Before We Could Count, We Could Connect

What if the deepest foundation of human intelligence is not the ability to count, speak, or remember more facts, but the ability to connect several simple perceptions into one moving model of the world?

A baby can look at two objects and distinguish them from three before learning the word “two.” A crow can choose the larger group. A bee can navigate among flowers using rough quantities. These abilities do not require arithmetic, written symbols, or language. They are ancient, distributed across the animal kingdom, and remarkably modest.

Yet from this modest beginning, some minds build mathematics, stories, tools, institutions, and scientific theories. Others may possess the same elementary perceptions without combining them into equally flexible structures. The crucial question, then, is not whether a brain can recognize quantity. It is whether it can hold several representations together, compare them, transform them, and use them in a new context.

This is where two seemingly distant facts begin to illuminate one another. The human brain has an intuitive number sense that appears before language. At the same time, differences in the organization of the cerebellum, especially in its hemispheres, have been associated with working memory, language processing, executive control, and the capacity to move fluidly among ideas. Together, these clues suggest a powerful thesis:

Intelligence may depend less on possessing sophisticated mental parts than on coordinating simple parts into a system that can keep possibilities alive.

The first intelligence is not counting

Imagine placing three berries on a rock, then adding one more. You do not need to recite “one, two, three, four” to notice that the group has grown. You can perceive a rough change in quantity immediately. This ability is often called a number sense, but the phrase can mislead us. It is not calculation in the formal sense. It is closer to a perceptual response to magnitude, pattern, and difference.

Small quantities are especially accessible. We can usually recognize one, two, or three objects almost instantly, without counting them one by one. This capacity appears in human infants before they understand language. It also appears in animals as different as monkeys, fish, birds, and insects. A monkey may distinguish two pieces of food from four. A bee may learn that one visual landmark is associated with a better reward than several others. A crow may compare groups while deciding where to forage.

This suggests that the earliest form of quantitative thought is not symbolic but relational. The mind does not initially ask, “What is the exact number?” It asks, “Is there more or less? Is this group larger? Did something change? Does this arrangement match what I expected?”

That distinction matters because much of intelligence begins with the ability to detect relationships before we possess names for them. A child may understand that one tower is taller than another before knowing the word “taller.” A musician may hear that one phrase resolves another before knowing the theory of cadence. A hunter may sense that a trail contains more recent tracks without assigning a numerical count.

The raw material of thought is therefore not necessarily language or symbols. It is structured difference. A mind notices that two things are alike in one respect and unlike in another. It detects sequence, proportion, repetition, interruption, and change.

But raw perception is not yet flexible intelligence. A bee can discriminate quantities in a narrow setting. A human can turn quantity into a recipe, a calendar, a budget, a theory of probability, or a metaphor for social power. What enables that expansion?

From perceptual fragments to cognitive fluidity

Consider the difference between seeing four objects and thinking with the idea of four. Seeing four apples is an immediate perceptual event. Thinking with four allows you to compare four apples with four days, four people, four beats, or four steps in an argument. The same relation can travel between domains.

This movement is what might be called cognitive fluidity: the ability to let information from one mental system interact with information from another. Quantity can connect with space, space with time, time with language, language with social intention, and social intention with planning. A simple perception becomes powerful when it can enter many kinds of mental traffic.

Working memory is central to this process. To compare two ideas, you must keep one available while examining the other. To understand a sentence, you must retain its beginning while processing its end. To plan a hunt, conversation, or experiment, you must hold the current state of the world alongside a possible future state.

The limitation is easy to underestimate. Working memory is not an infinite desk. It is more like a small table in a crowded workshop. If too many tools, parts, and instructions are placed on it, the craftsman loses the ability to see how they fit together. The problem is not merely forgetting individual items. It is losing the relationships among them.

This may help explain why the ability to combine ideas can be more consequential than the ability to store them. A person might know many facts yet struggle to use them because the facts remain isolated. Another person might know less but recognize that a principle from ecology applies to organizations, that a rhythm from music can clarify a writing problem, or that a spatial diagram can reveal a flaw in a verbal argument.

The distinction resembles the difference between a box of instruments and an orchestra. Possessing a violin, trumpet, and drum does not produce music. What matters is coordination, timing, and the ability to let one part alter the role of another.

The cerebellum and the hidden architecture of coordination

The cerebellum is often treated as a specialist in movement. It helps refine balance, timing, and motor coordination. Yet the same general functions are relevant to thought. Reasoning also requires sequencing, prediction, error correction, timing, and the smooth integration of multiple signals.

Comparative research on brain anatomy has raised the possibility that Neanderthals had relatively smaller cerebellar hemispheres than modern humans, particularly on the right side. Such a finding cannot serve as a simple measure of intelligence, and it cannot tell us exactly how any individual Neanderthal thought. Brain size and brain organization do not translate directly into a ranking of minds.

Still, anatomy can suggest constraints on cognitive architecture. Larger cerebellar hemispheres have been associated with executive functions, language processing, and episodic and working memory. If these systems depend partly on the ability to coordinate sequences and maintain several representations, then differences in cerebellar organization might have affected not raw intelligence but the ease with which different kinds of thought could interact.

That is a subtler possibility than the old image of one species being simply cleverer than another. Perhaps the critical difference was not that one brain could perceive more objects, recognize more patterns, or remember more isolated experiences. Perhaps it was that one brain was better able to bind perception, memory, language, and social inference into a single flexible workspace.

