Have you ever felt anxious just by seeing something that reminds you of a bad experience—even if it didn’t harm you directly? For example, a child stung by a wasp might start fearing the wasp’s nest, even though the nest itself never hurt her. This kind of response is called emotional inference—and for the first time, scientists have uncovered how our brain makes this happen.
Researchers Xiaowei Gu and Joshua Johansen at the RIKEN Center for Brain Science in Japan have made a significant discovery. They identified how certain brain circuits in rats allow them to "infer" emotions from past experiences, much like humans do. Their research, recently published in the prestigious journal Nature, provides the first clear explanation of how our brains build internal emotional models—a key part of higher-order thinking.
Let’s break down this fascinating study.
 |
| Cartoon showing an example of how inferred emotions are learned. A child often watches wasps fly in and out of their nest in the woods near her house. One day she is stung by one of the wasps for the first time. Afterward, seeing the wasp nest alone makes her feel anxious. In this study neurons in the medial prefrontal cortex of the brain, and their connection to the amygdala, were found to be essential for this type of learning to occur. Credit: RIKEN |
What Are Inferred Emotions?
Imagine this scenario:
A child plays near the woods every day and notices a wasp flying in and out of its nest. One day, she gets stung. That painful experience leaves her frightened. From then on, even if she just sees the nest (and not the wasp), she feels scared.
Why? Because her brain has created a link between the nest and the sting—even though the nest never caused harm directly. Her emotional reaction is based on inference, not direct experience. This mental connection between a harmless object (the nest) and a negative feeling (fear) is what scientists call emotional inference.
This type of thinking is very different from basic learning. It shows how the brain builds internal models—mental maps that help us navigate life based on experience, memory, and assumptions.
Until now, researchers didn’t know how the brain made these emotional inferences. That’s what makes Gu and Johansen’s study so important.
The Experiment: How Rats Learned Inferred Emotions
To understand how inferred emotions work in the brain, Gu and Johansen developed a smart and simple experiment with rats.
Step 1: Teaching a Basic Link
The rats were first trained to associate a harmless sound (a tone) with a neutral image (like a shape on a screen). This step was simple—just creating a mental connection between the two things.
Step 2: Creating a Negative Emotion
Later, the same image was shown again—but this time it was paired with something unpleasant, like a mild electric shock or another discomfort. This is known as aversive conditioning. The goal here was to teach the rat that the image now means something bad.
Step 3: Testing Emotional Inference
The next day, the rats were played the original sound, without the image or the shock. If the rats had learned to infer, they would feel scared just by hearing the sound—because it reminded them of the image, which was linked to the bad experience.
And that’s exactly what happened. The rats froze in fear when they heard the sound. This showed they could infer a negative emotion from an indirect experience, just like the child who feared the wasp's nest.
Digging Deeper: How the Brain Makes Emotional Inference Happen
After proving that rats can learn inferred emotions, the researchers wanted to discover which parts of the brain were responsible.
They focused on two main brain regions:
-
Medial Prefrontal Cortex (mPFC): This area is involved in planning, decision-making, and complex emotional thinking.
-
Amygdala: A small, almond-shaped structure that is crucial for emotional learning, especially fear.
Monitoring the Brain
To track brain activity, the scientists used calcium imaging—a technique that lights up neurons when they become active. They also used optogenetics, which allows scientists to turn specific neurons on or off using light.
What They Found
-
Before aversive learning, neurons in the mPFC responded in the same way to both the image and the sound—nothing special.
-
But after the rats experienced the unpleasant event, something changed. The neurons that responded to both the image and the sound became more active, but only if the two had been linked earlier.
This suggests the brain "tagged" these neurons during the first training, making them ready to react strongly later when the image became associated with something bad.
Why the mPFC Is the Key to Emotional Inference
To prove the mPFC was truly responsible, the researchers tried blocking this area during the learning phase.
Here’s what happened:
-
When the mPFC was turned off during the aversive learning, the rats could not make the emotional inference. Even though they saw the image and felt discomfort, their brains couldn’t link that to the earlier sound.
-
Later, when researchers blocked the mPFC’s connection to the amygdala during testing, the rats again failed to react with fear to the sound. This time, the memory had been formed—but they couldn’t recall it correctly.
-
Importantly, the rats still froze in fear when shown the image itself, meaning their basic memory was intact. Only their inference-based emotion was affected.
The Bigger Picture: What This Means for Human Emotions
According to Johansen, “Decades of research have shown that the amygdala stores simple emotional memories that come from direct experience. But our new findings reveal that the mPFC is crucial for complex, human-like emotional processing that involves thinking, linking, and inferring.”
This discovery has massive implications.
It suggests that higher-level emotions, like those involved in anxiety, trauma, or even empathy, rely on this system of internal modeling. Understanding how this system works could lead to better treatments for mental health conditions, such as:
-
Post-Traumatic Stress Disorder (PTSD)
-
Generalized Anxiety Disorder
-
Phobias
-
Depression
Many of these conditions involve emotional responses that are out of proportion or based on indirect experiences. With more research, scientists could find ways to correct or adjust faulty emotional inferences—offering hope for people suffering from emotional disorders.
Why Rats? And Why This Matters
Some people may wonder why this study was done in rats. The answer is simple: rats and humans share many brain structures and basic learning mechanisms. What we learn from animal models often gives us the foundation to understand the human brain.
By identifying how emotional inference works in rats, scientists can now explore similar mechanisms in humans using tools like brain scans, behavioral studies, and neural stimulation.
This kind of research helps bridge the gap between basic neuroscience and clinical psychiatry, allowing us to design better therapies rooted in how the brain actually works.
Conclusion: A New Frontier in Understanding Emotions
Gu and Johansen’s discovery is more than a scientific breakthrough—it’s a paradigm shift in how we understand emotional thinking. By showing how the brain makes inferences from experience, they’ve opened the door to exploring the deeper layers of emotion, ones that are central to human behavior.
Their work reveals that:
-
Emotional inference is a real and measurable process.
-
It depends on specific brain circuits linking the mPFC and the amygdala.
-
Understanding these circuits could transform mental health research.
As Johansen says, this study marks the beginning of a new era in emotional neuroscience. It shows that our emotions aren’t just reactions—they’re predictions, inferences, and part of a rich internal model that our brain constantly updates to help us survive, learn, and relate to others.
With this knowledge, we can begin to understand—and perhaps heal—the emotional worlds inside us all.
Reference: Gu, X., Johansen, J.P. Prefrontal encoding of an internal model for emotional inference. Nature (2025). https://doi.org/10.1038/s41586-025-09001-2
Comments
Post a Comment