The linguistic and cognitive factors associated with lexical-gustatory synesthesia: A case study

Brain and Cognition 106 (2016) 23–32

Contents lists available at ScienceDirect

Brain and Cognition journal homepage: www.elsevier.com/locate/b&c

The linguistic and cognitive factors associated with lexical-gustatory synesthesia: A case study Vijayachandra Ramachandra Language and Cognition Lab, Department of Communication Sciences and Disorders, Marywood University, Scranton, PA 18509, USA

a r t i c l e

i n f o

Article history: Received 5 June 2015 Revised 23 December 2015 Accepted 15 April 2016

Keywords: Memory Metamemory Paired-associate learning Judgment-of-learning Word–taste synesthesia

a b s t r a c t Past research on lexical-gustatory synesthesia (people who associate words with tastes) has shown that linguistic factors underlie the inducer (word)–concurrent (taste) mappings in these individuals. The developmental cognitive model envisioned by Simner and Haywood (2009), and an extension of it proposed in this paper can be used to understand these linguistic associations. The first aim of the current case study was to test these models by examining the linguistic factors associated with this type of synesthesia. The influence of lexical-gustatory synesthesia on memory was investigated using a pairedassociate learning task. Further, the present case study also examined if word-pairs with same or similar tastes and word-pairs with dissimilar tastes enhanced learning relative to word pairs that had no tastes. As predicted, the findings revealed possible phonological, phonological–lexical, and lexical–semantic factors linking the inducer–concurrent pairings. These findings are in line with the developmental cognitive models of LG synesthesia. There were no effects of synesthesia on memory as demonstrated by the lack of any significant difference between the synesthete and non-synesthete controls on a paired-associate learning task. Moreover, no significant differences emerged between the ‘‘no taste” and ‘‘taste” conditions (although she performed slightly better on the ‘‘no taste” condition). Interestingly, a metamemory task (judgment-of-learning) revealed the opposite. That is, the synesthete predicted that her learning would be better in the ‘‘taste” condition when compared to the ‘‘no taste” condition. This indicates that the word-pairs, which produced tastes, could have created a ‘‘foresight bias”. This is attributed to the unidirectional nature of this individual’s LG synesthesia. This finding should, however, be treated with caution because it is a preliminary finding based on a single subject and needs to be corroborated with future studies on other lexical-gustatory synesthetes. Ó 2016 Elsevier Inc. All rights reserved.

1. Introduction Synesthesia refers to a condition in which otherwise normal individuals involuntarily experience sensation in a certain unstimulated sensory modality when some other modality is stimulated. The modality in which the stimulation occurs is called the ‘inducer’ and the modality in which the sensation is experienced is referred to as the ‘concurrent’ (Grossenbacher & Lovelace, 2001). For example, a person with grapheme-color synesthesia may see the color ‘‘red” when presented with the letter L or the number 5. Lexicalgustatory (LG) synesthesia, which is the focus of the present study, is a very rare form of synesthesia in which individuals typically experience tastes in their mouth (sometimes smell odors) when they see or hear words. For example, the LG synesthete who participated in the current study tastes ‘‘cherry coke” when presented (reading or hearing) with the word chair, and when presented with E-mail addresses: marywood.edu

[email protected],

http://dx.doi.org/10.1016/j.bandc.2016.04.005 0278-2626/Ó 2016 Elsevier Inc. All rights reserved.

ramachandra@maryu.

the proper name Ian, she experiences the taste of ‘‘bread with marinara sauce and locatelli cheese”. Synesthesia was first clearly described by Sir Francis Galton in 1880. He reported the experiences of several individuals who visualized numbers as being represented in a fixed order in space. Some individuals felt that the numbers had certain shapes and these shapes in turn had specific colors. He described these experiences as idiosyncratic, familial, and varied from one individual to another (Galton, 1880). Synesthesia is not listed in the DSM or ICD classification systems as it generally does not adversely affect the daily activities of a person having this neurological condition (Hubbard, 2007). It runs in families and it could be polygenic in origin (Brang & Ramachandran, 2011). 1.1. Synesthesia related to taste There have been several reports in synesthesia literature where taste has been the ‘inducer’ rather than the ‘concurrent’ (e.g. Cytowic, 1993; Downey, 1911; Myers, 1911). There are 3 published

24

V. Ramachandra / Brain and Cognition 106 (2016) 23–32

studies in historical literature that have described individuals who experience taste (concurrent) when some other sensory modality is stimulated (Ferrari, 1907, 1910; Pierce, 1907). Eleven papers on LG synesthesia have been published in the contemporary literature (Bankieris & Simner, 2014; Cytowic, 1989; Gendle, 2007; Gray & Simner, 2015; Jones et al., 2011; Richer, Beaufils, & Poirier, 2011; Simner & Haywood, 2009; Simner & Logie, 2007; Simner & Ward, 2006; Ward & Simner, 2003; Ward, Simner, & Auyeung, 2005). Some of these studies, however, have been conducted on the same subjects, which is an indication of the rarity of this form of synesthesia. More recently, several studies on LG synesthesia have revealed the influence of linguistic and conceptual factors on synesthetic experiences. Studies on LG synesthesia have clearly shown the connection between phonology of the words and the tastes triggered by them. For example, words such as prince and cinema elicit the taste of ‘‘mince” (the critical phonemes in this word–taste association are /I/, /n/, and /s/). Orthography does not seem to have much effect on the taste experience. For example, /g/ resulted in the taste of ‘‘yogurt” not only in words with a phoneme and grapheme match such as ‘‘g” in begin but also when there was a mismatch as in ‘‘x” in exactly. The studies also showed that ‘‘fine-grained phonological distinctions” were important. For instance, the syllabic dark /l/ (as in bottle) produced the taste of ‘‘Rice Krispies” while the nonsyllabic dark /l/ (as in deal) resulted in the taste of ‘‘fingernails”. The word–taste associations are not always phonological in nature. The words post, past, coast, most and must led to the taste of ‘‘toast” but the words host and boast on the other hand which are phonologically similar to the words that induced ‘‘toast” resulted in totally different tastes. Further investigation has clearly demonstrated the role of learned lexical knowledge in synesthetic experiences. For example, the word blue tasted ‘‘inky” and the word bar tasted like ‘‘milk chocolate” (see Ward & Simner, 2003 for a detailed description of their case JIW). The lexical influence is also evident from findings that suggest that tastes are triggered by high frequency rather than low frequency words, and by real words rather than nonwords. The frequency of words is also important in determining the intensity of the synesthesia experience (Gendle, 2007; Simner & Haywood, 2009; Ward et al., 2005). Functional brain imaging studies have shown that the brain activation in individuals with LG synesthesia is different from that of non-synesthetes. The insula is highly active in response to the valence (pleasant/unpleasantness) of the tastes, and the medial parietal regions are active in response to the intensity of these experiences (Jones et al., 2011). It has also been noted that the primary gustatory cortex is activated in response to listening to words but not tones in a person with LG synesthesia (cited in Ward et al., 2005).

