Acquired and inherited forms of cross-modal correspondence

Coloured hearing synaesthesia (from the Greek syn [union] and aisthesls [sensation]), has been known to the scientific community for over 300 years and yet has gone relatively unlnvestigated. In this revlew we chart recent research Into the various forms of synaesthesia and contrast accounts of acquired and developmental forms of the condition. We also revlew the competing theories proposed to account for developmental forms, and conclude that preserved neonatal neural pathways between brain areas subserving audition and vision provide a likely explanation of the Condition. Cross-modal correspondences In a recent issue of Neurocase, Halligan et al. (1996) report the case of an individual (DN) with left hemianaesthesia, who is seemingly only able to report tactile stimulation when allowed to observe the stimulus being applied. Halligan and colleagues suggest two possible explanations of this phenomenon, the first of which proposes that incoming visual information raises the stimulus to suprathreshold, whereas purely tactile stimulation fails to reach the threshold for detection. This may be analogous to the circumstances reported by Valler et al. (1993) in which the detection of tactile stimulation in patients with hemianaesthesia can be promoted by the introduction of water into the ear (caloric imgation). Their second hypothesis to explain the facilitation of tactile stimulus detection through vision is that their subject has acquired a form of synaesthesia in which visual information gives rise to a sensation of touch. This second explanation begs the question of what synaesthesia might be. In the following review we give an account of recent research with individuals with apparently inherited (developmental) forms of synaesthesia and contrast these findings with studies that have investigated acquired forms. Definitions of synaesthesia We, along with others (Marks, 1975; Cytowic, 1989a,b; Motluk, 1994) have chosen to define synaesthesia as occurring when stimulation of one sensory modality automatically triggers a perception in a second and in the absence of any direct stimulation to the latter. So, for example, a sound might automatically and instantly trigger the perception of vivid colour, or vice versa. Many combinations of synaesthesia are reported to occur naturally, including sound giving rise to visual percepts ‘coloured-hearing’ and smell giving rise to tactile sensation, as in Cytowic’s (1993) subject MW. Our experience suggests that coloured-hearing synaesthesia is by far the most common form and that certain combinations of synaesthesia almost never occur (e.g. touch to hearing). Acquired synaesthesia A variety of neuropathological conditions have apparently given rise to acquired synaesthesias. In one of the earliest accounts Carnaz (1851) [cited in Krohn (1893) p. 331 speculated that synaesthesia of all forms was ‘pathological and due to some optical lesion’ and could therefore be seen as due to ‘hyperaesthesia of the sense of colour’. Whilst Correspondence to: J. Harrison D ow nl oa de d by [ U ni ve rs ity o f C am br id ge ] at 0 2: 19 2 3 M ar ch 2 01 6 246 J . Harrison and S. Baron-Cohen being an unlikely explanation for all forms of synaesthesia, a number of more recent studies have reported instances of synaesthesia as a consequence of visual loss involving anterior visual pathways. For example, Lessell and Cohen ( 1979) reported three patients with unilateral eye or optic nerve disease who saw phosphenes (patches of light) only when they heard noises. Bender et al. (1982) also reported a patient with optical neuritis who saw a blue light whenever he heard a loud noise or experienced sudden pain. Additionally, Vike et al. (1984) report a man who when auditorily stimulated with 65 dB clicks saw kaleidoscopic and spiralling lights in his left eye. On removal of a large cystic mass extending from his left medial temporal region to the midbrain, his synaesthesia stopped. Rather more complex synaesthesia was reported by a subject studied by Rizzo and Eslinger ( 1 989). This 17-year-old male, who had suffered perinatal visual deprivation as a consequence of retrolental fibroplasia, showed unusual associations between musical notes and coloured shapes. Perhaps one of the most extraordinary cases of acquired synaesthesia is that reported by Coriat (1913). The subject of report was said to experience coloured pain as the consequence of an attack of ‘hysterical hemicrania’ (possibly a description of migraine). Thus i t seems that most cases of acquired synaesthesia arise as a consequence of damage to anterior portions of the brain, often the optic nerve. In acquired forms the synaesthesia is usually little more than photisms or phosphenes, and typically lacks the complexities found in developmental synaesthesia. For example, subject EP, a developmental case studied by Baron-Cohen and colleagues ( 1987), reported seeing ‘Darkish grey, with spinach green and pale blue’ on hearing the word ‘Moscow’. This description is markedly more complex than the descripLions given in the studies of subjects with acquired synaesthesia who