Henry Wheeler Shaw once observed "The trouble with people is not that they don't know, but that they know so much that ain't so." This quotation appears in Malcolm Gluck's 'Brave New World,' a book about the wines of Australia, California, New Zealand and South Africa. It is obviously intended to prepare the reader for a claim Gluck makes - based on a paper presented at the Royal Economic Society's annual conference in 2005. It is not terroir which makes a wine, Gluck tells us, it is human intervention.

I suspect that some of the difficulty in pursuing this discussion is linguistic. What 'makes' a baby - copulation under propitious circumstances or its DNA? Many generations of French producers believe (to serve their own agendas, says Gluck) that place (the DNA) is the crucial determinant. This position effectively argues that you can minimise (though not eliminate) the human role just as you can replace copulation by in vitro fertilisation.

South Africa's wine of origin legislation rests entirely on the importance of terroir in finger-printing a wine. If Gluck is correct, wine certification for place will have to be abandoned. Even in a country where purposeless activity has been raised to an art form, there can be no justification in employing officials to rubber-stamp claims to place if origin itself is irrelevant.

How could the wine have got it so wrong? If we want to make Lafite Rothschild in Stellenbosch, we simply need to transplant Charles Chevalier and his team from Pauillac and voila! our dreams of producing South Africa's first true icon wine would be fulfilled.

Gluck may be right about the so-called New World style. There is a difference between how traditional (ie Old World) growers and winemakers manage their vineyards and vinify their grapes, and what has increasingly become the practice in California and Australia, New Zealand, South Africa, Chile and Argentina.

But there is still a nexus between human intervention and the role of place. Just look at the new range of Spier wines made by Frans Smit under the Creative Block label. The brand draws its inspiration from a project supported by Spier which provides funding for up-and-coming artists in exchange for illustrations on framed blocks about 25cm square.

The wines are unashamedly pure Frans Smit and reveal exactly why he enjoys the critical acclaim of many of his peers. There are two reds. The first, simply labelled 3, is a blend of Shiraz, Mourvedre and Viognier. The second, labelled 5, comprises Merlot, Cabernet Sauvignon, Malbec, Cabernet Franc and Petit Verdot. Both are opulent, with weight, density, richness and lashings of polish. None of this is a coincidence: Smit has sourced grapes from slightly warmer sites, with a preference for Paarl. His intention is to ensure that his wines show developed fruit flavours, plush textures, almost creamy tannins. He has made origin subordinate to style - though place is still inescapably part of the formula. He could not have produced the same wines if he had been compelled to buy his grapes in Elgin and Elim.

A slightly different aesthetic is reflected in the only (at this stage) white wine in the range - unsurprisingly called 2 (since it has two component parts: sauvignon blanc and semillon). Here his focus has been on elegance, on easy drinking, on bright, almost savoury, flavours. The sauvignon comes from Durbanville, and brings a zesty edge to the wine. It is typical of the appellation - and the sense of origin remains because Smit has retained this fruit purity through the vinification - and in decisions which include keeping this component away from oak. The lightly wooded semillon adds creaminess and complexity to the blend. The winemaking is more evident than the vineyard here. Since it is complementary and adds flesh to the somewhat austere mouthfeel of the Durbanville sauvignon blanc, you can hardly complain.

Malcolm Gluck thinks terroir is dead: he's ploughing a lonely furrow. Provenance is certainly evident even in these very winemakerly wines.

From Business Day, 24 February 2010