association is not a simple and straightforward ‘‘merging of the senses” as defined in the synesthesia literature (Simner, 2012). Like most types, LG synesthesia is also not a result of a direct association between two sensory centers in the brain (crosstalk). It is influenced by linguistic factors such as phonology and lexical– semantics (Gendle, 2007; Ward & Simner, 2003; Ward et al., 2005). Simner and Haywood (2009) have used a cognitive model to explain these complex word–taste associations. They have proposed that the word–taste mappings may not be formed during infancy but may occur during early childhood as a consequence of language acquisition. At the early stages of LG synesthesia development, food words get mapped onto the lemma (abstract concept associated with words) of food names and this in turn is linked to gustatory sensations (see Fig. 1). This can explain some of the direct lexical connections between words and tastes (chocolate evokes the taste of chocolate/chocolate milk shake). During later stages, words that are phonological neighbors of food words may start eliciting gustatory sensations by activating the lemmas associated with these food words. This can explain some of the phonological–lexical associations between inducers and concurrents (for example, tickle can elicit the taste of ‘‘pickle” by activating the food word ‘‘pickle” (or the lemma associated with it)). By extending the model proposed by Simner and Haywood (2009), it is reasonable to suggest that during the later stages of synesthesia development, words that share other linguistic associations (e.g. phonological and indirect lexical–semantic or conceptual) with food items produce gustatory sensations by triggering the lemma associated with those food items (see Fig. 2). For example, words that share certain phonemes with a food word can produce the taste of that food item (stroll producing the taste of strudel) and words that are conceptually linked to a food item can elicit the taste of that food (pantene producing the taste of vanilla pudding perhaps because of their similarity in consistency). Words that evoke thoughts

gustatory sensation

lemma of food names

1.2. Purpose of the study Synesthesia is not purely a sensory-perceptual phenomenon; higher-level cognitive processes are involved in these experiences. For example, variations of a letter (change in font, size, etc.) do not change the resultant colors experienced by most grapheme-color synesthetes (Simner, 2012). The components of language such as numbers, letters, days, and months are associated with different personality types and gender by some individuals with synesthesia (Simner & Holenstein, 2007). Interestingly, one of the earliest publications on synesthesia by Galton described a man who associated different numbers with different personifications. To this individual, the number 7 was ‘‘masculine”, 6 was of ‘‘no particular gender but gentle and straightforward”, 2 was ‘‘young and sprightly”, etc. (see Galton, 1880 for a detailed description). These are just a few illustrations to indicate that in most cases of synesthesia, the

food words

Fig. 1. Lexical-gustatory synesthesia acquisition during the initial stages based on the cognitive model envisioned by Simner and Haywood (2009).

25

V. Ramachandra / Brain and Cognition 106 (2016) 23–32

words with shared phonemes (phonological) words with indirect semantic links (indirect lexicalsemantic)

Life experiences lemma of food names

words that sound similar (phonological- lexical)

words with direct semantic links (direct lexical-semantic) gustatory sensation

Fig. 2. An extension of the lexical-gustatory model by Simner and Haywood (2009) to account for the word–taste associations experienced by LG synesthetes.

related to food items because of life experiences can also lead to gustatory sensations by activating the food lemmas. LG synesthesia is heterogeneous and idiosyncratic in nature. No two people with this type of synesthesia are exactly the same in terms of their inducer–concurrent pairings (Gendle, 2007; Ward et al., 2005). People with LG synesthesia, however, are similar in that their word–taste associations can be traced to phonological and/or lexical factors but are rarely related to orthographic aspects of words (Simner, 2007). It is important to re-examine the phonological and lexical–semantic factors in this rare form of synesthesia. The first aim of the current study is to test Simner and Haywood’s model and the extended developmental model of synesthesia that is proposed here by examining the linguistic factors underlying the LG synesthesia experienced by the synesthete, HG who participated in the current study. It is hypothesized that like the previous investigations by Ward and colleagues, HG’s word–taste mappings can also be explained by phonological, phonological–lexical, and lexical–semantic factors. In their case study on JIW, Ward and Simner (2003) identified the pattern of association between the phonemes in the inducers and the tastes that resulted. They selected the tastes produced by three or more different words from a large corpus of word–taste mappings and calculated the probability of certain phonemes triggering specific tastes with a consistency above chance levels. For example, from their analysis they found that the presence of the phoneme /m/ in a word would elicit the taste of ‘‘cake” for JIW and this association was above chance level expectations. Such detailed analysis was not possible for the current study because we did not have a large corpus of word–taste mappings for HG. There were only 14 tastes produced by two or more different words, and out of these 14 tastes only three were triggered by three different words. As well, there were some words that produced similar tastes for HG. The paper aimed at determining the common phonological/lexical links between the inducers that produced the same tastes as well as similar tastes and their resultant concurrents. This will further our understanding of the linguistic associations between inducers and their concurrents in people with LG synesthesia.

Several single subject case studies have shown that synesthesia can have a positive effect on memory (Baron-Cohen et al., 2007; Luria, 1968; Mills, Innis, Westendorf, Owsianiecki, & McDonald, 2006; Smilek, Dixon, Cudahy, & Merikle, 2002). Several group studies have also shown some positive effects on memory. These investigations have been conducted mostly on grapheme-color synesthetes and enhanced memory has been reported in the domain of both inducers (graphemes) as well as concurrents (colors) (for example, Gross, Neargarder, Caldwell-Harris, & Cronin-Golomb, 2011; Rothen & Meier, 2009; Yaro & Ward, 2007). Time–space synesthesia is the only other type where the effect of synesthesia on memory has been investigated. These studies have reported enhanced public memory (remembering dates of public events) and autobiographical memory in people having this type of synesthesia (Parker, Cahill, & McGaugh, 2006; Simner, Mayo, & Spiller, 2009). Memory in people with LG synesthesia has not been systematically investigated. There are two studies, however, that have reported the performance of LG synesthetes on a free-association task. TD, the LG synesthete described by Gendle (2007) performed superior to controls on a free-association task where she had to make taste associations for words that did not elicit any synesthetic experiences; this task was repeated after a gap of 14 days and TD showed a high level of consistency in these word–taste mappings when compared to a control group. The author concluded that TD’s superior ability to form cross-modal associations could possibly underlie her synesthesia. TD reported, however, that she used her synesthetic experience as a ‘‘mnemonic aid”. So, it is plausible that TD was using mnemonic strategies based on her perceptual experiences to form novel associations. Such superior performance on a similar free-association task was, however, not seen in the case of JIW, the LG synesthete studied in great detail by Ward and Simner (2003). The current study is a preliminary attempt to explore the effects of LG synesthesia on memory by using a paired-associate learning task. A paired-associate learning task is representative of how learning new information occurs in the everyday life of a person where words and other concepts are acquired in a particular context. Further, this study aimed to

