What Is Working Memory and Mental Imagery?

Title: What Is Working Memory and Mental Imagery?

Research Article Summary: This research article proposes a model to explain working memory and mental imagery as cognitive processes. The model is based on Turing's machine, which is represented as a system (Robot, World) consisting of a robot's sensory-motor devices (D, B) and an external world (W) represented by memory. The robot learns to perform mental computations by simulating the internal states of a Turing machine through "talking to itself." This is achieved by training the robot to perform examples of an algorithm with different input data, which leads to two effects: the robot learns to perform the shown algorithm with any input data using the memory, and it learns to perform the algorithm mentally using an "imaginary tape." The model provides a simplified, neurobiologically plausible explanation of working memory and mental imagery and is implemented as a user-friendly program called EROBOT. The program is designed to enhance the robot's sensory-motor devices and brain, allowing it to learn and perform mental computations using an "imaginary" memory device. The article includes educational experiments with EROBOT and discusses basic questions related to the brain, outlining possibilities for further development of the basic ideas.

Main Research Question: How can working memory and mental imagery be explained as cognitive processes using a model based on Turing's machine?

Methodology: The research uses a system-theoretical approach and neurocomputing background to design a neural brain for Turing's robot. The design involves associative neural networks that can serve as programmable look-up tables, replacing the "neurobiological" model with a more understandable "psychological" model. This allows the robot to learn to perform mental computations using an "imaginary" memory device.

Results: The research shows that the robot can learn to perform mental computations by simulating the internal states of a Turing machine through "talking to itself." This leads to the development of working memory and mental imagery as cognitive processes.

Implications: The model provides a simplified, neurobiologically plausible explanation of working memory and mental imagery and has educational and theoretical implications. It outlines possibilities for further development of the basic ideas and contributes to the understanding of cognitive processes.

Link to Article: https://arxiv.org/abs/0309009v1 Authors: arXiv ID: 0309009v1