Sci Review: When music is salty

Few words about of this science paper

Rachel Guetta

Current employment: Duke University, USA

Rachel Guetta is a 2nd year Duke Clinical Psychology doctoral student. She completed her BA at Wesleyan University, where she researched auditory perception and cross-modal associations between sound and taste. She then worked for two years at the National Center for PTSD at the Boston VA, where she developed an interest in comorbidity patterns between emotional dysregulation and multi-sensory reactivity. Rachel is interested in research at the intersection of emotional functioning and sensory processing. She is working to develop and validate a semi-structured clinical interview to investigate the etiology and maintenance of Misophonia in an effort to identify transdiagnostic intervention targets for this sound-intolerance condition and related emotional disorders.

Psyche Loui

Current employment: Northeastern University, USA

Psyche Loui is Assistant Professor in at Northeastern University. She is Director of the MIND Lab and enjoys asking questions about how the mind enables music, and what we can learn from musical functions of the brain. When not teaching and conducting research, she enjoys playing the violin. Research in her lab aims to understand the networks of brain structure and function that enable musical processes: auditory and multisensory perception, learning and memory of sound structure, sound production, and the human aesthetic and emotional response to sensory stimuli. Tools for this research include electrophysiology, structural and functional neuroimaging, noninvasive brain stimulation, and psychophysical and cognitive experiments as motivated by the following questions:
1. How do humans perceive and produce music?
2. How do expectations develop?
3. How do structural and functional connectivity in the brain enable perception and action?
4. How can music be used to understand the brain and to help people with neurological and psychiatric disorders?

Paper abstract

Here we investigate associations between complex auditory and complex taste stimuli. A novel piece of music was composed and recorded in four different styles of musical articulation to reflect the four basic tastes groups (sweet, sour, salty, bitter). In Experiment 1, participants performed above chance at pairing the music clips with corresponding taste words. Experiment 2 uses multidimensional scaling to interpret how participants categorize these musical stimuli, and to show that auditory categories can be organized in a similar manner as taste categories. Experiment 3 introduces four different flavors of custom-made chocolate ganache and shows that participants can match music clips with the corresponding taste stimuli with above-chance accuracy. Experiment 4 demonstrates the partial role of pleasantness in crossmodal mappings between sound and taste. The present findings confirm that individuals are able to make crossmodal associations between complex auditory and gustatory stimuli, and that valence may mediate multisensory integration in the general population.

Main takeaways

The present findings confirm that individuals are able to make crossmodal associations between complex auditory and gustatory stimuli, and that valence may mediate multisensory integration in the general population.

The authors point out several hypotheses debate whether cross modeling associations are:

  1. universal and innate
  2. learned from development
  3. impacted by culture or region
  4. influenced by subjective preference or pleasantness

Authors also point out that there is a growing need for researchers to incorporate complex and naturalistic stimuli in taste and gustation research.

This paper investigates the presence of crossmodal associations between complex sounds and complex tastes, and the role of preference as a mediator of these associations.

The attribute ‘‘intensity’’ is

The auditory stimuli consisted of recordings of an original violin composition, as opposed to isolated pitches or timbres. The gustatory stimuli consisted of custom-made chocolate ganache instead of simple stimuli, beverages, or merely the names of food or flavors. We chose chocolate, as it can adapt to sweet, sour, salty and bitter tastes, and because it can elicit a wide range of pleasantness ratings.

The research team conducted 4 experiments:

Experiment 1
The first experiment validates our auditory stimuli, and tests whether participants without any prior exposure to the musical stimuli are able to match the music clips to the intended taste labels.

Participants were 160 Amazon Mechanical Turk workers. Subjects were unselected for gender, age, ethnicity, and musical training. Participants first listened to a single music clip, and then selected a taste word (sweet, sour, salty or bitter) that best matched the clip. Participants were allowed one minute to complete the assignment. A response was considered correct if the participant’s chosen word matched the target taste category of the music clip. Experiment confirmed that there was an association between music clips and chosen taste labels. participants were able to identify the sour clips with greatest accuracy.

Experiment 2
Research team points out that taste has different dimensions, and that flavors cluster together. The question they try to answer here, by means of multidimensional scaling (MDS), is if analogous dimensions can organize musical stimuli as well.

Participants were 16 Wesleyan University undergraduate students participated in this experiment (4 male, 12 female). All participants reported that they had normal hearing, and no impairment of smell or taste functions. Number of years of musical training ranged from 0 to 15 years. The same eight music clips were used (two samples of each musical category: sweet, sour, salty and bitter). The test was administered using custom software written in Max/MSP. Each trial consisted of two randomized music clips played consecutively. After each pair of stimuli, participants rated how similar or different the two stimuli were from each other on a slider-controlled visual analog scale from “very different” to “very similar”, from 1 (most different) to 100 (most similar). Participants were also able to clearly distinguish differences between music clips within the four distinct taste categories. Within each category, average similarity ratings differed: ratings were highest for sweet and salty pairs, and lowest for sour and bitter pairs. The two-dimensional multidimensional scaling solution shows clear clustering of the musical stimuli by taste categories (Fig 1). Participants reacted most pleasantly to the sweet stimuli, with its warmth, vibrato and legato nature.

