There is a familiar contempt we extend towards our ancestors for using lead in water pipes. Roman engineers built it into their aqueducts; Victorian engineers threaded it through cities; and the last of it was only removed from British homes well into the twentieth century, some decades after the neurological damage it caused was understood. We regard this as a failure of knowledge, or perhaps of will. What it was, more precisely, was a failure to include the cost.
The people who specified lead pipe were not unintelligent. Lead was workable, durable, resistant to corrosion. It performed well on every criterion the frame of engineering and commerce made visible. The slow accumulation in the body, the effects on developing minds, the eventual cost of replacement — these were not weighed because the design frame of the time had no mechanism for including them. The harm was diffuse, long-delayed, and fell on people who had no voice in the original decision. So it was excluded. Not with malice. Structurally.
Microplastics are that pattern, repeating now.
Plastic was, and remains, an extraordinary material: cheap, light, mouldable, adaptable. Each individual application made sense within the frame that evaluated it. What that frame could not see — or more precisely, what it had no requirement to see — was the behaviour of the material across its full life cycle, at scale, across decades. Plastics break down but do not disappear. They fragment into particles small enough to cross biological membranes, and they are now found in human blood, lung tissue, and placentas. The home, which we think of as a place of shelter, turns out to be a significant source of exposure — through synthetic textiles, plastic packaging, coated surfaces, and the simple friction of daily life.
The ecologist Richard Thompson at the University of Plymouth coined the term “microplastics” in 2004. The science has been accumulating ever since. What has accumulated more slowly is the governance.
This is not, at root, a scientific failure. The knowledge arrives faster than it once did. What has not kept pace is the capacity to incorporate that knowledge into the design of products before they are made and distributed at scale. The competitive logic of commercial development means that whoever excludes the cost can price lower and move faster. The firm that internalises the long-term harm is at a structural disadvantage against the one that does not. No individual decision-maker chooses harm. The frame selects for exclusion.
The result is a familiar trajectory. The harm accumulates quietly, outside the frame. Then it crosses a threshold — epidemiological evidence, a high-profile study, regulatory attention, litigation — and suddenly it is no longer external. It becomes exposure. The cost that was excluded from the design decision reappears on the risk register, now far harder to address because it is embedded in global supply chains, billions of products, and decades of infrastructure. The lead piping, once valued, must now be dug out at enormous cost. The parallel is not decorative.
Systems narrow to what is measured and rewarded, shedding what falls outside the metric — this is the mechanism described in overoptimisation. The cost of microplastics was never in the metric. It accumulated in the gap between what was measured and what was actually happening. The evidence is not subtle makes a parallel point about larger frame failures: the evidence tends to arrive clearly and early, but the frame in use has no receptor for it.
Microplastics is a relatively modest example. Climate change is the same structure at a greater scale. So are certain construction practices, agricultural run-off, antibiotic residues in water systems, and the long-term effects of particular pharmaceutical compounds. Each was, at the time of its adoption, a local optimisation. Each became a societal-scale problem. In each case, the frame of commercial design had no requirement to look that far forward, or that far outward.
The question this raises is not whether we are more intelligent than the Romans who chose lead. We have progressed considerably in our technical capacity to understand harm. Where we have not progressed is in the governance of design — the institutional capacity to ask, before scale, what this becomes when it is everywhere and permanent. That question sits outside the competitive frame. It falls to regulators, researchers, and advocates who arrive after the fact, working against an embedded system.
The externality that will matter most in twenty years is forming now, in some design decision being made without that question being asked. Identifying it is not a technical problem. It is a frame problem.
Related
- Overoptimisation — the mechanism by which design frames shed unmeasured costs while optimising for the measurable
- Three mechanisms that keep it in place — proxy capture and separated knowledge compound the structural exclusion of long-term harm
- The evidence is not subtle — the same pattern at macro scale: evidence of frame inadequacy arrives clearly, but the frame lacks a receptor for it
- What gets removed does not come back — accumulated harm, like lost capability, does not wait to be addressed