Even in the same blend, which odors are easy to recognize can depend on what you smelled just beforehand. A 2017 study found that smelling part of a mixture for five seconds first reduced identification of those components, making subsequently added components relatively easier to recognize.
A second finding was that when two different compounds shared an odor note, that note stood out in the mixture. Recognition of a blend depended not only on its ingredients, but also on the order of smelling and on odor characteristics shared between components.
Four compounds smelling of cherry, caramel, smoke, and grape
The main aim was to test an earlier finding that brief adaptation changes how readily components are recognized in a mixture, using four different compounds. This time, the researchers also included a control condition in which participants first smelled the headspace of a bottle containing only water.
The compounds were cherry-like benzaldehyde, caramel-like maltol, smoky guaiacol, and grape-like methyl anthranilate. Fourteen nonsmokers participated and learned the compounds and their corresponding names before testing. For maltol, “cotton candy” was also used when participants found that label easier to recognize.
Participants smelled individual compounds and mixtures of two to four compounds, then reported which of the learned names they could identify. The main task tested identification using component labels, not numerical ratings of odor intensity. Preliminary ratings placed the compounds at moderate intensity, but their intensities were not perfectly matched.
Adapting to an existing odor leaves the added odor relatively prominent
Participants smelled the first bottle for about five seconds, then smelled a second bottle. For example, if the first contained cherry and smoke, and the second contained those two plus caramel, the first two were the “ambient” components and caramel was the “extra” component added in the second bottle.
In the control condition, the first bottle contained water and the second contained the same test mixture. Each compound had the same concentration in solution whether presented alone or in a mixture. Experimental and control conditions took place on separate days, with condition order varied between participants.
After water, correct identification was about 50% for both the components nominally designated “ambient” and those designated “extra.” After selective adaptation, identification of ambient components declined relative to the water control, and extra components were identified better than ambient components. This relative advantage occurred in mixtures of two, three, and four compounds.
However, the increase in identification of extra components relative to the water control was not itself significant. The central finding was that familiarized components became harder to identify, producing a difference in prominence between them and the extra components within the mixture.
Cherry was already a prominent cue
Not all components were recognized equally. In water-control mixtures, cherry was identified about 71% of the time on average, compared with about 43% for the other three components. Even compounds that were almost always named correctly alone became harder to recognize in mixtures.
In some combinations, however, cherry remained relatively recognizable even after prior exposure. Alongside adaptation, a component’s inherent prominence appeared to influence what participants noticed in the mixture.
The grape compound also had a smoky side
Participants learned methyl anthranilate as the grape compound. When presented alone, however, it elicited grape responses 61% of the time and smoke responses 29% of the time. Rather than dismissing this as a simple error, the authors investigated whether the compound shared an odor aspect with guaiacol.
When guaiacol and methyl anthranilate were mixed, smoke was identified more often than grape even in the water-control condition. Smoke responses also predominated when either component had been smelled beforehand.
For comparison, pairing each separately with maltol produced the usual pattern: the extra component was easier to identify than the previously smelled component. From this difference, the authors interpreted the shared smoky note as becoming more prominent when the two compounds were combined.
What changed here was the frequency of responses using the labels smoke and grape. This was not an intensity measurement showing that the smell of smoke became a particular number of times stronger. The limited set of available labels is also important when interpreting the findings.
Component information can still be retrieved from a mixture
The authors emphasized that components difficult to recognize in a mixture became recognizable again when the smelling sequence changed. In this view, component information was not completely lost on mixing; rather, the information that stood out changed over a short time.
They proposed mechanisms involving adaptation at the level of nasal reception and inhibition between signals in the brain, but the study directly measured people’s responses. It did not identify the mechanism through measurements of receptors or brain activity. It also tested only four compounds at specific concentrations after participants had learned their names, rather than establishing how many components people can identify in every natural blend.
Conclusion
Smelling part of a mixture for five seconds first reduced identification of those components, making subsequently added components relatively easier to recognize. Odors that were prominent to begin with, such as cherry, and a smoky note shared by two compounds sometimes departed from that general pattern.
Even with the same ingredient list, recent olfactory experience changes which side of a blend we notice. The study presents odor recognition not simply as a fixed impression, but as a process in which the information that stands out can shift rapidly.
The study was supported by the US National Institutes of Health, the University of Connecticut Foundation, and the university’s School of Dental Medicine and Clinical Research Center. No separate conflict-of-interest statement was found in the main text.
Reference
Frank ME, Fletcher DB, Hettinger TP. Recognition of the Component Odors in Mixtures. Chemical Senses. 2017;42:537–546. Original research article