Experiment 3
Experiment 3 further investigates the crossmodal associations between sound and taste, specifically using chocolate as taste stimuli to investigate its mapping with musical stimuli.

Participants were 50 Wesleyan University undergraduate students took part in the experiment (25 male, 25 female). Different stimuli were used: 4 Custom-made chocolate ganache was used as gustatory stimuli, as it can adapt to the four basic taste groups.

Participants were first presented with a familiarization phase, in which they sampled each of the four chocolates and listened to each of the four musical stimuli, all in randomized order. Next, participants were tested for their associations between the music and the chocolate. For each trial, participants ate a randomly assigned piece of chocolate, then listened to each of the four music clips in their entirety. The task was then to select which music clip best fit the chocolate they had just sampled. Participants were free to listen to the clips as many times as they wished before making their selection. As a palate cleanser between successive trials, participants were offered a bite of pita bread. No taste words were introduced while testing for the crossmodal association itself in order to avoid any semantic influences. Participants matched each ganache to the respective music clip above chance level of 0.25. Musically trained participants did not perform more accurately. Participants were able to associate across sound and taste modalities. This is a first demonstration of sound-taste matching with both complex sound and complex taste.

Experiment 4

Experiment 4 explores the mediating role of preference in sound-taste crossmodal associations. The use of preference as a cue was also confirmed in preliminary verbal interviews with participants. Participants were 23 Wesleyan University undergraduate students (15 male, 8 female) who took part in the previous experiment additionally participated in this follow-up study. Within this sample, 8 were musicians. Procedure. Participants were asked to rate each music clip and each ganache flavor on a 7-point scale, anchored by “highly unpleasant” (1) and “highly pleasant” (7).

Among the 23 participants 6 preferred the sweet music clip, 6 preferred the salty clip, 2 preferred the sour clip, and the remaining 9 participants gave similar preference ratings for two or more categories. No participants preferred the bitter music clip. Given the 7-point scale, average pleasantness ratings for the ganache samples ranged from 3.70 for bitter samples to 5.89 for sweet samples. Sour samples averaged 4.18 and salty samples averaged 4.06. Pleasantness ratings for music were also lowest for bitter and highest for sweet, with and in between. While most participants preferred sweet music and sweet chocolate overall, a participant who rated the salty ganache stimulus as pleasant was also likely to find the salty music clip to be pleasant. This similarity between pleasantness ratings across modalities suggests that participants could have associated sound and taste based upon their personal preferences. Fig 2A shows a correlation between preference ratings for music and chocolate samples, whereas Fig 2B shows taste preference profiles for sweet, sour, and salty music preferring subjects.

While Experiment 3 shows that individuals are able to map sweet, sour, salty and bitter across taste and sound modalities, Experiment 4 explores a mechanism contributing to this effect. The similarity of preference ratings between sound and taste suggests that participants may match complex sounds and complex tastes based on their individual preference profiles that transcend sensory modalities.


The authors themselves stated that applications of these findings may apply to food businesses and restaurant entrepreneurs in marketing products and optimizing consumer experience, capitalizing on emotional congruency between sound and taste. We agree with that, but we see additional application of these findings as base for whole new gastronomy trends, were auditory stimuli could be utilized at concerts, music festivals etc., in combination with complex gustatory and olfactory stimuli. Besides previously mentioned, we can see awesome utilization of this information in every day cooking as well as in the professional culinary and gastronomic circles.

Overall review

This science paper caught our eye, and we didn’t regret reading it. It was very interesting and insightful.

What we like the most

What we really like about this paper was that authors choose to incorporate link to auditory and complex taste stimuli so that all readers could hear the experimental samples. Besides this, we like when science papers are concise, clear and get-to-the-point based. This Paper is great example of correct set-up scientific form and adequate methodologies. We also really liked that authors gave us inside into experiments details like precise description of samples used.  

What could have been better

What could be better in this paper is the main concept and hypothesis. Because we would like to get more inside into taste aspect of cross modeling mapping, which would be more precise if research team had recruited technologists or gastronomes to give their contribution. Also, we feel like Introduction and General discussion part of paper were little vague in the sense that it was unclear how did researchers proved and confirmed that their hypothesis is the right one when they didn’t analyzed the others.


It contains some utilizable theoretical and experimental content which is really interesting. Also it is based on 45 previous scientific sources and numerous experiments. Reading this paper you can gain insight into how complex auditory and complex taste stimuli are associated in our brains.

Very interesting material which can serve as foundation for future experimental research and implementation especially in food industiry & hospitality.

Overall rating:

If you are interested you can download and read it by clicking icon above. Hope you are going to enjoy it as much as we did.

If you are interested you can download and read it by clicking icon on the left. Hope you are going to enjoy it as much as we did.