Your mind is not like a computer

And no, Cognitive Load Theory does not claim that it is

---

A myth about cognitive load theory is gaining momentum. The myth is that cognitive load theory models the mind as a computer.

I first became aware of this myth in a book by Guy Claxton which I wrote about last year. However, since then, I have seen Russell Tytler and Vaughan Prain, Australian education professors, repeat it in the pages of The Age. Specifically, they claim:

“Sweller’s computer model of brain functioning is now perceived as reductive and dated. As noted by Guy Claxton, current brain-based researchers no longer view the brain as using distinct memory boxes.”

Similarly, I have seen people suggest on Twitter that cognitive load theory uses a computational model of the mind. None of this is true.

I emailed both The Age and Russell Tytler after publication of the Tytler and Prain piece. I received no response from Tytler but, as I understand it, a response from Sweller was included in a later print edition of the paper. I cannot find this response online and the original article has not been amended. A reader who stumbles upon it will therefore assume the myth is true.

So, what is going on? How did the myth arise and what is the truth about how cognitive load theory models the mind?

I have a theory of what happened in the case of Claxton. Throughout the chapter of his book that is ostensibly devoted to cognitive load theory, he does not describe any of the theory’s effects or discuss any of the randomised controlled trials that have been conducted to establish these effects. Instead, he mainly refers to research from the 1960s and Baddeley and Hitch’s 1970s model of working memory. I suspect Claxton has not read much cognitive load theory and instead simply assumed that it is based on this earlier research that he is more familiar with.

I am not sure whether Claxton is responsible for all of the confusion, or whether the fact that cognitive load theory discusses ‘information processing’ causes people to assume this means computers. It is a natural leap to make. What processes information? A computer does. Therefore, information processing must imply a computer model. However, this is flawed reasoning.

There are a number of problems with modeling the mind as a computer, but I will focus on just one - a problem shared by the Baddeley and Hitch model of working memory. A computer has a Central Processing Unit (CPU) that runs the main computer program. This way, the CPU coordinates all of the other components of the computer. You could say that it is the ‘brain’ of the computer.

Baddeley and Hitch’s model has something similar - the Central Executive. But can you see the problem? Baddeley and Hitch’s model is supposed to be representing a function of the brain - working memory - so how can it have another ‘brain’ inside it, coordinating what it does? And what tells this brain what to do?

In a computer, this issue is resolved by the fact that there is a program, written externally to the computer and usually by humans, that the CPU runs. Humans tell the computer what to do. However, I am not sure I could substantiate the idea that human minds run a program. Where would that program have come from? How does it get into the brain? So, the Central Executive in the Baddeley and Hitch model represents a problem. It suggests that the brain is made up of a set of memory stores alongside a mini-brain that coordinates them. Which is no explanation at all.

At this point, it may be hard to see a resolution. We have run up against a paradox. It’s turtles all the way down. However, there is a system that processes information that does not act like a computer and does not require a Central Executive - the process of evolution by natural selection.

Living organisms possess a genome that contains information that codes for the proteins that the organism produces and that therefore defines its structure and function. Between generations, these genomes change, either by borrowing information from other organisms (such as by sexual reproduction) or by random mutations. These are then subject to selection pressure - those changes to the genome that allow an organism to have more and/or fitter offspring are retained. Those that do not, wither. Over unimaginably vast periods of time, this process of evolution can create highly complex structures that exhibit all the features of something that has been intentionally designed. Life can proceed from single celled organisms to whales, flowering plants and narcissistic politicians. But it needs no designer in order to do this. Evolution needs no central executive coordinating its actions.

Cognitive load theory proposes that the human mind processes information in a similar way to evolution. Long-term memory is like the genome. It can borrow information from other minds or it can generate new information randomly and then test it for usefulness. Just as evolution is shaped by the constraints of the environment, so is the mind. We decide whether a problem-solving move is any good by seeing if it gets us closer to some goal. In human cognition, working memory mediates much, but not all*, of the interaction between the long-term memory and the environment, just as the epigenetic system mediates the expression of genes in evolution. And just as evolution requires no Central Executive, modeling the human mind as a similar kind of information processing system requires the human mind to have no Central Executive. In effect, the knowledge structures in long-term memory combined with stimuli from the environment perform this coordinating task.

I suspect that some people won’t like these ideas very much. When I posted on Twitter a 2006 academic paper that draws the parallels I have described between evolution and human cognition, I was told it was ‘hideous’. This is not much of a refutation, but it does indicate how emotive the issue is. To some with a religious or mystical sensibility, dispensing with a Central Executive may feel like dispensing with the human soul, in much the same way that evolution dispensed with the need for God as designer. I don’t think religious people need to view it this way, but clearly some do.

Nevertheless, feel free to dislike the cognitive load theory model, to dismiss ‘information processing’ as distasteful and to make New Age arguments for the general woo-woo nature of the mind, just don’t claim that cognitive load theory models the mind as a computer. That’s demonstrably false.

---

*For brevity, I have omitted yet another discussion of biologically primary versus secondary knowledge but there should really be one at this point.