26

V. Ramachandra / Brain and Cognition 106 (2016) 23–32

examine if word pairs producing same or similar tastes, and word pairs with dissimilar tastes resulted in better learning when compared to words pairs, which induced no tastes for HG (a baseline control condition). This will shed some light on how words are learned by a person with LG synesthesia in everyday context where tastes can serve as cues to remember and learn new words. In a paired-associate task, the first word in the pair serves as a cue to remember the second word in that pair. Overall, the memory for the target word (the second word to-be-remembered) in word pairs generating tastes should be better than the ones with no taste because the taste of the first word may serve as a cue to remember the second (target) word (taste > no taste). In case of words generating same/similar tastes, the taste evoked by the first word should act as a strong cue to remember the second word in that pair when compared to word pairs producing dissimilar tastes. Therefore, the present research predicts that HG would remember the second (target) word in the word pairs in the following order: same/similar > dissimilar > no taste. In addition, HG’s ability to monitor her own learning (metamemory) was also examined using a judgment-of-learning (JOL) task, which refers to making judgments on future recallability of learned items prior to the actual recall of those items. A mismatch between JOL and actual performance indicates that the person is not able to monitor his/her own learning (Nelson & Narens, 1990). Overall, the JOLs may be higher for words producing tastes when compared to the ones not producing any tastes because the taste associated with the first word (cue) may aid in remembering the second word (target). Therefore, the present paper predicts: JOL taste > JOL no taste. Previous investigations have shown that JOLs for paired-associate learning are much higher for related word pairs when compared to unrelated word pairs (Mueller, Tauber, & Dunlosky, 2013). Based on this finding, the current research predicts HG’s JOL scores in the following order: same/similar > dissimilar > no taste. Please note that although the word pairs were semantically unrelated, the relatedness in terms of tastes could lead to such JOLs described above. 2. Methods 2.1. Case history HG is an ambidextrous female (writes with the right hand but uses the left hand for many activities such as pouring coffee and opening doors), and a native speaker of American English. She was a 27-year-old health care professional working in a hospital at the time the current study was conducted. She has experienced LG synesthesia her entire life but realized that it was a sensory condition only in high school. She is not aware of anyone else in her family with this condition but mentioned that one of her younger cousins sometimes said that she didn’t like certain names because they tasted like ‘‘throw up”. When HG was 5 or 6 years old, she began noticing that she associated names and words with food. She told this to her family but they just acknowledged this without delving deeper into the matter. When she was in high school, people came to know that she could link names of people with food items/tastes. It became a ‘‘party trick” and they would ask her to do such associations whenever people got together. Sometimes, even strangers would come up to her and ask-‘‘What does my name remind you of”? People were amazed by the fact that she could come up with the same food item/taste in response to their names every time they met her. She could make these name-taste associations even after a gap of several months. Everyone (including her) thought that she had some kind of ‘‘superpower”. When she was in high school she read an article about a girl with

synesthesia and realized that her condition was very similar to the girl in the story. She researched on the topic and discovered that she indeed had a neurological condition called synesthesia. She tastes most words but some words do not induce any taste or strong enough taste to really identify them. She experiences these tastes while hearing speech as well as reading. She revealed that the intensity of her synesthetic experience was sometimes dependent on how hungry she was. She recalls vividly that several years ago she was sick and had not eaten for days. She would lie in bed and watch TV, and every word that was said on TV produced an intense taste in her mouth. She became a vegetarian a few years ago and the meat tastes associated with certain words have been diminishing in intensity although she still experiences them. Some words are not associated with just tastes but are linked to textures and or/smells as well. At times, the tastes in response to words are of different temperatures and thickness. The intensity of her experience varies across words and she sometimes has a negative taste (e.g. chunky vomit) associated with certain words. She also visualizes alphabet, TV channels/number lines, and days of the week/months as being arranged in a specific pattern in space. Her tastes are very detailed. Examples: the proper name Joshua tastes like ‘‘Peanut butter and jelly sandwich (slightly soggy)”, the word Handicap produced the taste of ‘‘French fries (crispy and salty) with ketchup” and Amanda elicited the taste of ‘‘rare roast beef hoagie with American cheese and mayo”. 2.1.1. Consistency test Consistency is the best test to establish the genuineness of synesthesia (Baron-Cohen, Burt, Smith-Laittan, Harrison, & Bolton, 1996; Simner & Ward, 2006; Ward & Simner, 2003). During the first phase of the study, HG was given a list of 100 words and asked to provide the tastes that they elicited. Only 77 of these words produced a taste for her. She was also asked to compile a list of word–taste associations based on her everyday life experiences (this list was completed and given to the experimenter during the second phase of the study). Finally, she completed a detailed synesthesia questionnaire (www. syn.sussex.ac.uk/syn_questionnaire.doc). During the second phase of the study (after a gap of 3 months), she was asked to provide tastes for 80 words taken randomly from the initial 100-word list and the synesthesia questionnaire (only words that resulted in a taste were selected). HG had no knowledge that she would be re-tested. A control group (n = 10) of non-synesthetes (mean age = 24.7; SD = 2.75) was given the same list of 80 words and asked to free-associate the taste of a food or a drink for each of the words. This group was re-tested after a gap of only 1 week (they were aware of the retesting). The percentage of consistency of responses between the two presentations was calculated for both HG and the controls. 2.1.2. Influence of lexical and phonological factors A total of 221 word–taste pairs were analyzed to determine the lexical and phonological factors associated with LG synesthesia. These pairs were mostly taken from the list that HG had compiled based on her day-to-day experiences but also included pairs from the consistency test and the questionnaire. The association between the target words (inducers) and the resultant tastes (concurrents) was classified into 5 types: Phonological, Phonological–lexical, Direct lexical–semantic, Indirect lexical–semantic, and Unidentifiable. This analysis was conducted by two independent speech-language pathology graduate students trained in International Phonetic Alphabet (IPA) and an inter-rater agreement between them was calculated. Phonological: When common phonemes are shared between the inducers and the concurrents.

