The hardest question in cooking is whether the salt already went in, and a bench can answer it

The plan and the record are two different problems

There are two things you have to know while cooking, and only one of them gets any help. The plan is what happens next, and every recipe, card, app and printed list in the world is aimed at it. The record is what has already happened, and almost nothing is aimed at that at all.

The record is the one that fails first, and it fails in a specific and maddening way. Did the salt go in. Is that the second spoon of cumin or the first. Has this pan had oil in it. Those questions cannot be answered by looking at a recipe, because a recipe describes the intention rather than the state of your kitchen. And they cannot be answered by tasting, most of the time, because by the time you can taste the difference the decision has already been made.

Qualitative work on meal preparation difficulties after acquired brain injury describes the barriers as cognitive, physical and psychosocial together rather than as one thing, and points toward support that is tailored to the individual rather than handed out as a standard set ^[4]. Sitting inside that is the observation this essay turns on: knowing what to do and knowing where you are are separate demands, and a person can have one intact while the other is not.

Which is why the usual advice underperforms. Simplify the recipe, use fewer ingredients, write it down in big letters: all reasonable, all aimed at the plan, and none of it touches the question of whether the salt already went in.

Sources for this section: [4] Potential of using an assistive technology to address meal preparation difficulties following acquired brain injury: clients' and caregivers' perspectives

Make the bench show what has already gone in

The fix is to arrange things so that the record keeps itself. Nothing has to be ticked, nothing has to be remembered, and nothing depends on you noticing at the right moment.

Start with a line. Everything the recipe needs goes out on the bench, left to right, in the order it will be used, before any heat is involved. As each thing goes into the pan, it moves across to the other side, or into the sink, or into a tray. Then the question is not a memory question at all. What is left on the left has not gone in yet. It takes about ninety seconds to set up and it removes the single most common failure in the whole task.

Extend the same idea to the things that do not come in packets. Leave the empty tin, the empty packet and the used spice jar standing on the bench until the meal is on the table, because each one is a receipt. Keep a separate spoon in each pot rather than a single spoon that travels, so that a spoon on the bench means something. Turn a lid upside down once the pot has been salted, if that is your particular trouble. Put the kettle switch and the oven dial where you look rather than where they were installed, insofar as you can.

The principle underneath all of it is that objects cannot forget to be updated. A checklist is only accurate if the person using it interrupts what they are doing to mark it, and interrupting what you are doing is the exact operation that is expensive here. A tin that is now empty is accurate whether anybody was paying attention or not.

One tray holds the meal and the bench holds nothing else

The second half of the problem is everything that is on the bench and has nothing to do with tonight. A crowded surface makes the record unreadable, because the difference between used and not yet used disappears into a general clutter of objects.

So stage a kit. Before starting, put everything this meal needs onto one tray, and take everything else off the working surface. The post, the fruit bowl, the tablets, the shopping that has not been put away and the three things belonging to the last meal all go somewhere else, even if somewhere else is a box on a chair. What remains on the bench is the meal, and the meal is now legible.

There is a second benefit that people notice more than the first. A tray with everything on it is also proof that nothing is missing, which removes the mid-task search: the moment where you go to the cupboard for the paprika, find yourself in front of the cupboard without knowing why, and lose the thread. The search is where a lot of meals actually die, and staging removes most of the occasions for it.

Consistent homes for things do the same work over a longer timescale, and this is where a household has to cooperate. Frequently used tools and ingredients need a stable location that matches how the person actually works, which means other people returning things to the same place rather than to a more sensible one. Home-based occupational therapy for people with dementia and their family carers, which has been examined in systematic review and meta-analysis with effects reported on daily functioning and on carer outcomes, is built around exactly this kind of joint work on the environment and the routine rather than on the person alone ^[2].

Sources for this section: [2] Occupational therapy for people with dementia and their family carers provided at home: a systematic review and meta-analysis

A recipe rewritten as a list of objects rather than a list of sentences

Recipes are written as prose because prose is compact, and prose is the worst possible format for somebody who has to re-enter a task repeatedly. Every re-entry means finding your place inside a paragraph, which means reading sentences you have already done to work out which ones they were.

