Why it's so difficult to describe smells

A little while ago, my friend Susannah, who is in solo self-isolation during the pandemic, texted me to say she missed the smell of other people. This led to a conversation about how difficult it is to describe smells; we just don’t seem to have the specific language for smells that we do for colours, timbres, or textures. She wanted to know why.

Smells… flowery?

Because I am a psychologist, I initially assumed it was something to do with how the brain is set up.

We have well over five senses, but that’s a story for another day. However, of the five you’ve heard about, vision and hearing are pretty dominant in most humans, both in terms of the amount of brain space that gets devoted to them and in terms of the way we’ve set up our societies, which can be a real pain in the arse for people who are blind, deaf or both. Vision and hearing also seem to be ‘special’ in that when we hear or read language about those senses, the parts of the brain that are involved in perception and action get involved. This is important because it means we’re not just thinking about vision and hearing in an abstract way; our brains are trying to ground them in reality. However, we don’t seem to do the same thing with touch, taste or smell, so maybe we have a harder time linking language and the real world in those senses than we do with vision and hearing?

Maybe! There are several places where the chain from perceiving a smell to naming it could be disrupted, and -

OK, you know what? That’s enough about brain biology. It turns out it’s not about brain biology.

The problem with brain imaging studies like the ones I’ve just been talking about is that they tend to be done on people in Western countries and who therefore speak a fairly restricted set of languages. That’s important because it’s much easier to describe smells in some languages than others. Not everyone struggles to describe smells.

Oranges! (Orange you glad I didn’t say banana?)

An aside about the International Phonetic Alphabet I’m going to be using the International Phonetic Alphabet, or IPA, to describe some of the words that other languages use. The IPA is a useful way of describing what noises people make as they speak, so it can do things which written language can’t always do, like distinguish between how a Briton, an Indian, and a Tanzanian would each say “orange”. However, it’s not necessarily very easy to read if you aren’t familiar with it. IPA has some crossover with written English, but there are symbols you may not recognise, like ʔ (this is a glottal stop, i.e. the noise you make in the middle of uh-oh). If you’d like to know what noises the symbols represent, then here is a handy website with sound clips.

Ways of naming smells

Imagine I told you that you could no longer use the word blue. How would you describe things that were that colour? Maybe you’d say that they were periwinkle or sky or sapphire – all words that refer to a specific thing that is in the category of [redacted], but not to the category itself.

Not being able to use category words is the exact problem that English speakers have when describing smells. There are almost no category words for smells in English, except perhaps fragrant – which isn’t a very precise category word since it just means “a pleasant smell”. But that’s not the case for every language.

Now, consider a horrible smell: rotting fish.

Look, I needed a clear example. Consider yourself lucky that this picture is of fresh fish.

In English, we describe rotting fish smell and most other odours in a source-based way. That is, we refer to the source of the smell when we’re describing it, so we would probably call this a “fishy smell”, and speakers of most European languages would do something equivalent, like visgeur (literally “fish smell”) in Dutch, and odeur de poisson (literally “smell of fish”) in French.

What about other languages?

Some of them use verbs rather than adjectives. Thai has a lot of verbs that are used to talk about smell, like khaaw (“to have a fishy smell”). So do several other south east Asian languages, including some of the Aslian languages spoken by the Orang Asli, the indigenous people of the Malay peninsula. Unlike English, many of these languages do have category terms – for example, Jahai, an Aslian language, has odour verbs like pʔus, a term which includes the smells of old buildings, mushrooms and stale food.

One other way that we could talk about odours is through ideophones. You’re probably familiar with onomatopoeic words, which are a type of ideophone where the idea of a noise is evoked by a word - like thud, which evokes the sound of one thing hitting another. Ideophones, however, can evoke an idea in any sense, including smell. One language in which this happens is Huehuetla Tepehua, a language spoken by an indigenous people of north-eastern Mexico, in which the odour of rotting fish might be described as p’uks (“a strong, stinky smell”) – again, a category name for a smell which can also describe dirty nappies and dead animals. Generally speaking, in Huehuetla Tepehua odour ideophones, high-pitched vowels like i and consonants made at the front of the mouth like s tend to indicate less intense perceptions than low-pitched vowels like e and consonants made at the back of the mouth like ʃ.

The language we speak can have far-reaching effects, something I’ll return to in a later blog. In the particular case of smells, language can influence things like how accurately you can name a mystery smell and how likely you are to agree with other people what that mystery smell is. This is the case even though our initial emotional reactions to smells appear to be very similar across speakers of different languages.

Changing how we talk about smells

Speakers of a single language might talk about smells in multiple different ways depending on who they are and what they’re describing. One group who exemplify this are the thousand or so members of an indigenous people of Mexico called the Seri. They were traditionally nomadic hunter-gatherers and now live in two small coastal villages in north-west Mexico, where they mainly make a living from small-scale commercial fishing and buy their food from shops, though they do occasionally gather plant products to make food or medicine. In Seri, unpleasant smells are mostly described through verbs which cover a category of smells, like the verb -heemt (for skunks and rotten food, among other things), while pleasant smells are usually described in a source-based way, as in English.

So, the smell itself matters… but so does the speaker! Many older Seri can still remember their nomadic lifestyle, while younger Seri cannot, and they describe smells in quite different ways, with older Seri tending to use specific plant names for smells and younger Seri using words loaned from Spanish. It’s unclear whether the Seri smell verbs will remain in use by future generations – language is a living thing, and it is shaped by its users.

-heemt, or olor a zorrillo?

The role of the speaker’s experiences in describing smells is there in Thai, as well. Thailand is (generally speaking; not everyone in a country has the same attitudes!) a country in which odours are important: in food, in traditional medicine, and in religious practice. Potentially, these needs to talk about odoura have created a need for a vocabulary to describe them. In turn, that vocabulary allows more precision than she other languages about the experience of smell.

Using smell vocabulary matters

We can see this in action in another way among wine experts. In a study of wine experts and wine novices in New Zealand, the experts were better at naming odours than the novices, even though they could both detect whether a smell was present or absent at the same level of intensity. That latter piece of information is important, because it tells us that wine experts aren’t just overall better at smelling things than other people.

Wine talk (i.e. the way in which people who love wine talk about it) is a very useful example of why frequent use of smell words can help us become more attuned to smells. If we compare wine experts with coffee experts, we see that wine experts are more consistent about describing wine smells than coffee experts are about coffee smells – which is probably because ‘wine talk’ is a very well-established thing but ‘coffee talk’ is a much less well-established thing.

You know what I mean when I say wine talk: “Ah, Pamela, this Syrah is very open and has a peppery nose, but the key thing is that the granite soil at Le Petit Clos de la Merde d’Âne has made a notable contribution. The terroir really is terribly important. Now, let’s go and evict our tenants; I’m sure we can charge a higher rent than this.”

One last example! In Iran, both expert cooks and expert attars (herbalists) have extensive knowledge of herbs and spices, but while the Iranian cooking tradition is learned by watching and doing and very little talking about odours, attars need to be able to communicate their knowledge to customers. As you might expect from what’s come before, attars are more accurate and more consistent than naming both medical and culinary herbs than the general populations, but experienced cooks are not.

So, you want to get better at recognising and talking about smells? You’ll either need to develop a specialised vocabulary within the language(s) you already speak – if you don’t fancy learning about wine, perhaps you could become a perfumier! – or you’ll need to learn to speak a language which has a more extensive vocabulary of smells. But whether you pick Option A or Option B, the key is to spend lots of time communicating about odours with other people.