Learning a trade by standing there
A Socratic walk-through of learning a trade by standing there — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why must a trade be learned standing beside someone doing it when the manual is right there?
The manual exists. It is accurate, it is thorough, and a determined reader could get through it in a fortnight. Yet no trade anywhere accepts reading as a substitute for years on site — and the requirement survives in industries with every commercial reason to shorten it.
So either an entire sector is being sentimental, or the manual is not the curriculum. If it is the second, the sharp question is not "what is missing from the book?" but: what does an apprentice have access to by being present that no reader has, however good the text?
Reasoning it through
REASONING #Begin with the obvious answer and walk past it. Being there lets you watch someone's hands — but a camera does that, and nobody becomes an electrician by watching footage.
What presence adds is the whole job, in its real order, with everything the write-up had no reason to include. A procedure lists steps. The job also has sequencing that is not in the steps: what must be done first because it becomes physically impossible later, what is left loose until the end, which of the day's four unfinished things deserves attention now. Ask a tradesperson why they worked in that order and you often get a shrug, because the order is not stored as a rule. It is stored as a preference already paid for in mistakes.
Then there is when to stop. A text says "secure the connection". On site, done is a threshold enforced by other people — the inspector, the colleague who works on it next week, the customer who phones if it rattles — and an apprentice discovers where it sits by having work handed back. That is not information transfer; it is calibration against a standard held collectively and never written down at the resolution at which it is applied.
Now look again at the tasks a newcomer is given. Sweeping, fetching, holding the light, tidying at the end. The usual reading is that these are dues — work too trivial to entrust to anyone else. Jean Lave and Etienne Wenger read them the other way: such tasks are peripheral, in that little is lost if botched, but legitimate, in that they are genuinely needed and place the newcomer inside the job's field of view. Fetching the right fitting already requires knowing which one is right. They called this legitimate peripheral participation, and their claim is that it is the mechanism of learning rather than the waiting room before it.
Their tailoring case makes the design visible. Among the Vai and Gola tailors Lave studied in Liberia, apprentices learn roughly backwards from the order of production: finishing and ironing first, then sewing, and cutting last. That looks inefficient and is not. You spend your first year handling completed garments, so the target is in your hands long before you are allowed to spoil expensive cloth — and every stage you are finally taught, you have already seen the output of a thousand times.
Why, then, can the manual not simply be improved? The obstacle is structural rather than editorial. The people qualified to write down the sequencing and the thresholds are exactly those for whom both have gone automatic and unavailable to introspection. Asked to explain, an expert produces a tidy reconstruction that is not what they did.
One further thing travels through presence that no text addresses: membership. You learn whom to ask, what this trade finds interesting, what counts as a reasonable excuse. You are not only acquiring a skill; you are being admitted.
The analogy
THE ANALOGY #Learning a trade this way is like learning a language by living where it is spoken rather than from a grammar. The grammar holds the rules, correctly. The street holds what people actually say, when, and the expression on a face when you get it subtly wrong.
immersion surrounds you with a whole community speaking constantly, whereas a workplace shows an apprentice only the jobs that happened to come in that year — so a trade learned purely by presence carries exactly the gaps that year's work had, which is precisely what a syllabus exists to fill.
Clarifying the model
THE MODEL #None of this requires the missing knowledge to be mystical. It is ordinary and specific — sequencing, thresholds, cues, what done means here. What makes it resistant to writing is that it is stored across situations rather than as propositions, so the only reliable medium for carrying it is another situation.
Nor does it make the manual useless. Text is good at codes, tolerances and safety limits: the things that are precise, stable, and far too expensive to learn by trial. Presence and text are complements, and a modern apprenticeship pairs them deliberately. The pairing is the design, not a compromise.
Two honesty points. Lave and Wenger's framework is descriptive, built from a small set of ethnographies, and has been fairly criticised for making apprenticeship sound benign, when real ones ration access, lean on cheap labour, and gatekeep along lines unrelated to skill. Presence also transmits bad practice as faithfully as good. And the account says little about speed: for the parts of a trade that can be articulated, explicit instruction with feedback usually beats waiting for absorption.
A picture of it
THE PICTURE #How to readRead each box as a kind of participation, not a rank — the apprentice does real work at every stage, only work that costs less if it goes wrong. The downward arrows are widening access, and each is granted by other people rather than earned by study. The two upward arrows are the actual teaching moment: work handed back is how the threshold of done gets learned, since it is nowhere stated. The first box has no task at all and the last has stopped being inspected — the trajectory is a move from the edge of the work to its centre.
What became clearer
WHAT CLEARED #The manual is not an inadequate version of the curriculum; it is a different document with a different job. What has to be learned standing there is the material that only exists in situations — the order of operations, the cue that says stop, the collectively enforced threshold of good enough, and membership in the group that holds it. An apprenticeship is not a slow way of delivering a book's contents. It is the only available way of joining the practice the book is a summary of.
Where to go next
ONWARD #- Why formal vocational qualifications so often fail to certify the part of the trade that actually varies between workers.
- What happens to this learning when the work moves onto a screen and the newcomer can no longer see the expert's hands.
Key terms
TERMS #| Term | What it means |
|---|---|
| Legitimate peripheral participation | Lave and Wenger's term for newcomers doing real but low-stakes work at the edge of a practice, as the means by which they learn it. |
| Community of practice | the group of practitioners whose shared standards, vocabulary and judgements a learner is joining, not merely learning from. |
| Tacit knowledge | knowledge a competent practitioner uses reliably but cannot fully state, often because it has become automatic. |
Every term the collection defines is gathered in the glossary.