tend to report simple, one-colour flashes of light. Since testing EP we have come across other developmental cases in which equally rich colour experiences are triggered by pain, touch and taste. In the following section we discuss some other characteristics of developmental synaesthesia which distinguish it from acquired forms. Developmental synaesthesia Developmental forms of synaesthesia often prompt scepticism in the minds of many neuroscientists. Given that the primary data are usually the phenomenological accounts given by synaesthetes, such doubt seems understandable. Establishing the existence of synaesthesia has been dependent upon the presence of ‘symptoms’. However. the remarkable similarity in the accounts given by individuals with synaesthesia about their experience (see Baron-Cohen rf al., 1993; Harrison and Baron-Cohen, 1994. 1995) suggests that the condition is real and deserving of serious scientific attention. As Korb ( 1996) has recently observed: “The probability that some large number of people, independently reporting that they have synaesthesia, would go on to report such a close match in their phenomenologies and whose reported cross-modal linkages would be close to invariant after such long periods without there being a common underlying physical structure (innate o r acquired) to explain it is vanishingly small.” Korb’s observation testifies to the existence of the condition, but relies upon the phenomenological accounts given by subjects. Recent investigations of the developmental form of the condition have sought to go beyond the phenomenological accounts and to obtain objective data. However, it is worth remarking at this point that reports ol’ acquired forms do not seem to have been greeted with the same scepticism as developmental synaesthesia. This seems remarkable given that the data in both instance\ have been the subjective accounts given by the subjects studied. Admittedly, patients with acquired forms of synaesthesia have a demonstrable lesion, though, as we discuss later in this review, it is by no means apparent that the acquired forms can be ‘explained’ by the presence ot’ brain damage. Psychological investigations of developmental synaesthesia Recent investigations of developmental synaesthesia have aimed to provide evidence more analogous to ‘signs’ than ‘symptoms’ of the condition. The study of subject EP (Baron-Cohen et al., 1987) capitalized on the fact that for individuals with coloured-hearing synaesthesia, the colourword or colour-sound correspondence is consistent and invariant across time. In order to provide objective evidence of the condition, Baron-Cohen and colleagues made detailed notes of subject EP’s colour descriptions for over 100 words and letters. Ten weeks later, and without prior warning, EP was retested on the entire list and a comparison made between the descriptions given on each of the two occasions. Her performance was found to be totally consistent, a marked contrast to the performance of a control subject who was able to manage only 17%) consistency one week later, in spite of having been prewarned of a retest. A later group study of a further nine synaesthetes using a paradigm similar to that used with subject EP also found remarkably good performance in synaesthetes a s compared with a group of controls. However, even such striking results as these could have been due to superior recall memories, though neuropsychological testing suggests that memory performance in these subjects is unremarkable, falling well within the normal range. These two studies provided the first objective evidence in support o f the existence of developmental synaesthesia. D ow nl oa de d by [ U ni ve rs ity o f C am br id ge ] at 0 2: 19 2 3 M ar ch 2 01 6 Acquired and inherited forms of cross-modal correspondence 247 Functional neuroimaging studies of developmental synaesthesia Recent developments in neuroimaging have allowed us to examine the human brain in vivo in a bid to uncover the functional anatomy of synaesthesia. Paulesu et af . (1 995) have recently used Positron Emission Tomography (PET) to look for differences in regional Cerebral Blood Flow (rCBF) in synaesthesic subjects as compared with nonsynaesthesic controls. This study showed that subjects with synaesthesia exhibited increased rCBF, as compared with control subjects, in two visual association areas (posterior infero-temporal cortex and the junction of the occipital and parietal cortices) when listening to words as compared to listening to tones (all subjects were blindfolded when being scanned). This study and a subsequent single case replication using functional magnetic resonance imaging (MRI) suggest that posterior cortical structures are involved in developmental synaesthesia. This provides a further difference between acquired and developmental forms of the condition, in that it is generally damage to anterior pathways that gives rise to the former. Whilst these studies testify to the existence of the condition, they are mute regarding the causes. There has been no shortage of theories proposed to explain synaesthesia and in the following section we review the major accounts.