Adaptive Control of Thought

The Basic Idea

The Adaptive Control of Thought (ACT) is a theory about cognition and memory. Over the years, it has evolved and its most recent version is iterated as ACT-R, with the added letter standing in for ‘Rational’. According to ACT theory, we can create a model of the human brain through which we can analyze and predict human behavior, most recently based on the idea that people act rationally. ACT-R is therefore a cognitive architecture that maps out the functions of our higher cognitive processes to demonstrate how people process information and then act accordingly.1

ACT posits that multiple higher cognitive processes have the same underlying system; all of our thoughts, no matter what kind of thought it is, arises as a result of the same brain function. One of its most central tenets is that knowledge be divided into two kinds: declarative knowledge and procedural knowledge.2

We know that human agents tend to be generally well-adapted to their environment, and hence a careful analysis of the cognitive task encountered by the mind, coupled with an assumption of the optimality of human behavior, results in a putatively powerful methodology of prediction and explanation.

– Samuli Reijula, professor in philosophy at the University of Helsinki, in his paperThe Upside of Irrationality: The Unexpected Benefits of Defying Logic at Work and at Home “What, when and how do rational analysis models explain?”1

Theory, meet practice

TDL is an applied research consultancy. In our work, we leverage the insights of diverse fields—from psychology and economics to machine learning and behavioral data science—to sculpt targeted solutions to nuanced problems.

Our consulting services

Key Terms

Declarative Knowledge: Declarative knowledge (also known as descriptive knowledge) is knowledge that we are conscious of and can verbalize. It is information we directly encode from the environment and doesn’t require much synthesization.3 It emphasizes what one needs to do to solve an issue rather than how to solve it. Examples of declarative knowledge include facts, world history, or rules for solving mathematical equations.4 In essence, it is content that can be recited or memorized.

Procedural Knowledge: Procedural knowledge (also known as imperative knowledge) is knowledge you use while performing a task, but may not be able to verbalize. It is information encoded from synthesizing and observing transformations of the environment (behaviors).3 Procedural knowledge is about how we do something. Examples of procedural knowledge include behaviors we do habitually, such as riding a bike or driving a car.

Human Information Processing: Insights into how people receive, store, integrate, retrieve and apply information; in other words, how people learn. ACT is a theory of human information processing.5

Serial Processing: Serial processing suggests that only one piece of information can be processed at a time and that we encode information one object after the other.6

Parallel Processing: Parallel processing suggests that multiple pieces of information can be processed at a time and that we encode objects simultaneously.6

Rational Analysis: Rational analysis is the assumption that people react rationally to their environment. Traditional economics uses rational analysis to predict consumer behavior.7


Canadian psychologist John Anderson has an extensive background researching and developing his Adaptive Control of Thought model. His underlying assumption is that knowledge can be reduced into a theory or model, and therefore that a system can be created to perform human cognitive tasks.8 This assumption rests on the belief, articulated by Anderson in his 1990 book, that “we can understand a lot about human cognition without considering in detail what is inside the human head. Rather, we can look in detail at what is outside the human head and try to determine what would be optimal behavior given the structure of the environment and the goals of the human” (3)1. Such systems operate based on rules like ‘If x, then y.’

With the goal in mind of creating a model to depict human knowledge, Anderson developed the Human Associative Memory (HAM) model alongside cognitive psychologist Gordon Bower in 1973. This model computed the mathematical theories of human cognition prevalent in the 1950s and 1960s. However, the model only accounted for human memory and did not accomplish Anderson’s mission of showing that all higher cognitive processes (memory, language, problem solving, imagery, deduction and induction) have the same underlying system. Replacing the HAM model, Anderson developed the ACT model in 1976, which was able to account for these higher cognitive processes.9

In 1990, Anderson developed another version of ACT which he named ACT*. His hypothesis was that ACT*, alike ACT, showed that the mind is unitary (all of thoughts and mental faculties can be explained by the same underlying system) and that our experiences are stored in different facilities (linguistic, geometric, etc.) depending on their subject matter.10 Information comes through a ‘buffer’, known as working memory, which determines whether the information should be stored as declarative and later retrieved, or as procedural knowledge and executed in the moment to match current activity. The model looks as follows:2

Cognitive Architecture by John Anderson

The original ACT model only accounted for one kind of ‘cognitive unit’ that the mind would process, also known as ‘chunks’. These units were words and statements like ‘hate’ or ‘my mom is nice’. and Anderson speculated that working memory is able to process around 5 chunks at once before becoming overwhelmed.

