I've read thousands of tasting notes over the years and I'd estimate that fewer than one in twenty told me anything I could act on. The rest were lists of fruits, minerals and household objects arranged in a confident order, describing a wine I couldn't picture from the description.
This isn't a complaint about pretension exactly, though there's plenty of that. It's that the standard format of a tasting note is answering a different question to the one a reader has.
What the notes are actually for
The descriptive tradition comes from professional tasting, where the purpose is to distinguish one wine from another and to record impressions for later comparison. In that context, "crushed raspberry and dried thyme" is genuinely useful — it's a memory aid and a fingerprint.
The problem is that this format got exported to consumer writing, where the reader's question is completely different. They want to know: will I like this, does it go with dinner, and is it worth the money.
A list of aromas doesn't answer any of those. And worse, it implies a level of precision that the human sense of smell doesn't support, which puts people off wine entirely because they can't find the violets.
The uncomfortable research
There's a body of work on the reliability of tasting that's worth knowing about, and it's not flattering.
Studies have repeatedly found that even experienced tasters show limited consistency when the same wine is presented multiple times in a blind setting. Judges at wine competitions have been found to score the identical wine differently across a single session. Descriptor vocabulary varies enormously between tasters describing the same liquid.
There's also the well-known finding that adding odourless red colouring to a white wine causes trained tasters to describe it using red-wine vocabulary. That's not evidence that tasters are frauds — it's evidence that smell perception is heavily influenced by expectation, which is true of everybody and always has been.
The honest conclusion is that specific aroma descriptors are much less reliable than the confidence of their delivery suggests.
What is reliable
Here's the useful part. Some aspects of a wine are perceived consistently across tasters, and they happen to be the ones that predict whether you'll enjoy it.
Sweetness. Straightforwardly detectable and hugely important to whether someone likes a wine. Yet it's often omitted from notes entirely, or buried in jargon.
Acidity. The sensation that makes your mouth water. High acid wines feel fresh and food-friendly; low acid wines feel soft and can seem flabby. Easy to perceive, easy to describe, rarely mentioned in plain language.
Tannin. The drying, grippy sensation on your gums, mostly in reds. This is what people mean when they say a wine is "harsh," and it's the single biggest reason someone dislikes a red wine.
Body. How heavy the wine feels in the mouth. Compare skimmed milk to cream. Correlates loosely with alcohol.
Alcohol. On the label, and it tells you a lot about the style and the warmth of the finish.
Oak. Whether the wine tastes of vanilla, toast, coconut, or smoke. People have strong preferences about this and it's binary enough to be useful.
A format that works
What I write now, for anything I'd recommend, looks roughly like this:
"Dry, high acid, light body, no oak. Tastes broadly citrusy. It's sharp — if you find Sauvignon Blanc too aggressive, you'll find this worse. Excellent with anything fried or fatty. Twelve pounds, and it's a good twelve pounds."
That tells a reader everything they need. It gives the structural facts, the general flavour direction without pretending to a precision I don't have, a warning about who won't like it, a food context, and a value judgement.
Compare that to "aromas of gooseberry, cut grass and elderflower, with a saline finish and a whisper of struck flint," which is more evocative and much less actionable.
On the fruit descriptors
I'm not saying abandon them entirely. Broad categories are genuinely useful and reasonably consistent between people.
For reds: is it red fruit (raspberry, cherry, cranberry — usually lighter, higher acid) or black fruit (blackberry, plum, cassis — usually richer, riper)? That distinction is real, it correlates with grape variety and climate, and most people can perceive it.
For whites: citrus, orchard fruit (apple, pear), stone fruit (peach, apricot), or tropical. Again, a real spectrum that tracks ripeness and climate.
What I'd avoid is the specific single descriptor. "Morello cherry" versus "Bing cherry" is a distinction almost nobody can make reliably, and claiming it damages the credibility of everything else you say.
The confidence problem
The deeper issue with elaborate tasting notes is what they do to people who are new to wine.
Someone opens a bottle, reads a note about tobacco leaf and forest floor, smells wine, and concludes they have a defective nose. They don't. They have a normal nose, and the note was written by someone who had the bottle in front of them for twenty minutes with a specific vocabulary in mind and a professional incentive to say something distinctive.
I'd much rather someone drank a wine, decided they liked it, and knew that it was dry with high acidity — because that's a fact they can take to a shop and use to find another one they'll like. That's the entire practical purpose of describing wine, and the flowery version actively obstructs it.
What I'd suggest trying
Next time you open something, write four things down before you look at any note. Is it sweet or dry. Does it make your mouth water a lot or a little. Does it feel light or heavy. Do your gums feel dry afterwards.
Do that for a dozen bottles and you'll have a genuinely useful map of your own preferences, which is more than most people get from a decade of reading tasting notes. And you'll never have to find the violets.