Chef Grant Achatz has been writing a series for The Atlantic magazine over the previous weeks, outlining his approach to food and wine pairing. Achatz's is a unique perspective: he is chef and owner of haute American restaurant Alinea in Chicago and has previously worked as a winemaker. He brings both food and wine production experience to the table, and runs the breed of restaurant where pairing food and wine is treated with seriousness. According to Achatz, what makes a great pairing?

With a menu of 12 to 26 courses, it is impossible, and not terribly desirable, to have just one or two wines bridge the courses. Instead, we believe that each course can be enhanced by the wine--and visa versa: that the food can enhance the wine as well. That is the essence of a great pairing.

There are no hard and fast rules on pairing wine with food. The romantic notion that a sommelier with amazing taste-memory knows exactly what to pour with whatever dish has some occasional truth. But more often than not, that merely informs a direction. Then we begin opening dozens of bottles to find just the right complement to a course.

I tend to sit in the camp that you should drink whatever you enjoy most and not get too hung up on pairing foods with wine. It would be a rare and possibly injurious luxury to have the chance to open dozens of bottles of wine to choose which one works with a particular meal and I'd applaud a sommelier who could take the liver damage on my behalf. My memory for tastes is both as decent and as flawed as anybody elses.

In the instance of sitting down for a twenty six course dinner, there is no possible way that you could personally pick the best wine to pair with each course as you'd struggle to do so even if you had eaten everything on the menu before. Choosing wine before you've eaten something is a gamble; choosing one after you've eaten something just shortens the odds.

Much of what we perceive about food and wine comes down to expectations. When we expect that we're in the capable hands of an accomplished sommelier, we tend to make ourselves believe that they're making the right choices. Cornell University's Food and Brand Lab has research to this effect. They labelled the same wine as from California (probably favourable) and North Dakota (not yet known for its fine vintages). People who came in with a favourable impression of the wine not only rated it better but also the food that they ate it with. They dwelled at the table longer and ate more of the food on their plate, simply as a result of primed expectations on the label and waitstaff.

Michael Siegrista and Marie-Eve Cousin from Zurich's Swiss Federal Institute of Technology confirm with similar results in a lab setting. Both negative and positive reviews of a wine, when given before a tasting, will affect people's negative or positive perception of the wine, post-tasting. When we're primed to enjoy wine by a restaurant, surroundings and dining companions, we enjoy it with a lesser regard to the physical properties of the wine and food.