Greenness - underripe aromas and flavours - has long been recognised as a problem in South African red wines. Especially in the Bordeaux varieties (cabernets franc and sauvignon, and merlot). And especially for non-South Africans: it seems we locals have to some extent got used to it (though I remember Michael Fridjhon, for example, in the 1990s yammering on about how green our wines were and how we had to learn from the Australians, etc, etc).

As far as local wine critics are concerned, criticism of greenness has mostly been directed at merlot. For me, the most awful manifestations are usually in merlot, especially when that minty greenness is combined with overripe sickliness.

But the excellent tasting/workshop generously put on for a few dozen local wine judges by Charles Hopkins at De Grendel estate in Durbanville focused on cabernet. Charles's interest in the subject has been growing since the time when, at  some international get-together years back, the foreigners could all identify the Cape wine because of the greenness (if it's not burnt rubber, it's greenness, it seems), and Charles seems to be particularly sensitive to the characteristics - or is it that he's become sensitised?

Anyway, Charles has got hold of an objective measure for greenness - analysing for something called 2-methoxy-3-isobutylpyrazine. Happily this can, and most certainly should, be shortened to IBMP. It is the substance that, widely called pyrazines, gives the green characters to sauvignon blanc that so many people seem to enjoy (not me) and to cabernet sauvignon.

I can't report here in full on this tasting, or Charles's presentation. His essential objective point was, I think, that cooler areas (the sides of the Cape mountains subject to those famous cooling sea-breezes, etc) were going to deliver higher concentrations of IBMP. His essential subjective claim was (in line with at least some international opinions he referred to) that any level over 10 nanograms (ng) of IBMP in a wine was iffy; anything over 30ng was effectively a wine fault.

So, (quickly "trained" on a pair of sauvignon blancs) we tasted through some samples of cab or Bordeaux-style blends, mostly local from diverse origins. The tasters present were remarkably all over the place in identifying which in a line-up of six wines were the most "green". There was also no consensus on which were the preferred wines, though there were clusters of preferences.

One of those clusters (which included me) was for the Hillcrest Hornfels 2008, which apparently has an IBMP count of 41ng. Another seriously faulty wine (if you accept Charles's idea of faultiness here) was Meerlust Rubicon 2007, which was even "worse" at 42ng.

Now, this latter wine at least has been pretty well received overseas as well as here, so it's not just a question of us having seriously problematical palates.

What conclusions can we draw? That Meerlust Rubicon is a seriously faulty wine and we must learn to hate it? Or that, as with most "numbers" and factors in wine, balance is everything. Defining many faults is a problem. You could decide, I suppose, that an alcohol level over 14.5% in a cabernet is a fault just as much as an IBMP level over 30ng. But then you will taste a whole range of wines, and those aspects can be revealed as, indeed, faults in some wines, but not in others.

And I can't accept that "greenness" (which I think is not such a simple, unique thing as Charles suggests) is a fault in the same way the brettanomyces is. I'm not hostile to a bit of brett myself, but I can see that the tendency of brett is to obscure varietal and terroir characteristics. That is not true of something like greenness, it seems to me.

The reason I preferred the Hornfels (and liked the Rubicon too) was because they were more elegant and, in my opinion, better balanced than, say, the lush Rust en Vrede Cabernet 2008. We are, after all, tasting wines and not numbers.

The R&V was, incidentally, the wine with the second-lowest level of IBMP, at 10ng. The lowest was a smart Bordeaux, the Pichon Longueville Baron 2008 - a wine much admired by the overseas Bordeaux lovers on record and, I'd like to observe, probably always tasted in extreme youth in the context where the judges knew it was a top-level Bordeaux. To me, this wine showed all that is wrong in much modern Bordeaux, being, basically, too ripe (like the Rust en Vrede) and lacking elegance.

Things come in packages, you see, not as distinct, discrete accumulations. Greenness might well tend to come with elegance, not-greenness might tend to come with lush, hollow ripeness.

But that's also too simple, perhaps. This tasting was a very useful experience - not, in my opinion, because it answered anything definitively, but because it reminded us of some important questions, and gave us a useful tool with which to help us understand, and hopefully hone and improve, our own responses.