27

V. Ramachandra / Brain and Cognition 106 (2016) 23–32

Please see Table 1 for examples of the classification system used in the current study. To further confirm the phonological links between the words and the tastes they elicited, phoneme co-occurrence scores (PCC) were calculated for the list of word–taste pairs that had a possible phonological association (Ward & Simner, 2003; Ward et al., 2005). This was done by converting both the word and the taste into IPA and then counting the number of phonemes that were common to both the word and the taste it evoked. For example, the word stroll elicited the taste of ‘‘strudel” for HG; they both were converted into IPA and the common phonemes shared between them gave the PCC score. In this case, /strol/ ? /strudel/, the PCC score was ‘‘4” because /s/, /t/, /r/, and /l/ were the 4 phonemes shared between the inducer and the concurrent. This score was compared with a PCC score calculated for a control list, which was created by pairing each word with a random taste taken from the same list of word– taste mappings with a phonological association. These random pairings were generated using a list randomizer (www.random. org/lists). For example, ‘‘stroll” was randomly paired with ‘‘cream” which was the taste produced for the word ‘‘pretend”. The PCC score for this association was ‘‘1” as there was only 1 common phoneme between ‘‘stroll” and ‘‘cream” (the sound /r/). A few word–taste mappings were removed from this analysis because of the complex nature of their relationship. For example, the word suggestions (inducer) resulted in the taste of ‘‘tuna fish sandwich” (concurrent). Here, the syllable /sa/ in the inducer could be associated with the /sa/ in ‘‘sandwich”, the sounds /ʤ/ and /tʃ/ in the inducer could be related to the /ʃ/ in ‘‘fish” (/ʤ/, /tʃ/ and /ʃ/ belong to the category of sibilants) and to the /tʃ/ in ‘‘sandwich”. Such mappings with complex associations where different parts of the inducer were connected to different words in the concurrent were eliminated from the analysis because it is difficult to determine which phoneme in the inducer is actually producing the concurrent taste. Is /sa/ in the inducer producing the resultant taste or the phonemes /ʤ/ and /tʃ/ in the inducer producing the taste? Words that produced the tastes of brand names such as ‘‘Ruffles potato chips” and ‘‘Reese’s peanut butter cups” were also removed from this analysis (see Ward & Simner, 2003). There was a total of

Table 1 The classification system used to determine the word–taste associations for HG. Linguistic category

Examples of word taste associations

Phonological

ride ? rye bread with cucumber and hummus (/raɪ/ is the critical phonological segment) tickle ? pickle

Phonological– lexical Direct lexical– semantic Indirect lexical– semantic Unidentifiable

shake ? chocolate milk shake Pantene ? vanilla pudding (both have same consistency?) could not be classified into any of the categories above

85 word pairs with possible phonological associations but only 77 word–taste mappings were included for the final analysis.

2.1.3. Common linguistic associations between words producing the same as well as similar tastes The current paper aimed at determining the common phonological/lexical links between the inducers that produced the same as well as similar tastes and their concurrents. The association between these inducers and the concurrents and the probable phonological/lexical links are described in Appendices A (same tastes) and B (similar tastes).

2.1.4. Paired-associate learning HG was randomly presented with 45 word pairs (15 pairs which induced no taste, 15 that induced dissimilar taste, and 15 that produced either similar or same tastes). The presentation of each word pair was followed by a judgment-of-learning (JOL) task where HG was shown the first word of the word-pair and was asked‘‘how confident are you that in 10 min from now you will be able to recall the second word of the item when prompted with the first one?” The JOL ratings ranged from 0% to 100% for each word pair. Following the presentation and JOL ratings for all 45 pairs, she was tested on the paired-associate learning task by presenting her with the first word from the pair and she had to write the second word associated with that word (Nelson & Narens, 1990). Since the current study aimed at understanding how the tastes experienced by HG in response to words affected her judgment about learning word pairs, the JOL task was given only to HG. The pairedassociate learning task, however, was repeated on a control group of 10 non-synesthetes (mean age = 24.7; SD = 2.75).

3. Results 3.1. The consistency test The consistency of tastes experienced by HG across the two presentations (3 months gap) was significantly higher than the consistency shown by the control group across the two presentations separated by a gap of only one week (Z = 8.32, P < 0.0001). HG’s consistency score was also significantly higher than that of the control subject with the highest score-HS (Fisher’s exact test, P < 0.001). The results of the consistency test are depicted in Fig. 3.

100

Consistency of word-taste associations

Phonological–lexical: The relationship between the inducer and the concurrent is at the level of the whole word and the sound structure as well. Most of the words in this category were phonological neighbors of the food items/tastes. Direct lexical–semantic: the inducer directly activates the concurrent (both are directly related in meaning). Indirect lexical–semantic: the inducer indirectly activates the concurrent (the relationship between the two is conceptual in nature). Unidentifiable: When the relationship between the inducers and concurrents did not fall into any of the above-mentioned categories.

90

HG

80 70 60 50 HS

40 30 20

Control

10 0

Fig. 3. The consistency in percentage of word–taste associations for the synesthete (HG), the control group, and the subject with the highest score (HS).

28

V. Ramachandra / Brain and Cognition 106 (2016) 23–32

3.2. Phonological and lexical/semantic factors associated with synesthesia

3.3. Common linguistic associations between words producing the same as well as similar tastes Common linguistic associations between words producing the same as well as similar tastes are described in Appendices A and B respectively. The common linguistic links (phonological or lexical) are reported in the last column of the appendices. Appendix A indicates that common linguistic links exist for words producing same tastes but on the other hand, Appendix B clearly reveals that such common links were not observed for words producing similar tastes. The descriptions for Appendices A and B are provided in Appendices A.1 and B.1 respectively.

3.4. Performance on paired-associate learning As depicted in Fig. 4, HG’s performance on paired-associate learning was not significantly different from the control group (Z = 0.13, P = 0.8966). There was also no difference in performance on the 3 types of word pairs for HG (P > 0.05, Fisher’s exact test). The performance differences on the different word pair types are shown in Fig. 5.

