Care
What we can learn from the Gardener
Reading Alison Gopnik’s[^1] The Gardener and the Carpenter (2016), a philosophical book on parenting, I found myself asking whether Gopnik’s approach to Care could serve as one foundation of a triadic framework for contemplating education and learning for the future in an AI and uncertain world. Curious to explore these connections further, I am examining her approach alongside Martha Nussbaum’s examination of Compassion[^2] and Rupert Wegerif’s exploration of Dialogic Education[^3]. That investigation is ongoing; this essay is where it begins.
Why the Gardener and the Carpenter?
Caring deeply about our children is an attribute of what makes us human. One could argue we share this with the animal world, but our engagement in care goes well beyond that of the animal world.
In her book, Gopnik focuses on “parenting”, something she considers a surprising invention, which of course it is. Although she writes in a US context, her reflections stray well beyond its borders, as she highlights how, in the past thirty years or so, parenting as a concept has grown into a multibillion-dollar industry. But this has not come without cost. The author argues that a transformation has taken place and that under today’s societal and market-driven pressures and expectations, caring for children has been redirected into anxious goal-oriented work, as parents aspire for their child to thrive in tomorrow’s world.
The significance of this premise goes well beyond parenting. It equally permeates a child’s education, in and outside the home.
Gopnik, a developmental psychologist and philosopher with her own AI research lab at Berkeley, puts the magnifying glass on the twenty-first-century image of parenting, an image she considers fundamentally wrong. Not only is our contemporary concept of parenting based on bad science; more importantly, it is bad for children as well as their parents. Being bold, one could argue it has a profound adverse effect on society, which appears to promote it.
To support her assertions, Gopnik draws on the study of human evolution and her leading research into how children learn. Caring for children, she argues, is fundamental to their development and flourishing. This seems obvious. However, its goal should not shape them to turn out in a predictable, pre-defined and specific way, responding to our aspirations. On the contrary: children are meant to be messy, unpredictable, playful and imaginative. These are attributes we are increasingly ignoring, perhaps even forgetting. Children are very different from their parents and very different from one another. Being one of six, I can certainly vouch for that. They are also different from what school often desires and expects them to be.
Difference, flexibility and self-discovery nurtured in childhood will let children innovate, create, imagine and develop to navigate an uncertain and unpredictable world. They will survive uncertainty by having the resilience to adapt and by inventing strategies to plot their own route through it. According to Gopnik, “parenting” won’t make children learn; instead, caring parents create a secure and loving environment where insecurity and unpredictability have their place. They will let children learn, allowing children to explore, imagine and discover, finding their own sense of agency and security. They let their child be a child, encourage curiosity and risk-taking that feed exploration, letting them know they care.
Caring parents let children discover that life and the world aren’t always predictable, just like the gardener knows that her propagated seeds may grow and flower but often have a tendency not to do so as expected. It is inherent in the beauty of growing a garden. She continues to nurture her garden, despite never being able to predict fully what it will turn into. It may pleasantly surprise, or at times disappoint. Thriving on that uncertainty and surprise is exactly what makes a good gardener.
That uncertainty compels a gardener to explore, discovering what thrives and what doesn’t, even stumbling across blooms appearing she never planted, and clearing weeds so the garden’s beauty can emerge. Uncertainty and surprise are not the price of ongoing trial and error; they are the reward. They tell the gardener what the garden needs.
Where is the Gardener in education?
Reading Gopnik takes me beyond parenting, to education. And the place to start is where her own research lives: the early years, the foundation of human development.
Modern society is obsessed with measuring and benchmarking. It starts from the moment of birth. At times, this happens for extremely valid reasons. Yet, almost from the first developmental observation record, we measure young children against what they are meant to do and the developmental stage they should have reached for their age, rather than attending to the unexpected that might surprise us, and that should be given its place in the child’s development. However, the unexpected is less measurable and less desirable as it doesn’t fit the rubric. But it would tell us so much more about who the child really is.
To be fair to the science, developmental psychology sets out to describe what children tend to do; yet somewhere between the laboratory and the reception classroom, the description hardens into a prescription and turns into the checklist of what children ought to do. That logic, once set, tends to stick and turn into a template the child should fit into.
Like “Parenting”, school and formal education are heavily goal-oriented, aspiring to create a particular learner, adolescent and young adult who conforms to an image, in which expectations around knowledge acquisition and readiness for life beyond education are pre-set and used to mould a child’s schooling.
We expect children to pass standardised tests that are treated as the benchmark of good education, what education should look like for children to turn into the adults society aspires to.
Is Education too much of a Carpenter?
Are we standardising our children?
