| Grapheme-Colour Synesthesia |
| One individual's colour perceptions of graphemes with interesting patterns. Retrieved from: https://www.youtube.com/watch?v=dNy23tJMTzQ |
Grapheme-colour synesthesia is when graphemes, which include letters and numbers, consistently induce a particular colour sensation. The specific colour associations differ between individuals and are constant throughout life, usually starting from a very young age. The graphemes are referred to as the inducers, and the colour sensation as the concurrent [1]. Individuals with any type of synesthesia are referred to as synesthetes. Grapheme-colour synesthesia is one of the most common types found across populations. Key regions of the brain consistently activated during grapheme-colour synesthetic experiences are the colour area V4, grapheme areas, and specific regions of the superior parietal lobe (SPL) [2]. Although these areas of the brain are strongly implicated in all research papers, the pathways connecting these regions are currently under debate. Understanding grapheme-colour synesthesia may help us to understand what makes people different, and how our brains shape our individual perceptions of the world.
1. Projector vs. associator synesthetes
 |
| Figure 2. Comparison of increased grey matter in projector (blue) and associator (red) synesthetes. [4] |
 |
| Figure 1. Schematic diagram of projector (left) and associator (right) synesthetic pathways with respect to the superior parietal lobe (SPL), colour area V4, and letter shape area (LSA) in the fusiform gyrus. [2] |
Though applicable to most types of synesthesia, the categorization of projectors and associators is especially important in grapheme-colour synesthetes. Projector synesthetes experience colour sensations that are projected in the physical world when looking at graphemes [2]. For instance, projectors will see colourful letters on a sheet of paper, but the same set of letters appear achromatic to a non-synesthete. Associator synesthetes, on the other hand, experience colour sensations in the space of their mind when looking at graphemes, but seeing achromatic graphemes on paper. The space of the mind mentioned is not physical, but akin to where you would conjure up an image of a colourful cupcake in your mind.
In light of these differences, it is important to take into account which type of synesthetes the subjects are when conducting studies, since results often differ between projectors and associators [3]. For instance, it has been found that projector synesthetes activate the colour area V4 from grapheme areas (bottom-up) whereas associators activate V4 indirectly from the superior parietal lobe (top-down) [2]. Furthermore, there is a difference between regions of increased grey matter in the two types [4]. Projector synesthetes have increased grey matter in the calcarine sulcus involved in visual experience and the precentral gyrus involved in planning. Associator synesthetes showed increase grey matter in the hippocampus involved with spatial memory, and the angular gyrus associated with language processing and metaphors.
2. Neural mechanism models
 |
| Figure 3. Activation of PTGA (light blue) and V4 (dark blue) as seen in grapheme-colour synesthetes when shown achromatic graphemes. [13] |
The cross-activation and the disinhibited feedback model are the two most common models used to theorize the neural mechanisms behind grapheme-colour synesthesia. Other models have also been suggested, but they are usually branched off from the two models mentioned. Regardless of the model, the regions of the brain consistently activated during the phenomenon of grapheme-colour synesthesia and studied as key regions of interest are the colour area V4 and the posterior temporal grapheme areas (PTGA) [5]. PTGA is stimulated in both synesthetes and non-synesthetes when viewing achromatic (black and white) graphemes since it is involved in processing letters and numbers, whereas V4 is only activated in synesthetes because only the synesthetes experience the sensation of colour. Even synesthetes who later become blind still get activation in V4 when using braille. Other specific areas activated include the anterior lingual gyrus associated with classifying colours, and intraparietal sulcus regions of the brain associated with attention and binding of senses into a representative world [5][6].
Anatomically, grapheme-colour synesthetes have been shown to have greater anisotrophy in diffusion tensor imaging (DTI) studies, differentiating from non-synesthetes in the orientation and density of axons as well as myelination level [7][8]. This translates to increased structural connectivity seen primarily in the inferior temporal cortex (part of the PTGA), especially in projector synesthetes.
