Meaning of "human minds and machines both manipulate symbols and internal representations"

To understand the phrase "human minds and machines both manipulate symbols and internal representations," it helps to break down exactly what cognitive scientists meant by this "Computer Metaphor" during the 1950s.

Before this shift, Behaviorist psychology treated the mind strictly as an input-output machine: a stimulus happens, and a behavior is produced. They ignored whatever happened inside. The cognitive revolution challenged this by arguing that between the input and the output, there is an entire world of hidden data processing happening.

Here is what that means in practice:

1. Internal Representations (The "Data Structures" of the Mind)

A computer does not physically contain the things it processes. If you are developing a chess game in Python, there are no physical wooden pawns or checkered boards inside the CPU. Instead, there is an internal representation. You might use a GameState class containing a 2D array to map the board, where specific strings or integers represent the pieces. The computer holds an abstract, internal model that stands in for the real world.

The Cognitive Revolution proposed that the human mind does the exact same thing. When you plan a route to a friend's house, or when you close your eyes and picture a chess board to calculate a move, you are interacting with a mental model. Your brain has encoded the outside world into neural data structures—concepts, images, memories, and variables—held in your working memory.

2. Symbol Manipulation (The "Algorithms" of the Mind)

Once you have representations, you need to do something with them. In a computer program, you write functions to manipulate the data. If a player wants to move a knight, the program takes the current state of the board (the symbols), applies a strict set of rules (like checking for legal L-shapes or the 50-move rule), and updates the array to a new state. The computer does not physically "know" what a knight is; it simply follows logical rules to manipulate symbols from State A to State B.

Early cognitive scientists, like Newell and Simon, realized that human reasoning could be described the same way. When we solve a math equation, parse a sentence, or make a logical deduction, we are:

  1. Taking internal symbols (e.g., the numbers in a math problem, or the words in a sentence).

  2. Applying cognitive algorithms (e.g., the rules of arithmetic, or the grammar rules Chomsky proposed).

  3. Producing a new internal state or an outward behavior.

Why This Was Revolutionary

By framing the mind this way, scientists could finally map psychological processes using the same logic used in computer science. They could start asking: What is the algorithm the brain is running to parse visual data? What is the computational complexity of human decision-making? Where is this mental data stored, and how is it retrieved? It allowed researchers to stop looking at humans as just animals reacting to stimuli, and start studying them as sophisticated information-processing systems. The brain is the hardware, and cognitive processes—our thoughts, memories, and logic—are the software running on top of it.