Do not start with the first photograph. Start with the whole, then the boxes, then the folders. Item-level description comes last, and for most of an archive it may never be needed.
The quickest way to fail at cataloguing a photographic archive is to open the first box and begin describing the first photograph. At a careful ten minutes an image, a collection of 50,000 frames would take more than four working years. The method that works runs the other way: describe the whole archive in a paragraph, then the boxes, then the folders, and go down to individual photographs only where there is a reason to.
That is not a shortcut. It is how archival description is designed to work, and it produces something useful at every stage.
What standard should a photographic archive follow?
In the UK, the reference point is ISAD(G), the General International Standard Archival Description, which The National Archives describes as the international standard for archival cataloguing and which is maintained by the International Council on Archives. Its central idea is hierarchy. A catalogue moves from broad to narrow through four levels:
| Level | What it describes | In a photographer's archive |
|---|---|---|
| Fonds | The whole body of records created by one person or organisation | The photographer's complete output and papers |
| Series | A group arising from one activity or of one type | Editorial commissions; advertising work; Polaroid tests; job files |
| File | A unit kept together by the creator | One folder, sleeve set or job bag |
| Item | A single thing | One transparency, negative strip, print or digital file |
Descriptions become more detailed further down, and information given at a higher level is not repeated below it. The National Archives adds that item level is the less common one in archive catalogues. For photographs, the folder is the natural working unit: it is usually what the photographer labelled, and it usually corresponds to one job or one subject.
ISAD(G) will not tell you how to caption a picture. For the photograph itself, the working standard is the IPTC set of photo metadata fields, which is what picture desks and search engines read. A sound catalogue uses both: archival hierarchy for the collection, IPTC fields for each scanned image.
Where do you start with unlisted boxes?
Survey first. Before recording anything in detail, go through the whole collection and note the number of containers, the formats, the approximate dates and the condition. Look for the photographer's own system: job numbers on sleeves, a ledger, a card index, dividers by client. That system is the skeleton of your catalogue.
Number the containers as found. Give every box, drawer and binder a running number and write it on the outside in pencil or on a tag. Photograph each one before unpacking. Do not consolidate half-empty boxes or re-sort folders by subject.
Record ownership and source. The National Archives advises recording, for every body of material, the date it arrived, its source, the basis on which it is held and who holds rights in it. For a single-owner archive this is one entry, but it should exist in writing.
Make a box list. One row per box: its number, location, formats, rough quantity and the labels on the outside. A box list for several hundred containers takes days, and at the end of it the archive can be searched for the first time.
What should a folder-level record contain?
The next pass opens each box and lists the folders inside. Keep the fields few and fill every one.
| Field | Example | Notes |
|---|---|---|
| Reference | ARC/0217 | Unique, permanent, never reused |
| Title | "Red shoes / white bg" | Transcribed exactly from the label, in quotation marks |
| Supplied title | Red court shoes on a white background, studio | Your clarification, kept separate |
| Dates | 1994; or c.1993 to 1995 | Say when a date is estimated |
| Format and base | 6 x 7 colour transparency | Also negatives, 35mm slides, prints, Polaroids, files |
| Extent | 84 frames in 9 sleeves | Count frames if quick; otherwise count sleeves |
| Description | Two or three sentences | Who, what, where; what is unknown |
| Condition | Good; some sleeves yellowed | Flag odour, mould, sticking |
| Rights note | People identifiable; brand names visible | Facts only, no conclusions |
| Location | Box 31 | So it can be found again |
The reference and title in that table are illustrations of form, not entries from a real catalogue.
Two habits protect the catalogue's value. First, keep transcription and interpretation apart. What the sleeve says is evidence; what you believe it shows is inference, and a future researcher needs to know which is which. The same discipline applies when writing captions from contact sheets and sleeve notes. Second, use the same words for the same things. A short controlled list of formats, places and subject terms, of the kind described in the guide to keywording with controlled vocabularies, is worth more than a long, inconsistent one.
When is item-level cataloguing worth it?
Go to item level when one of four things is true.
- The item is being digitised. Each scan needs its own identifier, and the file name should carry the folder reference. The conventions are covered in the guide to file naming for photo archives.
- The item is being published or licensed. It needs a caption, date, creator, credit line and copyright notice embedded in the file. The Library of Congress recommends recording title, creator and creation date in the file where the format supports it, using a common embedded schema such as IPTC.
- The item is at risk. Deteriorating film should be identified individually so that it can be isolated and copied.
- The item is exceptional and will be asked for by name.
For born-digital files, item-level work can be partly automated. Cameras write the date, time and technical settings into each file, and the EXIF viewer shows what is already there before anyone types a word. Captions and keywords can then be written into batches of JPEGs with the tool to add captions and keywords to photos. For film, a quick visual index helps cataloguers and researchers alike: photograph each sleeve on a lightbox, or lay out scans with the contact sheet maker, and attach the sheet to the folder record.
What does a finished inventory look like?
A worked example is public. The Robert Harper Photograph Library's inventory lists 503 folders and can be searched and downloaded as a CSV file at /archive-inventory/. It accounts for 95,487 unique photographs, or 177,143 frames including duplicates and variants, with the digital files held as Nikon NEF raw files, TIFFs and JPEGs and the analogue material mainly on 6 x 7 colour transparency film. The 503 inventory lines add up exactly to those totals, though they have not been checked frame by frame against the physical material.
On top of that inventory sit 330 described collections, arranged under 24 principal categories and browsable in the photography archive index. That is the hierarchy in practice: one fonds, one rights holder in 825 Ventures Ltd, folders as the working level, and item-level description still to come for the film, which is largely unscanned. Publishing the folder list before the scans exist is a deliberate choice. It shows what is there, and what is not yet known.
From list to catalogue
A folder-level inventory is the point at which an archive becomes usable by other people: researchers can ask for a folder, and a scanning or cataloguing company can quote against real quantities. Archive.London is looking for partners to take its film from folder level to item level, and cataloguers and digitisation specialists can find the scope on the digitisation partners page.
Frequently asked questions
What is the first step in cataloguing a photographic archive?
A survey. Walk the whole collection before describing any of it: count the boxes, note the formats, date ranges and condition, and work out how the photographer organised the material. The National Archives stresses understanding what you have and how it fits together before entering data. Number the boxes as found and do not rearrange anything at this stage.
Do I need to catalogue every photograph individually?
Usually not. Archival practice describes from the general to the specific, and item-level description is the exception. A folder-level record with a title, dates, format and quantity makes material findable at a fraction of the effort. Reserve item-level captions for photographs that are being scanned, published or licensed, where a caption, date and credit travel with the file.
What information should I record for each folder?
A unique reference, a title taken from the original label, dates or an estimated range, the format and base (for example 6 x 7 colour transparency), the quantity, a short description of the subject, the location of the box, a note on condition and anything known about rights. Add who made the record and when.
What software do I need to catalogue photographs?
A spreadsheet is enough to begin, provided each row has a unique reference and the columns are used consistently. The National Archives offers a spreadsheet template for contributing catalogues to its Discovery service. Move to a collections database or digital asset management system once the structure is settled, and make sure whatever you choose can export your data.
Should I rearrange the photographs into subjects?
Not physically. The order in which a photographer kept the work, by job number, client or date, is evidence in itself and often the only clue to what a sleeve contains. Keep the original order on the shelf and create subject access in the catalogue, where one folder can sit under several headings at no risk to the material.
Sources
- The National Archives: Cataloguing archive collections
- The National Archives: Ownership and rights
- Library of Congress: Recommended Formats Statement, still image works
Published by Archive.London. Corrections are welcome under our editorial policy.

