Language of Smell

One of our five senses, it is the most complex, the most evocative, and the most mysterious sense – But we don’t know how to talk about it

Think about that for a moment – what words have you got to describe flavour except for comparing it to something else? Strong, weak, rich, complicated – there are some, but not many. Compared to the plethora of words we have to describe colour, shape, movement, sound – the flavour landscape is desolate. Most people duck away from vivid descriptions, preferring hedonic terms, like “It’s good.” And yet food is incredibly important to us; it evokes memories, creates atmospheres, is used to bond with other people and change our mood.

Smell plays a bigger role in our lives than we might realise Image credit: Public domain

Different people respond to different flavours differently, partly because of memories, and partly because of different sensitivities to flavours.

Sensitivity can be mapped: our sensitivity to bitterness is gradually lost as we age, allowing us to drink stronger tea. Strangely, this changes differently for men and women: the decline in scent and in bitterness sensitivity is gradual for men, but for women, doesn’t kick in until the menopause. This may be because many bitter things are natural poisons. Our ability to detect them directly impacts our ability to survive in the wild. It’s more important to refine this sense early so we learn lessons we remember our whole lives. Similarly, children are mad about sugar. This sensitivity to it, which we lose as we age, could be another survival tactic, encouraging children to quickly hone in on the densest sources of energy.

But different people lose sensitivity to different smells. It's possible this may depend on the pollutants and viruses you’re exposed to during your lifetime. More mysterious is the fact that by actively studying scents, you can halt sensitivity loss entirely. You can also teach yourself to like flavours or get sick of them by exposing yourself to them – adjusting your internal regulator.


These differences make it harder than ever to find the words we need to describe flavour. Market researchers use people specially trained to articulate the nuances of flavour to decode consumer experiences. One such company is MMR, a market research company with a sensory science centre based in Reading. Part of MMR’s job is to link what goes on in our mouths and noses to what goes on in our brains. Producers want to know more about this to create better products and market brands.

However, a single successful product could take as long as 8 weeks, over 500 people, and a lot of dud samples – especially if alcohol is involved. (Asked to judge an alcoholic drink, most people get so excited they rate the first drink ridiculously highly. This is why MMR always let them have a first “wash out sample” – to get them ready for serious product assessment.)

Caroline Withers from MMR explained to me how understanding consumer perception of products can be challenging, mostly not because of differences between people, but because of language limitations that make why they like or dislike something hard to gauge.

Pairwise comparisons often give the most detail. Is it saltier? Crunchier? A darker colour? MMR combine consumer liking data with a sensory map produced by a trained panel of experts to understand what attributes drive consumer liking. However, people can be easily influenced by their surroundings. This is why the experts work in blank rooms, often in individual booths, with no colours or pictures, strictly controlled temperature and lighting, and a precise air pressure. They limit themselves to 2-3 hours before their nose and palate are overwhelmed. Wine tasting competitors equilibrate their palates by all eating or drinking the same thing in advance of a session. Another solution for palate cleansing is to sniff your own clothes or skin – your usual “background”.

To show how easily non expert consumers can be misled, researchers gave members of the public a piece of velvety material to hold whilst they were describing food – and it changed their language. People started saying things like “creamy” and “soft” much more often, or where they didn’t do so before. In one training experiment, MMR provided sugar solutions to University of Reading students and asked them to put them in order of sweetness. All the participants agreed that the second solution was the sweetest, the third the least sweet and the first in between – but they were wrong. Overwhelmed by the clawing sweetness of the second sample, a 10% sugar solution, they had failed to realise that both the first and third samples were 5%.

Sensory panellists work together to develop a collaborative vocabulary. Language can be very personal: “This note reminds me of my son’s backpack,” said one sensory panellist – and after a lot of teamwork decoding the strange memory, they decided this was the odour of old bananas.

When they’re not working on studies, the panellists are training, building up a chemical database of delicate differences in smells and tastes. By going through the same training experiences, this also means their results are more consistent, everyone uses the same scales of intensity and knows what that term means when somebody uses it. But sensory panellists don’t just have to be good at articulating notes; they also have to be systematic: consumers want abstract descriptions, but producers want plottable, statistical, numerical science. Is it 62 sour, or 63? 87 bitter, or a bit less? Panellists provide all of these values by training on the samples and scoring them individually to generate averages.

MMR have trained sensory panels in their sensory science centres in the UK, USA, China and Singapore, and collect consumer data farther afield. Sometimes they’ll compare the same product in different centres to make sure they get the same result – but they have to use strict protocols to ensure everybody treats the samples similarly. Even so, surprises happen. When a biscuit proved very different in the UK and China, MMR’s sensory scientists explored. The difference wasn’t in the biscuit – it was in the way it was eaten. UK panellists bit into the biscuit, whilst Chinese panellists would take it whole in their mouths. This led to surprisingly different experiences, and so brought biscuit biting into the protocols.

Bitten biscuit. Image Bitten biscuit. Image © Ryan Fung via Flickr.

The Chinese also use language differently, such as talking about teas as “lively” – a flavour English panellists had to learn. Translators work to bridge these gaps between lexical and cultural traditions about food. Similarly, MMR have tricks to warm people up, such as asking them what a brand would be like as a person, comparing it to animals, and delving into emotional associations. However, when working with the Japanese, who can be very reserved, this proved tough, especially when asking about emotional characteristics!

I also asked Caroline if she expected any future changes in the way we relate to flavour and aroma – she said yes. “People are becoming braver and more challenging flavour combinations are appearing,” she told me, from birch or maple waters to 'There’s a beer for that'. This brings new challenges to the table for industries, such as adult soft drinks and exciting prospects of new products just around the corner.

For more about smell and the science we don’t know about it, read our blog.