In a rural clinic in Uganda, Manu Prakash once saw an expensive centrifuge being used as a doorstop. The machine itself was not broken. The problem was simpler and more fundamental: without reliable electricity, it could not do the thing for which it had been designed.

A centrifuge is, at heart, a machine for spinning. In laboratories it spins samples fast enough to separate substances of different densities, including the components of blood. Modern versions can be expensive, heavy and dependent on infrastructure. Prakash and his colleagues began with a different question. If spinning was the essential function, how much of the conventional machine was actually required?

Their answer came partly from a toy. A whirligig, or buzzer, can be made from a disc threaded onto string. Pull the string and the disc spins one way, winding the string around itself; release and pull again and it reverses, repeating the cycle at extraordinary speed. Versions of the toy are ancient. The researchers cite examples dating to around 3300 BCE.

Saad Bhamla, Brandon Benson, Chew Chai, Georgios Katsikis, Aanchal Johri and Prakash tested yo-yos, tops and whirligigs, then modelled the mechanics of the string and disc. They borrowed equations used to describe the supercoiling of DNA. The resulting Paperfuge weighed about two grams, cost roughly twenty US cents in materials and could reach 125,000 revolutions per minute, generating forces of about 30,000 g. It could separate plasma from whole blood in less than a minute and a half and isolate malaria parasites in about fifteen minutes, without electricity.

The mechanism had been available for thousands of years. What changed was what somebody knew enough to ask of it.

This is one of the stranger features of technological progress. Advancement can mean adding capability, but it can also mean discovering what can be removed. The useful question is not always, How do we reproduce the machine? It can be, What is the machine actually doing? Once the function is separated from the familiar object that performs it, other materials and mechanisms become visible.

Prakash's lab applied a related principle to microscopy. The Foldscope, developed by James Cybulski, James Clements and Prakash, folds from paper, weighs less than nine grams and in its original form cost less than a dollar in parts. It can provide more than 2,000 times magnification with submicron resolution. Paper, normally the thing on which scientific knowledge is printed, became part of the instrument itself.

Work of this kind is often described as frugal innovation: finding ways to produce useful performance with fewer resources. Another useful term is bricolage, the practice of making do by recombining resources already at hand for new problems. Neither idea means that scarcity is desirable. Constraint can provoke invention, but it can just as readily prevent it. A missing medicine remains a missing medicine. A hospital without electricity is not secretly fortunate. Poverty, war and collapsed infrastructure kill people and waste talent on a scale that clever improvisation cannot redeem.

That distinction becomes unavoidable in Gaza.

When Tala Mousa's family evacuated their home before it was bombed, she filled a bag with whatever she could reach. The last thing she put into it was her technology textbook. Afterwards, the family was living in a tent. The day after the bombing, Tala opened the book and found a lesson about making blocks from mountain stone. Around her was the material left by destroyed buildings.

She asked whether the mountain stone could be replaced with rubble.

Tala discussed the idea with her younger sister, Farah. They took it to their mother, Samar, an industrial engineer, and their grandfather also helped with the practical work of developing prototypes. The sisters eventually created Build Hope Palestine, a method for turning selected rubble into non-load-bearing blocks. Safe debris is crushed and sieved, mixed with locally available binders such as clay, ash or powdered glass, moulded and dried. The blocks are intended for uses such as partitions, pavements and garden beds rather than structural walls. In 2026, when Tala was seventeen and Farah fifteen, the project won the Middle East regional award in The Earth Prize.

It would be obscene to suggest that the destruction of their home somehow bestowed a creative opportunity upon them. They should not have needed to invent building materials from the remains of bombed buildings. Their achievement belongs beside the conditions that made it necessary, not in place of them.

What survived the destruction was not merely a book. Tala still possessed something the bombing had not removed: knowledge she could recognise as transferable. A lesson about one material became a question about another. Her mother could contribute engineering knowledge. Her grandfather could help turn an idea into something testable. The sisters could experiment, fail, alter the mixture and try again. Physical resources had been devastated; intellectual and practical resources had not vanished with them.

This is where bricolage becomes more interesting than the slogan of 'making something from nothing'. Nobody makes something from nothing. People make things from what remains: materials, memory, observation, skill, relationships, old techniques, half-understood mechanisms, books carried out of houses, questions that have not yet been exhausted.

Ingenuity depends on more than scarcity. It requires an ability to distinguish between the part of a problem that is genuinely load-bearing and the part that merely arrived with the conventional solution. The centrifuge needs rotation. It does not inherently need an electric motor. A microscope needs controlled optics. It does not inherently need a heavy metal body. A building block needs appropriate material properties. It does not inherently need mountain stone.

