Pairing

Red wine and cheese is one of the most confidently repeated pairings in food, and it is the one place in this wing where I can point at an actual peer-reviewed experiment — and where the experiment turns out to say something more interesting, and more careful, than the headline everybody took from it.

The study people half-quote

Galmarini, Dufau, Loiseau, Visalli and Schlich, “Wine and Cheese: Two Products or One Association? A New Method for Assessing Wine-Cheese Pairing,” Beverages 2018, 4(1), 13 (doi 10.3390/beverages4010013).

Sixty consumers in Dijon, recruited for actually drinking wine and actually eating these cheeses. Three wines — one white (Pouilly-Loché) and two red (Maranges, Beaujolais) — and three cheeses (Comté, Époisses, Chaource), tasted individually and then in all nine combinations, using Temporal Dominance of Sensations with a simultaneous liking rating, so the panel recorded what dominated when across a whole portion rather than a single bite.

The sentence that got picked up, verbatim from the conclusions:

“There was a wine effect on the liking of the combinations, showing that in the present case, white wine was a better companion for the evaluated cheeses than the red wines.”

Which is where most retellings stop. It is also the least useful sentence in the paragraph.

⭐⭐ What the same conclusion actually says

Two things sit either side of that line and they change its meaning.

First, the authors hedge it themselves — “in the present case”, and “the evaluated cheeses”. One white wine, two reds, three cheeses, one region, sixty people. That is a real result and it is not a law of nature, and the paper does not pretend otherwise:

“Wine-cheese interactions were found when describing the combinations, revealing that the perception of a combination of products is not the result of an additive or subtracting effect, which can be predicted based on their individual perception, but that they are complex associations that need to be deeply studied. This is one of the reasons why establishing a rule of thumb can be difficult and sometimes even contradictory.”

🔴 The paper cited as the source of a new rule of thumb explicitly says rules of thumb are the wrong shape of answer here.

Second — and this is the part worth actually taking home — it gives the mechanism. The white wine did better for a stated reason:

“This liking was explained by a reduced duration of astringency or bitterness as dominant.”

and

“astringency, bitterness and sourness in wine and in these wine-cheese combinations were perceived as negative drivers of liking by consumers.”

⭐⭐ So the finding is not “white beats red.” It is that a pairing succeeds when the cheese shortens the time astringency and bitterness spend in charge of your mouth. Colour is a proxy for tannin, and tannin is the thing under test. A low-tannin red is not the villain of this experiment; a big tannic one is.

That reframing is worth more than the headline, because it predicts rather than prescribes — and it immediately explains three pairings that a red-wine rule cannot.

Which is why the reliable pairings are reliable

Sweet wine with blue cheese. Sauternes with Roquefort, Port with Stilton — probably the most consistently praised pairing in the Western repertoire, and it is a sweet white with a savage salty cheese. Sugar suppresses bitterness directly, salt suppresses it too, and there is no tannin in the frame at all. The mechanism above says this should work, and everyone’s mouth agrees.

Beer, very often better than wine. Most beer carries far less astringency than a tannic red, the carbonation physically scrubs fat off the palate between bites, and the malt brings a sweetness that does the Sauternes trick in miniature. A washed rind against something Belgian, a cheddar against a bitter, a blue against a stout. Jeff’s brewing and distilling side is a whole other project, and this is the point where the two touch.

Cider with alpine and with cheddar works for the same reasons plus one more — the acid cuts fat — and it is the cheapest way I know to improve a cheeseboard.

⚠ And the corollary nobody likes: a big tannic red is the hardest thing on the table to pair with cheese, and it is the bottle most people open for a cheeseboard. Not because it is bad. Because it brings the exact attribute the panel scored as a negative driver of liking, and asks the cheese to fix it.

”What grows together goes together”

The other rule you will meet, and I think it holds up better than the red-wine one — but for a reason worth stating rather than mystical.

The Burgundy study is itself an example: the wines and cheeses were chosen because they “shared terroir.” Comté with a Jura wine, Chaource with Champagne, a Loire goat cheese with Sancerre. These combinations were arrived at by a very large number of people eating and drinking what was in front of them over a very long time, and selection pressure of that kind is a real search algorithm.

⚠ But it is a heuristic and not a mechanism, and it fails cheerfully. Stilton and Port share no terroir at all — one is Nottinghamshire, the other Portugal — and it is a better pairing than most things that do. Treat “what grows together” as a strong starting guess, not an explanation.

The cheese does something to the wine, too

One detail from the same study that I find genuinely surprising: the cheeses had very different capacities to carry a pairing. Tasted alone, Chaource had practically no temporal drivers of liking at all; Comté’s mushroom note was a positive driver; Époisses’ saltiness was a negative one.

⭐ Which means “pair the cheese to the wine” and “pair the wine to the cheese” are not the same instruction, and the asymmetry runs the way most drinkers do not expect. A cheese can rescue a wine by shortening its astringency. The reverse — a wine improving a cheese — is a much weaker effect in this data.

What I would actually do

Nothing on this page requires you to buy anything specific, so here is the short version.

• Serve the cheese properly first. An hour out of the fridge does more for a pairing than switching bottles. The board covers the timing.
• If in doubt, pour something with less tannin — a white, a good beer, a dry cider — and stop treating that as the compromise option.
• For a blue, go sweet. It is the nearest thing to a guaranteed result in this whole section.
• Don’t over-produce it. Bread, cheese, and one good glass beats a board with six accompaniments and a bottle that is fighting all of them.

Sources

• M. V. Galmarini, L. Dufau, A. Loiseau, M. Visalli, P. Schlich, “Wine and Cheese: Two Products or One Association? A New Method for Assessing Wine-Cheese Pairing,” Beverages 4(1), 13 (2018), doi:10.3390/beverages4010013. Study design, panel size and all quoted passages read from the paper itself, not from secondary coverage.
• ⚠ Note that the wider literature is not settled: other work has reported reds as the better accompaniment, and a rule stated in either direction is overstating what has been shown.