It’s happened again. There I was, leaning back, my feet on the fender, toasting my toes in the firelight of comfortable certainty – then along comes a wine which singes my socks.

DeMorgenzon, Stellenbosch. Andrew Jefford ©

I’d been handed a glassful of pale, fine-beaded, softly seething sparkling wine with lacy, intricate scents, and a fresh, poised palate with a little of the elusive ‘mineral’ wash one looks for in a Blanc de Blancs. Champagne for sure, I felt; ten or twelve years old, from a house which avoids oxidative handling. Pol Roger? Perhaps even Roederer? Comprehensively wrong. It was a well-stored bottle of the 1992 Graham Beck Blanc de Blancs, grown and vinified in South Africa’s Robertson. Robertson does have the limestone soils generally rare in the wine-growing southern hemisphere, but it’s also considered a warm, inland location. How was it possible to make a wine like that there?

It was a prelude, in fact, to realising that I had misunderstood South Africa more generally. It’s not, as I’d always assumed, a nation with a red-wine vocation which happens to make a lot of white wine, thanks to the varietal legacy of its brandy-drinking past. Instead — like New Zealand, though in an utterly different way — its white-wine aptitude may well outweigh that of its reds. Quite why this should be is a terroir puzzle.

Latitudes are low here: Stellenbosch lies almost exactly one degree north of Adelaide. South Africa has very little land at the same latitude as most of wine-making Victoria and New Zealand’s North Island, and no land at all in an equivalent location to Tasmania and New Zealand’s South Island. Its raw degree-day totals often look frightening (Robertson has 2181 and Stellenbosch itself 1945, which theoretically makes both warmer than the Barossa, and hardly a good white-wine prospect).  

There are many different protocols for calculating degree days, though, and in any case the way weather stations measure temperature is not the same as the way in which vines experience temperature. Trying to understand a growing season’s ripening inputs, and thus calibrate the climatic aspects of all of the world’s great wine terroirs on a single scale, seems to me to be the biggest challenge facing wine science. It’s our version of establishing the Hubble Constant. We’re not there yet.

The buffering of this giant tongue of land at whose tip the Indian Ocean and its Agulhas current mingles and churns with the Atlantic’s Benguela current is considerable. Anyone who has ever visited the Cape, where bookcase mountains draped with silvery wigs of cloud often punctuate the horizon, will realize that the potential for positioning vineyards on cool slopes, and courting substantial diurnal temperature differences, is almost limitless.  

Rainfall levels vary widely. There are growers here who manage to dry farm (using bush vines at low planting densities) on less than 400 mm a year. At the other extreme, Boekenhoutskloof, at the closed end of the Franschhoek valley and just 30 km or so from Stellenbosch, is drenched in 2,300 mm of rain a year.  

I suspect, too, that there is no southern-hemisphere nation whose vineyards are more wind-exposed (often positively, but sometimes catastrophically) than South Africa. Robertson, for example, has a train of freshening south-easterly wind that arrives with Swiss punctuality at 2.30 every afternoon, as the lopsided avenue of jacarandas next to the De Westhof winery testifies. The showcase estate of Vergelegen in Stellenbosch had to be comprehensively replanted following wind damage, and it’s commonplace for winemakers in many regions here to tell you that the wind sees to yield control, making green harvesting unnecessary.  

South African light, too, has its own quality. Winemaker Gerhard Smith of La Vierge in the Hemel-en-Aaarde valley spent seven years working in New Zealand’s Martinborough. “If I’d spent as much time working in the vineyard in New Zealand as I do here, I’d have been burnt to a crisp,” he told me. Yet Martinborough is cooler than Hemel-en-Aarde, and one would normally expect UV levels to be significantly higher in these South African latitudes, lying as they do almost seven degrees closer to the equator. Once again, something isn’t working quite as you’d expect.

The vine, remember, is a scholar among plants: it will notice all of this. It then writes a report every year — in its grapes. Restrained and sensitive winemaking enables us to sniff and to sip the report.

There are many welcome new plantings in South Africa of Roussanne, Marsanne, Grenache Blanc, Viognier and other Mediterranean whites, and complex white blends are a delicious feature of the work of South Africa’s avant-garde, about which more in due course. The four main white varieties here, though, are Chenin Blanc (around 17 per cent of South Africa’s vineyards), Colombard (11 per cent), Sauvignon Blanc (7 per cent) and Chardonnay (7 per cent). Which is the most successful? Which tells the terroir story best?

Colombard can’t do it, and no one asks it to. The country has some great Chenins, and the variety often furnishes the backbone of those singular blends, yet rightly or wrongly the estates and producers choosing to commit resources to make their ‘top white’ with Chenin are still a minority. Sauvignon performs outstandingly well here, but varietal character is almost always centre stage, making it hard to detect the stones and the soils and the slopes behind — a problem (or a privilege: most Sauvignon Blanc consumers love its varietal style) New Zealand’s Sauvignon growers are also amply familiar with.  

That leaves Chardonnay – and South Africa’s Chardonnays now seem to me to be some of the most enjoyable, the most drinkable yet also the most refined outside Europe. I’ll describe them in more detail next week.

Written by Andrew Jefford