From Ori Berkowitz:
I'm able to identify reductive white wines - as a style - by regions specifically. All of the Paardeberg and southern Swartland producers are characteristically reductive. Palladius, Lammershoek, Vondeling, Scali all from these regions display this reductiveness. This begs the question, is there terroir involved?

Tim James responds:
There certainly is something distinctive about these white wines - mostly blends from the slopes of the Paardeberg where so many fantastic wines are being made now (some of them WO Swartland, some WO Voor-Paardeberg). We could add to the list of fine white blends the Chenin-Chardonnay from The Observatory and the Black Rock White, amongst others. But whether it is a reductive character that binds them is not all that likely.

As Eben Sadie (the virtual inventor of this style of wine from the area) says, all these wines are made in a notably oxidative manner, so they shouldn't show 'reductive' characters. (Again - for an approachable discussion of reduction, see Jamie Goode's article). The tendency is for low sulphur use, for ageing in barrel, and for the use of grape varieties that are notably 'oxidative': chardonnay, chenin, grenache blanc, viognier.

All of which amounts to saying that the wines should be characteristically oxidative, rather than reductive. I must say that that is the character I generally get from at least some of them (notably Palladius and The Observatory white, both of which can be a little 'funky'. But I do recall discussing the Observatory wine with another winemaker, who also noted a reductive note as well as the oxidative one....

From Rod Easthope (of Craggy Range winery in New Zealand; Rod has also had plenty of experience in SA as a consultant, and was winemaker at Rustenberg for a few years from 1996; it's very good to know he keeps up with Grape, and to have his immensely knowledgable comments occasionally)

Handling grapes, juice and wine oxidatively does not necessarily prevent reductive characters. Small amounts of these characters may be produced by the yeast during the last stages of ferment when they have exhausted nitrogen based nutrient supply. This would be compounded in the wines you have mentioned – i.e. I presume Eben and the others are minimal interventionists and therefore do not add nitrogen based supplements to their ferments, which can prohibit the formation of reductive characters. Once formed, these reductive notes can be difficult to notice early in the wine's life due to other ferment derived aromas. As these dissipate, the reductive notes become more apparent and can strengthen due to the reductive atmosphere when wines are aged on lees (whether there is SO2 present or not).

Many great white burgundies (and indeed many great white wines from the old world) possess these reductive notes and are variously described as flinty or minerally. I for one, like faint whiffs of these characters - although my technical training says I shouldn't.

So you could say that reductive characters are due to site-derived influences, whereby the site produces ripe grapes but without sufficient assimilable nitrogen. However, I would guess that most sites in the world have this "quality" but it is commonly ameliorated in the cellar by those nasty winemakers. They can also add copper sulphate to take it out prior to bottling. But copper sulphate can also remove other sulphide based aromas that are positive for the wine's character.

A last thought: it is only very good wines that have the inherent complexity and flavour that can make reductive notes a positive part of the wine's bouquet.