Unlocking Social Learning: How Your Preferences Shape What You Believe

Mira Elwood in Mind & Education August 2025 • 5 min read.

"Forget Unbounded Beliefs: Discover the surprising interplay between your personal preferences and the information you absorb from others, and how it impacts your ability to learn and adapt."

Imagine navigating a world of constant information, where every choice you make is influenced by the actions and opinions of those around you. This is the realm of social learning, a process where we observe others, gather information, and adjust our own beliefs and behaviors accordingly. But what if the way we learn isn't just about the information itself, but also about our own deeply held preferences? Recent research is challenging long-held assumptions about how societies learn, revealing the surprising role that our individual tastes and biases play in shaping collective knowledge.

For years, the prevailing theory suggested that 'unbounded beliefs' were essential for effective social learning. This meant that people needed to be open to even the most unlikely possibilities, capable of drastically changing their minds based on new information. However, this idea doesn't quite align with the reality of how we process information. We all have our preferences, things we like and dislike, and these biases inevitably color how we interpret what we see and hear.

A new study is turning this conventional wisdom on its head, introducing a concept called 'excludability.' This novel framework highlights the crucial interplay between our preferences and the information we receive, demonstrating that societies can still learn effectively, even without everyone embracing unbounded beliefs. In fact, the research suggests that in many cases, our preferences can actually guide us toward the truth.

Excludability: The Key to Social Learning?

A network of interconnected minds, symbolizing how individual preferences influence social learning.

The researchers identify a simple condition for learning they call 'excludability,' which is a joint property of agents' preferences and their information. Excludability suggests that it's not necessary for individuals to be completely open to all possibilities, but rather that they can effectively rule out incorrect actions, even if they don't immediately know the 'correct' one. This subtle distinction has profound implications for our understanding of social learning.

Excludability requires that for each pair of actions, a single agent must be able to receive a signal that makes her arbitrarily convinced that one action is better than the other, no matter what full-support belief she starts with. Put differently, information must be able to distinguish the set of states in which one action is better than another. Excludability implies that society can never get stuck on a wrong action: if an action is suboptimal at the true state, then some agent will receive a private signal convincing her not to take that action. We establish that this property of displacing wrong actions leads to social learning.

Social learning
Herds
Information cascades
Single crossing
Euclidean preferences
Location-shift information
Unbounded beliefs

To illustrate, imagine a group of friends trying to decide which new restaurant to try. Some might be adventurous eaters, willing to try anything, while others might have strong preferences for certain cuisines or aversions to others. Excludability suggests that even if some friends are hesitant to step outside their comfort zones, the group can still make a good decision if at least one person is willing to try new things and can effectively communicate their experiences, good or bad, to the rest of the group.

Why Preferences Matter

These new findings highlight the importance of considering individual preferences when analyzing how societies learn. It's not enough to simply assume that everyone is a perfectly rational information processor. Our biases, tastes, and predispositions play a significant role in shaping what we believe and how we act. Understanding this interplay is crucial for designing effective communication strategies, promoting informed decision-making, and fostering a more nuanced understanding of social phenomena.

About this Article -

This article was crafted using a human-AI hybrid and collaborative approach. AI assisted our team with initial drafting, research insights, identifying key questions, and image generation. Our human editors guided topic selection, defined the angle, structured the content, ensured factual accuracy and relevance, refined the tone, and conducted thorough editing to deliver helpful, high-quality information.See our About page for more information.

Everything You Need To Know

How does 'excludability' differ from the idea of 'unbounded beliefs' in the context of social learning?

The concept of 'unbounded beliefs' suggests that effective social learning requires individuals to be open to all possibilities, even unlikely ones, and drastically change their minds based on new information. 'Excludability', on the other hand, proposes that individuals don't need to embrace all possibilities but rather be able to rule out incorrect actions. 'Excludability' focuses on displacing wrong actions and requires that for each pair of actions, a single agent must be able to receive a signal that makes them arbitrarily convinced that one action is better than the other, no matter what full-support belief they start with, while 'unbounded beliefs' suggests that individual should be open to all information.

What implications does the concept of 'excludability' have for understanding how societies learn?

'Excludability' suggests that societies can still learn effectively, even if not everyone is open to all possibilities, because learning to exclude incorrect options facilitates that learning. This challenges the conventional assumption that 'unbounded beliefs' are necessary for effective social learning. It also highlights the role of individual preferences and information diversity in the learning process. According to 'excludability', an agent will receive a private signal convincing them not to take that action if an action is suboptimal at the true state. The 'excludability' is a joint property of agents' preferences and their information, and information must be able to distinguish the set of states in which one action is better than another.

How do personal preferences influence the process of social learning, and what are the consequences of these influences?

Personal preferences significantly shape the process of social learning by influencing how individuals interpret and absorb information from others. These biases, whether conscious or unconscious, affect what we believe and how we act. These preferences, along with the information individuals receive, are central to 'excludability'. A consequence is that people aren't perfectly rational information processors. This understanding helps in designing communication strategies and promoting informed decision-making, fostering a nuanced understanding of group dynamics like 'herds' and 'information cascades'.

Can you provide an example illustrating how 'excludability' works in a real-world scenario?

Consider a group deciding on a restaurant. Some may have strong culinary preferences, while others are adventurous. According to 'excludability', the group can still make a good decision if at least one person is willing to try new things and effectively communicates their experiences to the group. The adventurous person might have 'location-shift information' about one action being better than another, no matter the full-support belief of the other individuals. This information, combined with the group's 'single crossing', allows the group to exclude certain actions and decide on a restaurant.

How does considering individual preferences, such as 'Euclidean preferences', impact the analysis and effectiveness of social learning models?

Considering individual preferences in social learning models provides a more realistic and nuanced understanding of how societies learn. Traditional models often assume everyone is a rational information processor, neglecting the role of personal tastes and biases. Incorporating preferences, such as 'Euclidean preferences', acknowledges that people's predispositions shape their beliefs and actions. This affects how individuals perceive and internalize new information, which in turn alters collective knowledge. It's important to consider individual preferences because it is not enough to simply assume that everyone is a perfectly rational information processor.