The homunculus bias

Original work: Jennifer Garcia (User:Reverie)Derivative work: User:Pbroks13Derivative work of derivative work: User:Was a bee [CC BY-SA 2.5 (

My youngest daughter is a fan of the film Inside Out, which is set inside the head of a little girl called Riley Andersen. The film focuses on five characters who each represent one of Riley’s emotions and who observe the outside world on a screen in a control room. As such, it represents a common idea. When faced with the mystery of human consciousness, how many of us as children have imagined a little person or homunculus inside our head, pulling levers and turning dials? The trouble is that it doesn’t solve the problem it addresses, it just shifts it. What is in the tiny person’s head? Is it ever tinier persons all the way down?

As educated adults, we probably don’t imagine our minds in this way, but that doesn’t stop us from perhaps making a similar error.

In 1974, building on previous work on short- and long-term memory, Baddeley and Hitch proposed replacing the concept of a short-term memory store with ‘working memory’.  The original model of working memory contained three components, a ‘central executive’ and two ‘slave’ systems known as the ‘phonological loop’ and the ‘visuo-spatial sketchpad’. The need for these two separate systems stemmed from the fact that experiments seemed to show that people could process verbal and visual information at the same time, with little loss of efficiency. In 2000, Baddeley added a third slave system, the ‘episodic buffer’.

The notion of working memory, alongside the experimental finding that it appears to be extremely limited in its capacity, is a key concept in Cognitive Load Theory. But what of Baddeley and Hitch’s central executive? How does it know which levers to pull and dials to turn? What is directing the central executive? Is it another central executive?

In their 2011 book, Sweller, Ayres and Kalyuga presented a solution – ditch the central executive. It doesn’t sound like much of a solution at first because we must have some kind of system for directing our attention. We live in stimulus-rich environments and yet, due to working memory limitations, we can only process a few items of information at a time. So how do we know what to attend to?

Sweller et al. claim that our actions are directed by knowledge held in long-term memory. They argue that this is the only system that can logically serve the role of a central executive:

“If knowledge held in long-term memory does not have the attributes normally attributed to a central executive, we are immediately faced with the question of how a knowledge-independent central executive determines its actions. We require another central executive governing the first central executive that in turn requires a third central executive, etc.”

If true, who you are and why you do what you do is a result of what you know. In some ways, this is perhaps a profound idea. In others ways, it is perhaps blindingly obvious.

Yet there seems to exist a kind of homunculus bias. Although we would be embarrassed to suggest that there are little people in our heads, we often find ourselves proposing the existence of similar entities. In education, one example might be the suggestion that there are trainable general-purpose skills such as critical thinking, problem solving and creativity. To the extent that such skills exist, they must consist of knowledge held in long-term memory. When faced with a problem, we can either retrieve some information relevant to solving it or we are forced to use guess-and-check. Even mildly useful heuristics such as ‘draw a diagram’ consist of things we know. How could they not?

Nevertheless, people constantly suggest that variants on problem-based learning or inquiry learning provide a gym workout for the little critical thinking, problem solving and creativity people in our heads.

Someone should make a movie.


7 thoughts on “The homunculus bias

  1. Tom Burkard says:

    Sweller and his colleagues seem to have made their names by producing evidence to support ideas that most people outside the world of education would regard as “blindingly obvious”. This is not to belittle their achievements–the notions that children are naturally good and can learn from nature are beguiling, especially for romantics who believe that any form of systematic teaching will crush ‘creativity’.

    • Interestingly, I don’t think *they* have made their names at all. I think teachers looking for an antidote to the nonsense have sought them out. Certainly, that’s my story.

      • Tom Burkard says:

        Well, there certainly are a lot of teachers who’ve heard of them! KSC (2006) is one of the most frequently cited education papers to appear in the last generation. I just did a check on Google Scholar, and it has been cited six times more frequently than Yeager and Dweck’s 2012 paper on ‘growth mindset’.

  2. Michael Pye says:

    Tom that is because the cohort that read and quote papers are very different to the cohort of teachers. The first is predominantly researchers. Most teachers who use research will never cite anything. Also science often proves the obvious (as well as often proving the strange) the important bit is figuring out which one applied to the situation at hand.

    • Tom Burkard says:

      If that were the case, Michael, we all might as well give up now. There are numerous conduits such as this blog that connect research findings with working teachers. I’ve always cited specific studies while delivering INSET, and I have frequently encountered teachers who know enough about the relevant research to comment. The number of times that a paper has been cited is a pretty good indicator of which way the wind is blowing.

  3. Greg,
    I am trying to understand the distinctions here. Many people think of long term memory as the part of our mind that stores and retrieves previous experiences and knowledge. But memory is a more generic term than that. If we have a develop an allergy our body is in a sense remembering some chemical. If we develop an immunity our body is remembering a particular response to a virus.

    These are very different to recall of facts we learned some time ago. But some of our mental actions operate a lot more like these than the model of someone looking up some facts and then making a decision based on them. The question becomes a nature verses nurture or blank slate verses evolutionary constraints very quickly. My current reading list is full of material on focus, attention and distraction. These are all mental activities that combine evolved biology with learned ideas that can tune this biology.

    I think I can sum it up with two interpretations of long term memory
    1. The recall of previous experiences during our present mental activity.
    2. The process where previous mental activity influences present mental activity.

    The distinction is whether there is just one process – the dragging of prior thoughts into the present consciousness or many, some that are well below our conscious thought processes.

    In both cases you need to pick some constraint on previous such as longer than one month ago to make the long term part clear.


  4. Pingback: Welsh students to know less than their English peers – Filling the pail

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.