Ways into the collection

Terminology in context#

Different sources can use the same word differently, or different words for related concepts. The vocabulary hub connects source terminology and controlled references while keeping their roles distinguishable.

Three small investigations#

The following are suggested research exercises, not claims of findings in the corpus. Their purpose is to show how the same interface can support different units of analysis.

Make exclusions visible#

A selection can look more consistent after uncertain records are removed. That can be appropriate for a bounded analysis, but the exclusion rule belongs in the method. Keep a separate count of records excluded because dates, roles or locations are unresolved. Otherwise the resulting map or chronology can imply a completeness the evidence does not support.

Source collections may also differ in focus. A catalogue of a particular builder, region or instrument type is not sampled in the same way as a broad inventory. When comparing groups, inspect source coverage before attributing a difference to the instruments themselves. A concentration of records can reflect documentation activity as well as historical activity.

Return from a term to its evidence#

Use a controlled vocabulary to relate terms, then return to the original wording before drawing a substantive conclusion. A mapping can make records discoverable together while leaving the interpretation provisional. The OMARO notebook explains how a reference label, an attributed classification and a reviewed statement play different roles.

Turn an interesting pattern into a research question#

A cluster of similar names, dates or map points is a useful starting observation. Before interpreting it, inspect representative records and ask whether the pattern follows instrument history, one source provider or a normalization choice. Keep the original labels and source attributions visible when comparing candidates. A shared workshop name needs identity review; a shared location needs a defined historical scope.

For a repeatable analysis, translate the exploratory question into an explicit cohort in a fixed dataset release. State what counts as one observation and which records are excluded. Return to the Navigator and source evidence to inspect exceptions rather than removing them solely because they complicate the chart.