Children drawing wiggly lines are told to draw straight lines. Play, imagination and scaffolded learning experiences that allow for exploration are too often belittled and deemed unfitting. Despite the overwhelming calls for personalisation, learning in education settings has been reduced to standardised knowledge acquisition. And underneath all of it sits a quieter mechanism. Our assessment frameworks, by their very nature, measure what children answer, never what they question, rarely how they apply what they have learned. Assessment wants certainty and measurability, resulting in its logical endpoint: the predicted grade.
In several school systems, learners are told their predicted grades before they even sit the test. It already defines their expected test achievement. This is not an oddity of the system; it is the system, perfected. If you only ever measure answers, answers become predictable, and once they are predictable, why wait for the child? The carpenter has the blueprint.
Do we ever ask ourselves what this does to a learner? Is this mechanism of assessing and predicting allowing learners to become curious explorers? Does it take learners into the meaningful and reflective construction of knowledge, and move learning forward?
Yet, the picture is not all doom-and-gloom, and glimpses of the Gardener can be found in educational settings all over the world, and we should treasure them.
An interesting example is Stanford’s Mobile Inquiry-based Learning Environment (SMILE)[4], which illustrates how our technology-driven world can enhance a nurturing and caring setting for learning and assessment. It is instructive for what it reveals. SMILE invites learners to generate the questions rather than answer them in an inquiry-based learning setting, and the depth of their inquiry becomes the object of interest. But SMILE’s research findings carry a sting. SMILE proved hardest to implement precisely in classrooms where rote learning was the norm. Children who had been assessed on answers throughout their whole schooling struggled when they were invited to ask questions. Not because the instinct to wonder was gone, but because they had systematically learnt that asking isn’t what’s wanted or rewarded: the right answer is. The question has been rubbed out; it has been silenced. And a capacity that remains hushed long enough is at risk of something worse than silence, of being extinguished entirely.
Care, not safeguarding
There is a word for what the gardener does dealing with the uncertainty of what the garden will yield. It is not “managing” or “monitoring” or “mitigating”. It is caring. Care is what makes uncertainty productive rather than merely chaotic. The gardener does not care despite not knowing what will grow, but because of it.
So where has Care gone in education? It certainly has not disappeared; too many good teachers prove that. It has been wrongly translated into safeguarding.
Allow me to set the record straight, because this is prone to being misread: safeguarding is not wrong, it is necessary.Children must be protected from genuine harm, and the duty schools carry is real and demanding, often dealing with highly challenging contexts and unpredictability of behaviour. The problem arrives when institutional safeguarding replaces relational care, when the school stops being a place where a child is really known and nurtured in her difference. Instead, it becomes a place where risk is strictly managed and boxes are ticked, where the safeguarding file becomes the child’s story.
And the aversion to risk is where parenting and schooling may have something in common. We can watch this happen most visibly in the physical world. The roaming radius of childhood and how far a child may wander unsupervised has collapsed within a few generations, from miles to metres, each generation kept closer “for its own good”[^5]. Gopnik also refers to this in her book. However, this is not solely a parental phenomenon. Schools mirror this trend.
Research into children’s physical play in schools finds teachers caught in a genuine bind: many want to let children explore, climb, tumble and test themselves. They recognise the developmental value of these explorative and risk-taking activities that help children discover their world. Yet reluctance, regulation and the fear of liability often hold them back[^6].
An example? A school plans a study day on water in their wonderful climate garden. The weather is forecast to be sunny and warm. The children are asked to bring their wellington boots. They aren’t really allowed to play barefoot as this is not in line with safeguarding. Their physical exploration and experience of mud and water have already been reduced. The requested spare set of clothing is of course a good idea when it comes to playing in and exploring water. What is interesting in this example is that the school understands the importance of play and exploration, giving the children some freedom to explore and discover, under careful supervision and within safeguarding constraints. School safeguarding anxiety often mirrors examples of parental anxiety, making sure some children are wearing life jackets, worried about roaming towards a pond that turned out to be a very large puddle.
But the climbing frame, at times the source of the odd accident or the odd swollen lip, is only the visible edge. It has become the metaphor for the institutional reflex that not only removes physical risk, but worse, removes intellectual risk: the unpredictable question, the unmarked exploration, the lesson that might not “work”, the answer that wasn’t on the mark scheme. We safeguard the body and, almost without noticing, we safeguard the mind from venture, from failure, from surprise. From precisely the conditions under which, as Gopnik highlights, children actually learn.
A gardener would recognise these conditions as a paradox to Care. Care holds the space for the fall and the recovery, for the plant that struggles to take and the one that takes over. Care accepts that the child will grow beyond you, beyond prediction; that this is the point. Safeguarding, pursued as an end in itself, quietly reverses that logic. It succeeds when nothing unexpected happens. Which is to say, it prospers when nothing is happening at all.
Assessing for measuring or caring?
