SOME foods are made for one another. From the comforting cuddle of mozzarella, tomato and marjoram atop a pizza to the tantalising trinity of ginger, garlic and soy sauce that make East Asian dishes sing, some combinations seem so natural that it is difficult to imagine a world without them. And yet for centuries, gourmands and academics have been confounded by why some foods harmonise so well.
In 1992, chefs Heston Blumenthal and hit the lab to try to solve this culinary riddle. They happened upon the idea that foods that taste good together also share many volatile flavour compounds – the aroma-carrying chemicals that rise up into the back of the nose to create the perception of flavour on the tongue. Their findings were validated in 2011, with a study that .
Advertisement
Ahn at Indiana University and his colleagues used the data to construct a network model, a complex map of the relationships between all of the recipes’ ingredients and the flavour compounds they shared. This confirmed that recipes from North America and western Europe do tend to pair ingredients that share flavour compounds.
“Flavour pairing theory” made waves in the culinary world, with dedicating resources to applying the idea to their products and tapping into open-source data on flavour compounds to predict what the next big food trend would be.
And yet, the concept was fundamentally flawed: it . Now, a new network model offers a solution that could unite all food flavours once and for all. As a doctor turned food specialist, I decided to put it to the test by creating a festive Christmas meal like no other.
The original flavour pairing theory neatly explained why coffee and chocolate is a match made in heaven: they share a legion of roasted, nutty and earthy flavour compounds, such as pyrazines. And it helped us create a range of whacky food pairings. Chocolate and blue cheese share at least 73 flavour compounds, for instance, which is why a blob of Roquefort on a cube of dark chocolate really does taste good.
But the data also revealed that Indian, South-East Asian, West African and some southern European cuisines typically don’t pair ingredients with shared flavour compounds. In fact, they do the opposite: they tend to combine ingredients that are least likely to share such compounds, such as lamb and cinnamon or fish sauce and tamarind. These are combinations the flavour pairing hypothesis says should clash, but are truly delectable, as everyone who has tried an authentic biryani or pad thai will know.
What is flavour bridging?
The answer came, surprisingly, from a major Spanish communications firm, TelefÓnica, that had long been exploring ways to use its expertise in network science in other sectors. It tasked researcher , now at the Central University of Technology, Arts and Design in Madrid, Spain, and his colleagues with looking again at the vast, sprawling network of flavour connections seen in the world’s cuisines. They subsequently teamed up with legendary Spanish chef , who ran the three Michelin-starred restaurant, El Bulli, until it closed in 2011.
The team reanalysed the original flavour pairing network. Within the model, each ingredient is represented as a “node”, and the connections between nodes indicate the sharing of flavour compounds, thus mapping out the relationships between ingredients across various dishes. The strongest nodes are those that share numerous flavour compounds with many other ingredients, giving them a versatility that makes them likely to be key components in multiple dishes.
Simas and his team discovered that recipes with incompatible ingredients use a third ingredient that shares numerous flavour compounds with both the incongruent foods. They called it . Their analysis suggested that the most harmonious pathway between two ingredients may not be the most direct one. In the original analysis focusing on direct connections, familiar ingredients like cocoa, coffee and cheeses had emerged as the most crucial nodes in Western cuisines. The bridging analysis elevated teas, wines, tomatoes and a plethora of less familiar ingredients as key nodes that connect seemingly incompatible ingredients.
Unusual food combinations
For example, while garlic and strawberry share few flavour compounds with each other, both have strong relationships with roasted beef. Intriguingly, a more circuitous path via roasted onion and Bantu beer, common throughout Africa, could further enhance their compatibility, potentially making for a more flavourful and complex dish.
Problem solved? Not quite. Up against the flavour bridging enthusiasts are an . One major issue is the bridging network’s simplicity. It ignores each ingredient’s texture, fat content, relative saltiness, sweetness, sourness and savouriness, and doesn’t consider the specific concentrations of aroma compounds in each ingredient. Another problem is that the concept doesn’t define the minimum number of shared compounds that are required for greatest palatability.
There was only one thing for it: I needed to test the idea for myself. So I decided to cook up an experimental festive feast. The idea was to incorporate pairs of foods that clashed horribly and see if they could bridged with a third ingredient to create delicious new combos. I enlisted adventurous New Scientist staff Obomate Briggs, Finn Grant, Joshua Howgego and Clare Wilson to judge my culinary creations, and they brought a blend of scientific rigour and festive cheer to the table. The results speak for themselves.
WELCOME DRINK
Champagne with chestnuts
Every Christmas party needs a welcome drink, and champagne seems befitting. Boasting more than 600 different flavour compounds, its bright acidity and hints of pear, green apple and citrus mean that champagne is awash with potential fruit pairing partners. The festive chestnut, however, is way down the list, making it a suitably unnatural partner for my experiment. Scrutinising flavour compound databases revealed, however, that both champagne and chestnuts share many volatiles with pineapple, which should harmonise the two.
