Different Forms of Learning: From Basic Associations to Problem-Solving

How do children – or anyone – actually learn? Is it always a gradual build-up of knowledge, or can understanding arrive all at once? Psychologists and educators have explored these questions for over a century, and what they’ve found is that learning is not a single process. It takes many different forms, ranging from the most basic associations we make as infants to the complex reasoning we use to solve real-world problems. One of the most influential frameworks for making sense of this range comes from American educational psychologist Robert Gagné, who organized the different forms of learning into a clear, logical hierarchy.

Gagné’s hierarchy of learning: from simple to complex 🔗

Robert Gagné, best known for his 1965 book The Conditions of Learning, proposed that learning tasks can be organized in a hierarchy according to complexity – from basic stimulus recognition all the way through to sophisticated problem-solving. His core idea was straightforward: you cannot master a higher-order skill without first acquiring the subordinate skills beneath it. In other words, learning is cumulative. Each new capability depends on what has already been learned.

Gagné identified eight types of learning arranged in hierarchical order, beginning with the simplest forms and progressing to the most complex. These eight types are: signal learning, stimulus-response learning, chain learning, verbal association, discrimination learning, concept learning, rule learning, and problem-solving. The lower four types focus primarily on behavioral responses, while the upper four shift toward cognitive processing – thinking, understanding, and applying knowledge.

For teachers and curriculum designers, this hierarchy has a direct practical implication: higher levels build on lower ones, requiring progressively greater amounts of prior learning for success. A child cannot be expected to solve mathematical word problems before they have grasped basic number concepts. The hierarchy guides how instruction should be sequenced – from foundational to advanced.

Classical and operant conditioning: the building blocks of behavioral learning 🔗

The first two levels of Gagné’s hierarchy – signal learning and stimulus-response learning – correspond closely to the two most foundational theories in behavioral psychology: classical conditioning and operant conditioning. Both represent forms of associative learning, but they work in fundamentally different ways.

Classical conditioning: learning through association 🔗

Classical conditioning is a learning process in which a neutral stimulus becomes associated with a reflex-eliciting unconditioned stimulus, eventually producing the same response on its own. Ivan Pavlov’s famous experiment demonstrated this: after repeatedly pairing a bell (neutral stimulus) with food (unconditioned stimulus), his dogs began to salivate at the sound of the bell alone. The bell had become a conditioned stimulus, triggering a conditioned response.

In the classroom, classical conditioning appears in subtle but important ways. When teachers consistently play a particular piece of music before a quiet reading time, children begin to associate that sound with a calm, focused state. The routine itself becomes a signal. Classical conditioning focuses on involuntary responses to stimuli – the learner is largely passive, reacting to what the environment pairs together.

Operant conditioning: learning through consequences 🔗

Operant conditioning, developed by B.F. Skinner, works differently. Here, it is the consequence of a behavior – not an associated stimulus – that shapes whether that behavior is repeated. Operant conditioning occurs when a behavior is associated with the occurrence of a significant event, such as a reward or punishment. A child praised for completing their homework is more likely to do it again; a child who loses screen time for disruptive behavior learns to avoid that behavior.

The key distinction is this: in classical conditioning the response is a reflex and involuntary, while in operant conditioning the response is voluntary behavior. Classical conditioning shapes emotional and physiological reactions; operant conditioning shapes deliberate choices. Both are constantly at work in learning environments, and both inform how teachers structure routines, feedback, and reward systems.

Concept learning and rule-based learning: moving into cognitive territory 🔗

As we move higher in Gagné’s hierarchy, learning shifts from behavioral responses to cognitive understanding. Two particularly important stages here are concept learning and rule learning.

Concept learning 🔗

Concept learning involves the ability to make consistent responses to different stimuli – it is the process by which a learner categorizes or groups stimuli based on their common properties. Rather than responding to a single specific stimulus, the learner recognizes a class of things that share defining features. A child learns that a robin, an eagle, and a sparrow are all “birds” even though they look quite different. This requires the learner to identify abstract properties, not just surface features.

Concept learning has two important dimensions. A concrete concept involves identifying objects based on their observable characteristics, such as shape or color – for example, identifying which objects are round. A defined concept, on the other hand, requires understanding more abstract ideas, such as “democracy” or “justice,” that cannot be directly perceived but must be explained in relation to other ideas and experiences. As students progress through school, more and more of what they are expected to learn falls into this second, more demanding category.

Rule-based learning 🔗

Rule learning is a step above concept learning in cognitive complexity. Rule learning involves being able to learn relationships between two or more concepts and apply those relationships in different situations, including situations not previously encountered. It forms the basis for learning general procedures, formulas, and principles. When a student learns that the area of a rectangle equals length multiplied by width, and can then apply that rule to different shapes and contexts, they are engaging in rule-based learning.

