A child cannot quite solve the problem alone.
The pieces on the table are familiar. The instruction is clear enough. The child tries one arrangement, then another. An adult does not take over. Instead, a question is offered, a gesture points toward a relationship the child has not yet noticed, and the task changes shape. A few minutes later the child is doing something that was not available at the beginning.
Where, exactly, did the new capacity exist?
Not fully in the child before the interaction. Not simply in the adult. Not in the words alone. It emerged through a relation structured by language, tools, attention, history, and participation.
Lev Vygotsky's work remains influential because he made development difficult to imagine as an isolated mind unfolding by itself. Higher psychological processes, in his account, are formed through social and cultural activity. Language and other symbolic tools do not merely express thought after it develops. They participate in the development of thought.
Vygotsky did not write a modern theory of Inner Technology. The legitimate connection is narrower: his work shows why human capacities cannot be understood only as properties inside individuals. Development is mediated.
The social does not sit outside the mind
Vygotsky's cultural-historical approach challenged accounts of development that treated social context as background. In the selections published in Mind in Society, he argued that higher functions appear first in social interaction and are later reorganized within the individual.
That claim has often been simplified into the slogan that learning is social. The original idea is more radical.
Culture supplies tools for thinking. Language, writing, number systems, diagrams, categories, gestures, maps, and institutional routines change what people can remember, plan, communicate, and regulate. These tools are not neutral containers. They reorganize activity.
A child using private speech to guide a difficult task is not merely talking out loud. Speech is participating in self-regulation. A student learning algebra acquires not only answers but symbolic means for representing relationships that would otherwise be difficult to hold. A team using a shared decision protocol changes the distribution of attention and responsibility in the room.
Capacity, from this perspective, develops through mediated activity.
The zone that became a slogan
Vygotsky's zone of proximal development is among the most repeated ideas in education. It is also among the most flattened.
In broad terms, the zone concerns the distance between what a learner can do independently and what becomes possible with guidance or collaboration. The concept directs attention away from static performance toward developing possibility.
A test of what someone can do alone tells us something important. It does not tell us everything about what the person is ready to learn.
That insight helped inspire later ideas such as scaffolding, although "scaffolding" itself was developed by Wood, Bruner, and Ross rather than coined by Vygotsky. Contemporary scholars have warned against treating the zone as a fixed band into which the right instructional technique can simply be inserted. Vygotsky's developmental theory is more dynamic, involving the relation between instruction, emerging functions, and cultural activity.
This distinction is important for the Institute. The useful principle is not "give just enough help." It is that human capability can be revealed and developed in supported relations that change what becomes possible.
Internalization without a little person moving inward
The language of internalization can sound as though social material is copied from outside to inside. That is too mechanical.
Vygotsky's deeper claim is transformational. Socially organized processes are reconstructed as psychological processes. The form changes.
A child first participates in a shared act of attention and later becomes capable of directing attention more independently. External reminders can become internal strategies. Dialogue can become inner speech. Cultural tools can become part of how a person thinks.
This has direct implications for practice architecture. A training environment does more than present content. It organizes forms of participation that may, over time, become capacities the person can carry elsewhere.
But transfer cannot be assumed. A person may perform well with support and struggle without it. A strategy learned in one cultural or task context may not generalize cleanly. The movement from supported performance to independent capability is precisely what must be studied rather than presumed.
Language as developmental infrastructure
Vygotsky's work on thought and language is especially relevant in a time when language itself is increasingly generated by machines.
Human language is not only a channel for reporting finished thought. It helps organize categories, plans, memory, reflection, and social coordination. The words available in a culture shape what can be jointly attended to and discussed.
This is one reason educational language matters. If a learner acquires a precise distinction, such as evidence versus inference, the distinction can become a tool for future thought. If a team acquires language for uncertainty, conflict, or ethical risk, experiences that were previously vague can become discussable.
The same principle creates a new question for generative AI. If people increasingly receive language before they have formed a thought, what is being mediated, and by whom?
AI can expand access to explanation, examples, translation, and feedback. It can also place ready-made formulations into the developmental sequence before the learner has performed the organizing work those formulations once required. Vygotsky does not tell us whether this is beneficial or harmful. He gives us a better unit of analysis: the person using cultural tools inside a social activity.
From learner to institution
The social mediation of capacity does not end when childhood ends.
Workplaces, professions, scientific communities, civic institutions, and digital platforms all provide symbolic tools, norms, roles, and forms of participation. They train people in what deserves attention and how problems should be represented.
A hospital teaches clinical reasoning through rounds, records, categories, seniority, devices, and interaction with patients. A newsroom teaches verification through routines and editorial challenge. A laboratory teaches evidence through instruments and collective scrutiny. A software team teaches problem-solving through code review, issue systems, documentation, and informal norms.
The capacity is distributed across person and practice before it becomes available as individual expertise.
This is why the Institute's concern with human development cannot be satisfied by asking whether individuals possess traits. It must ask what developmental relations institutions make possible.
Who gets access to more capable peers? Which tools clarify the work? Which forms of language sharpen perception? What kinds of help foster independence, and which create dependence? Which activities let emerging capacities become visible before formal performance metrics would detect them?
These are Vygotskian questions even when the setting is adult and technological.
Measurement and interpretation
Vygotsky's historical concepts do not correspond to one modern test battery.
Researchers have operationalized the zone of proximal development through dynamic assessment and forms of mediated learning, but methods vary. Static tests estimate current independent performance. Dynamic approaches examine how performance changes with prompts, feedback, teaching, or graduated assistance.
This can reveal learning potential that a single unaided score may miss. It also creates measurement complexity. The amount and quality of assistance must be standardized or documented. Improvement during mediation can reflect task familiarity, motivation, rapport, prior knowledge, language, or strategy use. A dynamic score is not a context-free measure of an inner growth capacity.
