In a groundbreaking study, researchers from the University of Trento, the University of Lincoln, the University of London, and the University of Vienna have shown that rats possess the ability to distinguish between two varieties of white wine, Riesling and Sauvignon Blanc. Published in Animal Cognition, the study challenges our understanding of animal cognition, suggesting that non-human animals can categorize complex olfactory stimuli without relying on linguistic or human-specific cognitive abilities. This discovery opens the door to exploring the broader cognitive capabilities of non-human animals, particularly in complex sensory domains such as smell.
Study Overview and Objectives
The study titled “Rats Can Distinguish (and Generalize) Among Two White Wine Varieties” aimed to investigate whether rats could accurately discriminate between wines of two different grape varieties, something traditionally thought to require a higher level of cognitive processing. Researchers employed an innovative olfactory discrimination method based on a go/no-go task, which has been used in the past to test the sensory capabilities of various species. The experiment’s goal was to determine if rats could learn and generalize the categorization of complex smells, like those found in Riesling and Sauvignon Blanc, in a way that mirrors the sensory distinctions made by human wine tasters or sommeliers.
Nine adult male domesticated rats were used in the experiment. These rats had already been trained on an unrelated odor categorization task, which meant they had some prior experience with differentiating scents. The researchers used an automated odor delivery system and a set of chambers where the rats were required to perform a specific action (a nose poke) to trigger a lever presentation. Depending on their response, they either received a food reward for the correct choice or a penalty for an incorrect one. The rats had to distinguish between two wine varieties: one (Riesling or Sauvignon Blanc) was associated with a food reward (S+), while the other (the other wine variety) was associated with no reward (S−).
Training and Methodology
The training process was designed to be rigorous and involved repeated exposure to both wine varieties. To ensure the rats could generalize their learned responses, the researchers presented not only multiple samples of each wine variety but also wines from different vintages and geographic regions. This helped ensure that the rats were not just associating a single sample of each wine variety with the rewards but instead were learning to recognize the wines as categories.
Each rat underwent extensive training until they achieved an accuracy of over 80% correct responses for three consecutive sessions. The experimental setup included automated odor delivery systems to present the wine odors in controlled conditions, ensuring that the rats’ responses were based solely on their ability to discriminate the smells, not external distractions. In the go/no-go task, the rats were required to perform a nose poke to initiate the task, after which the lever would be presented. If they correctly identified the rewarded wine variety (S+), they were rewarded with food, while an incorrect response led to a timeout penalty.
Findings and Results
The results of the experiment were both surprising and informative. All nine rats successfully learned to distinguish between the two white wine varieties, with a majority also able to generalize their learned categorization to novel wines that were not part of the initial training samples. In other words, the rats demonstrated that they were not just memorizing specific wines, but rather recognizing and applying the category-based distinction between the two wine varieties to new and unfamiliar examples.
On the testing phase, the rats were presented with novel wines that were from the same grape varieties but from different samples that the rats had not encountered during training. The rats were able to press the lever more frequently when they were presented with a wine that matched their trained rewarded category (S+), indicating that they had generalized the concept of wine variety and successfully applied it to novel instances. In contrast, they responded less frequently to wines that fell into the non-rewarded category (S−).
Statistical analysis of the rats’ responses showed a significant difference between their behavior toward novel wines that matched their trained categories and those that did not. Using a generalized linear mixed model, researchers were able to confirm that the rats had generalized the learned categorization across different samples of the wines. Notably, the rats correctly identified 94% of the trained wines, and even in novel cases, they correctly identified 65% of the wines, with the accuracy improving with repeated exposure.
This ability to generalize categorization across different wine samples was a key finding of the study, suggesting that rats can not only differentiate complex odors but also apply that discrimination to new situations, much like humans might apply the learned categories of wine types in a tasting scenario.
Peanuts’ Unusual Behavior
One interesting aspect of the experiment involved a rat named Peanuts. Despite successfully completing the training phase like the other rats, Peanuts exhibited an unusual response pattern during the testing phase. While all other rats generalized their responses to novel wines, Peanuts showed a higher tendency to respond incorrectly to S− wines (the non-rewarded category) in the testing phase. This means that Peanuts seemed to press the lever more frequently for the wrong wine variety, deviating from the behavior of the rest of the rats.
This anomaly could suggest a few things about olfactory processing in rats. It’s possible that Peanuts had a different olfactory sensitivity or a more complex response to the wines, which led to inconsistent choices. Another explanation could be that Peanuts, in a manner similar to human connoisseurs, might have found the wines interesting in their own right, appreciating the distinct attributes of each wine variety. This suggests that individual differences in olfactory processing or memory consolidation might play a role in how rats respond to complex smells, especially when it comes to categorization tasks.
The Role of Mixed Training in Olfactory Learning
A key methodological innovation of the study was the use of mixed training. Unlike traditional scent discrimination tasks, where rats are trained on one odor at a time, the researchers used a mixed odor presentation approach. This approach involved introducing multiple wine samples of both varieties within the same session, rather than training the rats on just one wine at a time. This technique likely accelerated the rats’ ability to generalize their learning, suggesting that mixed training may be an effective way to enhance olfactory categorization skills. In contrast, sequential training methods, which focus on one target odor before moving on to another, may not foster the same level of generalization.
The mixed training approach allowed the rats to process the wines simultaneously, increasing the cognitive complexity of the task. This method could be a valuable tool in future studies of olfactory learning, especially when investigating how animals process complex sensory inputs and make distinctions in environments that require simultaneous comparisons of different stimuli.
Broader Implications and Future Research
The findings of this study suggest that rats, like other animals, are capable of categorizing complex olfactory stimuli without needing linguistic or human-like cognitive abilities. This has important implications for our understanding of animal cognition. It shows that even species with relatively small brains, like rats, are capable of sophisticated mental processes like generalization, which has typically been considered a hallmark of higher cognitive functions in humans.
Further studies may explore whether rats are capable of more complex forms of olfactory discrimination, such as distinguishing wines based on terroir, winemaking style, or even the region of origin. Additionally, researchers could investigate whether other species, such as dogs or primates, share similar abilities in distinguishing and generalizing between complex scents, expanding our understanding of olfactory processing across different mammals.
In conclusion, the ability of rats to distinguish between two complex varieties of wine underscores the remarkable olfactory capabilities of non-human animals. This research not only advances our understanding of animal sensory perception but also suggests that complex olfactory categorization may be more widespread in the animal kingdom than previously thought. The study opens up many exciting avenues for further research into how animals process smells and how those abilities compare to human olfactory cognition.
Reference: Elisa Frasnelli et al, Rats can distinguish (and generalize) among two white wine varieties, Animal Cognition (2025). DOI: 10.1007/s10071-025-01937-2