Do Letters or Colors Help Us Recall an Odor’s Name?

You have smelled it before, but its name will not come to mind. In such a situation, which is a more useful clue: a color suggesting the scent’s source, or some of the letters in its name?

In 2025, Eda Nur Capkan and Funda Yildirim presented 40 Turkish speakers with color or letter cues after they smelled odors. The aim was to examine how cue type relates to naming accuracy and speed, and to coordinated electrical activity in the brain. They also considered familiarity, liking, and the feeling of knowing an odor but being unable to retrieve its name.

Color or two-letter cues for 12 odors

Participants were aged 18–32, with a mean age of 23.35, and 65% were women. Odors were presented using standardized pen-shaped testing devices. Based on preliminary testing, the researchers selected 12 odors: coconut, anise, lemon, lavender, mushroom, grass, caramel, cola, banana, eucalyptus, onion, and rose.

The letter cues left only two letters of the odor’s name visible, with blanks replacing the others. All were presented in Turkish. The color cues used representative colors selected from images of the odor sources. Both cue types therefore supplied visual information related to the correct answer.

Each participant named six odors with letter cues and the remaining six with color cues. These assignments were reversed for half the participants so that the ease of naming particular odors would not favor one cue type alone. Each odor was tested six times, yielding 72 trials per person and 2,880 trials overall.

Each trial proceeded through smelling the odor, waiting briefly, seeing the cue, and thinking for four seconds. When the response prompt then appeared, participants pressed the space bar and said the odor’s name aloud.

Letter cues produced more correct names

Correct names were given in 763 of 1,440 letter-cue trials and 258 of 1,440 color-cue trials. As proportions of all trials, these were approximately 53% and 18%. These were repeated responses by 40 people, not observations from 2,880 different participants.

A statistical model accounting for repetition within individuals and differences among odors likewise found that, among trials in which an answer was given, correct responses were more likely with letter cues. In contrast, the probability of giving any answer, whether correct or incorrect, did not differ significantly between the conditions.

The letter cues therefore appear to have brought the answers closer to the correct names rather than increasing the number of attempts. However, letters that directly reveal part of the answer and a color that can evoke many objects do not necessarily provide equal amounts of information for narrowing the possibilities.

Greater accuracy did not mean shorter response times

Average response times were approximately 2.27 seconds with letter cues and 2.30 seconds with color cues, with no significant difference. Here, response time was measured from the prompt appearing after the four-second thinking period until the keypress, not from the start of smelling.

It therefore does not establish that the entire process of thinking of a name took the same amount of time. Nevertheless, with this measurement method, the accuracy advantage from letter cues did not translate into a difference in response speed. The authors argue that retrieving the correct name and initiating a response should be considered separately.

Familiarity, liking, and the feeling of almost knowing also mattered

Participants rated familiarity, liking, and the feeling of knowing the odor but being unable to retrieve its name on nine-point scales. Records with higher familiarity or liking ratings tended to show greater accuracy and shorter response times. However, associations with speed were not large, and the correlation with liking was particularly weak.

The feeling of being unable to retrieve a name was stronger for records in which an incorrect name was given than for those with no answer. Individual comparisons between correct responses and the other response categories were not significant. A strong feeling of almost knowing does not necessarily lead to the correct answer.

The authors interpret this feeling as a clue that some associated information is available but the correct name cannot be retrieved, rather than a complete failure to access memory. Recognizing an odor as familiar and being able to name it are not the same thing.

Color cues showed more widespread coordination in brain-wave signals

The researchers recorded brain waves using electrodes at 32 scalp locations and analyzed the four seconds after the cue disappeared. They estimated the direction of coordinated activity using a method that tests how well past values of one signal predict future values of another.

This analysis found stronger coordination between more electrode pairs for color cues than for letter cues. Letter cues showed relatively limited combinations linking frontal and lateral locations or frontal and parietal locations, whereas color cues showed more widespread combinations that included the back of the head.

This coordination does not count actual connections between neurons. It is a statistical relationship among signals recorded on the scalp. On its own, it cannot establish that neural activity at one location directly drove another or that a particular brain area alone retrieved the name.

More widespread activity does not necessarily mean better performance

The authors discuss the possibility that seeing part of a word offers relatively direct access to candidate names, whereas a color may require several steps: associating it with an object, linking that object to an odor, and searching for its name. Yellow, for example, can suggest many things besides lemon.

The combination of more widespread coordination and lower naming accuracy in the color condition led to the interpretation that broader activity does not necessarily mean better performance. Using letters to access a more restricted set of relevant information may have been more efficient in this naming task. This is a hypothesis explaining the measurements, not direct proof of processing pathways or cognitive load.

Analyses of brain-wave coordination remain affected by electricity from the same source reaching multiple electrodes and by signal-processing choices. The authors also call for confirmation using other analytical methods and brain-measurement techniques with better spatial localization.

Whether cues helped compared with no cue is a separate question

The study did not include a condition without cues or with cues unrelated to the answer. It established that, among two kinds of relevant cues, partial letters made correct naming more likely than colors. It did not demonstrate sharper olfaction itself or an improved ability to detect odors.

Scoring counted only responses matching the correct name, so associating the odor with something similar still counted as incorrect. Whether the findings from Turkish speakers and a limited set of test odors extend to other languages or complex everyday scents remains to be investigated.

Publication was supported by the University of Twente’s BMS Open Access Fund 2025, and the authors declared no commercial or financial conflicts of interest.

Summary

In an odor-naming task, cues revealing two letters of a name led to more accurate answers than color cues. However, the time from the response prompt to the keypress did not differ, while brain-wave coordination was more widespread in the color condition.

The authors emphasize that naming an odor is not solely an ability of the nose: it also involves retrieving words from memories of experience and meaning. Different accompanying cues may support different parts of that process. Distinguishing knowing an odor from being able to name it may also help us understand differences in ability when assessing olfaction.

References

Capkan, E. N., & Yildirim, F. (2025). Cues for odor naming affect performance and brain connectivity. Frontiers in Human Neuroscience, 19, 1671670. DOI: 10.3389/fnhum.2025.1671670. Full text. © 2025 Capkan and Yildirim. The original article is published under CC BY.