Practices of assessment and monitoring children’s development should not be dismissed. But have we gone too far, forgetting that growing up and learning tend to be rather messy, and that this messiness is what nurtures children’s development?
We assess children on answers, not questions. As highlighted, several education systems predict their grades before they even sit the test using defined benchmarks and mock tests. We determine their risk instead of accompanying their learning endeavours and encouraging exploration. Whilst these practices may be defensible in terms of educational aspirations, together, they may describe a carpenter’s workshop, a system quietly optimised for the predictable child.
The world children will inherit is not becoming more certain, it is becoming radically less so. With the acceleration of AI, it is becoming less certain in a peculiar, even paradoxical way. We are surrounded by machines operating through prediction, whose workings we understand less and less. The old premise of education: acquire the knowledge, demonstrate it and be rewarded, is dissolving. What children will need is not a better blueprint. They will need what the gardener cultivates: adaptability, curiosity, the nerve to explore and resilience to be surprised by what they learn. They will need to be their own agents, applying their knowledge and become forever learners in a world that can’t be predicted for them.
Yet here is an uncomfortable twist: AI could easily reduce the gardener’s prowess and resilience, making the carpenter’s design more predictable and influential. The same predictive power that unsettles the old certainties also offers education an irresistible tool: hyper-personalised learning, every child’s pathway optimised, every outcome forecast. We may call it personalisation, but forget the most fundamental thing: knowing the child.
Differentiation by the algorithm. Prediction dressed up as achievement.
Unless, that is, we decide otherwise: SMILE is a reminder that technology can just as readily nurture the space for inquiry, discovery and exploration, reopening the question rather than closing it. The machine is not the carpenter. The intent is.
Care, then, is where a rethink of education should begin. But care alone is not enough. Care describes a stance towards the child in front of us; it does not yet tell us how education might cultivate a wider moral imagination. That is a different concept: Compassion. This deserves its own essay. Nor does care alone tell us what teaching and learning look like when they are built for openness rather than certainty. That, too, is to come.
Gopnik shows how modern society has systematically stripped the nurturing, exploratory, and even risk-taking instincts of childhood from parenting and schooling, in pursuit of our aspiration for the predictable child and her safety. For Gopnik, care is not the mother’s alone; it is shared by grandmothers, alloparents and the several adults who raise a child. And then, in the summer of 2025, Geoffrey Hinton, the man often called the godfather of AI and as responsible as anyone for the predictive machinery now reshaping our world, took the stage at Ai4 in Las Vegas. He suggested that what we most urgently need for artificial intelligence is maternal instinct. Conceding he doesn’t know how to embed such instincts in AI models, he urged researchers to prioritise the task[^7].
The instinct we seem to have engineered out of our schools is the one AI’s own architect believes should be engineered into the machine itself.
Inspired by Gopnik’s The Gardener and the Carpenter, I am still in awe of the beauty the carpenter can bring to our world, yet I’m even more struck by the care that has gone into the creation of a garden that evolves and looks different and offers new perfumes every time I am in it.
This is the first of three essays discussing focusing on the future of education and what education should consider. The second takes up Compassion, through the work of Martha Nussbaum; the third turns to what Dialogic Education might look like, with Rupert Wegerif.
Written with some editorial help and challenging by Claude.
[^1]: Alison Gopnik, https://www.alisongopnik.com and https://en.wikipedia.org/wiki/Alison_Gopnik. See also Jenny Anderson and Rebecca Winthrop, The Disengaged Teen: Helping Kids Learn Better, Feel Better and Live Better (2025), whose “explorer” learner echoes Gopnik’s argument.
[^2]: Martha Nussbaum, Upheavals of Thought: The Intelligence of Emotions (2001).
[^3]: Rupert Wegerif, Dialogic Education and Technology: Expanding the Space of Learning (2007), Rethinking Education Theory: Education as Expanding Dialogue (2025).
[^4]: Stanford Mobile Inquiry-based Learning Environment. See Buckner, E. & Kim, P. (2014), “Integrating technology and pedagogy for inquiry-based learning: The Stanford Mobile Inquiry-based Learning Environment (SMILE)”, Prospects, 44(1).
[^5]: “Letting Children Roam”, Resilience.org (2017), https://www.resilience.org/stories/2017-11-13/letting-children-roam/.
[^6]: Jerebine, A., Exploring risk in children’s active physical play in schools, doctoral thesis, Coventry University; and the preschool study summarised at https://research.childrenandnature.org/research/limits-to-risky-play-at-preschool-include-teachers-reluctance-lack-of-affordances-and-safety-regulations/ (pull the specific finding you want to cite — I have not verified the figures].
[^7]: Reported from Ai4, Las Vegas, August 2025, by multiple outlets