Number of words correct

Out of the 221 inducer–concurrent pairs analyzed, the possible underlying linguistic factors were as follows: 85 – Phonological, 26 – Phonological–lexical, 44 – Direct lexical–semantic, 10 – Indirect lexical–semantic and 56 – Unidentifiable. The inter-rater agreement between two independent judges who conducted the analysis was 88.24%. This indicates a phonological and lexical/semantic link that may exist between the inducers and the concurrents in case of LG synesthesia. PCC scores were calculated for a set of 77 word–taste pairs that had a potential phonological association and for a control list generated from the same 77 pairings (these lists were generated after removing some word–taste mappings from the analysis as explained in Section 2). A paired two-tailed t-test showed a highly significant difference between the mean PCC score for the actual word–taste mappings (Mean = 2.62, SD = 0.93) and the mean PCC score for random pairings of words and tastes [(Mean = 1.21, SD = 1.10); t(76) = 9.62, P < 0.0001]. This suggests that the phonological association between the words and the tastes is not a chance occurrence.

14 12 10 8

No taste

6

Similiar

4

Dissimilar

2 0

Fig. 5. The number of words correct for HG on the paired-associate learning task for the three different types of word pairs.

3.5. Judgment of learning ratings for paired-associate learning The item-by-item absolute prediction score was obtained by subtracting HG’s percentage of correct responses on the pairedassociate learning task from her mean item-by-item JOL rating percentage. This was calculated separately for each of the three word-pair types (no taste, similar/same, and dissimilar). A score of 0 indicates a perfect predictive accuracy, a negative score indicates underprediction, and a positive score indicates overprediction of performance (Kennedy, Yorkston, & Rogers, 1995; Ramachandra, Rickenbach, Ruda, LeCureux, & Pope, 2010). The results, which are shown in Fig. 6, revealed that HG underpredicted her performance on no taste items and on the other hand, she overpredicted her performance on the similar/same and dissimilar word pairs.

4. Discussion The consistency of tastes experienced by HG across the two presentations separated by 3 months was significantly higher than the consistency shown by the controls. The tastes experienced by HG were very detailed and complex. For example, Garfield tasted like ‘‘tomato broth in minestrone soup” and Julie tasted like ‘‘peanut butter and jelly sandwich (not soggy but with extra jelly)”. The consistency of tastes and the detailed descriptions given by the subject is consistent with other previous reports on LG synesthesia (Gendle, 2007; Ward & Simner, 2003; Ward et al., 2005). Phonology seems to play an important role in eliciting tastes (85 out of the 221 word taste pairs had a possible phonological

45

15

Dissimilar

35 30 25 20

HG

Controls

15 10

Number of words correct

Number of words correct

40

10

0 -5 -10 -15 -20

5

-25

0

-30

Fig. 4. The number of words correct on the paired-associate learning task for HG and the control group.

Similar

5

No Taste

Fig. 6. The mean item-by-item absolute prediction scores for HG on the three different types of word pairs.

V. Ramachandra / Brain and Cognition 106 (2016) 23–32

association). The influence of phonology is evident in many of the word–taste associations experienced by HG. For example, ride elicited the taste of ‘‘rye bread with cucumbers and hummus” (/raɪ/ being the critical syllable shared between the inducer and the concurrent); stroll resulted in the taste of ‘‘apple strudel” (/s/, /t/, /r/, and /l/ are the phonemes shared between the inducer and the concurrent). The highly significant difference in PCC scores between the word–taste mappings and a control list (random word–taste pairings) suggests that the phonological link between the words and the tastes is not merely a chance occurrence. Such significant differences in PCC scores were also reported in studies by Ward and Simner (2003), and Ward et al. (2005). There were several inducers that shared sounds belonging to the same category with the concurrents but they could not be taken into account during PCC calculations. For example, consider the word–taste pairings clang (/klæN/) ? clam (/klæm/) (PCC = 3), easily (/izəli/) ? vaseline (/væsəlin/) (PCC = 4), and crammed (/kræmd) ? cranberry (krænberi/) (PCC = 3). The PCC score does not consider the similar sounds shared between the inducers and the concurrents. The sounds /N/, /m/ (nasals); /z/, /s/ (fricatives); and /m/, /n/ (nasals) in the word–taste mappings above are related based on their manner of production but were not considered for PCC calculations because they were similar but not the same sounds. If this factor is taken into account, there could be a much higher degree of phonological overlap between the word–taste pairs in people with LG synesthesia. Twenty-six words had a phonological–lexical link. Sometimes a word that rhymed with a food name produced the taste of that particular food item. For example, tickle tasted like ‘‘pickles on a McDonald’s cheeseburger with ketchup and mustard” (notice that ‘tickle’ and ‘pickle’ rhyme with each other) and trip produced the taste of ‘‘white mint chocolate chip ice cream” (notice that the words ‘trip’ and ‘chip’ rhyme with each other). Please see Appendix C for some more examples of phonological–lexical association. Out of the 221 word–taste pairings, 44 of them had a direct lexical–semantic association. For example, shake tasted like ‘‘chocolate milk shake” and broil tasted like ‘‘broiled flounder”. There were 10 words that had an indirect lexical–semantic relationship (please see Appendix D for some examples). The life experience of HG could have also played an important role in her word–taste associations. For example, the word cat tasted like ‘‘French dressing”. During one of my conversations with HG, I had asked her the reason for this association and she told me that her grandmother had a ‘‘Catalina dressing” (a type of French dressing) when she was a kid. There were 56 words-taste pairs (out of the 221 analyzed) that were classified ‘‘unidentifiable” because they did not fall into the phonological, phonological– lexical, direct lexical–semantic, or indirect lexical–semantic categories. It is possible that many of the inducer–concurrent associations are due to these kinds of early childhood experiences. As predicted, the present study showed that linguistic factors underlie the inducer–concurrent links experienced by people with LG synesthesia. The strong influence of linguistic factors underlying HG’s synesthesia is in line with the studies by Ward and Simner (2003), and Ward et al. (2005) but contradict the findings by Gendle (2007) who did not find much phonological/lexical links between the words and the tastes (observed only in 5.7% of words and 0.9% of nonwords). LG synesthesia may not be purely a sensory-perceptual phenomenon; higher-level linguistic and cognitive processes may be involved. During the later stages of LG synesthesia development, words that share linguistic factors (phonological, phonological–lexical, direct and indirect lexical–semantic) with food items (directly or indirectly) produce gustatory sensations by triggering the lemma associated with those food items. This fits well with the model proposed by Simner and Haywood (2009) and its extension proposed in the present paper (see Figs. 1 and 2). Sometimes a

