One day, in the deep sea, I went into the blue. Scuba divers are supposed to stay together and follow the reef, but I had drifted away. Above, below and all around was a soft blue haze. The pressure gauge needle pointed outside the safe range but also indicated enough air which was surprising yet also somehow made sense.

I later learned that my scuba-diving buddy pulled me up as I was sinking fast and that hallucinations, such as me thinking my pressure gauge had gone up to 11 quickly arise under sensory deprivation.

Several years later, diving into social neuroscience, it seemed to me that the phenomenon of loneliness was similar to that experience of going into the blue. Beyond the emptiness, it was the lack of social others as reference points in our mental space. We often talk about other people as locations: “He and I got closer”; “They have drifted apart over the years”; “I have a tight social circle”; “She sure is climbing up the social ladder”; and so on. It is almost impossible to describe social relationships without borrowing a spatial metaphor.

Is this a language habit or something deeper about the way our brain handles social life?

We can think of social relationships as a navigation problem: Imagine arriving to a new town. You need to get acquainted with the townspeople in order to find a job and a place to stay. One person approaches you; she is friendly and likes to hang. Another you encounter is a top developer; he is domineering and vain. His assistant is on the admiring side, willing to do anything for you as well.

How would you act? We can envision a game board framed by power and affiliation, where every interaction moves these characters around. Within this space, the trajectories of our relationships evolve. Their paths tell the stories of friendships and foes. Does the brain keep track of social dynamics just as it tracks us moving here and about?

To answer that, we should consider how the brain navigates real, physical space. Studies of spatial navigation typically record the activity of neurons firing when an animal explores the environment. This approach reveals location-specific neurons in a brain region called the hippocampus: a single neuron fires when the animal visits a certain place, and a different neuron fires when it visits the next. The sheer beauty of superimposing neural firing on geographical maps shows how the outside world maps onto our neural landscape.

To examine whether a similar social landscape exists in the brain, my lab used a social interaction game. Participants chose their own adventure by going through a story line; they met different characters and decided how to interact with them. Their choices shaped relationships along the dimensions of power and affiliation. Choosing to abide by one character’s demands, for example, elevated her in power. Engaging another in personal conversation enhanced the affiliation between that character and the participant.

By translating the choices into coordinates on a two-dimensional social space, we could ask whether the brain encoded the characters’ location as the story progressed. Consider an imaginary line drawn between yourself and every character you interact with in this space. The orientation and length of that line, or vector, will signify the whereabouts of each character in your social space. Scanning participants’ brains using functional magnetic resonance imaging as they played the social game enabled us to search for neural signals that corresponded to the characters’ trajectories as the relationships in the story evolved.

We found that the hippocampus tracked that vector’s orientation and that the posterior cingulate cortex tracked its length. These two brain regions, members of a network that supports the mapping of movement in physical space, also track trajectories of the social kind—abstract trajectories that lack any physical reality but are structured and organized alike. Other labs have veered into spaces of odors, sounds and all manner of arbitrary domains. Organizing information along continuous dimensions and tracking the relationships between them constitute a cognitive map. This concept was formulated by American psychologist Edward Tolman in 1948 and is now matched with a neural code.

With multiplex spaces utilizing navigational computations, we can no longer think of the hippocampal formation as the brain’s earthly GPS. The navigation system of the brain organizes information in a relational manner in various realms, from physical arenas to fully abstract terrains. As we progressively build a detailed, mechanistic understanding of how navigational computations are implemented in single neurons and brain networks, we approach the ultimate question: How does the brain integrate maps with multiple dimensions and various domains into one coherent life space?

Editor's note: This post was produced in partnership with the annual Falling Walls Conference in Berlin, which coincides with the anniversary of the fall of the Berlin Wall and showcases the work of researchers from around the world, including the author of this essay.