Think about the experience of drinking good wine. You stick your nose in the glass. The aroma is elusive, enticing. You taste, remembering to swirl the wine so it touches your whole palate. Again there is that vague familiarity, but the flavours are hard to pin down.

If the wine is good, it will have you reaching for the bottle to see what the winemaker says you can smell and taste; you may even reach for a reference work to find out what makes this wine so delicious.

While wine labels seem to be ever more informative, not all wine writing is as helpful in answering this question. In some cases, the senses-swapping condition synasthesia appears to be a prerequisite for understanding - the writer if not the wine. Take this example from the 1968 catalogue of Gerald Asher: "Chateau Lynch-Bages, Grand Cru Classe Pauillac, Chateau-bottled. Just the wine for those who like the smell of Verdi. Dark colour, swashbuckling bouquet and ripe flavour."

It seems unlikely that any of us will know what Verdi or a swashbuckler smells like. This example may be more than 40 years old, but you don't have to look too hard even now to find writers who compare drinking a wine to a Mozart concerto or an epic opera.

The antidote to this irrational exuberance is a style sometimes dismissed as "nature notes" by its opponents. This is the more commonly found description of wines that refer to berries and spices and other familiar flavours - at least if you have grown up in the West. Even here a little obscurity can reign. The description of a wine as having a "cigar box" aroma assumes that an interest in drinking must mean an interest in smoking.

Everyday descriptions don't always work though. Forest fruits, for example, are wide-ranging and which ones are present depend on which forest you mean. Still, we can't hold wine writers accountable for what their readers may or may not know. If you are ready to make the effort, an aroma sampler kit may help in understanding the wine writers' references.

So where can we turn for advice that acknowledges the drink may have inspired poets, but keeps its feet long enough on the ground to remember that this is an agricultural product? Long-term readers of the may well recall the type of pithy advice often dished out by Kevin Sinclair. Here's a tip from 1994 on buying a Brown Brothers wine he thought would be ready to drink by 1997: "Around mid-June of that year, get a large leg of lamb, crust it with herbs, stick it to roast in a slow oven and open a bottle of this deep, ruby-hued wine. You'll be having so much enjoyment you won't worry about the change of sovereignty."

If you are looking for more depth, it seems likely that at some point you will be reading something by Robinson, author of the much-praised , which covers the entire spectrum of wine knowledge. For Robinson there are "two great variables" to understand in wine: terroir and grape.

She has tackled terroir in updates of Hugh Johnson's , first published in 1971, since updated five times. Now, with co-authors Julia Harding and Dr José Vouillamoz, she tackles grapes in a beautifully illustrated work likely to be the definitive word for many years to come.

But what about winemakers - surely the third "great variable" in the equation? Although the book is designed to help the wine industry, she allows little space for how their "philosophies" might influence the product. Perhaps she doesn't take those philosophies too seriously.

However, as a reference work partly intended for an industry audience, much of the information is technical. Patience is its own reward though - a useful motto for the reader of this book, which spins apparent trivia into insights, some more obviously useful than others.

If the intensely flavoured Croatian ranac bijeli grape is almost extinct, then it needs a chronicler - and who better than one who can tell us the grape's viticultural characteristics in one sentence and slip in a joke? "Early ripening and high sugar level in the berries, therefore at risk from wasps and children."

You may not have been losing sleep wondering what the currant grape is called (korinthiaki) or which variety is Armenia's most planted (mskhali), but if your inner wine geek is secretly thrilled by finding out, then you will find this book worth exploring. Who can resist the description of the magyarfrankos grape as a "dark-skinned escapee from experimental plantings during Hungary's communist era"?

A book intended as a complete guide to 1,368 vine varieties obviously also includes the big names - complete with the facts that make the 1,200-page book comprehensive. Why, one wonders, does Robinson tell us that two of the "varieties commonly mistaken for syrah" are the serena e zeza and shesh i zi from Albania? This isn't a sense of the word "common" most of us will be familiar with.

Six pages of close forensic analysis later and Robinson's dry wit and learning have taken us through DNA charts showing that the grape is a half-sibling of the floral viognier and a great-grandchild of pinot. There are references to Pliny the Elder, Agathocles the tyrant of Syracuse, and the Roman emperor Marcus Aurelius Probus.

Finally she demolishes the claims of those erudite Albanian farmers who thought their varieties shared a common heritage with syrah "simply because its name is similar to Serene, the old name for syrah in the Cote Rotie". And so Robinson and her co-authors firmly establish the real botanical and historical identity of the grape.

The more romantically inclined may regret that in the process she has also demolished the idea that the grape was brought to Europe from Shiraz in Persia by crusaders, incidentally also disproving a cherished Australian idea. "Since Shiraz is the Australian name for Syrah, some authors even argued that the Australians had maintained the original name while the French had Frenchified it. However, the Shiraz hypothesis is doubtful, not least because the Crusades were mainly focused on the Holy Land and did not go as far as Persia."

You could skip all this and go straight to the tasting notes - a section labelled "Where it's grown and what its wine tastes like" - but you would miss out on some lean, elegant and informative writing.

The notes for each region that the wine grows in are equally sparse but useful. Want to know the difference between French and Spanish syrah styles? Spanish "wines tend to be notably sweeter and plumper".

While the closely argued DNA evidence, with colourful charts, are of more use to the grower than the consumer - even one buying fine wines - Robinson's writing holds interest. There are also some beautiful paintings of bunches of individual grape types but no index.

I'm left with one conundrum. The length of time that a wine flavour persists on the palate is measured in caudalies, equivalent to seconds. Why is it that a pleasure that only lasts seconds is worth spending years investigating?

Yet somehow it is.