The one where Jennifer Aniston taught us how the brain works

Sarah Brockway
2 days ago
7 min read

Writer: Sarah Brockway

Editors: Sam Alper, Madelyn Miles, and Madeline Buhman

Illustrator: Lo Kronheim

Think about your weekday morning routine before you head to work. Your alarm goes off, maybe you hit the snooze button a couple times, then you get up and (hopefully) brush your teeth. Have you ever wondered how your brain is able to distinguish between a tube of toothpaste and the other containers in your medicine cabinet, seemingly in the blink of an eye? Even if your toothpaste and your shaving cream container are both red and white, cylindrical in shape, and sitting next to each other in your cabinet, your brain can quickly identify that your toothpaste is “the thing I use to clean my teeth” versus, for example, “the thing I use to shave.”

Throughout your day, your brain recognizes hundreds of representations: people, places, or things that mean something about the world we live in. While your toothpaste might just be an object in your medicine cabinet, your brain has also categorized it as an important tool to get your teeth clean. This ability to remember what things are important, and why they are important, is how the brain makes sense of the world around us. Your sister may mean more to you than just a familiar face, and you may associate a slice of Hawaiian pizza with everything that is good (or bad) about society.

Our brain is constantly assigning meaning to the things we encounter, which helps us quickly remember who, what, and where is important to us. But how do neurons – the cells in our brain – work together to help us recognize all the representations we need to in our everyday lives?

How does the brain encode representations?

A representation (for example, the things we use to clean our teeth) is much more abstract than a single person, place, or thing (like a tube of toothpaste). Because of this, neuroscientists have spent decades trying to figure out how neurons work together to identify and differentiate between representations – in other words, how neurons encode different representations. Two different strategies have been proposed to explain how the brain can take in all the items at the bathroom sink and recognize what we need to get our teeth clean. Strategy 1 is that the brain uses a large number of neurons that work together to encode a representation. Think about a stadium full of people talking at a conversational volume, combining to create a very loud sound. Strategy 2, on the other hand, is that only a few neurons are needed to encode a representation. Think now about the same stadium in which everyone is silent except for 10% of the audience scattered throughout the stadium that is shouting at the top of their lungs. In each of these scenarios, the total volume in the stadium is roughly the same, but the number of individuals contributing to the noise is vastly different.

So which of these strategies does the brain actually use to encode a given representation? Are many neurons acting in concert, or are only a few neurons needed to do the job? In 2005, neuroscientist Rodrigo Quian Quiroga and others at the California Institute of Technology published an article in Nature that was one of the first to answer this question.

Recording neuron activity in the human brain

To answer their question, Quiroga and colleagues needed a way to record activity from neurons deep inside the brain. This is typically very hard to do without causing harm to human subjects during the process. Luckily, the researchers had the opportunity to work with epilepsy patients who had electrodes implanted in their brains to help doctors track their seizures. The researchers identified eight patients with electrodes located in memory regions of the brain, such as the hippocampus. Each of these patients was shown many different photos of well-known celebrities and world landmarks. As the patients viewed each photo, the electrodes in their brain recorded neuron activity. This approach allowed the researchers to measure the activity of over a thousand neurons across the eight patients and find individual neurons that were active in response to the photos they presented.

What did the researchers find out?

Quiroga first identified a single neuron in the hippocampus of one epilepsy patient that repeatedly activated in response to photos of Jennifer Aniston, the beloved actress from the 90s and 2000s hit show Friends. This neuron activated when the patient was shown many different paparazzi-esque photos of Jennifer Aniston as she faced in different directions relative to the camera, wore different facial expressions, and featured different hairstyles (including the Rachel, of course). This Jennifer Aniston-loving neuron did not activate in response to photos of other celebrities like Julia Roberts or Kobe Bryant. This told the researchers that the neuron did not just activate in response to seeing any famous person, or even in response to seeing any white, brunette actress. What was really interesting was that this Jennifer Aniston neuron also did not activate in response to photos of the actress with her then-sweetheart, Brad Pitt. The researchers took this to mean that the neuron encoded the concept of “Jennifer Aniston” but not “Jennifer Aniston plus boyfriend” or “Jennifer Aniston plus Brad Pitt.” This suggested that a single neuron could encode an extremely specific representation, such as the actress who played Rachel Green in Friends.

