The Colour and the Shape

“You cannot trust this boy! His mind has been warped by colours, sounds and shapes!”

Edward Tattsyrup,  The League of Gentlemen – BBC

Synesthesia is one of the most fascinating things I have come across in my short journey into Neuropsychology. It is one of the phenomena that I cannot fully comprehend. That’s not to say I don’t understand it conceptually, but I find the idea of ‘seeing’ sounds and ‘tasting’ words so amazing that is it… well, incomprehensible. I was speaking to a friend recently about it (or rather I was trying to impress upon him just how awesome the human brain is) and he asked me if there was any benefit to being a Synesthete. It struck me that I didn’t have an answer to that question. One of the best things about not knowing the answer to things is that you get to search for an answer. It was that search that led me to this paper:

Grapheme-Colour synaesthesia benefits rule-based Category Learning

Watson, Blair et al (2012)

This study aims to see if the grapheme-colour Synesthete will utilise their grapheme colours in a rule to accurately categorise achromatic letter pairs, performing better in accuracy than non-synesthetes. The prediction was that the Synesthete would perform considerably better than the control group (and better than chance) on achromatic stimuli as they would perceive the graphemes in colour, which would aid the learning of the category rule.


The Method

There were 10 grapheme-colour synesthetes involved in the study and 80 non-Synesthete controls. The 80 controls were then separated into 10 groups of 8, one for each Synesthete stimulus set. Why a separate stimulus set for each Synesthete? Well, the reason for this is because the colours experienced in colour-grapheme are individual i.e. an ‘R’ to one Synesthete will blue but will be green to another, red to another and so on. It therefore follows each Synesthete needs a set designed for them and their colour experience. This ensures that when you are designing a set that follows the rule ‘members of category 1 contain a green followed by a pink grapheme’ you can make the letter on the left green and the right pink for all the participants. It is important to mention that the idiosyncratic nature of this synesthesia is not unique to grapheme-colour and is true for all types of synesthesia.


Figure 1. An example stimulus set

There were 3 groups per stimulus set: the Synesthete was presented with a stimulus set which was achromatic; a control-achromatic group where non-Synesthetes viewed the same achromatic stimuli as the Synesthete; and a control-colour group who were presented with the stimuli which was coloured according to the Synesthete’s colour scheme.

But how they could be sure that even the Synesthetes were using colour rules? That question was no doubt raised in the designing stage of the study as two more tasks were carried out to identify which participants were using colour rules and which were not. Immediately following the category learning task the participants were presented with 10 transfer test-trials and then a recognition task with 10 ‘foil’ pairs. He transfer test presented novel pairs which did not fit a category and the foil pairs were not present in the original blocks. Those using non-colour-based rules would perform differently on these from those who were.

The Results


It appears as though all the predictions of the study were correct and that colour did indeed benefit rule-based category learning. The synesthetes were not as accrate as the control-colour group on the categorisation accuracy task but their accuracy improved sharply and by the last epoch they performed better than the other groups. The control-colour group outperformed on the Transfer acuracy task and for the recognition task the control-achromatic group rejected more of the foils, as was predicted.

However, at this point it is still uncertain as to whether it is the use of colour-based rules that produced these results or that it was simply the presence of the stimuli in colour that helped. Is it possible that those with the achromatic stimuli weren’t as stimulated as those with colour ones? The transfer and recognition tasks showed that the colour made a difference but the authors had to go a step further to show that colour-based rules were the main reason for the effect. What they did was recreate the experiment with colour stimuli which had no connection and so was useless for categorisation. The results showed that each of the groups in this second experiment performed similarly to one another and similarly to the control-achromatic group in the first experiment.

So, what do the results actually say about how much grapheme-colour Synesthesia benefits rule-based category learning? In the discussion there are 3 main points raised:

1, Synesthetes learned more slowly. The categorisation accuracy was comparable in the Synesthete group to the control-colour group, although the control-colour group was more accurate in the first 2 Epochs (each Epoch was 2 blocks). The Synesthete group improved on Epoch 3 and was more accurate on the final one.

The authors point to the possibility that this is because a synesthete’s experience of the colours is not as vivid as the experience of real colours, therefore there are delays in acquisition. This is not to be dismissed but I wonder whether the synesthete’s experiences of the colours they see every day make them less remarkable and therefore at the start of the tests they find it harder to apply them to a colour-based rule? That is to say that the colours are always present in their experience and so begin by using other tactics to categorise the pairs. This could also explain the sharp rise in accuracy. Perhaps using some of the  graphemes which occur less frequently in day-to-day life would change the results?

2, Synesthetes were not as successful at transferring their learning to novel stimuli. It is suggested that this may be down to less vivid experiences, as in the point above. The paper then explains that one of the Synesthetes said that he saw the pair as one whole colour.

3, Synesthetes were slower to respond that participants viewing real colours. It has been suggested that synesthetic colours cannot be induced without conscious recognition of the grapheme (e.g. Laeng, 2009), which is given as one reason for the slower reaction times.

This need for conscious awareness was demonstrated in a study where synesthetes were presented with a series containing the number 5, which had an arrangement of 2’s within it. Thinking was that a synesthete would be able to spot the 2’s very quickly because they would present themselves a different colour. However, they did not spot them any more quickly than non-synesthetes. This showed that the number must be attended to for the colour to present itself (Ramachandran & Hubbard, 2001b)

5s and 2s

Overall I found this to be an interesting and thought-provoking study. I can see how investigations into synesthesia can be difficult to carry out due to the idiosyncratic nature of the phenomenon and I think the authors have designed some tasks which are solid. However, one of the things that struck me at the end was how although the Synesthetes were more accurate than non-synesthetes in categorisation accuracy, it wasn’t by a large amount. I think that the most interesting finding and the one which requires more investigation is the increase in synesthete accuracy.

Laeng, B. (2009). Searching through synaesthetic colours. Attention, Perception, and Psychophysics, 71(7), 1461 – 1467

Cytowic, R.E. and Eagleman, D.M. (2009). Wednesday is Indigo Blue: Discovering the Brain of Synesthesia. MIT Press.

Ramachandran, V.S., and Hubbard, E.M. (2001b). Synaesthesia: a window into perception, thought and language. J. Consciousness Stud. 8, 3-34.

The Colour and the Shape – Foo Fighters, 1997