29

single phoneme in a word may trigger a very specific and detailed taste but such associations may be inconsistent and context specific. Example, the sound /tʃ/ in chair led to the taste of ‘‘cherry coke” but not all words with /tʃ/ consistently produced ‘‘cherry coke” for HG. This indicates that speech sounds are generally not very strong triggers of gustatory sensations (Simner & Haywood, 2009). Synesthesia is a heterogeneous condition and this conceptual framework depicted in Fig. 2 may not fit well with all experiences and all individuals with LG synesthesia. The current study showed the presence of common phonological and lexical links for words producing the ‘‘same” tastes and the absence of any overt links in case of words producing ‘‘similar” tastes (Appendix A vs. B). These findings indicate that words that share common phonological and lexical–semantic variables produce the same tastes by triggering the same set of lemmas associated with food names. On the other hand, words that do not share any linguistic links generate different tastes by activating different set of lemmas associated with food names. This once again bolsters the linguistic factors underlying the word–taste mappings in people with LG synesthesia. Finally, HG’s performance on a paired-associate learning task was no different from that of the controls. In fact, there were 5 control subjects who performed better than HG. Free-association tasks have been used with LG synesthetes in two studies (Gendle, 2007; Ward & Simner, 2003), and superior performance has been reported by Gendle (2007). Contrary to the predictions made in this paper, the results of the paired-associate learning revealed that HG did not significantly differ in her performance between the three types of word pairs (no taste, similar/same, and dissimilar). The difference in her performance was also not in the predicted direction. First, this shows that her synesthesia is not due to some extraordinary ability to make word–taste associations. Second, this indicates that she was not able to use her taste experiences as a ‘‘mnemonic aid” to learn the associations between the word-pairs. It also suggests that the triggering of tastes in both the first and the second word (target word to-be-remembered) in the pair did not impede her learning (no significant difference between the no taste and the taste conditions). Finally her JOL responses were consistent with the predictions made in the present study (JOL taste > JOL no taste). Her performance on the paired-associate learning was, however, in the opposite direction (no taste > taste). The word pairs that produced ‘‘tastes” could have created a ‘‘foresight bias” resulting in higher JOLs relative to the ones that did not produce any tastes (Koriat & Bjork, 2005). When HG was presented with the word pairs producing the same or similar tastes and asked to make judgments regarding the likelihood of saying the second word when presented with the first word of the pair, she could have experienced a feeling of learning because of the tastes they produced. However, during the actual testing, presentation of the first word would have elicited the taste associated with that word but it would not have generated the second word associated with that taste. This account holds true even for dissimilar word pairs in which the first word would have elicited the corresponding taste and the taste associated with the second word but not the word itself (if she had learned the word pairs based on their tastes). This is because HG’s synesthesia was unidirectional. That is, the words produced the tastes but the tastes on the other hand did not elicit the corresponding words. The perception of tastes without being able to produce the words (inducers) is supported by findings in which LG synesthetes tasted some words, which they were not able to say on a picture naming ‘tip-of-tongue’ task (Simner & Ward, 2006). This explanation concerning the present study is speculative and the findings presented here should be interpreted with caution because they are based on a single subject and a small dataset. Please note that her JOLs for dissimilar pairs were higher than the JOLs for similar/same pairs

30

V. Ramachandra / Brain and Cognition 106 (2016) 23–32

(opposite to what was predicted) but the actual performance on the test was better for similar/same pairs when compared to dissimilar pairs. It’s not clear why this difference between JOL and performance emerged. As mentioned in Section 1, most studies that have sought to examine the effect of synesthesia on memory have been conducted in grapheme-color synesthesia. Although, the current study did not reveal any positive or adverse effects of LG synesthesia on memory, further studies are needed to gain a much better understanding of this relationship.

Acknowledgments This study was supported by a research initiation grant from Marywood University. The author would like to thank HG for her time and commitment to this study. The author also wishes to thank Colleen Meighan, Jillian Gradzki, Sydney Pauline, Christina Dixon, and Karen Buckley for their help with part of the data analysis.

Appendix A

List of words that shared the same tastes and the probable linguistic links.

a

Words (Inducers)

Taste (Concurrents)

Phonological/lexical link

folder, full Joshua, Julie Christian, Christopher door, endorsed

fruit rollup peanut butter and jelly sandwich Rice krispies really lame (cheap) milk chocolate

phonemes {f, l} phoneme /dʒ/ phonemes {k, r, I, s} /dɔr/ in door and endorsed may be related to ‘‘cote d’or” which is a Belgian chocolate a /lɪd/ (lid) ? /aɪst/ (iced) guaranteed ? tea {/t/ and /i/ are the common phonemes}

lid, guaranteed

lid led to ‘‘iced tea”

coat, couch, covert

guaranteed led to ‘‘ice tea” malted milk balls (whoppers)

priest, approved, improved enormous, Vincent, metropolitan flack, broil chatter, clang clap, clash graze, greatest easily, excellent task, basket

circular prune flavored snacks tomato soup broiled flounder clam chowder French onion dip grape nuts vaseline lip balm garlic bread

/koʊ/ in coat, covert, couch is related to ‘‘coat” or cover”, and malted milk balls have a chocolate covering/coating phonemes {p, r} ? ‘‘prune” No phonological/lexical link a {f,l}in flack ? flounder, {b,r}in broil ? broiled a /ʧ/ in chatter ? chowder, /kl/ in clang ? clam /kl/ ? no link with taste concurrent /gr/ ? ‘‘grape” {s/z,əl} ? ”Vaseline” ‘‘task” shares phonemes with ‘‘basket” and garlic bread is usually served in a basket

Indicates that no common phonological/lexical links were found.

5. Conclusion

A.1. Description of Appendix A

The current case study aimed at examining the linguistic and cognitive aspects of a person with LG synesthesia. Analysis of 221 word–taste pairs showed that linguistic factors such as phonological, phonological–lexical, and lexical–semantic factors may underlie her synesthetic experiences. In the beginning stages of development of LG synesthesia, food words get connected with lemma of food names and this in turn is linked to gustatory sensation. During the later stages of synesthesia development, words that share phonological or lexical/semantic features with the food words may result in a gustatory experience by triggering the lemma or the abstract concept associated with the food words (Simner & Haywood, 2009). Finally, the LG synesthesia had no effects on the synesthete’s memory when examined using a paired-associate learning task. The presence of taste (similar or dissimilar) did not have any facilitative effects on her memory but did positively influence her memory monitoring in that she predicted her performance to be better in the ‘‘taste” when compared to ‘‘no taste” condition. This can be attributed to the unidirectional nature of LG synesthesia. This is just a preliminary investigation of memory and metamemory in LG synesthesia and cannot be generalized to other individuals with this condition. Further research using different kinds of memory tasks and other types of memory monitoring tasks such as feeling-of-knowing and ease-of-learning is necessary to better understand the cognitive and metacognitive aspects of LG synesthesia.