But how do we really know that individual neurons encode the idea

of someone or something rather than their physical appearance?

We know that the Jennifer Aniston neuron found in Patient 1 was active in response to many different photos of the actress. But what if that neuron’s activity was associated with the patient seeing the physical traits of Jennifer Aniston, and not with the concept of Jennifer Aniston? The researchers suspected that visual information was not the primary reason that the Jennifer Aniston neuron was activating. If this were the case, the neuron would have been active when Patient 1 was presented with photos of Jennifer Aniston with Brad Pitt. To validate this conclusion, Quiroga took the experiment a step further by examining neuron

activity in additional epilepsy patients. They next found a neuron in the hippocampus of a second patient that activated only in response to the actress Halle Berry. The Halle Berry neuron not only activated when Patient 2 was shown photos of Halle Berry as herself and as Catwoman, one of her most iconic roles, but also in response to black and white drawings of Halle Berry, and even the words “Halle Berry” set in text. This convinced the researchers that a single neuron could encode the idea of someone rather than simply the way they looked.

The researchers then extended this major finding from the idea of a person to the idea of places and things as well. A third epilepsy patient had a hippocampal neuron that activated in response to photos of both the Taj Mahal and the Sydney Opera House. While these buildings are located in different places around the world and serve different purposes, Patient 3 identified them as the same landmark due to their somewhat similar appearance. The fact that a single neuron fired in response to both buildings – because Patient 3 personally categorized them as the same building – told the researchers that a single neuron could encode the abstract idea of the Taj Sydney Mahal Opera House (as I like to think it existed in Patient 3’s mind).

Taking all of these findings together, Quiroga and colleagues concluded that single neurons can encode very specific representations, such as distinct people and places, and that the representations encoded are unique to the individual depending on their own perception of the world around them.

Returning to the two strategies for neuronal representation

Given that a single neuron can encode specific abstract ideas, it is likely that that only a small number of neurons is needed for a person to identify and distinguish between representations. This means that the brain likely uses Strategy 2 to encode representations. Relating this to our stadium analogy, only a few scattered crowd members need to shout at one time to create the crowd noise. It’s the individual members of the crowd, including their location in the stadium, the pitch of their voice, and how long they shout, that create different sounds – or, in the case of the brain, create different representations.

Does this mean that every single neuron in the brain encodes the representation of only one person, place, or thing? The answer is almost certainly no. Quiroga’s study, while innovative and exciting, does not provide direct evidence for the idea that the Jennifer Aniston neuron in Patient 1 is the only neuron in their brain that activates in response to Jennifer Aniston. Realistically, there probably aren’t enough neurons in memory regions of the brain to have that sort of one-to-one specificity for each representation we need to remember. It’s likely that the Jennifer Aniston neuron activates along with a few other neurons scattered throughout the brain to help encode the idea of Jennifer Aniston. This is ultimately a good thing, because if the function of one Jennifer Aniston neuron is lost, you can still remember who Jennifer Aniston is thanks to a few other neurons. By using a small number of neurons at a time to encode a specific representation, the brain can keep its facts straight and make sure that all the people, places, and things that are important to us can be remembered when it matters.

We still have a lot to learn about how the brain encodes all the representations that we need to remember, but innovative studies like this one that take advantage of unique clinical situations have propelled memory research forward in exciting ways. I’m not sure if Jennifer Aniston knows that research using her face ended up teaching us a lot about how our brains work, but I’m sure she’d be excited to find out that one of her fans is a tiny neuron in the hippocampus of an anonymous epilepsy patient – and maybe she would take a tiny bit of satisfaction in knowing that this neuron was decisively anti-Brad Pitt.