What matters more than anything else is agriculture

The right support for traditional farming could help Africa more effectively than any amount of "dev

Gordon Brown and Bob Geldof are good people, but it seems as if all the best intentions of those in the highest places are misguided - destined to make things steadily worse in Africa. "Development" is a good idea, but horribly misconstrued. Technology is vital, but only if directed at the carefully identified problems of people at large - which, for the most part, it isn't. Extreme poverty is vile, but wealth per se does not necessarily reduce the vileness. New models are needed: for Africa and for the world as a whole. By being the first to enact such models, Africa could yet emerge as the world leader.

What really matters, far more than anything else, is agriculture. Build on that in its traditional form - the subtle exploiter of landscape and employer of people - and all of Africa could be both stable and enviable. But who, apart from the farmers themselves and a few eccentric academics, even knows what agriculture is?

Traditional farming is perceived as a disaster worldwide, dragging down the farmers themselves and their whole communities. Agriculture, the mantra has it, must "compete" with everything else, from diamonds to tourism, and so must focus on "commodities" to sell on the allegedly free world market. Biotech companies such as Syngenta are lauded for providing modern crops in the form of genetically modified organisms (GMOs) - perceived in high places to be both profitable and necessary. Governments such as Zambia's that have turned them down are seen as backward, if not downright wicked. But, in truth, the ignorance is all on the reformers' side.

Perhaps the most fundamental error, reflected in a thousand treatises and news reports, is that Africans starve because their farming is bad. It is innately unproductive, we are told. One tonne of cereal per hectare is typical for much of Africa, against the 12 or more one would expect in East Anglia. The native crops have not been bred for high yields. They are too vulnerable as well: they die in the field (typically a third or more is lost) and are spoiled in store (another 50 per cent of what is left). Contrast the misery and poverty of African farmers with the wealth of Norfolk. Clearly, Africa's farming must become more like ours . . .

But all the scholars and true aficionados (including farmers) I have spoken to worldwide these past 40 years say the complete opposite. Tropical soils are extremely variable and often poor, and tropical climates are fickle and extreme: traditional farming typically and subtly copes with both. The yields of East Anglia are simply not possible except in exceptional circumstances, and it is better to aim for reliable yields in bad (and typical) years than for highest yields in the best. High inputs are rarely justified, precisely because outputs are innately unreliable. Besides, high outputs are disastrous - they are simply gluts - if markets are not guaranteed. The mixed cropping typical of small traditional farms is highly desirable in countries with unpredictable climates and with so many pests - as with the 600 different conventional varieties of beans that are still grown in Angola. Angola is among the countries that have failed in recent decades, but not because it cannot produce food. It is two and a half times bigger than France, with a population roughly the size of Rio's, and with every kind of climate and landscape. A 30-year civil war and fields full of landmines do most to explain the shortfall.

In contrast, the industrialisation of farming can pay its way only through mass production, and that means monoculture - highly precarious, and potentially disastrous. The notion that countries such as Angola actually need GMOs to provide sufficient yields is simply a misunderstanding, or a straightforward lie. We should object to GMOs not primarily for reasons of health or environment, but because of economics and politics: their introduction suppresses local production and increases the dependency of poor countries on those who supply the new technologies. The argument in favour of GMOs, supported not least by Tony Blair, rests on the assumption that they are necessary. If they are not needed, there is no point in taking any risk at all.

Of course, traditional farms have drawbacks. The work can be intolerable, and far more crops are lost than is necessary. So new technologies are desirable - small machines of the kind that rich people use on their allotments, and also the highest-tech: IT is always useful. Once the technology is geared to the task, all African countries could feed themselves well, several times over. Yet the world is not geared to provide such appropriate (both low and high) technologies. Increasingly, egged on by governments (beginning in Britain and the US), science is increasingly financed by corporates, for corporates. Hence the illusion grows that without industrialisation and corporatisation there can be no science or modern tech at all: that small farms of traditional structure are bound to be backward. Again, this is just not true.

Then there is employment. In the developing world in general, 60 per cent of people work on the land: that's about two billion worldwide. In Africa, it can be even more. In Angola, it is 80 per cent; in Rwanda, 90. We can all agree that 90 per cent is too many - that leaves too few to do everything else. Even 60 per cent is perhaps too much. But it is ludicrous to suggest that the western model - with only 1 per cent in full-time rural work in Britain and the United States - is intrinsically desirable, or could ever be the norm.

Britain is able to employ so few people on the land because there are plenty of other things for Brits to do - if no longer making ships, then at least selling insurance and cutting each other's hair. If Africa ever develops big urban industries, that would be the time to take people off the land en masse. To force people to leave the land before an alternative is put in place seems very much like wickedness.

Then there is the huge and all-pervasive mistake, which says that Africa and the developing world in general could solve their problems if only world trade were truly free. What nonsense. First, third world farmers cannot compete at all unless they industrialise their farms. If Africa industrialised its farms, that would throw most of its people out of work, and it cannot industrialise without foreign investment, which entails foreign control. This means the cash will mostly leave the continent, and the economy will be administered from elsewhere. Finally, western markets are far from bottomless, and are already oversubscribed. Even if the playing field were level (which it never can be) Africans would soon find it too difficult to compete with the US and Europe and the enormous labour forces of Asia and South America.

Of course, exports of high-value crops from bananas to cardamoms have provided good incomes for many countries for many years - but only for some people. Push exports much further than usual, and we will merely create a global dogfight that will benefit western consumers and traders (such as Tesco), but will leave the world's farmers permanently on the brink. Given that farmers worldwide are the biggest single group of workers, it would be worse than can yet be conceived.

The task is not to twiddle with World Trade Organisation rules or even to ease up on debt, but to rethink. In the short term, the prime task for the world as a whole, and in Africa in particular, must be to build on traditional agriculture, which alone can maintain landscapes and provide good jobs for the billions who need them: with appropriate-tech, small-scale financial support, and the general ambition not to trash small farms, but to make agrarian life tolerable, and indeed positively agreeable and desirable. Yet most of what seems to be on the agendas even of the best-intentioned seems to be going in different directions altogether.

Colin Tudge's latest book, So Shall We Reap, is published in paperback by Penguin (£8.99)