However, in the ACT* model, information could come in the form of spatial images and temporal strings as well as abstract propositions.10 What counted as a cognitive unit was expanded. Temporal strings encode the order of a set of items; spatial images encode the spatial configuration of a cognitive unit; abstract propositions encode meaning. Another key element of ACT* was its suggestion that all information begins as declarative, and combines with method learning to produce procedural knowledge.11

Moreover, ACT suggests that information was serially processed, whereas ACT* suggests information could be processed simultaneously (parallel processing). The last notable difference is that the ACT* theory added a ‘sub-symbolic’ component: a feature that could determine what meaning is activated when we run into a cognitive unit with many possibilities. For example, if we encounter the sentence “The robber took money from the bank,” the word ‘bank’ has two meanings: a financial institution, or the land sloping down next to a body of water. The sub-symbolic component of ACT* activates the financial institution meaning because it knows that meaning is processed in relation to money or robbery.10

A few years later, Anderson, alongside his colleagues at Carnegie Mellon University, developed the most recent model of ACT by combining the original with rational analysis. This model is known as ACT-R and predicts behavior based on the idea that humans act in ‘optimal’ (see: rational) ways. The ACT-R is essentially a production system, like a machine, that operates according to ACT*. The theory was moved to a computer program, on which researchers can download the ACT-R code, input information about a specific task, and analyze people’s predicted performances.12



The Adaptive Control of Thought theory, in all its iterations and models, rests on belief in a unitary theory of mind. This theory of mind suggests that we acquire complex skills thanks to an underlying ‘human’ system that differentiates us from other creatures. Instead of suggesting that we have actually evolved specialized brain faculties, the unitary theory of mind suggests we have an innate neurological component that enables us to successfully acquire more skills as we grow and gain experience. Along these lines, becoming an ‘expert’ in something is not thanks to a special characteristic unique to an individual, but rather thanks to our brain’s ability to effectively maneuver its underlying system. What makes humans so impressive and unique then, is the plasticity (flexibility) of this system, which we can apply to different kinds of skills.10

ACT therefore provides evidence for skills that humans have acquired that have little to do with evolution. For example, there are many tech-savvy individuals who are great at computer programming. Computer programming would not be a skill anticipated in the evolutionary process, as it does not ensure our survival, and therefore there is no special faculty that enables us to perform that skill. Rather, the ACT theory applies since it suggests that the brain capacity required for computer programming uses the same underlying processes that we use for other skills as well, and we simply have to learn to apply our brain to the new context.

ACT theory also supports the idea that all behavior is a response to stimulus, as put forward by the behavioral perspective. This school of thought suggests that all behavior is learned through conditioning, meaning there must be an underlying cognitive system that is able to learn how to respond to different stimuli (or cognitive units) in our environment.


It seems odd that all of our cognitive processes can be reduced to a simple mathematical program or code. ACT theory suggests that all the skills we acquire are thanks to an underlying system of information processing, but it is unclear whether that system is impacted by emotions, cognitive biases, or other internal events.

There are other models of cognitive processes that suggest that our specific brain areas are in fact activated for different purposes, negating the unitary theory of mind. For example, the frontal lobe is thought to be responsible for movement, speech, and reasoning, while the temporal lobe is activated for memory, object recognition and understanding language.13 In particular, humans’ ability for language acquisition, which differentiates us from other creatures, is widely believed to be due to the existence of specific ‘language’ areas in the brain. This research suggests that we in fact do have special faculties responsible for different behavior. However, it could be argued that the underlying system proposed by ACT involves all these different parts of the brain.

