Synesthesia

 

Learning Outcomes

1. What is synesthesia?

2. How common is it, and what is the etiology?

3. Describe the characteristics of synesthesia.

4. What is the evidence that synesthesia is a bottom-up sensory phenomenon (and a top-down conceptual one)?

5. How can this evidence be reconciled?

6. Describe the current major neural theories of synesthesia that explain it in terms of cross-activation, disinhibited neural feedback, or reduced neural plasticity.

7. What are the implications of synesthesia for multisensory perception; the binding problem; and metaphor, creativity, and language?

 


 

Introduction to Synesthesia

 

Definitions

- definition: a stimulus in one sensory modality evokes an additional experience in an _________ sensory modality (intermodal synesthesia) or in a different aspect of the same modality (intramodal synesthesia)

e.g., hearing a word spoken causes a particular flavour sensation

e.g., seeing an achromatic letter or number that also produces a colour experience

- etymology: “joined sensation,” from Greek syn = together + aisthesis = perception

- synesthesia links two normally independent ______: the subjective qualities of conscious experience (synesthesia has been called “qualia becoming deranged”)

- the stimulus that triggers the synesthesia is the inducer, and the modality in which the resulting synesthesia is experienced is the concurrent

- various types are identified using “inducer → concurrent”

e.g, grapheme → colour synesthesia

- there are over ___ different types of synesthesia, including grapheme → colour, tone → colour, and even (rarely) letter → gender

 

Diagnosis/Objective Tests

• Gustav Fechner (1871): surveyed coloured letter experience among 73 synesthetes

 

• Francis Galton (1880, 1881, 1883)

- described a synesthete who perceived dates as coloured

- claimed that the colours were derived from a _______ ____ used for childhood instruction

- also found that synesthesia ran in families

- suggested that both environmental and genetic components were responsible for synesthesia

 

• Mary Whiton Calkins (1894): first used the term “___________” in an English publication

 

• in the 20th century, psychology was dominated by the rise of behaviourism; thus, there was little interest in conscious phenomena--especially ones as bizarre and inexplicable as synesthesia

• due to the cognitive revolution, interest began to increase in the late 1970s and 1980s, and surged with the use of functional ____________ in the late 1990s and 2000s

graph of synesthesia publications by decade (Web of Science, 2025)

 

• Baron-Cohen et al. (1987; 1996): Test of Genuineness (TOG)

- developed ____ for spoken word → colour synesthesia

- participants are given different spoken word inducers, and choose a colour square that most closely matches the concurrent

- retested after several weeks or months, and ___________ is determined:

synesthetes score 70-90%

controls score 20-38%

- Baron-Cohen & colleagues (2006): Revised Test of Genuineness (TOG-R) uses greater range of standardized colours (Pantone Matching System®)

 

• Cytowic (2004): diagnostic criteria

1. automatic and ___________: occurs without any conscious effort

2. _________ extended: some synesthetes experience the concurrent in the same physical location as the inducer; for others, it floats around them in space

3. consistent (over time), elementary (not pictorial, but generic in quality: e.g., not an entire visual scene, but just a colour or a shape), and ________ (e.g., not “green,” but “spring leaf green” or “lime pale green”)

4. highly _________: aids in remembering stimuli (e.g., grapheme →  colour synesthesia helps spelling); synesthetes are more likely to have eidetic memory

5. loaded with ______: associations can be intensely aesthetically pleasing--or displeasing (e.g., “Derek” tastes of earwax)

 

• Eagleman et al. (2007):

- problem: some research groups simply use a self-report _____________ to determine if a person has synesthesia

- developed a single protocol for testing several types of synesthesia

- has a standardized scoring system

- The Synesthesia Battery is available online at www.synesthete.org

 

Prevalence

- 1 in 23 people has some type of synesthesia (Simner et al., 2006); equal in males and females (Simner & Carmichael, 2015)

- 1 in 90 has grapheme → colour synesthesia (Simner et al., 2005)

- most common types are grapheme → colour, and days of the week → colour synesthesia

- those who have one type of synesthesia are more likely to have another type as well (about 40-50% chance)

- synesthesia is 8 times more common in _______, poets, & novelists

e.g., famous artists with synesthesia: Vladimir Nabokov, David Hockney, and Wasily Kandinsky; also, Nobel Prize-winning physicist Richard Feynman

 

Etiology

__________: present from birth

- heritable: up to 40% of a synesthete’s first- and second-degree relatives also have it

- can skip generations

- family members with it do not necessarily have the same kind of synesthesia

- if family members do have the same kind of synesthesia, they do not necessarily have the same associations (e.g., mom sees 5 as green, daughter sees 5 as red)