Rewriting helps more than people expect, and it only has to be done once per dish. One instruction per line, numbered, in the order they happen, with the quantity inside the line rather than in a separate list at the top. Each line starts with a verb. No line contains two actions. Anything that has to wait gets its own line so that waiting is a step in place of a gap.

Then put the numbers on the physical world too. Number the containers on the tray to match the lines, or set out the little bowls in the order they get used. Once the recipe and the bench agree with each other, finding your place stops being a reading task and becomes a looking task, and looking is far cheaper.

Keep the rewritten versions. A folder of six or eight dishes in this format, in a place that never moves, is worth more than any amount of general strategy, because the work of converting a recipe is exactly the work that is hardest to do on a bad evening. Do it on a good afternoon, once, with somebody else if that helps, and it is done for good.

Where a device fits, and what configuring one actually takes

There is a real research field here and it is worth knowing what it says, because the marketing around cognitive assistance is far ahead of the findings.

Work on an assistive technology for meal preparation implemented in a supported residence for adults with acquired brain injury reports promising results, and just as usefully reports the practical business of implementation: the setup, the configuration, the training of staff and users, and the ongoing adjustment that a real setting demands ^[3]. A configuration study of a cognitive orthosis for meal preparation looks at the same problem from the other side, examining what it takes to configure such a system with efficiency and effectiveness, which is itself an admission that the configuration is substantial work instead of an afternoon ^[5].

The honest summary is that these are small, context-specific studies of promising systems, not evidence that a device will suit a particular person or that it makes independent cooking safe. Anything sold on that second claim has gone past what has been shown.

What generalises from that literature is a way of choosing. Name the exact failure first: losing place in the sequence, forgetting a step, not starting at all, not noticing a cue, or needing another person to prompt. Then ask whether the thing in front of you addresses that specific failure, whether the person can operate it without a hidden gesture or a tiny screen, and whether anybody in the household can tell when it has stopped working. A device nobody can reconfigure becomes a problem within a month.

And a boundary that is not negotiable. No feature on a device settles whether somebody can safely cook unsupervised. That is an individual decision made by people who know the person and have assessed them, and technology is one layer inside a support plan in place of a verdict on it.

Sources for this section: [3] Implementation of an assistive technology for meal preparation within a supported residence for adults with acquired brain injury: a mixed-methods single case study · [5] Can we configure COOK, a cognitive Orthosis for meal preparation, with efficiency and effectiveness?

Being a prompt without becoming the cook

The instinct when somebody hesitates is to say the next step out loud. It is fast, it works, and repeated every evening it turns cooking into something done under supervision, which is the arrangement most likely to end it.

The distinction worth holding is between building the system and running it. Setting the line out, staging the tray, rewriting the recipe, agreeing where the tin opener lives: that is building, it happens once, and it leaves nobody in the room afterwards. Standing at the person's shoulder narrating the method is running, and it has to happen every single time, which means it is not really support at all but a permanent staffing arrangement neither of you wanted.

If you do have to prompt in the moment, prompt at the level of the record instead of the plan. Instead of telling somebody to add the salt, which does the thinking for them, point at the line of ingredients: what is still on this side. That gives back the information without taking the decision, and over time it teaches the system instead of replacing it.

Before choosing any of this, watch a real meal being made and see where it actually breaks down, because the answer is frequently not where the household assumes. Qualitative work on meal preparation after acquired brain injury describes barriers that are cognitive, physical and psychosocial at once, and support that has to be matched to the individual ^[4]. The step where somebody stalls may be the starting, or the searching, or the fact that cooking for one feels pointless, and each of those needs a different answer.

If it is worth a professional look, ask a GP for occupational therapy and name meal preparation specifically. It is standard territory for the discipline, used both to see how everyday function is holding up and to work on it ^[1], and a therapist watching the real task in the real kitchen will spot in twenty minutes what a family has been arguing about for six months.

Sources for this section: [1] Occupational therapy and cooking: A scoping review and future directions · [4] Potential of using an assistive technology to address meal preparation difficulties following acquired brain injury: clients' and caregivers' perspectives