When learning an unfamiliar scent, a short name might seem to make it easier to organize in our minds. But even if a name has no meaning, can sounding like one’s native language help with learning categories?
A study published in 2025 by Yaxiong Cao and colleagues examined this question in people who used Chinese and English. It sought to separate the effect of adding words to scents from the effect of those words sounding like a familiar language.
Using meaningless temporary names rather than real names
The participants were 80 native Chinese speakers with high English proficiency. Most were recruited from students majoring in English at Chinese universities. Those who passed preliminary screening, including an odor-discrimination test, were randomly assigned to four groups of 20.
Three groups heard pseudowords with Chinese-like, English-like, or Georgian-like sounds, respectively. Pseudowords are new, meaningless words constructed according to a language’s sound patterns. None of the participants was familiar with Georgian. The remaining group learned using odors alone, without accompanying words.
This was not an experiment teaching familiar, meaningful names such as rose or banana. Participants were not shown images identifying the odor source or given its actual name.
Learning assigned pairings rather than natural similarities
The experiment used 18 odors arranged in six sets of three. In each trial, participants smelled odors A and B, followed by a reference odor X, and chose whether A or B was more similar to X. In the word groups, a recording of the corresponding pseudoword was played before each odor.
Importantly, the correct answer was not determined by the odors’ inherent similarity. The researchers assigned in advance whether A or B would be paired with X. Participants learned the pairings for the task through feedback on the screen indicating whether their answer was correct.
They practiced 24 trials a day for four days, totaling 96 trials. The odor–pseudoword pairings were not completely consistent on every occasion. In the test after training, all participants responded without hearing words. This examined whether practicing with temporary names allowed them to retain the categories when only odors were available.
Categories were learned, but an added benefit from names was unclear
Accuracy improved in every group from about 44–48% on the first day. Final-test averages were approximately 68% without words, 68% with Chinese-like words, 69% with English-like words, and 77% with Georgian-like words.
Looking only at these figures, the group that heard the least familiar sounds performed best. However, differences in improvement between groups were not statistically clear. Comparisons of learning curves likewise did not show clear differences between any word group and the no-word group.
Another statistical approach provided moderate support for no difference in many comparisons. However, the comparison between the Georgian-like and no-word groups was inconclusive, supporting neither a difference nor its absence clearly. It is therefore not possible to conclude either that unfamiliar-language sounds were superior or that words can never help.
What is needed for names to help?
The authors consider several possible reasons why an advantage from words was not clear in this experiment. One is that the association between an odor and its temporary name was not consistent on every occasion. This may have made stable associations harder to form, weakening the effect of naming.
Processing both meaningless words and odors may also have been more demanding initially than learning odors alone. Another proposed explanation is that familiar sounds evoke existing words or written characters, making it more complicated to associate them with new odors.
These explanations remain untested hypotheses. The present results alone do not show whether the Georgian-like group’s higher numerical performance reflected fewer distracting associations or differences in attention.
With only 20 participants per group, the study may have missed small effects. It also involved an artificial categorization task over four days, so its findings cannot be directly generalized to situations involving meaningful scent names or long-term memory. The original article contains inconsistencies between the sample sizes reported in the methods and some individual improvement percentages, so detailed proportions also require caution.
Summary
Participants learned the odor categories assigned by the task, with or without words. However, giving odors temporary names or making those names sound like the native language did not demonstrate a clear additional benefit.
The authors emphasize that whether names help perception should not be considered solely in terms of whether words are present. How consistently is a name linked to an odor? Does it have meaning? What experience does the person using it bring? Systematically varying these conditions can help explain how language might be used in learning scents.
References
Cao, Y., Majid, A., & Vanek, N. (2025). Not all verbal labels grease the wheels of odor categories. Language and Cognition, 17, e33, 1–24. DOI: 10.1017/langcog.2025.1 (Full text ).
Original article: © The Author(s), 2025, published by Cambridge University Press, CC BY 4.0.