Importantly, rules are chains of concepts. To learn a rule, the learner must already understand the concepts that the rule connects. A student cannot apply the rule of subject-verb agreement without first understanding what a subject and a verb are. This is precisely why Gagné insists on a hierarchical approach: skipping steps creates gaps that make higher-level learning fragile or impossible.

Problem-solving: the highest form of learning 🔗

At the top of Gagné’s hierarchy sits problem-solving – the most cognitively demanding and educationally significant form of learning. Problem-solving involves developing the ability to invent a complex rule or procedure for the purpose of solving one particular problem and then using the method to solve other problems of a similar nature. It is not simply the application of a known rule; it requires combining multiple rules, adapting them, and generating new understanding in the process.

When a learner solves a problem, they emerge from the experience having learned something more than the solution itself – they develop a new, transferable capability. This makes problem-solving the richest form of learning in terms of long-term educational value.

Trial-and-error problem-solving 🔗

Trial and error is a fundamental method of problem-solving characterized by repeated, varied attempts which are continued until success – or until the learner stops trying. It is one of the earliest approaches humans and animals use when faced with an unfamiliar challenge. Thorndike’s experiments with cats in a puzzle box are a classic example: the cat would try random actions until it accidentally discovered how to escape, and over repeated trials, it gradually learned to repeat the successful behavior more quickly.

Trial-and-error learning has genuine educational value. When a correct solution is discovered, error generation forms part of the conceptual formation during the learning process – mistakes become scaffolding. However, this method has clear limits. Trial and error makes no attempt to discover why a solution works, merely that it is a solution, and it does not readily generalize findings to new problems. It tends to work best for simpler, more contained challenges.

Insight-based problem-solving 🔗

A more advanced form of problem-solving is insight learning – the sudden understanding of a solution that seems to arrive all at once. Insight is the sudden understanding of the components of a problem that makes the solution apparent. The German psychologist Wolfgang Köhler first documented this systematically in experiments with chimpanzees. In one well-known case, a chimpanzee named Sultan needed to reach a banana placed out of reach. After a period of apparent contemplation – not random trial-and-error – he suddenly connected two sticks together to retrieve the food, demonstrating a genuine “aha” moment.

Köhler argued that it was this flash of insight, not prior trial-and-error attempts, that represented a qualitatively different kind of learning. Insight involves cognitive restructuring – the learner reorganizes how they understand a problem, and the solution becomes apparent. With insight, the solution appears suddenly and completely, while trial-and-error requires gradual improvement through repeated attempts. Insight-based solutions also tend to be more durable and transferable, since the learner genuinely understands the underlying principle rather than just having stumbled upon an answer.

In practice, both approaches often work together. A learner may begin with trial-and-error attempts that build familiarity with a problem, and then experience a moment of insight that consolidates everything into a clear understanding. Teachers can support this by designing tasks that are challenging enough to require genuine thinking, providing time for reflection, and resisting the urge to immediately supply answers when students struggle.

What Gagné’s framework means for teaching 🔗

Gagné’s hierarchy is not merely a theoretical classification – it is a practical guide for how instruction should be designed. Teachers can use the hierarchy to design effective lesson plans: building foundational signal and stimulus-response learning first, then progressing step-by-step to more complex tasks, active engagement, and continuous assessment at each level. Before introducing a new concept, a teacher should ensure that students have already mastered the prerequisite skills beneath it. Before asking students to apply a rule, they must first understand the concepts the rule connects. And before expecting students to solve complex problems, they need a solid repertoire of rules and concepts to draw upon.

Understanding the range from classical conditioning to insight-based problem-solving also helps teachers recognize that not all learning looks the same. A child memorizing multiplication tables is engaged in a different cognitive process than a child figuring out why a bridge design keeps failing. Both are forms of learning – both are valuable – but they require different instructional approaches and different measures of success.

What do you think? If Gagné is right that each level of learning depends on mastering what came before, what does that mean for students who have gaps in foundational skills – and how should teachers respond? And considering that both trial-and-error and insight are legitimate pathways to problem-solving, do current classroom assessments do enough to value the process of working through a problem, not just the final answer?

References

1. https://www.irejournals.com/formatedpaper/1703362.pdf
2. http://www.vkmaheshwari.com/WP/?p=854
3. https://edusights.com/gagne-hierarchy-of-learning-made-easy-with-examples/
4. https://www.simplypsychology.org/classical-conditioning.html
5. https://nobaproject.com/modules/conditioning-and-learning
6. https://www.tutor2u.net/psychology/reference/similarities-and-differences-between-classical-and-operant-conditioning
7. https://twurobertgagne.weebly.com/eight-conditions-of-learning.html
8. https://educationaltechnology.net/robert-gagnes-taxonomy-of-learning/
9. https://en.wikipedia.org/wiki/Trial_and_error
10. https://study.com/academy/lesson/trial-error-learning-overview-features-examples.html
11. https://opentextbc.ca/introductiontopsychology/chapter/7-3-learning-by-insight-and-observation/
12. https://www.mindthatbear.com/posts/what-is-insight-learning