Whenever the association is phonological/lexical in nature, the common critical phoneme/s or the lexical connections are listed in the last column of the appendix. For example, the words folder and full elicited the taste of ‘‘fruit rollup”. The common phonemes shared between the inducer (folder, full) and the concurrent (fruit rollup) are /f/ and /l/. Therefore, the phonemes /f/ and /l/ in the words folder and full could have led to the taste of ‘‘fruit rollup”. Although some words shared a common taste, the phonemes producing this taste were different. For example, both flack and broil produced the taste of ‘‘broiled flounder” but the /f/ and /l/ in flack were associated with ‘‘flounder” and the phonemes /b/ and /r/ in ‘‘broil” were associated with ‘‘broiled”. Similar separate associations were seen for the words chatter and clang as described in Appendix A. In case of the words lid and guaranteed, it is interesting to note that lid led to the taste of ‘‘iced tea”. The emphasis here is on the word ‘‘iced” because of the link between /id/ in the word lid and /aist/ in iced but guaranteed led to ‘‘ice tea” because of the sounds /t/ and /i/ in the inducer and the resultant taste (tea) as well. Some words had a phonological–lexical association. For example, door and endorsed elicited the taste of ‘‘milk chocolate”. The phonemes /dɔr/ in the words door and endorsed may be related to ‘‘cote d’or” which is a Belgian milk chocolate available throughout the United States. The words coat, covert and couch produced the taste of ‘‘malted milk balls (whoppers)”. Whoppers are chocolate covered malted milk candies and the words covert and coat

V. Ramachandra / Brain and Cognition 106 (2016) 23–32

could be directly connected to the taste of ‘‘chocolate covered candies” because of their semantic relationship (cover/coat ? chocolate covered). The word couch shares the phoneme segment /koʊ/ with covert and coat. Therefore, the word couch could be eliciting the taste of these candies covered with chocolate indirectly through the words coat/cover because of the common phonemes shared between them. The words enormous, Vincent, metropolitan led to the concurrent ‘‘tomato soup” but there were no identifiable phonological/lexical links between them. The words clap and clash produced the taste of ‘‘French onion dip”. /k/ and /l/ were the common phonemes shared between the two words (inducers) but they had no obvious phonological/lexical links with the taste concurrent they produced. On the other hand, the word clang which also shared the same phonemes (/k/and /l/) with clap and clash led to ‘‘clam chowder”. This could be a phonological association because both ‘‘clang” (/N/) (inducer) and ‘‘clam” (/m/) (concurrent) have nasal sounds as opposed to clap and clash, which do not have any nasal phonemes in them. Therefore it is easy to see how clang resulted in the taste of ‘‘clam chowder” (clang ? clam).

31

pork” because of the relation between ‘‘tension” and ‘‘pulling”. So, in this case the word pull could be a common lexical–semantic link between the words pit, tension and the taste they produced ‘‘pulled pork”. It could be argued that the taste of ‘‘pulled barbecue pork sandwich” is not very different from that of ‘‘pulled pork”. Therefore, similar lexical–semantic links between the inducer and the concurrent could be identified in this case. Appendix C. Some examples of phonological–lexical association between the inducers and their concurrents Most of the associations in this category included phonological neighbors (tickle, tackle-pickle, trip-chip, etc.). Some other examples are described here. An interesting example of a phonological–lexical link between the inducer and concurrent was the word Martin. This elicited the taste of ‘‘fluffernutter sandwich”, which is a sandwich made with peanut butter and marshmallow crème. Perhaps the syllable /mar/ in Martin triggered the word ‘‘marshmallow” which in turn could have resulted in the concur rent-‘‘fluffernutter sandwich”. Other examples under this category

Appendix B List of words that shared similar tastes and the probable linguistic links. Words (Inducers)

Taste (Concurrents)

Phonological/lexical link

discount, closer

discount ? number fruit snacks closer ? yellow fruit snacks

a

McDonald’s, cooler

McDonald’s ? Hi-C orange chips cooler ? Hi-C citrus drinks Jennifer ? tuna fish salad with celery suggestions ? tuna fish sandwich

Jennifer, suggestions

aboard, Melvin, ottoman

pit, tension

a

aboard ? popsicle stick Melvin ? orange strawberry popsicle ottoman ? lime popsicle pit ? pulled barbecue pork sandwiches tension ? pulled pork

The count in discount ? number fruit snacks closer ?? a McDonald’s ? Hi-C orange chips served there? Cooler ? Hi-C orange drinks kept in the cooler? a Jennifer ? fish (/f/ is common phoneme). It’s also likely that the connection is based on life experience suggestions ? sandwich (/sa/ is the common phoneme). Also, /tʃ/ in suggestions belongs to the same category as /ʃ/ in fish (both are sibilants) No links identified for any of the words

pit associated with bull and bull rhymes with pull? tension produced by pulling pork?

Indicates that no common phonological lexical links were found.

B.1. Description of Appendix B As it can be seen in Appendix B, none of the inducers had a common association with the concurrents. For example, the words discount and closer produced similar tastes (two different varieties of fruit snacks) but there was no common association between them. The word discount could have produced the taste of ‘‘number fruit snacks” because of the association between count in discount and the number in number fruit snacks (fruit snacks shaped like numbers). On the other hand, there were no identifiable associations between closer and ‘‘yellow fruit snacks”. The only two words that could be described as having a common lexical–semantic link are the words pit and tension. The word pit led to the taste of ‘‘pulled barbecue pork sandwiches” and tension resulted in ‘‘pulled pork”. As described in Appendix B, the word in pit is associated with bull (pit bull is a type of dog) and bull rhymes with pull. Therefore, the word pit could have indirectly led to the taste of pulled pork via the word bull. On the other hand, tension elicited the taste of ‘‘pulled

include the word better eliciting the taste of ‘‘butter” and Marty producing the taste of ‘‘sweet tarts” (the relationship between the words mart and tart).

Appendix D. Examples of indirect lexical–semantic association between the inducers and their concurrents The word insurance resulted in the taste of ‘‘biscuits with old bay seasonings”. Old bay is an American insurance company and therefore, the word insurance could have triggered the taste related to ‘‘old bay seasonings” (an American seasoning company). It is interesting to note that both the old bay insurance company and the old bay seasoning company are based in Maryland (a state located in the eastern part of the United States). However, the location of both the companies could have been just a coincidence. The word funny led to the taste of ‘‘apple jacks cereals”. This could be due to an indirect lexical–semantic link between ‘funny’ and ‘jokes’ (sounds like ‘jacks’).

