Cassava processing

Why would a community put its staple root through days of soaking and drying before anyone eats it?

A household short of food takes its cassava roots and, instead of cooking them, grates them, packs them into a sack, and waits three days. Then presses them. Then roasts the pulp dry over a fire, spending fuel and labour on a crop already harvested. The labour falls on people who have least to spare, and boiling is available and far cheaper. So the sequence must do something boiling cannot — and if we can work out what, we can predict which steps are negotiable.

Start with what is in the root. Bitter cassava contains cyanogenic glucosides, chiefly linamarin — not itself the poison, but a sugar-capped, water-soluble, fairly stable molecule. The root also contains an enzyme, linamarase, which strips that cap.

Why does the plant not poison itself? The two are kept apart. The glucoside sits in the cell vacuole; the enzyme sits outside it, in the cell wall region. They meet only when tissue is broken — by an insect's mandibles, a rodent's teeth, a spade. This two-compartment arrangement recurs across plant defence; the mustard family runs the same trick with glucosinolates and myrosinase.

So the root is a loaded device with the trigger held separate, and damage is the trigger. What follows when it fires? Linamarase splits linamarin into glucose and acetone cyanohydrin, and it matters enormously that this intermediate is conditionally unstable: it falls apart into acetone and hydrogen cyanide readily above roughly pH five and at warm temperatures, and is comparatively persistent when cold and acidic. Hydrogen cyanide boils at about twenty-six degrees Celsius — a gas at the temperature of a tropical afternoon.

Now derive the processing rather than describe it. The goal cannot be to destroy linamarin directly; nothing in a village kitchen destroys it. But the root carries its own demolition enzyme. So the strategy that suggests itself is: fire the plant's own reaction deliberately, drive it to completion, and let the volatile product leave.

Every step now has a reason. Grating maximises cell rupture, so enzyme meets substrate everywhere rather than at a cut face. Heaping the wet mash for a day or three gives hydrolysis time to finish and the cyanohydrin time to decompose, at ambient temperature in a moist medium — exactly what the chemistry wants. Pressing removes the liquid, and both the glucoside and the cyanide products are water-soluble, so the drained juice carries a share off. Roasting drives residual hydrogen cyanide off as vapour and dries the product for storage. Soaking whole roots for days — the route to fufu — reaches the same place differently: microbial softening ruptures cells from within, the endogenous enzyme acts, and the water carries the products away.

And now the sharp prediction that makes this an explanation rather than a story. Boiling should be worse than waiting, not merely less good. Heat denatures linamarase: cook the root quickly and you destroy the enzyme before it has worked, leaving intact linamarin in the food — hydrolysed later by gut microbes, releasing cyanide inside the person rather than over the fire. The slow, cold, wet route is not a primitive substitute for cooking; it is the only route that uses the enzyme instead of killing it.

That prediction holds: methods that heat fast leave far more bound cyanogen than methods that grate and hold wet, at equal total time. The wetting method developed for at-risk communities exploits exactly this — flour is re-wetted and spread thinly in the shade for a couple of hours before cooking, so residual enzyme acts and the cyanide evaporates (recalled).

The obvious alternative readings each fail, and specifically.

"They soak it to soften it for pounding." Softening happens and is welcome. But sweet, low-cyanogen varieties are eaten boiled without days of soaking, and the elaborateness of processing tracks the bitterness of the variety rather than the hardness of the root. A practice about texture would not be graded to toxicity.

"Fermentation is for flavour and keeping." True as a co-benefit — gari's sourness is prized and the dry product stores. Same objection: it scales with variety, not with taste.

"It is tradition without a reason." Refuted by a grim natural experiment. Where processing is truncated — in droughts, in wartime displacement, when fuel runs out and roots are sun-dried in a hurry instead of soaked — outbreaks of konzo follow, an abrupt and permanent spastic paralysis of the legs. Practice and disease are time-locked, about as close to a controlled test as an ethnographic question gets.

And the question to ask of the whole account: if bitter cassava is this much trouble, why grow it? Because the cyanogens are a defence and the household is borrowing it. Bitter varieties resist insects, rodents and casual theft, tolerate poor soils, and can be left in the ground as a living store. The labour is not the price of a bad crop but of a well-defended one, paid deliberately.

Where I will not pretend to precision: I decline to give a threshold intake at which cyanogenic cassava becomes dangerous, because no clean one exists. The body detoxifies cyanide to thiocyanate using sulphur from the amino acids cysteine and methionine, so the same dose is far more harmful to someone whose diet is almost entirely cassava and nearly free of protein — precisely the situation in the famines when konzo appears. Dose and nutritional status interact, and a single number would misrepresent that. Regulatory limits for cyanogen in flour are administrative ceilings, not physiological thresholds. The pH and temperature figures above are recalled approximations; the boiling point of hydrogen cyanide is the one hard constant here.

• How the wetting method was trialled in konzo-affected villages, and what made it adoptable where earlier advice was not.

Every term the collection defines is gathered in the glossary.