The cleanest approach is to describe the support, task, population, and outcome precisely.
For Institute work, Vygotsky should guide design questions more often than it supplies a ready-made metric.
Support can reveal capacity and also conceal it
The relational nature of development creates a measurement paradox.
Assistance can reveal what a learner is ready to do, but assistance can also make performance look stronger than the capacity that will remain when support disappears. This is why dynamic assessment is methodologically interesting and difficult. The researcher has to record not only whether performance improved, but what kind of prompt produced change, how much help was required, whether learning transferred, and whether support could be reduced over time.
The distinction is especially important with AI. Generative systems can provide unusually rich mediation: hints, examples, reformulations, explanations, plans, and completed responses. A person may perform at a high level while the system is present. That performance cannot automatically be read as independent capability.
Nor should dependency be assumed simply because assistance was used. Human expertise has always been tool-mediated. Mathematicians use notation. Writers use dictionaries and editors. Scientists use instruments. People with disabilities may gain real capability through assistive technology that should not be judged by an ideal of tool-free independence.
The better question is functional and developmental: what capacities are being redistributed between person and tool, which of those distributions increase meaningful agency, and which create fragility when conditions change?
A Vygotskian lens also asks whether mediation becomes progressively appropriable. Can the learner understand the reasoning, choose when support is needed, detect when the tool is wrong, and eventually perform critical parts of the task without opaque dependence? Or does the tool remain a black box whose contribution cannot be inspected?
This is not a question Vygotsky could have answered about generative AI. It is a contemporary application of his insistence that development must be studied in the relation among person, cultural tool, social activity, and emerging capacity.
Culture changes the form of development
A final boundary concerns universality. If psychological tools are culturally organized, then developmental research cannot assume that one form of participation, dialogue, schooling, or independence represents the endpoint for everyone. The same task can recruit different histories of language, authority, collaboration, and prior practice.
This does not make comparison impossible. It makes context part of the explanation. Researchers should report language, educational setting, task familiarity, forms of assistance, and cultural conditions that could shape mediated performance. For a Northern European institute working across populations, that discipline matters more than attaching a Nordic example after the fact.
Vygotsky's value here is methodological humility: the developing person is never encountered outside a history of tools and relations.
Mediation changes what independence means
Vygotsky's framework complicates the modern ideal of the independent learner. Independence is often imagined as what remains when support is removed. A more developmental view asks what forms of support have been incorporated well enough that a person can now act with greater range, flexibility, and self-direction. The social does not disappear when competence becomes individual. It has helped organize the means through which the individual can proceed.
This distinction matters around AI. A tool can function as scaffold, collaborator, shortcut, authority, or substitute, sometimes within the same task. The relevant question is not simply whether assistance was used. It is whether the relation expands what the learner can subsequently understand and do, or whether performance stays coupled to the support. That is an empirical question, not a slogan about technology. It requires delayed assessment, transfer tasks, observation of strategy, and attention to who controls the interaction. A Vygotskian lens does not tell us that AI is educationally good or bad. It asks what developmental relation the tool has actually entered.
What the evidence does not show
Vygotsky's influence does not show that every form of collaborative learning is effective. It does not show that independent practice is unimportant, or that any adult guidance automatically develops capacity.
The zone of proximal development is not a license to infer hidden potential without careful operationalization. Nor is it a precise universal distance that can be measured the same way across tasks and cultures.
The historical texts also cannot settle current empirical debates in developmental psychology. Vygotsky worked in a particular intellectual and political context, wrote across rapidly developing ideas, and died at thirty-seven. Much of his work reached English readers through edited and translated collections. Interpretation therefore requires textual and historical care.
The evidence also does not support treating "social mediation" as an explanation for every mental process. Biology, individual differences, material conditions, and direct interaction with the physical world matter. A cultural-historical lens is powerful precisely when its scope remains visible.
The human beside the tool
The AI age may make Vygotsky newly relevant for a simple reason: more cognition is becoming visibly mediated.
People think with search engines, recommendation systems, shared documents, navigation tools, messaging platforms, and generative models. The old fantasy of an isolated individual mind was already weak. It is becoming harder to sustain.
The question is not whether tools belong inside cognition. They always have.
The developmental question is what kinds of people different tools help create.
A good prompt may solve today's problem. A good developmental environment changes what the person can do when the prompt is gone.
The child at the table eventually reaches for the next piece without help. The gesture offered earlier has disappeared. Something of the relation remains, reorganized as capability.
That is the possibility Vygotsky keeps asking us to see: what begins between people can become part of what a person is able to carry within.
Research references
Vygotsky, L. S. (1978). Mind in Society: Development of Higher Psychological Processes. Harvard University Press. https://doi.org/10.2307/j.ctvjf9vz4
Vygotsky, L. S. (1934/1987). Thinking and Speech. English translation archive. https://www.marxists.org/archive/vygotsky/works/words/
Hedegaard, M. (1990). The zone of proximal development as basis for instruction. In Vygotsky and Education. https://doi.org/10.1017/CBO9781139173674.017
Valsiner, J., & van der Veer, R. (2014). Encountering the border: Vygotsky's zona blizhaishego razvitya and its implications for theories of development. https://doi.org/10.1017/CBO9781139028097.009
Chaiklin, S. (2003). The zone of proximal development in Vygotsky's analysis of learning and instruction. https://doi.org/10.1017/CBO9780511840975.004
Junefelt, K. (1990). The Zone of Proximal Development and Communicative Development. Nordic Journal of Linguistics, 13(2), 135-148. https://doi.org/10.1017/S0332586500002201