32

V. Ramachandra / Brain and Cognition 106 (2016) 23–32

References Bankieris, K., & Simner, J. (2014). Sound symbolism in synesthesia: Evidence from a lexical-gustatory synesthete. Neurocase, 20(6), 640–651. Baron-Cohen, S., Bor, D., Billington, J., Asher, J., Wheelwright, S., & Ashwin, C. (2007). Savant memory in a man with colour form-number synaesthesia and Asperger syndrome. Journal of Consciousness Studies, 14, 237–251. Baron-Cohen, S., Burt, L., Smith-Laittan, F., Harrison, J., & Bolton, P. (1996). Synaesthesia: Prevalence and familiality. Perception, 25, 1073–1079. Brang, D., & Ramachandran, V. S. (2011). Survival of the synesthesia gene: Why do people hear colors and taste words? PLoS Biology, 9(11), e1001205. Cytowic, R. E. (1989). Synaesthesia: A union of the senses.Berlin: Springer. Cytowic, R. E. (1993). The man who tasted shapes.London: Abacus. Downey, J. E. (1911). A case of colored gustation. American Journal of Psychology, 22, 528–539. Ferrari, G. C. (1907). Una varieta nuova di sinestesia [A new variety of synesthesia]. Rivista Psicologia, 3, 297–317. Ferrari, G. C. (1910). Un nuovo caso di sinestesia uditivo-gustativa [A new case of auditory–gustatory synesthesia]. Rivista Psicologia, 6, 101–104. Galton, F. (1880). Visualised numerals. Nature, 21, 252–256. Gendle, M. H. (2007). Word-gustatory synesthesia: A case study. Perception, 36, 495–507. Gray, B., & Simner, J. (2015). Synesthesia and release phenomena in sensory and motor grounding. Cases of disinhibited embodiment? Frontiers in Psychology, 6, 54. Gross, V. C., Neargarder, S., Caldwell-Harris, C. L., & Cronin-Golomb, A. (2011). Superior encoding enhances recall in color-graphemic synesthesia. Perception, 40, 196–208. Grossenbacher, P. G., & Lovelace, C. T. (2001). Mechanisms of synesthesia: Cognitive and physiological constraints. Trends in Cognitive Sciences, 5(1), 36–41. Hubbard, E. M. (2007). Neurophysiology of synesthesia. Current Psychiatry Reports, 9 (3), 193–199. Jones, C. L., Gray, M. A., Minati, L., Simner, J., Critchley, H. D., & Ward, J. (2011). The neural basis of illusory gustatory sensations: Two rare cases of lexical-gustatory synesthesia. Journal of Neuropsychology, 5(2), 243–254. Kennedy, M. R., Yorkston, K. M., & Rogers, M. A. (1995). Self -monitoring abilities of two adults with traumatic brain injury during verbal learning. American Journal of Speech-Language Pathology, 4, 159–163. Koriat, A., & Bjork, R. A. (2005). Illusions of competence in monitoring one’s knowledge during study. Journal of Experimental Psychology: Learning, Memory, and Cognition, 31, 187–194. Luria, A. R. (1968). The mind of a mnemonist.New York: Basic Books. Mills, C. B., Innis, J., Westendorf, T., Owsianiecki, L., & McDonald, A. (2006). Effect of a synesthete’s photisms on name recall. Cortex, 42(2), 155–163.

Mueller, M. L., Tauber, S. K., & Dunlosky, J. (2013). Contributions of beliefs and processing fluency to the effect of relatedness on judgments of learning. Psychonomic Bulletin & Review, 20(2), 378–384. Myers, C. S. (1911). A case of synaesthesia. British Journal of Psychology, 4, 228–238. Nelson, T. O., & Narens, L. (1990). Metamemory: A theoretical framework and some new findings. In G. H. Bower (Ed.), The psychology of learning and motivation (pp. 125–173). New York: Academic Press. Parker, E. S., Cahill, L., & McGaugh, J. L. (2006). A case of unusual autobiographical remembering. Neurocase, 12, 35–49. Pierce, A. H. (1907). Gustatory audition: A hitherto undescribed variety of synaesthesia. The American Journal of Psychology, 18, 341–352. Ramachandra, V., Rickenbach, B., Ruda, M., LeCureux & Pope, M. (2010). Fast mapping in healthy young adults: The influence of metamemory. Journal of Psycholinguistic Research, 39(3), 213–224. Richer, F., Beaufils, G.-A., & Poirier, S. (2011). Bidirectional lexical-gustatory synesthesia. Consciousness and Cognition, 20(4), 1738–1743. Rothen, N., & Meier, B. (2009). Do synesthetes have a general advantage in visual search and episodic memory? A case for group studies. PLoS One, 4(4), e5037. Simner, J. (2007). Beyond perception: Synesthesia as a psycholinguistic phenomenon. Trends in Cognitive Sciences, 11, 23–29. Simner, J. (2012). Defining synaesthesia. British Journal of Psychology, 103, 1–15. Simner, J., & Haywood, S. L. (2009). Tasty non-words and neighbours: The cognitive roots of lexical-gustatory synaesthesia. Cognition, 110(2), 171–181. Simner, J., & Holenstein, E. (2007). Ordinal linguistic personification as a variant of synesthesia. Journal of Cognitive Neuroscience, 19, 694–703. Simner, J., & Logie, R. H. (2007). Synaesthetic consistency spans decades in a lexicalgustatory synaesthete. Neurocase, 13(5–6), 358–365. Simner, J., Mayo, N., & Spiller, M. J. (2009). A foundation for savantism? Visuospatial synaesthetes present with cognitive benefits. Cortex, 45(10), 1246–1260. Simner, J., & Ward, J. (2006). The taste of words on the tip of the tongue. Nature, 444, 438. Smilek, D., Dixon, M. J., Cudahy, C., & Merikle, P. M. (2002). Synesthetic color experiences influence memory. Psychological Science, 13(6), 548–552. Ward, J., & Simner, J. (2003). Lexical-gustatory synaesthesia: Linguistic and conceptual factors. Cognition, 89, 237–261. Ward, J., Simner, J., & Auyeung, V. (2005). A comparison of lexical-gustatory and grapheme-colour synaesthesia. Cognitive Neuropsychology, 22, 28–41. Yaro, C., & Ward, J. (2007). Searching for shereshevskii: What is superior about the memory of synaesthetes? The Quarterly Journal of Experimental Psychology, 60 (5), 681–695.

Web references Accessed 01.06.15. Accessed 02.03.15.