You Cannot Learn This Alone
Why judgement can’t be taught in complex work, and what that means for how engineers develop.
By Chris and Penny Hamlin
One line has come back to us more often than any other in the research behind the Beyond the Blueprint programme. It was said by a mid-career engineer, technically strong, well regarded, in a job most people would consider a success.
“I chose to become an engineer to help make the world a better place, but I am struggling to understand how to get there from where I am.”
Most people hear the first half of that sentence. The half that matters is the second. This is not somebody who has lost their reasons, or who needs reminding why the work is worth doing. The reasons are intact and stated plainly. What has gone is any visible route from where they are standing to the thing they came into engineering to do.
It is worth being clear what that is not. It is not a request for promotion, and it is not an admission of technical weakness. The person who said they can do the work in front of them to a high standard, and knows it. It is a statement about position rather than about motivation, and about being able to see the destination and not the path.
The instinctive organisational response is a course. Some of those courses are good. Almost none of them touch this, and the reason is worth setting out carefully, because it explains a great deal else besides. The capability that sentence is reaching for is not a body of knowledge. It cannot be handed over. And in more than fifty years between us, we have never seen anybody acquire it on their own.
Hard is different from unpredictable
Most of an engineer's training is aimed at problems that are hard but knowable. Size the exchanger, tune the loop, run the calculation, check it against the standard. These can be extremely difficult, but they share a property that makes them tractable: there is a right answer, it exists before you find it, and once someone has found it the method transfers. That is what expertise is for, and engineering is very good at building it.
The problems that arrive with experience do not behave that way. They involve people with competing intentions, conditions that shift while you are deciding, and consequences that only become visible after the decision is irreversible. There is no right answer waiting to be found. There is a judgement about which action is worth taking now, and a readiness to discover you were wrong and adjust.
Judgement of that kind does not travel as content. You can be told exactly what someone else decided and precisely why, and be no better able to make the decision when it is your turn. Knowing the reasoning is not the same as having the judgement. That is not a failure of explanation. It is the nature of the thing being explained.
How engineers used to pick it up anyway
Historically this mattered less, because judgement was acquired without anybody arranging it. It came by proximity.
Early in Chris's career, at ICI, the facility was old and the question of whether it had a future was live and unresolved. The meetings were not where the useful material was. The useful material was the twenty minutes afterwards, when two senior people stood in the corridor and worked out what had actually been agreed, what they were going to do about the part nobody had said out loud, and which risks they were genuinely worried about rather than formally recording. If you were nearby, you heard it. If you heard enough of it over enough years, you began to be able to do it yourself.
Something larger worked the same way. For years the internal conversation about that business had been about survival: how bad things were, how old the technology was, what the market was doing to margins. Then it was sold to Huntsman, and the new owners arrived with their attention on the people rather than on the market or the ageing plant. They made it clear they thought they had acquired some of the most capable engineers in the industry, who had kept a difficult asset running against the odds, and they reinforced it consistently over a long period rather than announcing it once. The attitude in the organisation shifted, and with it what the place believed it could do.
It is worth being honest that this did not land the same way for everyone. It reached the people who already had something invested in the future of the place, whose reasons for being there went beyond the wage. For others, whose attachment was closer to transactional, it changed less. Whether what the new owners activated was motivation or something more like relief is a question we still find genuinely difficult, and we raise it here only to make the point that this material is not simple.
What is not in doubt is that nobody learned any of it from an announcement. It was witnessed at close range over months. None of it was on a development plan. It was not designed and it was not measured. It was ambient, and because it was ambient it was invisible, which is why its removal from today’s working environment has been so poorly noticed.
What has actually changed
The easy explanation is on-line working, but it is too easy. Remote and hybrid patterns have certainly thinned the ambient contact, but engineers were losing this well before 2020, and an argument that blames the technology invites anyone who prefers or needs to work remotely to dismiss the whole thing.
What has really changed is the pace, intensity and variety of communication, and the fact that organisational structures have not kept up with any of it. There is more contact than ever and less of it is the kind that develops anybody. The meeting is compressed. The decision arrives as a message. The twenty minutes in the corridor does not happen, because everyone is already in the next call. The channels multiplied and the sensemaking did not.
So if you are finding this difficult, the difficulty is structural. It is not a deficiency in you. The scaffolding that used to hold engineers up while they developed judgement has been dismantled without replacement, and being asked to manage without it is not a fair test of anything.
