Information Processing, Human Error and the Learning Process — Page 147, Lesson 235

I want to walk you through a really important concept in how your brain handles information — something called “bottom-up” and “top-down” processing. This is central to understanding how we perceive, remember, and solve problems, especially in the cockpit where you're constantly bombarded with sensory input. Let’s start with the core idea. Many mental activities — things like remembering, perceiving, and problem solving — involve a combination of information from two sources. The first source is incoming information from the outside world — that is, the input received by your sense organs. Your eyes, ears, skin, and so on are constantly gathering raw data from the environment. The second source is the information already stored in your memory — prior knowledge derived from past experience. So your brain is always blending what it senses right now with what it already knows. Now, the analysis of that sensory information coming in from the outside is known as “bottom-up” processing. It’s also called data-driven processing, because it relies entirely on the data received via the senses. Think of it as building up from the raw pieces — the light patterns, the sounds, the pressures — without any help from memory or expectation yet. But here’s the catch: the sensory information is often incomplete or ambiguous. Your eyes might not catch every detail, or the lighting might be poor, or the sound might be muffled. That’s where the second source comes in. The information already stored in your memory — your prior knowledge — influences your expectations and helps you interpret the current input. This influence of prior knowledge is known as “top-down” processing, or conceptually-driven processing. It works from the top — from your concepts, your stored knowledge — down to the sensory data, helping you make sense of what’s incomplete. In practice, these two sorts of processing operate in combination. Let me give you the example from the book. Imagine you see a moving black shape of medium size with a smooth texture. That’s the bottom-up part — raw sensory data: black, moving, medium-sized, smooth. But on its own, that’s just a shape. Your top-down processes, based on already stored knowledge, enable you to identify it as a Labrador dog. You’ve seen dogs before, you know what a Labrador looks like, so your prior knowledge fills in the gaps and gives the sensory data meaning. The top-down processes interact with the information provided by the bottom-up processes. This combined operation is sometimes known as interactive processing. So to sum it up: bottom-up is data-driven, starting from the senses. Top-down is conceptually-driven, starting from your stored knowledge and expectations. And in real life, they always work together interactively. This is a foundational idea for understanding how you perceive your instruments, interpret radio calls, and even how errors can creep in when your expectations override what your senses are actually telling you.