To extract and concentrate flavour from the chestnuts, I opted to make a simple syrup by simmering a cup of chestnuts with equal amounts of water and sugar. I added a tablespoon of this syrup to a 50:50 mix of champagne and pineapple juice poured into a champagne glass and voila: pineapple-chestnut fizz!
Knowing beforehand that this was an experimental menu, my guests seemed surprised to be greeted with a pleasantly refreshing cocktail. Let’s hope it isn’t downhill from here.
Result: A charming boozy introduction
Average score: 7 out of 10
MAIN COURSE
Roast turkey with chocolate cake
The only option for our feast’s centrepiece was turkey, the meat everyone loves to hate. Notoriously dry owing to broiler breeds’ lack of fat, I set myself the challenge of trying to pair it with my all-time favourite Christmas dessert: a chocolate Yule log. Fortuitously, sweet-toothed researchers had already analysed the volatile compounds in chocolate cake – and sure enough, it makes for a terrible natural pairing with turkey according to the flavour pairing theory. However, referring to flavour data, I discovered that peanuts pair well with both turkey and chocolate, partly thanks to the nutty, toasted flavour compounds created through the roasting of cocoa and turkey skin.
To combine the three, I chose to blend soft peanut butter into a turkey gravy to give a satay-like twist. Texture would come from slicing and baking the chocolate cake into crunchy croutons to garnish the turkey.
Result: An almost unanimous hit. “The turkey and chocolate was an unexpected delight,” says Josh.
Average score: 8 out of 10
SIDE DISH
Brussels sprouts and Scotch whiskey
Love them or loathe them, no Christmas meal would be complete without Brussels sprouts. Their divisiveness is in part due to variations in the gene encoding the TAS2R38 bitterness receptor on the tongue, resulting in half of us finding their taste intolerable. I decided to pair this classic festive vegetable with the most famous of winter warming liquors, Scotch whiskey. Sharing scant flavour compounds, a bridge came in the form of apple, which shares 70 flavour compounds with whiskey, and more than 100 with the sprouts.
Sautéing Brussels until just tender is probably the most flavourful way to cook them. Left too long in a pan of boiling water, bitter-tasting glucosinolates turn into eggy smelling, sulphur-containing volatiles. Adding chunks of apple to the pan and moving them to one side, I sprinkled in a little sugar and a splash of whiskey to make a glaze, which I then used to coat the greens.
Result: A triumph for all except Brussels sprout-hating Obomate.
Average score: 7 out of 10
SIDE DISH
Cabbage with gingerbread
With my creative juices in full flow, I attempted an ambitious brassica and dessert combo in the hope of pepping up a rather dull vegetable. The peacemaker between the sweet and fiery gingerbread and bitter cabbage came in the form of golden raisins, which I simmered with small pieces of gingerbread in a little sugar water to make a sticky compote to serve atop gently braised, shredded cabbage.
Result: Too sweet for some, it was generally seen as a pleasing twist on this seasonal vegetable.
Average score: 6 out of 10
DESSERT
Cloves with Emmental cheese
As I considered the meal’s conclusion, I turned my attention to the cheeseboard. Emmental has a place in my heart for its mild, slightly nutty flavour that goes great on a cracker. Working down my list of seasonal ingredients, cloves struck me as a particularly unlikely pairing. Cross-referencing with other ingredients, I was encouraged to see that beer would bridge the two admirably, sharing the sweet, creamy compound vanillin with the cloves, and chocolatey tetramethylpyrazine with Emmental. I needed a neutral medium to combine them into a dessert and my inspiration came from Betty Crocker’s traditional bread pudding. I folded grated Emmental cheese into slices of bread soaked in egg, sugar and beer, then baked for 45 minutes before drizzling with a homemade clove syrup.
Result: Mostly positive reviews were drowned out by howls of revulsion from Josh, who found the “disgusting” clove syrup too medicinal.
Average score 7 out of 10
While not every pairing won universal acclaim, my four guests seemed largely impressed. “Some of the flavour combinations blew my mind – sometimes in good ways and sometimes not so much,” says Clare. Initially sceptical, Finn is now a convert to food bridging after finding himself “tucking into second helpings” of nearly everything. Obomate was also won over by the sweet main and the sticky clove syrup. Josh offered the most effusive praise, labelling the meal “a massive win”, remarking on how each dish’s flavours coalesced into “a delicious whole”.
Whether these culinary successes were due to luck or intentional design remains an open question, but it was certainly a fun evidence-based twist on a traditional dinner party that anyone can try. I am willing to keep on experimenting. So, who’s ready for my next gastronomic gamble: an avocado, broccoli and chocolate surprise?
Stuart Farrimond is now thinking about the peppery potential of Easter eggs