____________: acquired after birth

- may result from stroke or closed-head trauma, temporal lobe epileptic seizures, or blindness, and can be permanent

- may be temporary (e.g., due to hallucinogenic drugs like LSD, meditation, or sensory deprivation)

 

Characteristics of Synesthesia

 

_____________: if two synesthetes both have grapheme → colour synesthesia, they are not likely to have the same response

e.g., one synesthete may see the letter X as white, but another may experience it as red

 

Bottom-up processing (Ramachandran & Hubbard, 2001a):

Expt. 1:

- constructed displays of randomly placed graphemes (5s), but other graphemes (2s) were placed in the display, forming a certain shape (e.g., a triangle)

visual search

- difficult for non-synesthetes to detect the shape (~60% correct); for grapheme → colour synesthetes, the shape popped out (~90% correct)

- ___-___ is due to early visual processing

 

Expt. 2:

- graphemes were placed at point of fixation on a screen and gradually moved outwards to the periphery

- the size of the grapheme was scaled up to ensure it was still identifiable

- beyond 11° of  eccentricity, the grapheme was clearly visible, but lost its ______

eccentricity

 

Expt. 3:

- individual graphemes presented in the periphery (< 11°) are easily identified

- however, when other letters flanked the target, it was difficult to identify

- these _______ graphemes still evoked the appropriate colour

flanking

 

- this process must take place at an early _______ level, rather than a higher cognitive level

Conclusion: synesthesia is a _______ sensory phenomenon?

 

Perception vs. imagination (Nunn et al., 2002):

- spoken word → color synesthetes placed in PET scanner

- spoken words caused activation in V4 (but not other visual areas)

- non-synesthetes scanned while they imagined colour: no activity in V4

- synesthesia is like __________, not imagining

 

______________ (Mills et al., 1999):

- participants given Stroop task: coloured numbers briefly shown to grapheme → colour synesthete

- task: name each colour as quickly as possible

- conditions:

• matched condition: number (e.g., 2) presented in actual colour (e.g., blue) of inducer → concurrent relationship (i.e., 2 → blue)

• mismatched condition (e.g., 2 shown in red)

• control condition: all numbers presented in black ink

- results: mismatched condition produced significantly ______ performance (25 s) than matched condition (15 s)

 

_____________ (Johnson et al., 2007):

- grapheme → colour synesthetes reported experiencing a one-way relationship

e.g., seeing a number would result in a colour experience

e.g., seeing a colour would not cause a number to be seen

- on Stroop task, colour-naming identification was slower in the mismatched condition

- however, in the mismatched condition, digit-naming was also slower

- this implies that synesthesia can be _____________

i.e., graphemes → colours, and colours→ graphemes

 

Top-down processing: attention (Ramachandran, 2000):

- created large graphemes (e.g., a 5) out of different smaller graphemes (e.g., 2s), called “_____ figures” (Navon, 1977)

- when grapheme → colour synesthete attended to figure’s component graphemes (i.e., the 2s), it was seen in the colour evoked by 2 (e.g., red)

Navon figure

 

- when figure attended as a global form (i.e., a 5), it was seen in the colour evoked by 5 (e.g., green)

- _________ is important for binding colours to forms

 

Top-down processing: _______ (Dixon et al.., 2006):

- synesthete shown ambiguous character in two different contexts:

black digits: 3 4 5 6 7

black letters: M U S I C

- the digit 5 and the letter S were exactly the same stimulus

digit 5 seen as pink

letter S seen as green

- top-down processing (i.e., _______ and meaning) affects synesthetic experience

 

Top-down processing: concepts (Dixon et al., 2000: “Five plus two equals ______”)

- tested grapheme → colour synesthete

- she saw 7 as yellow; identified 7 slower when it was printed in blue ink

- she was shown a series of slides, each containing one symbol

e.g., “5” then “+” then “2” then “=”

- final slide showed a colour patch

e.g., yellow or blue

- task: name colour as quickly as possible

- result: was over 250 ms slower in mismatch condition (blue patch) than in match (yellow)

- conclusion: synesthetic experience can be elicited by ________--not just by stimuli that are physically present

 

Categories of Synesthetic Experience

 

Ramachandran et al. (2003):

- proposed that number grapheme → colour synesthesia has two subtypes

- based on either perceptual or conceptual stimulus attributes:

_____ synesthetes: processing is primarily bottom-up

- thus colours are elicited by the actual visual appearance of a grapheme

e.g., 7 → yellow

 

______ synesthetes: experience is affected by top-down processing

- thus colours are elicited by the concepts conveyed by the stimuli

e.g., Arabic numerals, Roman numerals, and days of the week, which all represent ordinality

 

Dixon et al. (2004):

- proposed two subtypes, based on experience in external or internal space:

__________ strongly experience colours as overlays “projected” and bound to a grapheme, in the outside world

 

___________ experience weaker synesthetic colours “in the mind’s eye” only

- proposed that projectors = lower synesthetes, and associators = higher synesthetes; however, this is the subject of some dispute (see Ward et al., 2007)

 


 

Theories of Synesthesia

 

Early Theories

• people are just _____

- but this is not an explanation

• learned associations (e.g., coloured alphabet blocks or refrigerator magnets), or memories

- but color-blind synesthete (S.S.) with S-cone deficiency--which makes it hard to discriminate blues and purples--describes seeing numbers in “_______ colors,” meaning colors he is unable to see in the real world

• they are potheads and acid junkies who have been on drugs (compare synesthesia to ___ experience)

- but how is this experience explained?

• they are just using _________

- but what is a metaphor, in terms of the brain?

 

Current neural models: What causes breakdown in the modularity of perception?

 

1. Local Cross-Activation

(V.S. Ramachandran & Ed Hubbard, 2001a; Hubbard & Ramachandran, 2005)

- the gene in synesthesia fails to _____ the excess synapses that are present in all infants

- the result is cross-activation from a certain stimulus due to direct cross-wiring between different brain areas

cross-activation

 

- specifically, visual word form area (VWFA) in the ________ gyrus responds to visual letters, words, and numbers; this is adjacent to V4

hypothesis: in lower synesthetes, there is cross-activation between VWFA and V4

(this could explain grapheme → colour synesthesia)

- areas around the _______ gyrus (specifically, the temporal/parietal/occipital (TPO) junction) implicated in representing more abstract numerical ideas--concepts like ordinality

- anterior inferior temporal (AIT) cortex also found to encode conceptual representations of words, letters, and numbers

hypothesis: in higher synesthetes, there is cross-activation between areas around the angular gyrus and V4 or another higher colour-processing area in the ________ ________ gyrus

(this could explain month → colour synesthesia: each month has an ordinal value)

- the difference between lower and higher synesthesia results from _____ the genes for synesthesia are expressed

higher vs. lower synesthesia

 

Evidence:

- visual presentation of graphemes (e.g., ____) affects lower, not higher synesthetes

- accounts for the fact that upper/lowercase make no difference, because neither VWFA nor AIT respond differently to capitalization

 

Hubbard et al. (2006):

- visually identifiable graphemes presented at lower contrast (e.g., light grey on white background) failed to produce synesthetic _______

- VWFA less activated by low-contrast letters, more by high contrast

 

Rouw & Scholte (2007):

- used diffusion tensor imaging (DTI) to map white matter (tracts of myelinated axons)

• found increased cross-wiring in the right inferior temporal cortex (near VWFA and V4)

• additional anatomical connectivity in left posterior IPS (intraparietal sulcus thought to be involved in _______), and frontal lobes

 

General problems:

- what about adventitious synesthesia? (no new ___________ are created by hallucinogens, for example)

- why is synesthesia not apparent until mid-childhood? (should be present from birth, because the brain of a neonate is more highly interconnected)

 

 

2. Long-Range Disinhibited Feedback

(Peter Grossenbacher & Lovelace, 2001; Richard E. Cytowic & David Eagleman, 2009)

- posit that the degree of neural connectivity is identical in synesthetes and nonsynesthetes

- however, activity is usually ________ by equal excitation and inhibition

- synesthesia is due to (genetically determined) reduced inhibition in existing feedback pathways

disinhibited feedback   disinhibited feedback

 

- feedback may come from a ____________ nexus (like the TPO junction)

 

Evidence:

- accounts for ___-____ effects, because feedback comes from multisensory areas (“more connections” theory also accounts for this)

- acquired synesthesia may be due to brain damage, meditation, hallucinogenic drug use; cannot result from more connections

- LSD blocks receptors for _________ (which is primarily inhibitory); the resulting disinhibition allows neural targets to be more easily activated, hence synesthesia

- the visual cortex can take on a representation of tactile space (e.g., in the newly blind, or even in blindfolded sighted individuals after a few days): evidence for activation of ________ somatosensory-visual connections

 

General problems:

- adventitious synesthesia and LSD-induced hallucinations are often complex, including visualizations of animals and complex scenes, in contrast to the generic experiences of congenital synesthesia, which generally are simply of colour, for example

- also does not account for synesthesia emerging in mid-childhood

 

 

3. Reduced Plasticity:

(Julia Simner, Jamie Ward, et al., 2005)

- if a certain letter is seen in a certain colour, this will strengthen the connection between them

- because we see letters in various colours, this tends to average out

- but if plasticity is reduced, an initial letter-colour pairing may not ______

- accounts for emergence of synesthesia in mid-childhood

 

