Catalogues, Databases & Registers

Three source families, three different jobs

The terms often overlap in everyday use, but their differences matter because they reveal why the record exists and what it was designed to capture. Format alone does not determine reliability; purpose, timing, source basis and object specificity do.

What different catalogues can contribute

A catalogue's evidential value depends on the activity it documents. The collector should extract the event that is native to that catalogue, then avoid stretching the entry into claims it was not designed to support.

Databases are powerful because they connect - and risky because they compress

Aggregated databases can reveal repeated appearances, title changes, collection dispersals, dealer networks and movements between cities that would be difficult to reconstruct manually. They can also place several interpretive steps between the collector and the original evidence.

At each layer, wording may be shortened, uncertainty may disappear, dates may be converted, names may be merged and relationships between adjacent entries may be lost. A field reading “Owner: John Smith” could mean confirmed owner, consignor, estate association, index heading or a former attribution. Read methodology notes, field definitions, punctuation conventions, abbreviations and source citations before treating a field label as a factual verdict.

Registers: strong evidence of an act, not a universal certificate

Registers can be especially valuable because they are often created close to the event and as part of routine administration. Their strength lies in documenting that the administrative act occurred. Their limit is that the recorded information may have been supplied by another party and may not extend beyond the register's operational purpose.

Build an object identity profile before searching

Search quality depends on the quality of the object record. A loose description invites false matches; a structured identity profile provides multiple ways to test a candidate and helps distinguish the individual object from another example of the same type.

Keep identifiers separate. A dealer stock number, auction lot number, catalogue raisonné number, accession number and your own inventory number may all be “numbers”, but they refer to different systems and events. Recording each with its issuing body, meaning, date and source prevents later confusion.

Search like a provenance researcher, not a shopper

Search systems reward exact terms, but historical records are unstable in language. Names, titles, dimensions, categories and identifiers change. A failed exact search is often evidence about your search terms, not evidence that the object was absent.

Match the record to the individual object

Finding a plausible record is not the same as identifying the present object. The match must be built from a combination of features, with greater weight given to characteristics that are difficult to duplicate or transfer.

State the result at the level the evidence warrants: confirmed, highly probable, probable, possible, unlikely, disproved or unresolved. A good conclusion names the decisive features and the remaining mismatch rather than presenting confidence as intuition.

Trace the entry back to its evidential lineage

Preserve the hierarchy between an original record and later representations of it. A scan of a contemporary ledger is not the same as a scholarly transcription; a transcription is not the same as a database abstraction; a seller's paraphrase is not the same as any of them.

When the finding matters, inspect the original or the closest available reproduction. Compare omitted qualifiers, marginal notes, crossed-out entries, ditto marks, column headings, page sequence and adjacent records. These features can determine whether a name belongs to the object, the transaction, the consignor or merely the surrounding section.

Respect scope, silence and changing records

Every resource has boundaries: period, geography, object category, institution, auction house, record series, value threshold, voluntary participation or digitisation status. Absence becomes meaningful only when there is a strong reason the object should have appeared within that defined scope.

Preserve a PDF print or full-page capture, screenshot, stable identifier, access date, URL and visible source notes. This is not merely administrative housekeeping: it allows a later researcher to reconstruct what the public record said at the time you relied on it.

Resolve conflicts rather than choosing the most convenient source

Conflicting catalogues and databases often preserve the history of changing knowledge. A later source may incorporate technical research; an earlier one may be closer to the transaction; a label may merely repeat a sale description. The conflict must be mapped before it can be judged.

Common traps and seller-facing red flags

A practical research workflow

Document the research trail so another collector can test it

A provenance conclusion becomes more useful when its search path, source basis and limitations are preserved. This allows later correction, independent review and efficient continuation when new records appear.

Use language that matches the evidence

Final judgement

Catalogues, databases and registers move provenance research from unsupported story to testable chronology. Their greatest value is not instant certainty. It is the ability to expose names, dates, identifiers, intermediaries, record locations, changes of description and points of conflict that direct the collector towards better evidence.

The disciplined collector never stops at “I found it in a database”. They explain the record's purpose and scope, trace its source, match it to the individual object, distinguish a documented appearance from a complete provenance, preserve the search trail and keep uncertainty visible. That is what turns a search result into defensible provenance research.