Start with an identity, not a verdict
A jar labelled “successful” tells a future reader almost nothing. Begin instead with a stable batch identifier, the recorder’s name, the ingredient source and the question being investigated. The identifier should connect the vessel, photographs, observations and any later analysis. It need not be elaborate; it must remain consistent when the notebook changes hands.
Distinguish a culture’s lineage from a batch’s history. A starter may have passed between several kitchens, while the current batch began with a particular delivery of flour. Record what is known about each, and mark gaps explicitly. “Reported by the donor” and “documented in the original notebook” describe different levels of evidence. Neither should silently become an established historical fact.
Keep original labels when possible, alongside a readable transcription. A spelling, local name or language choice may carry information that disappears when everything is standardised for a spreadsheet.
Describe the conditions of observation
Useful notes say how an observation was made. “The room felt warm” is a sensory recollection. A temperature reading is a measurement, but it still needs a time, a location and an instrument context. Was the reading taken in the room or in the material? Was the instrument checked according to its manufacturer’s instructions? Precision in a display is not the same as accuracy.
The same distinction applies to acidity. A pH entry without a sampling method or calibration record is difficult to interpret. Titratable acidity and pH are different measurements, and neither should be casually substituted for the other. A studio archive can preserve these details without presenting itself as a laboratory or claiming that a recorded value establishes safety.
| Field | What to preserve | Important limitation |
|---|---|---|
| Identity | Batch code, keeper, source culture | Lineage may be incomplete |
| Ingredients | Supplier, lot, amounts, substitutions | Similar names do not imply equivalence |
| Observation | Time, appearance, instrument context | Appearance does not establish safety |
| Revision | Author, change, reason, previous wording | A correction should not erase its history |
Give images a job
A photograph is most useful when the reader knows what it was meant to show. Include a date, the batch identifier, the viewpoint and a short description. Keep lighting and framing reasonably consistent if comparing visible changes. Do not adjust colour so heavily that a documentary image becomes misleading. A photograph can establish that a surface looked a certain way; it cannot identify microorganisms or rule out hazards.
People also belong in the image record, but permission matters. Ask before publishing a participant’s face, private notes or identifying workplace details. A contributor may agree to an internal research photograph without agreeing to its use on a public website. Record the scope of permission separately from the image caption.
Keep source knowledge attached
Record who taught a method, which publication informed a choice and whether the studio adapted it. Sandor Ellix Katz’s The Art of Fermentation and Wild Fermentation can open a wide reading path, but citing a book does not replace acknowledging a person who demonstrated a practice. Likewise, a restaurant interpretation should not displace the community knowledge on which it depends.
The Dublin Core metadata terms offer useful language for creator, source, date and rights. A small studio does not need a museum database to borrow that discipline. A carefully maintained plain-text index may be more durable than a complicated system that nobody updates. Store at least one backup separately and test whether the files can actually be opened.
Make disagreement and revision visible
Do not overwrite yesterday’s conclusion without explanation. Add a dated correction describing what changed and why. If an ingredient identification was uncertain, preserve the original uncertainty rather than polishing it away. Negative results, abandoned trials and missing observations all help a future researcher understand the limits of a collection.
A useful handover ends with unresolved questions: Which records remain inaccessible? Which permissions need renewal? Which interpretations rely on one person’s memory? This is where a fermentation library becomes more than a recipe folder. It preserves the conditions under which knowledge was made, and gives the next keeper a responsible place to begin.
Next entry: knowledge across kitchens →