2.1 Cross-activation model
 |
| Figure 4. Nearly simultaneous activation of V4 and PTGA in grapheme-colour synesthetes (blue). Repeated activation of V4 is hypothesized to be a result of hierarchical analysis of graphemes and thus colour. [13] |
The cross-activation model is the primary model to explain grapheme-colour synesthesia [5][9]. This model theorizes that grapheme-colour synesthesia results from incomplete pruning of neural connections between different brain regions starting prenatally, which leads to hyperconnectivity within and between affected regions throughout life [2]. Specifically, there is decreased pruning between PTGA in the inferior temporal lobe and colour area V4. In the brain, these regions are adjacent to each other, which make it easy for connections to be made and retained, though connections are sparse in a normal brain. That is not to say that areas of the brain must be adjacent to each other in order for the occurrence of hyperconnectivity. If this model is correct, activation of V4 and PTGA are expected to be close in time because of the horizontal connections between them [10].
Many studies use functional magnetic resonance imaging (fMRI) studies looking at connectivity to see whether their results support this model, and several have shown evidence of hyperconnectivity between V4 and grapheme areas [11][12], but this method cannot analyze the temporal aspect of activation important in distinguishing the models [5]. Thus, magnetoencephalography (MEG) and electroencephalography (EEG) studies have been conducted more recently. Using MEG, it has been found that V4 and PTGA activation during grapheme-colour synesthesia are almost simultaneous [13], consistent with what is expected based on the cross-activation model, though this has only been shown so far in projector synesthetes, with no studies looking at associator synesthetes.
2.2 Disinhibited feedback model
The disinhibited feedback model is not commonly used to describe grapheme-colour synesthesia, but it is still a major model to consider. This model suggests that synesthetic experiences are the result of reduced or loss of inhibitory signals that usually prevent communication between different regions of the brain [2], most significantly from areas of multisensory integration such as the temporo-parietal-occipital junction [5]. This theory takes prevalence when considering that synesthetic experiences can be drug-induced, though the hallucinations are often more complex and not as consistent as congenital synesthesia.
2.3 Other models
Other models branch off of these two key models, most notably the re-entrant processing model and hyperbinding model. The re-entrant processing model takes inspiration from both models [1] and suggests that synesthesia occurs because signals in the brain may sometimes flow backwards compared to the normal processing of sensory data, from the anterior inferior temporal lobe and posterior inferotemporal cortex to the V4, producing coloured graphemes [5][13]. This theory came from the fact that context is often important in determining synesthetic experiences. The hyperbinding model, on the other hand, is not exactly a model on its own, as it requires one of the other models to form a useful explanatory mechanism. This model theorizes that synesthesia occurs through higher than normal activation of multi-sensory neural binding in the parietal region of the brain, which is what determines how we view the world around us taking into account all of the sensory inputs going into our brains. Despite having distinct models to explain grapheme-colour synesthesia, it is likely that these models actually work in conjunction to produce synesthetic experiences which is supported by the fact that different forms of synesthesia may arise from the inherited trait.
3. Development
 |
| Figure 5. Consistency test averages of synesthetes (square), high-memory performing individuals (triangle), and the average child (circle) in percentages across three time frames. Both letters and numbers were tested. [15] |
Few studies have been done on the development of grapheme-colour synesthesia [14], where most studies focus on the mechanisms of already-developed synesthesia. One study conducted by Simner and Bain in 2013 endeavored to increase our knowledge on this topic. Their paper was the first study to conduct a randomly sampled longitudinal study on the development of grapheme-colour synesthesia in young children directly for over four years [15]. They tested colour consistency, one of the most common observable features of synesthesia [16], by asking the children to match each letter of the alphabet and the numbers 0 to 9 to 13 different colours. 615 children were tested in their previous 2009 study looking at children 6 or 7 years old in the first time frame, then tested the high performing children a year older in the second time frame to measure what they called delayed consistency [17]. In each time frame, they were asked to do the test twice with a 10 second break between to measure immediate consistency of colour choices. Finally in the 2013 study, they did the same test again with the same individuals now 10 or 11 years old, and compared the results to the two previous age brackets.
