The mental association of color and temperature

April 1, 2014 by George Van Doorn, The Conversation
Are you feeling warm or cold by the colours? Credit: Flickr/Joe, CC BY-ND

In a typical kitchen or bathroom you often find the hot and cold water taps labelled red and blue.

It's common practice in industrial and interior design in many parts of the world to present information about by means of colour cues. People often talk about painting a room in warm hues such as yellows and oranges.

This association of a with a temperature is known as "crossmodal correspondence" and has been known about for many years.

But why do we make these associations? Are they learned just because that's the way hot and cold have always been labelled, or is there some deeper association?

Much of the earlier research on this has been subjective and focused on asking people to rate colour patches or coloured stimuli as being either warm or cool, or by giving people a thermal stimulus and asking them to report which colour they were reminded of.

We wanted to examine the colour-temperature correspondence using more objective measures, and the results were published last month in PLoS ONE.

The colour temperature test

We used a number of tests including an Implicit Association Test (IAT). The IAT is designed to assess the strength of automatic associations between different concepts in a given individual. (You can take a test yourself here if you'd like.)

A total of 42 people from Japan and Australia were run through one of three experiments. We manipulated the combination of colour and thermal stimuli presented in each trial.

In two experiments, the thermal stimuli were words such as "cold" or "warm" presented on a computer screen. In the third experiment people could feel a "warm" or "cold" stimulus on the skin of their index finger using a Peltier device.

Typical hot and cold taps - marked red and blue. Flickr/Anjum, CC BY-ND

In half of the trials people were shown congruent combinations (they were asked to associate red and warm) while in the other half of the trials people were shown incongruent combinations (this time they were asked to associate red and cold).

The results of our IAT confirmed our prediction that congruent pairings would give rise to shorter response times than incongruent pairings. This shows that the speed of response to a colour or a thermal word is influenced by the colour-temperature correspondences.

A one way effect

Some of the results also show that this colour-temperature correspondence might only work one way. Specifically, there was no difference in response times between congruent and incongruent pairs when people were identifying the colour of a stimulus in the pair.

But when people were asked to identify a temperature, the congruent pairings gave rise to shorter than incongruent pairings. People were quicker at identifying a "warm" stimulus when it was combined with red, relative to a "warm" stimulus paired with blue.

This asymmetrical effect is not unexpected given that colours are often used to indicate temperature, whereas temperature is seldom (if ever) used to indicate colour.

The colour of sound

A similar one-directional effect is seen in people who have a condition known as synaesthesia. They often perceive numbers, months or even sounds as colours; but fail to perceive colours as sounds, for example.

The correspondences between colour and temperature are presumably based on our natural observation of the environment around us with correlations that are likely learnt through experience.

The warm glow of an orange sunset over the cold blue water – a typical colour temperature association. Credit: Flickr/Vincent_AF, CC BY-SA

Fire and the sun are both warm; hence all the colours of fire and the sun might be associated with warmth. Water and forests are cool; hence blue and green may be associated with coolness.

The association might also be linked to the fact that our skin gets redder when we are warm and blue when we are very cold.

It's also been argued that many other crossmodal correspondences are based on natural, rather than learnt, correlations present in the environment, such as sound pitch and size. Large falling objects usually land with a deep thud, whereas smaller objects are more likely to make a high-pitched "ping" as they hit the ground. The calls of small animals are usually high-pitched relative to their larger counterparts.

So the reason we use red for the hot tap and blue for the cold tap is explained by our objective test. This crossmodal association is likely learnt via our encounters with colours and temperatures that correlate in the natural world. It also explains why we are slower to react when the colour combinations are reversed (such as hot and blue).


In those cases where the crossmodal association is learnt through experience, different cultures and environments could possibly influence the colour-temperature association.

Our tests were done on people in Australia and Japan, and bathroom/kitchen taps in Japan are tagged with the same red (hot) and blue (cold) as they are in Australia, so our results were consistent for the colour-temperature association.

But in some African countries the association is reversed with blue signifying hot and red cold. Given the natural observed association of colour and temperature described earlier, it would be interesting to see if our test would produce different results in such countries.

Explore further: Research shows wallabies lose on the pokies

Related Stories

Research shows wallabies lose on the pokies

January 30, 2014
( —Biologists have discovered that a wallaby's perception of colour is more similar to a dog than a quokka, sparking questions as to why marsupial colour vision has evolved so selectively.

Recommended for you

Researchers devise decoy molecule to block pain where it starts

January 16, 2018
For anyone who has accidentally injured themselves, Dr. Zachary Campbell not only sympathizes, he's developing new ways to blunt pain.

Scientists unleash power of genetic data to identify disease risk

January 16, 2018
Massive banks of genetic information are being harnessed to shed new light on modifiable health risks that underlie common diseases.

Blood-vessel-on-a-chip provides insight into new anti-inflammatory drug candidate

January 15, 2018
One of the most important and fraught processes in the human body is inflammation. Inflammatory responses to injury or disease are crucial for recruiting the immune system to help the body heal, but inflammation can also ...

Molecule produced by fat cells reduces obesity and diabetes in mice

January 15, 2018
UC San Francisco researchers have discovered a new biological pathway in fat cells that could explain why some people with obesity are at high risk for metabolic diseases such as type 2 diabetes. The new findings—demonstrated ...

Obese fat becomes inflamed and scarred, which may make weight loss harder

January 12, 2018
The fat of obese people becomes distressed, scarred and inflamed, which can make weight loss more difficult, research at the University of Exeter has found.

Optimized human peptide found to be an effective antibacterial agent

January 11, 2018
A team of researchers in the Netherlands has developed an effective antibacterial ointment based on an optimized human peptide. In their paper published in the journal Science Translational Medicine, the group describes developing ...


Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.