Before food ever reaches the mouth, the brain has already made a decision.
That idea sits at the centre of Professor Charles Spence’s work, and quietly challenges the way we think about taste, pleasure and eating itself.
Spence is a prize-winning experimental psychologist at the University of Oxford, and also the author of international best-seller Gastrophysics: The New Science of Eating and many others. For decades, his research has explored how all of our senses interact to shape flavour, revealing that what we taste is influenced as much by expectation, colour and context as by ingredients.
“Eating and drinking are amongst the most dangerous things we do,” he says. “We’re putting something into our bodies that could poison us. That’s not something to be taken lightly.”
It’s an unusually stark way to begin a conversation about food, but it captures Spence’s approach: grounded, evolutionary, and focused on what the brain is trying to protect us from.
Why colour matters more than you think
Taste buds, Spence says, aren’t the brain’s first line of defence. Relying on taste alone would be slow and inefficient. Instead, the brain looks for signals before food ever reaches the mouth.
“Our brains use other cues to predict what things will taste like,” Spence says. “Those cues help us decide whether something is likely to provide energy, or whether it’s bitter and to be avoided because it might be poisonous.”
Colour plays a key part in this prediction system. Over time, the brain learns patterns from the world around it. “Our brains learn through experience that certain cues, like colour or appearance, tend to predict sweetness, energy content or edibility,” he says. He gives an example of warmer colours such as reds and oranges, common in ripe fruit often signal energy and sweetness. Green, by contrast, is typically unripe or sour. Over hundreds of thousands of years, the brain has become excellent at forming these shortcuts.
What this means is that before we even taste food, the brain has already formed an expectation of what that food will taste like. Taste buds still matter, but they become a form of confirmation rather than discovery. The flavour experience becomes less about sense and more about prediction.
By the time we actually eat, the work has largely already been done.
“When we taste something, we’re really just checking: does it taste the way my brain predicted it was going to?” Spence says. “If it’s roughly what you expected, then you don’t really think about it much. We live in a world of flavour predictions rather than flavour experiences.”
This simple idea that expectation shapes flavour is central to the field of gastrophysics and underpins much of Spence’s work.
If you’ve ever been disappointed by a dish that looked amazing, or surprised by something that looked plain but tasted great, you’ve experienced expectation in action. Vision, Spence says, is the fastest and often the most influential sense in forming flavour predictions.
Spence describes an experiment he did in the lab where participants were given drinks deliberately coloured in ways that didn’t match their flavour. “We gave people numerous different coloured drinks and told them, ‘we’ve just made these colours up. Ignore them they don’t tell you anything.’”
Even with that, people struggled. “Your brain kind of automatically sets these expectations,” he says. “People couldn’t override it.”
The same thing happens with wine. In blind tests where white wine is dyed pink or red, even experienced tasters insist they can smell and taste red wine characteristics.
“The more expert you are,” he says, “the more you’re tricked by the colour of the wine.” Expertise doesn’t free us from bias, it can deepen it.
From fruit trees to supermarket shelves
While many of these associations began in nature, Spence argues that modern food environments now shape them just as powerfully, if not more so.
“We’re not picking our fruit from trees much anymore,” he says. “So I think the associations we learn probably come from the supermarket.”
Pink is his favourite example. “I don’t think pink really exists in nature much,” Spence says. “But it does in strawberry milk, strawberry ice cream, strawberry bubblegum. Over time, the brain learns to associate pink with sweetness, even when that link is entirely artificial.”
Packaging reinforces these expectations. Crisp flavours tend to follow strict colour codes, and Spence has tested what happens when those codes are broken. In one experiment, prawn cocktail crisps were placed into a cheese and onion bag.
“People tasted the colour of the bag rather than the flavour of the crisp,” he says.
He recalls a similar reaction when Coca-Cola released its white Christmas can in 2011. “People said it tasted different,” he says. “Especially when you drink straight from the can, all you see is the white.”
Description and price are also factors which are powerful expectation setters, and the brain treats them seriously.
“If you tell someone before they taste a drink that it’s going to be very bitter, it will taste more bitter to them,” Spence says. “And it’s not just that they say it’s bitter, you can actually see enhanced bitterness responses in the brain.”
The power of labelling, pricing and suggestion
Expectations don’t just change what we think we taste, they change what happens in the brain itself. When we expect something to taste a certain way, the brain prepares for that flavour before we even eat or drink it.
“It’s setting a prediction that gets anchored in the brain,” Spence says.
This helps explain why blind taste tests so often surprise people. Without labels, prices or branding, confident opinions quickly fall apart. “As soon as you know something is expensive, it tastes better,” Spence explains. “That’s the expectation doing the work.”
Pleasure before the first bite
Despite the precision of his research, Spence is quick to admit that food pleasure isn’t something that can be pinned down neatly.
“Great food experiences last a couple of hours at most,” he says. “A lot of the pleasure comes from the anticipation, and the memory afterwards.”
That insight helps explain why presentation, storytelling and setting matter so much. On aeroplanes, for example, Spence points out how frustrating the limited menu descriptions are and how they undermine enjoyment. “If all they can say is ‘chicken or fish’, they’ve lost it right there,” he laughs. “They’re not setting the right expectations.”
Even at home, expectation can be engineered. “Just put wine in a heavy bottle and tell me it’s expensive,” he jokes. “I’ll love it.”
What we tend to get wrong about food science
Spence is wary of how his research can be sometimes interpreted in the media. One common misconception, he says, is the idea that bright colours automatically mean something is artificial or unhealthy.
“Cochineal, for example, a type of scale insect, gives a brilliant bright red colour,” he says. “It’s entirely natural, but it’s insect-based, which people find unpleasant.” Consumer fear around artificial colourings has led brands to dull down palettes, even when natural alternatives exist. “It’s about what people think is natural,” Spence says, not necessarily what is.”
For Spence, this understanding of flavour isn’t about tricking people or manipulating choice. It’s about recognising how little control we actually have over our sensory experiences. He’s careful to stress that expectation isn’t something we can simply switch off once we’re aware of it. “Even when you know the colour is misleading, your brain still goes there,” he says. “These associations are automatic.”
That inevitability, he suggests, should change how we think about food judgement and guilt. If taste is always filtered through memory, context and prediction, then eating becomes less about willpower and more about understanding the systems shaping our choices in the first place.
The future of flavours
Spence turns to what still excites him: colour-changing foods such as butterfly pea flower tea, which shifts hue when lemon juice is added, augmented reality to alter the appearance of food in real time, and the possibility of altering flavour without changing ingredients at all.
“Why bother?” he asks, anticipating the scepticism. “Well, maybe because it’s hard to get certain visual cues naturally. You could create metallic-looking drinks, liquid silver, purely through perception.”
It’s a fitting place to end. Not with taste as a fixed property of food, but as something flexible, constructed, and deeply mental.
Because if Spence’s work shows us anything, it’s this: taste doesn’t begin on the tongue, it begins in the mind.