Of the 615 children aged 6 or 7, only 8 had immediate and delayed consistency test scores that indicated they were grapheme-colour synesthetes at the of the second time frame a year later[17]. When tested again four years later, 5 of the 8 children plus one of the previously categorized high-memory children (thus a total of 6 children) passed all of the consistency tests, making them almost certainly genuine synesthetes [15]. This shows that it is possible to have late development of synesthesia, or the fading of the experience as a child grows older. Furthermore, it was found that throughout the development of grapheme-colour synesthesia, children may have changing colour perceptions at younger ages, but then start to fix their perceptions as they get older at greatly varying rates. In this particular study, at 6/7 years of age, 1/3 of colour associations persisted [15]. At 7/8 years of age, it became 50% consistent, until finally the children averaged 70% consistency in colour associations at 10/11 years of age. Adults typically have 80-100% consistency, so fixation likely persists past the ages tested.
4. Synesthesia in non-English writing
Grapheme-colour synesthesia is not restricted to English graphemes only, and is a universal phenomenon experienced across languages and cultures [18].
4.1 Transfer to novel inducers
 |
| Figure 6. Reaction timing of congruent and incongruent graphemes before and after training of novel Glagolitic inducers. Reaction timings were significantly faster in congruent graphemes after training. [18] |
It has been shown that grapheme-colour experiences can be transferred to a different written language, such as the Glagolitic alphabet, after only 10 minutes of a simple repetitive writing task [18]. They tested this with a Stroop task, which presents the synesthete with a physically coloured Glagolitic grapheme that is either congruent or incongruent with the colour association they developed. It was found that synesthetes were able to report the colour of the actual grapheme more quickly when it was congruent with their synesthetic colour association, thus suggesting that synesthesia can indeed be transferred to a novel inducer through a semantic level of representation. This ability is likely due to the involvement of the lateral inferotemporal (IT) cortex involved in determining what is represented by the things you see [19].
4.2 Chinese characters
 |
| Figure 7. Grapheme-colour synesthetic experience of Chinese characters separated by radicals. |
Chinese is different from the English alphabet and the Glagolitic alphabet mentioned previously in that it consists of logographic characters instead of a phonetic alphabetic system. There is no Chinese equivalent of the letter 'A' in English, although there are individual number characters. Furthermore, Chinese characters cannot be sounded out by their strokes like English can be sounded out from its syllables. Although Chinese characters are not made up of letters, they can be dissected into radicals with different meanings, positions, and sometimes indicating pronunciations and tones of the compound character. Findings show that both the function and position of the radicals in a Chinese character play a role in determining colour associations, but pronunciations and tones do not have evidence of any role [20]. In Chinese, radicals containing some meaning usually appear on the left side of the character (about 90% of the time), which is associated with the saturation, and to a lesser extent, the hue of the colour association. Radicals on the right side vary greatly in how much area it takes up in the character, and is associated with the luminance of the colour association. Furthermore, compounds are seen to be dim and strongly saturated in colour. Thus, Chinese characters also elicit synesthetic experiences in a consistent, pattern-oriented fashion like a phonetic alphabet language, even though the writing systems are entirely distinct [21].
Japanese characters are another type of logographic language when looking at Kanji, which is based off of Chinese characters. However, the radicals in Kanji don't seem to have specific colour associations in Japanese synesthetes, perhaps because the Japanese language system usually makes use of Hiragana and Katagana which are not logographic character-based [22].
Hey, I just read your wiki and I think you did a really great job! It's really well written. I found your last section on grapheme-colour synesthesia in non-English writing especially interesting.
Thanks Michelle!
You have really explained the concept well. Going through your wiki, it is notable you have done intensive research with very careful planning of images on your page. I really agree with your thought: that hopefully by understanding this, we can come to a better understating of how human perception works :)
This is such an amazing read! I agree with Michelle! The last section was particularly interesting! Great organization with the text and the pictures as well. Good job!
The figures were used really well to compliment the data. I found it particularly interesting the way it was formatted as well in regards to the hyperlinks. Nice work!
Great choice of images/videos to supplement your topic. Your write-up seemed very concise as well as being very friendly to readers that are unfamiliar with the topic.
I really liked your wiki! It was really interesting and the images complimented the text really well.