Six Ingredients Blend into the Scent of Red-Fruit Cordial

Smelled separately, the ingredients have different odors. Mixed together, they become a single scent reminiscent of red-fruit cordial. A 2013 study used a six-component mixture of this kind to investigate how a unified odor depends on the identity of its ingredients and their proportions.

In humans, removing vanillin or substantially changing the overall proportions led the mixture to be classified as smelling different from the original. Newborn rabbits also responded differently to mixtures depending on their ingredients and proportions. The study explored both the similarities and the differences in how humans and rabbits perceive a mixture as a whole.

Six components form one unified scent

The mixture contained six components: vanillin, frambinone (raspberry ketone), isoamyl acetate, β-ionone, ethyl acetate, and β-damascenone. Earlier research had shown that, mixed in particular proportions, they produce an odor reminiscent of red-fruit cordial for humans. The original paper calls this mixture RC.

The researchers asked whether RC forms a unified odor distinct from its individual ingredients, whether every component is necessary, and what happens when their proportions change. They also compared other six-component mixtures to check whether simply having six ingredients produces the same effect.

People grouped similar odors without having to name them

The main human study involved 198 adults with no specialist training in smell, divided into three groups of 73, 52, and 73. In each task, they smelled seven to nine samples and freely grouped those they considered similar. They were not prompted from the outset to identify a red-fruit odor.

The samples were coded, and participants could smell them again as needed. The sorting results were used to create maps in which samples placed closer together had been grouped as more similar in odor. Intensity was adjusted in preliminary tests, although not every sample was perfectly matched in strength. Participants used more descriptions of odor quality than of intensity to explain their groupings.

RC was classified separately from its individual components. In contrast, the two comparison mixtures, also containing six components each, were placed near some of their individual ingredients. Having six ingredients did not necessarily produce a unified scent distinct from the original odors.

Removing vanillin changes the scent; large shifts in proportions change it further

When five-component mixtures made by omitting one RC ingredient at a time were compared, the mixture without vanillin was particularly far from the original RC. Omitting frambinone also had a relatively large effect, whereas mixtures without ethyl acetate or β-ionone remained close to RC. The same operation of removing one ingredient had different effects depending on which ingredient was removed.

Proportions mattered too. Mixtures in which all components had equal proportions, or in which the proportions were substantially rearranged, were classified as having odors distinct from the original RC. The mixture with substantially rearranged proportions was placed farther from RC than the mixture without vanillin.

There was nevertheless some tolerance for small changes. The condition with less β-ionone and more ethyl acetate occupied a position overlapping with the original RC. Rather than every small adjustment producing a different scent, some changes preserved the unified odor while others disrupted it.

Testing rabbits’ responses to learned odors

The rabbit experiments used 382 pups born to 99 females. The analyses excluded 42 pups for which successful conditioning could not be established.

On the first day after birth, pups learned a test odor through five minutes of exposure together with a pheromone associated with nursing. The following day, researchers presented the test odor alone and checked for searching movements of the mouth and head. Each pup participated in one experiment and encountered a small number of samples in counterbalanced order.

Pups that had learned just one component of RC responded to that component alone but showed little or no response to the whole RC mixture. Yet they responded readily to other six-component mixtures containing the same learned component. This suggests that RC produced a change beyond simply adding unfamiliar ingredients.

An asymmetry: learning the whole also produces responses to the ingredients

The researchers also tested learning in the opposite direction. When pups first learned the whole RC mixture, a high proportion responded the next day not only to RC but also to each component presented separately. The result was asymmetric: learning one component rarely produced a response to the whole, whereas learning the whole produced responses to its individual components.

The authors interpret this as pups perceiving the mixture as a unified whole while also retaining information about its ingredients. The experiment did not test whether RC smelled like red-fruit cordial to rabbits. Instead, it investigated mixture processing through the stimuli to which learned responses generalized.

In tests following learning of one component, removing another component from RC increased responses to the mixture, regardless of which component was omitted. Responses also returned when the proportions were made equal or substantially rearranged. Some small changes in proportions, however, still elicited no response, a pattern that partly overlapped with the human results.

Similar processing, but different tasks

Humans and rabbit pups shared a pattern: the particular mixture called RC had a unified odor distinct from its individual ingredients, and that unity depended on both component identity and proportions. In humans, removing different components had effects of different sizes. In rabbits, every omission produced a large change when assessed by whether a response occurred.

Species differences may not be the only explanation; developmental stage and task may also matter. Humans sorted similar odors, whereas rabbits underwent conditioning associated with nursing. The authors also note that a response/no-response measure may miss subtle perceptual differences. Whether humans would perceive the ingredients more readily after comparable conditioning remains an open question.

Summary

Humans classified the six-component RC mixture as a unified scent distinct from its individual ingredients. Removing vanillin or making large changes in proportions altered that unity, but other six-component mixtures and small adjustments did not produce the same pattern.

Learned responses in rabbit pups also varied with mixture identity and proportions, and learning the whole mixture produced responses to its individual components. An odor is not merely the sum of its ingredients: perceiving a whole and retaining information about its parts can coexist, depending on the conditions. This study illustrated that possibility in two species.

The research received grants from the Burgundy Regional Council, the European Regional Development Fund, and IFR92, together with fellowship support from the French ministry responsible for higher education and research. The funders had no role in study design, analysis, the decision to publish, or preparation of the manuscript. The authors declared no competing interests.

Reference

Sinding C, Thomas-Danguin T, Chambault A, et al. Rabbit Neonates and Human Adults Perceive a Blending 6-Component Odor Mixture in a Comparable Manner. PLOS ONE. 2013;8:e53534. Original research article