While the ACT theory explains different higher cognitive processes, it fails to explain why some people are ‘experts’ and others not, in particular areas. Additionally, while it informs us of the basic system that underlies memory, it does not get into detail about why some cognitive units are remembered better than others. Other theories, like the levels of processing effect, better explain how we use levels to encode information.

Lastly, ACT-R suggests that people always perform rationally. However, according to many behavioral scientists, people actually perform according to bounded rationality. Bounded rationality suggests that our ability to act rationally is impeded by our limited cognitive capacity, time, and available information, all components not factored into ACT theory. In other words, although we ideally act rationally, in practice, we rarely do so.

ACT in Education

The ACT theory suggests that we process information in cognitive units in our working memory, which distinguishes between declarative and procedural knowledge. The theory also suggests that our working memory can only handle a certain number of cognitive units (5) at a time before becoming overloaded. This information can be important for teachers to know, so that they can break down their material into chunks of knowledge that children can more easily process.2

Since the model also divides knowledge into two distinct kinds, it can inform teachers that the best way to ensure information is encoded and easy to retrieve is making that information both declarative and procedural. Students should be encouraged to combine their knowledge with actions. For example, combining an action or activity with a learned fact means that information will be encoded as both kinds of knowledge and therefore deepen the level of learning.2

Related TDL Content

Algorithms for Simpler Decision-Making: The Case for Cognitive Prosthetics

ACT-R basically transforms our cognitive processes into an algorithm that can be used to predict human behavior. In this article, writer Jason Burton examines society’s increasing shift towards algorithmic decision-making processes and the outsourcing of our cognitive processes to computers.

Unitasking: How to Get More Done in Less Time

The ACT-R model suggests that the working memory buffer through which we process information can get overwhelmed with cognitive units and therefore inhibit our performance. This finding might be in support of unitasking over multitasking, as unitasking limits the amount of information our working memory must process at once. In this article, our writer Ipsitaa Khullar outlines a growing field of research suggesting that unitasking is more effective than multitasking.


  1. Poyhonen, S. (2016). What, when and how do rational analysis models explain? [Contributing Paper].,when_and_how_do_rational_analysis_models_explain.pdf
  2. Heick, T. (2021, March 22). Learning theories: Adaptive control of thought. Teach Thought.
  3. Yengin, I., & Ince, I. F. (2014). Applying the Adaptive Control of Thought-Rational Theory into the Design of Mobile Worked Examples Applications. International Journal of Robots, Education and Art, 4(2), 21-28.
  4. Craft, K. (2018, April 26). What Procedural Knowledge Is and How to Use It. Tettra.
  5. Human information processing. (n.d.). EduTech Wiki. Retrieved March 30, 2021, from
  6. Li, K., Kadohisa, M., Kusunoki, M., Duncan, J., Bundesen, C., & Ditlevsen, S. (2020). Distinguishing between parallel and serial processing in visual attention from neurobiological data. Royal Society Open Science, 7(1), 191553.
  7. Anderson, J. R. (1991). The adaptive nature of human categorization. Psychological Review, 98(3), 409-429.
  8. ACT-R Home. (n.d.). Carnegie Mellon University. Retrieved March 30, 2021, from
  9. Anderson, J. R., & Lebiere, C. J. (2014). The atomic components of thought. Psychology Press.
  10. Anderson, J. R. (1996). The Architecture of Cognition (1st ed.). Psychology Press.
  11. Adaptive control of thought theory. (n.d.). EduTech Wiki. Retrieved March 30, 2021, from
  12. ACT-R. (n.d.). Psychology Wiki. Retrieved March 30, 2021, from
  13. Stroke Education Manual | Barrow Neurological Institute. (n.d.). Pinterest. Retrieved March 30, 2021, from

Read Next

Notes illustration

Eager to learn about how behavioral science can help your organization?