Psychologist Robert Sternberg has argued that creativity involves, among other things, redefining problems and questioning assumptions embedded in the way they have been presented. That sounds abstract until you watch a child play. A cardboard box becomes a house, then a boat, then a counter in a shop. A stick acquires half a dozen functions in an afternoon. Research on pretend play and creativity describes precisely this capacity for substitution and recombination, although the evidence is more complicated than the comforting claim that play simply causes creativity.

When I was about six, a friend gave me a handmade wire cyclist that had come from her grandfather's shop. It was the kind pushed along from behind with a long wire handle, although mine was more detailed than many of the examples I have since found, with a male cyclist whose legs pedalled as the bicycle moved. What fascinated me was the mechanism. I remember examining it, trying to work out what connected to what and how pushing the toy could make the wheels turn and the cyclist's legs move. I wanted to know who had made it, and how they had known how to make all those separate movements work together.

Handmade push-wire vehicles have been documented across southern Africa, and related moving wire toys appear in collections from elsewhere in sub-Saharan Africa. A Smithsonian push-toy bicycle from the Free State, dated 1984–1987, is made from wire, plastic bag and paint. Another Smithsonian bicycle with rider, collected in the Democratic Republic of the Congo in 1988, uses bent wire, cloth, rubber around the wheels and a long pushing stick. Object collections and accounts from Zimbabwe, Botswana and South Africa describe children and makers using wire, rubber, plastic, wood and other found materials to build vehicles with moving parts.

Calling such objects evidence of creativity born from poverty is too easy. Poverty explains why bought materials may be unavailable; it does not explain the mechanical intelligence required to make a wheel drive a figure's legs. The maker still has to observe motion, understand relationships between parts, manipulate unforgiving material and keep adjusting the arrangement until movement travels through it correctly.

As a child I did not have the vocabulary of transmission, linkage or mechanical advantage. I could still follow a causal chain with my hands and eyes: push this, that turns; when that turns, this moves. The toy made the world inspectable.

Psychologists have a useful language for part of this. Viktor Gecas's review of self-efficacy places it within a broader literature on agency, mastery and control: the experience of oneself as capable of producing effects. He also discusses Robert White's idea of effectance motivation, an intrinsic drive to interact with the environment and discover that one's actions can make something happen. Play, tinkering and experimentation can provide repeated encounters with that fact.

Age produces an odd reversal. Children possess little formal power. Adults can vote, organise, study, earn, buy, build, contact institutions, join campaigns and acquire specialised knowledge. Yet perceived agency can contract as actual agency expands. Experience teaches us where the obstacles are. Much of that learning is valuable. Some constraints really are immovable from the position in which we stand. Other conclusions may be sediment left by disappointment: attempts that failed, authorities that ignored us, systems that did not respond, expertise that told us how things are done.

Albert Bandura's distinction between efficacy expectations and outcome expectations helps here. I cannot do this is different from even if I do this, nothing will change. The first concerns belief in one's own capacity. The second concerns the responsiveness of the world around us. Confusing them can be cruel. A person confronting an unresponsive institution does not necessarily need more confidence; the institution may need to change.

That is why hope, in this sense, cannot be an accusation. It cannot mean that every problem yields to ingenuity, or that people who remain trapped simply failed to imagine hard enough. Some systems are designed to resist the people most affected by them. Some shortages cannot be improvised away. Some losses cannot be converted into lessons.

But constraints and assumptions are not the same thing, and they often arrive disguised as one another. One marks an actual boundary. The other marks the point at which somebody previously stopped asking questions.

Knowledge matters partly because it helps us tell them apart. The Paperfuge did not emerge from ignorance of centrifuges but from understanding centrifugation deeply enough to strip away much of the machine. Tala's question did not come from knowing nothing about building blocks. It came from carrying a lesson about them into a landscape the lesson had never anticipated. The wire cyclist did not move because wire is magical. Somebody understood enough about movement to persuade one motion to produce another.

The whirligig returns us to an object so simple that it can be mistaken for trivial: a disc, a string, two hands pulling in rhythm. For millennia it could remain a toy. Then someone looked closely enough at what it was doing to see a laboratory instrument hiding inside the motion.

Some problems have boundaries that are brutally real; others contain assumptions we have stopped noticing. Wisdom may consist partly in learning the difference, although we will often get it wrong. What seems worth preserving is the impulse that turns an object over, watches what connects to what, and asks another question before deciding that the arrangement in front of us is the only one possible.

The more useful question may not be whether a problem can be solved, but whether we are certain we have found the part that cannot be changed.

Sources and further reading