Part two: Do horses see in colour?

Here's our resident scientist Ilse who brings us the second part of her colour vision blog and finally answers the mystery question: do horses see in colour? 

In my last blog, I introduced the mind-bending concept that colours don’t actually exist; they’re all in our head.

Colours are how our brains interpret different parts of the spectrum of light. We have three types of colour receptor, which allow us to see the “colours” red, green and blue. All other colours (I’ll carry on using colour for want of a better word!) are combinations of these three.

Horses have two types of colour receptors, giving them colour vision, but not as we know it! In order to try and understand how colours might appear to a horse, we need to carry out some experiments. However, it gets very complicated very quickly.

For starters, horses can’t speak, so we can’t just ask them outright. Human colour vision tests that rely on a person reporting a coloured number or letter hidden in a pattern of dots of another colour clearly won’t work on horses (unless their trained to count of course…).

So we need to come up with a behavioural experiment that gives us a way of testing whether a horse can tell the difference between two colours.

Researchers have tried various methods, but the two most successful are boxes and trapdoors that require the horse to make a choice of some sort.

Typical tests for colour blindness in humans involve reporting the number formed by dots of one colour in a background of dots of another colour. Clearly we can’t us this type of test on horses.

Typical tests for colour blindness in humans involve reporting the number formed by dots of one colour in a background of dots of another colour. Clearly we can’t us this type of test on horses.

In the experiment with the boxes, they were painted different colours, and the horse could choose which box to go to.

The other experiments used a board divided in two with a differently coloured region either side of the divider that the horse could push with its nose to open the trapdoor and get a food reward.

In both experiments, horses were trained to associate a certain colour with a food reward.

The idea is, after enough training sessions, a horse will look at a coloured box or region on a board and expect there to be food. In the experiment, the horses are presented with a choice between the colour they’ve been trained to and a different colour.

he two-choice trap-door food rewarding experimental set up (adapted from [1][2])

he two-choice trap-door food rewarding experimental set up (adapted from [1][2])

If they can tell the difference between the two colours, they should choose the one they associate with food (as we all know, food is the main reason for living according to a horse, and is therefore a great motivator).

For example, if a horse is trained to associate blue with food, we can test to see if they can tell the difference between blue and other colours like red and green.

At the risk of repeating myself a lot, “blue”, “red” and “green” are human concepts; they’re how our brain processes what we see. These colours probably look completely different to a horse!

However, this type of ‘two-choice’ experiment is not without its pitfalls. In addition to the two choices having a different colour (e.g. red and blue), they may appear to be different brightnesses (i.e. the red may appear brighter than the blue).

It’s possible, therefore that the horses could be telling the difference between the two not by colour, but by brightness. We therefore use different shades of grey both in training and in the experiments to test to see how sensitive horses are to different brightnesses (e.g. what is the smallest difference in brightness that horses can see?).

So what is the verdict; how colourful is a horse’s world? Well, the answer is that we can never truly know how the world looks to our horses. However, we can try and understand using our perception of the world. This is the likely extent of colour in the horse’s world:

isle3.png

A combination of blue and green to almost yellow. This isn’t to say that red is invisible to a horse. A pure red pole will look black because it won’t reflect any blue or green light. A pole that is a mix of green and red will only look green to a horse, which may be a problem if it's seen against a grassy background.

Although trying to understand the colour vision of the horses may, at times, appear to be a mind-bending philosophical exercise, we can still apply the scientific method to build up a picture of the horse’s world.

We can use this information to put ourselves in our horses’ shoes, to try and see the world through their eyes.

Read part one of this blog here.