Evidence:

Ward & Simner (2003):

- JIW’s spoken word → taste (actually, flavour) synesthesia is likely to have originated during vocabulary acquisition

- ________ that trigger a certain taste also tend to appear in the name of the corresponding foodstuff

e.g. college → “sausage”, Sydney → “kidney”

- there is often a semantic association between triggering word and taste

e.g., blue → “inky”, bar→ “milk chocolate”

- synesthetic tastes are generally childhood foods, rather than foods first eaten in adulthood

- implies that synesthesia is not due to innate connections, but can be influenced by __________ knowledge

 

Simner et al. (2005):

- tested grapheme → colour synesthetes

- classified responses into 11 common colour terms

- found non-random patterns of colour associations with letters

e.g., 43% linked A → ___, 58% linked B → ____ or _____, 29% linked C → ______, 49% linked O → _____, and 43% linked Y → ______

- also, people who lack synesthesia showed some evidence of the same trends

e.g., 43% chose ___ for A, 47% chose ______ for Y, and 49% chose ____ for B

- high-frequency letters are generally associated with high frequency colours

- the specific grapheme-colour links in synesthesia are not arbitrary--perhaps due to the orderly topographic mapping of many cortical areas

 

Witthoft & Winawer (2013)

- 11 grapheme → colour synesthetes evaluated

- colour-letter associations closely matched those in Fisher-Price® magnetic letter sets, which were owned by 10 of the synesthetes

- colours matched closely to at least 14 of the 26 letters in the set--a probability of less than 1 in 1 _______ (one synesthete matched to all 26 letters)

- follow-up found strong prevalence of matches to toy colours in 400 synesthetes who were children when the toy was available (Witthoft et al., 2015)

- in addition to genetic and perceptual factors, synesthesia must also involve ________ and memory

 

General problems:

- unclear how this theory accounts for other forms of synesthesia (e.g., day of the week → gender)

- does not account for adventitious synesthesia

 


 

Implications of Synesthesia

 

• “synesthesia may be the very first perceptual condition for which science can map its ____” (Cytowic, 2009, para. 11)

 

• synesthesia may not be ______; interaction across the senses is actually the rule rather than the exception (Ward, 2008)

e.g., watching a movie binds vision + audition

- ____________ perception (a unitary experience resulting from combined stimulation of more than one sensory system) occurs in everyone

- in a food, for example, sensory attributes tend to reliably co-occur

e.g., flavour is comprised of taste, smell, touch, temperature, sight, sound, and sometimes even pain/irritation

e.g., a strawberry smells like strawberries, tastes sweet, feels soft, etc.

- why do some odourants smell “sweet”?

- is flavour the result of olfaction and gustation failing to develop as fully segregated, independent senses?

- in other words, is flavour a universal type of synesthesia found in ________?

- this goes against the notion of strict modularity in the brain (Doctrine of Specific Nerve Energies)

 

• synesthesia may help us understand the binding problem: how are separate aspects of a single stimulus (e.g., colour, form, motion, etc.) reintegrated into a coherent, holistic percept?

- synesthetic experiences are not __________ but are synchronized and separate

- binding of qualia may be accomplished by coordinating subsystems using feedback connections, which requires interconnections between differently specialized areas, resulting in a holistic percept

- a “____________” model suggests that synesthesia arises through an overactivation of parietal binding mechanisms

- applying TMS to parietal binding regions disrupts synesthesia (Esterman et al., 2006)

(hyperbinding may use one of the above neural models--cross-activation or disinhibited feedback--to create simultaneously active areas which are anomalously bound together)

 

• synesthesia may reveal the ____________ basis of metaphor, creativity, and language (Ramachandran, 2004)

- metaphor:

▸ can be defined as “seeing the similar in the dissimilar”

▸ metaphor is an important potential component of creativity

▸ perhaps this makes the synesthesia gene adaptive (synesthesia is more common in artists)

• damage to the left angular gyrus (implicated in higher synesthesia) leaves patients with an impaired understanding of the dual nature of _________ (e.g., “all that glitters is not gold” is treated literally)

- creativity:

▸ “______” synesthesia allows novel connections to be made

▸ emphasizes the subjective nature of individual experience

- language:

▸ in both English and Tamil speakers, 98% named a rounded shape “_____” and a jagged one “____”; also shown by children as young as 2½ (Ramachandran & Hubbard, 2001b; Köhler, 1929)

▸ this effect is not obtained in patients with damage to the left angular gyrus

▸ implies that objects are not named arbitrarily, but are due to synesthesia-like mapping (called “_____ _________” in linguistics)

▸ thus, a type of synesthesia that we all share may be part of the basis for language