Lone experience is not enough either
If the answer is not content, and the informal route has closed, the question becomes what conditions actually let judgement develop. The first step is the one most often skipped. The ideas that matter in this kind of complex work are simple to state and phenomenally nuanced in application. You do not learn to ride a bike by studying angular momentum, nor by watching somebody else ride one. You learn, and this is the part usually left out, by riding it, falling off, and having someone alongside you who fell off for the same reason last week.
Experience without reflection produces habit. Experience with reflection produces judgement. And the most useful reflection is not offered from outside by someone observing your situation. Coaching has a real contribution and we use it, but the reflection that changes what you do next tends to come from peers with something at stake in a similar context, who will say the thing a polite observer will not, and who are close enough to the problem to notice which part of your account you have quietly edited.
Which means the cohort is not the delivery mechanism for the Beyond the Blueprint programme. It is the mechanism. We did not choose it because it is fashionable or because groups feel supportive. We chose it because in this kind of work it is the only structure that matches how the capability actually develops. The design is a consequence of the diagnosis, not a justification arrived at afterwards.
Why the group outlasts the programme
Cohorts have been fundamental to most of the personal growth and development programmes Chris and Penny have benefited from over the years, most recently Iconic for Chris and Celebration of Being for Penny. The consistent finding across all of them is slightly awkward: the curriculum is rarely what lasts.
Penny went through a Civil Service management development programme years ago and can remember almost nothing of the learning, with the exception of having to memorise a poem. The relationships formed on it were stronger and lasted longer than most of those formed at work over the same period.
Chris's MBA cohort is the same story with a longer run on it. What that group remembers is the pub lunches between lectures and skipping a session to watch a rugby World Cup final in the common room, with the consequences that followed. Out of it came a relationship with Alistair Bowden that has now run for more than twenty years, through periods of intense contact and periods of years with none, through families growing up alongside each other and children playing rugby against each other. In all that time it has never stopped provoking original thought in either direction. Alistair now sits on the Beyond the Blueprint Advisory Board, which is not a coincidence, it’s exactly the point.
Two ingredients are necessary. Shared human experience, the unscheduled part that happened alongside the learning rather than in it. And a live intellectual thread, something the relationship can keep working on. Warmth alone decays into fondness. A shared interest alone is a mailing list. Together they produce something still generating value two decades later with nobody maintaining it.
What a cohort has to allow
Some of the clearest evidence comes from outside engineering entirely. Penny's experience of Celebration of Being was that people arrived with no history of each other and no preconceptions, and simply were themselves. Almost none of the early conversation touched anyone's job or professional background. That surfaced later, almost as an afterthought. People engaged with what was in front of them rather than with a reputation that had arrived ahead of the person.
The value of that inversion is practical rather than sentimental. Hierarchy suppresses the content of a room. The person holding the observation that would change the discussion does not offer it, because their standing does not grant them permission. The person with the seniority does not receive it, because their standing obliges them to transmit rather than listen. Both suppressions lift at once when people arrive as people, and what was already in the room becomes available.
Hierarchy here is not only the organisation chart. It is anything that tells someone their contribution is worth less before they have made it: accent, gender, background, discipline, years served, familiarity with and prior knowledge of others, the confidence of whoever spoke first. A room that suspends all of that is not being accommodating, it is stimulating and inspiring.
What we are actually trying to build
It would be dishonest to describe the six-month Beyond the Blueprint programme as the whole ambition, so here it is plainly. We want EvolveCollective to become something that outlasts any cohort's end date: a community where some of the most experienced engineers in the world support others earlier in their careers, where the conversation covers the balance of professional, personal and societal obligations rather than career progression alone, and which people return to at different stages of a working life rather than graduating out of.
We cannot install that. It is not something that can be specified and commissioned. It has to form, around people who care about the same things, and our job is to convene it, hold the conditions and then get out of the way while it becomes whatever it becomes. We are clear-eyed that this is an aspiration and not yet a track record.
It is, though, the only honest answer we have to the engineer we started with, whose difficulty was never a shortage of reasons or motivation. The route to making a difference tends to become visible in community.
The first cohort of Beyond the Blueprint begins in October. Its members are founding members, which is a plain description rather than a flourish: they are joining something unfinished, and what it becomes will be shaped by them rather than presented to them.
There is a harder question sitting immediately behind this one. If more of the technical work is going to be done by systems, what remains of the engineer is precisely the part that is least automatable and least learnable alone. That is worth talking about next.
Image credit: AI generated using OpenAI’s image-generation tool in ChatGPT, prompted by PJ Hamlin, 29 July 2026