Imagine two people watching a herd move across a valley. Both can see that one cluster is larger than another. Both can remember where the animals were a moment earlier. But one person can more readily connect the scene to a verbal warning, a plan for surrounding the herd, an estimate of how many people are needed, and a prediction about where the animals will move next. The difference lies not in one isolated faculty. It lies in the coordination layer that lets many faculties operate together.

Language may amplify this coordination. A pre linguistic number sense can register “more.” Language can turn that perception into “There are too many animals for us to approach from this side.” Working memory can preserve the statement while visual perception tracks movement. Planning can then transform the representation into action.

The words do not create the underlying perception. They make it portable, stable, and shareable. Language allows a fleeting impression to become an object that can be inspected, questioned, revised, and transmitted to another mind.

Why small number sense can lead to big worlds

The connection between number sense and cognitive fluidity becomes clearer if we think of quantity as a primitive mental coordinate system.

A quantity is not just a count. It can organize experience. We use it to represent distance, frequency, duration, intensity, risk, scarcity, and social scale. “More” and “less” are among the first abstractions that can migrate across domains. Once a mind can compare groups, it has a foundation for comparing almost anything that varies by degree.

A child who understands that five blocks make a larger group than two does not yet understand multiplication. But the child has encountered a general relationship: accumulation changes scale. Later, that relationship can attach to money, sound, time, votes, evidence, or emotional intensity.

This is one reason basic cognitive capacities can have effects far beyond their original function. A primitive system for detecting quantity may become the seed of mathematics. A system for detecting sequence may become the seed of music, syntax, or procedural skill. A system for predicting movement may contribute to both throwing a spear and anticipating another person’s behavior.

The power comes from reuse. Evolution does not need to invent a separate mental organ for every abstract achievement. It can modify existing capacities and connect them more densely. The same machinery that predicts the next step in a movement may help predict the next word in a sentence. The same sensitivity to relative magnitude may support both choosing the larger food pile and estimating the likelihood of an event.

This creates a useful model of intelligence with three layers:

1. Detection: noticing a difference, pattern, quantity, or change.
2. Maintenance: holding that information in mind long enough to compare it with something else.
3. Transfer: applying the relationship in a new setting.

Animals and infants clearly possess parts of the first layer. Human culture depends heavily on the third. The second layer is the bridge. Without enough working memory and executive coordination, perception remains tied to the immediate situation. With them, a relationship can be detached from its original context and reused.

This model also explains why education sometimes produces knowledge without understanding. Students may learn that a formula works in one kind of problem but fail to recognize the same structure in another. They have detection and perhaps memorization, but not transfer. The knowledge has not become fluid.

The modern threat: too much information, too little coordination

The architecture of cognition has practical implications beyond prehistory. Modern life floods the small workshop of working memory with fragments: notifications, open tabs, competing instructions, partial conversations, and rapidly changing goals. We may have access to more information than any earlier generation while becoming less capable of combining it.

The result is a peculiar form of cognitive poverty. We can retrieve facts but cannot hold the larger pattern. We can count metrics but cannot understand what they measure. We can collect perspectives but cannot compare them. We can communicate constantly while losing the ability to maintain a shared model of what is happening.

The remedy is not simply to consume more information. It is to design conditions for coordination.

When facing a complex problem, begin with the simplest relations you can observe. What increased? What decreased? What repeated? What changed after an intervention? These questions return thought to its perceptual foundation and prevent premature abstraction.

Then externalize the pieces. Draw a map. Arrange objects on a table. Write competing explanations in separate columns. Convert a verbal sequence into a diagram. External tools are not a sign that thinking has failed. They enlarge the workspace in which thinking can occur.

Next, force transfer. Ask where else the same structure appears. If a project behaves like a queue, a market, an ecosystem, or a game, what does that analogy reveal? The goal is not decorative metaphor. It is to test whether a relationship survives when its surface details change.

Finally, protect the coordination layer. Do difficult reasoning before opening a stream of interruptions. Group related tasks. Reduce the number of active goals. Leave enough silence for representations to interact rather than merely arrive and disappear.

Key Takeaways

• Start with relationships, not labels. Before asking what something is called, ask what changed, what differs, and what remains constant.
• Treat working memory as a limited workspace. Use diagrams, lists, physical objects, and written comparisons to keep relationships visible.
• Practice transfer deliberately. After learning an idea, apply it to a different domain, such as using a biological concept to examine a team or a musical concept to improve writing.
• Separate detection from understanding. Noticing a pattern is only the first step. Understanding requires maintaining it, testing it, and using it elsewhere.
• Design for coordination. Fewer simultaneous inputs often produce richer thought than a larger volume of information.

The most revealing contrast is not between a primitive mind and a modern mind. It is between a mind that experiences events as disconnected impressions and one that can make them participate in a common structure.

A baby’s recognition of three objects, an animal’s comparison of food groups, a language user’s sentence, and a scientist’s equation may seem to belong to different intellectual worlds. In one sense, they do. In another, they may be variations on the same foundational act: preserving a relationship long enough for it to connect with something else.

That act is easy to overlook because its simplest form feels effortless. We see that one group is larger. We remember what happened before. We understand a phrase. Yet civilization may depend on the gradual transformation of these small, ancient capacities into a system capable of combining perceptions across time, people, and domains.

The question is therefore not merely how much a brain can hold. It is what its contents are able to do together.

The future of intelligence may belong not to the mind with the most information, but to the mind that can make the simplest observations enter the richest relationships.