and any normalized date used for sorting.
Time zones, approximate dates, date ranges, and missing dates should remain visible. If an email was sent near midnight in one zone and received the next day in another, both timestamps may matter. If a witness says an event occurred in early spring, a false day should not be invented merely to make the software sort cleanly.
Build a contradiction register beside the chronology.
Contradictions deserve their own structure. The chronology may show that two sources disagree, but a contradiction register explains the proposition in dispute, the competing accounts, the relative source quality, any contextual explanation, and the research needed to resolve it. This keeps conflict from being buried in a note cell or smoothed away in prose.
Not every difference is a contradiction. Two people may describe different intervals, use different definitions, or possess different access. The researcher should test whether the sources truly answer the same question before calling them inconsistent. That act of classification is one of the most valuable forms of judgment in record work.
Design separate views for separate uses.
The working chronology and the briefing chronology should rarely be identical. The working version may contain thousands of rows, duplicate source links, research notes, exclusions, and provisional inferences. A deposition view may emphasize the witness, knowledge, and prior statements. A motion view may emphasize admissible support and disputed propositions. An executive view may contain only the events necessary to understand the decision.
Each view should remain generated from or reconciled against the same underlying record. That preserves accuracy while allowing the final presentation to respect the attention and responsibility of its audience.
Audit the chronology as evidence and as an object.
Before delivery, sample rows back to the source, confirm date normalization, test links, review quotation boundaries, and search for unsupported declarative language. Then inspect the object itself. Can the reader distinguish established fact from assertion? Are the most consequential conflicts visible? Do filters and labels behave consistently? Can the source record be produced for examination?
Federal Rule of Evidence 1006 concerns summaries used to prove voluminous material and requires access to the underlying originals or duplicates. A research chronology is not automatically a Rule 1006 summary, but the rule illustrates an important discipline: a summary should never become a substitute for a source system nobody can inspect.
Official reference points
These sources frame the relationship between summaries and the underlying record. Admissibility and local practice require counsel review.