Working Memory Models

There are a number of conceptual models of working memory, but arguably the two most important are the Working Memory Model and Cognitive Load Theory.

The Working Memory Model

According to the Working Memory Model,^​1,2​ working memory is made up of four parts:

1. Central Executive (directs attention)
2. Phonological Loop (language)
3. Visuospatial Sketchpad (visual & spatial)
4. Episodic Buffer (interface & integration)

Central Executive

The Central Executive System (CES) allocates attention and is made up of four component parts, allowing us to:^​4​

1. Focus attention
2. Divide attention
3. switch attention
4. Link between working memory/short term memory stores (Phonological Loop, Visuospatial Sketchpad) and long-term memory.

Phonological Loop

The Phonological Loop is the part of our working memory that deals with sounds and language, both written and verbal. Among other things, it is also used during the acquisition of vocabulary, and learning to read. ^​3​

Phonological Store (Inner Ear)

This is the short-term store for verbal & auditory information. Auditory stimuli can enter the phonological store directly, but visual verbal material (written language) must enter after being recoded in the articulatory control/loop.^​5​

Articulatory Control/Loop (Inner voice)

The articulatory control (or articulatory loop) does three things:^​5​

1. Offsets memory decay of information in the phonological store by performing subvocal rehearsal (i.e. when you repeat a word or number in your head as you find a pen to write it down so you don’t forget).
2. Recodes the visual information from written language so it can go into the phonological store.
3. The “assembly and production of speech sequences.”

Visuospatial Sketchpad (Inner eye)

The visuospatial sketchpad (VSS) is where visual and spatial (and perhaps tactile & kinesthetic (positioning of one’s body in space)) information is combined, stored and maintained.^​6​ The VSS also allows us to bring up images in our “mind’s eye.”^​7​

Episodic Buffer

The episodic buffer is a temporary storage system that acts as an interface between the different subsystems of WM, as well as interfacing with long term memory. This is where information is integrated, made coherent, and manipulated. It’s controlled by the central executive.^​2​

Cognitive Load Theory

Cognitive load is the “amount of time and effort required for the processing component of working memory.” Cognitive Load Theory thus “emphasizes the limited cognitive capacity of working memory and how easily working memory can become overloaded.”​^​8​

Since all “conscious, effortful cognitive processing and temporary storage of information draw from the same working memory capacity,” when there is a high demand on processing, fewer items can be held in WM, and fewer items to be properly encoded into long-term memory.^​8​ This is because “there is not enough time to devote to focusing on the information to be retained, resulting in the information not being rehearsed frequently enough to prevent decay,” and information is lost (forgotten).^​8​ This also works in the other direction, and if attention is focused on maintaining or rehearsing information, processing can be impeded, “slowing it down or causing processing errors.”^​8​

Furthermore, heavy cognitive load and its subsequent reduced attentional control reduces ones “ability to exert inhibitory control over extraneous, irrelevant processing and information,” and makes cognitive “switching and updating difficult.” In consequence, “focus is lost, the mind wanders, and the task is abandoned.”​^​8​

When it comes to reducing cognitive load, the most important thing is automaticity.^​8​

“A task or procedure is said to be “automated” when it is overlearned or mastered to the point where it can be performed without conscious, mental effort. Automaticity speeds up processing, reduces cognitive load, and increases retention of information because the processing involved requires little attention. A prime example is the acquisition of reading fluency. A fluent reader has automated word decoding processes that “free up” working memory capacity for processing, such as making inferences, and for retaining more information, leading to better reading comprehension.”^​8​

Sources: