I’m Not a Math Person or Brain Freeze?
Slow down, witty!
It’s all too common to hear someone say of themselves “I’m not a math person.” Some of us might even suspect that our child is not a math person. Let’s start by unpacking the phrase math person. No one is a math person. Sure, some of us might be blessed with more of the cognitive capacities that undergird math skills, just as those of us who are seven feet tall have been blessed with a body type that lends itself to basketball skills. But both basketball and math were invented–and are continually reinvented–by humans. Math skills, then, cannot be selected for, cannot be part of one’s genetic endowment, so one cannot be or not be a math person. Math is a cultural artifact, and like those of basketball, its requisite skills are 100% acquired, genetic pre-dispositions (note the “pre”) be damned.
Then why do so many of us struggle with math? The answer is brain freeze, by which, note well, I am not referring to a headache from inhaling Rocky Road. Brain freeze is, in less colloquial terms, cognitive overload–the overwhelming of short-term memory.
John Sweller’s (1988) foundational work determined that when working memory is forced to juggle too many steps at once, our brains simply run out of space. Impossible! You scoff. What about those mnemonists who can recite pi to tens of thousands of digits or memorize the names and birthdates of hundreds of people in the time it takes to recite the information?
Yes, we will return to these superhuman feats in a moment. For now, consider George Miller’s famous 1956 paper, “The Magical Number Seven, Plus or Minus Two,” in which Miller establishes that short‑term memory has a sharply limited capacity–roughly five to nine things at once (this is why phone numbers are not a string of dozens of digits–people used to memorize them (no kidding)).
Math, especially the kind first encountered in Algebra I, demands students hold in mind several different pieces and kinds of information simultaneously. According to Chinnapan (2010), “Algebraic tasks place heavy demands on working memory because of the need to coordinate multiple representations.” In other words, Algebra’s symbolic complexity creates inherently high intrinsic cognitive load, especially when students lack schemas for variables and relationships like the following:
multi‑step problem-solving
symbolic juggling
switching between representations
remembering what the problem is about
These procedures all compete for the same limited mental workspace, and when that workspace overflows–you guessed it—brain freeze. This is why multi‑step Algebra problems feel impossible and students stare blankly at the page even when they “know” the content. And this is why word problems are particularly intractable–they tax two systems at once, the mathematical and the verbal, triggering a dual processing load (Teachfind, n.d.). Students must read, translate, visualize, and calculate.
What’s to be done? Let’s return to the mnemonists. They are just as likely as the rest of us to forget where they left their phones. When they perform their memory feats, they aren’t exercising a superpower but a supertechnique, often called the method of loci or the memory palace. The method of loci is a memory technique in which you place pieces of information along a familiar mental path — a house, a street, a daily route — and then “walk” that path in your mind to retrieve them. It works because it converts abstract items into spatial, visual scenes, which the brain remembers far more easily than raw symbols or lists.
Relevant here is not the specific technique but what is generalizable from all methods deployed to overcome the limits of working memory–representational redescription, a term I borrow from Karmiloff-Smith (1992). The method of loci redescribes, say, the digits of pi as a narrative of a walk through a familiar terrain.
Specifically, Karmiloff‑Smith’s theory argues that learning becomes robust when the brain re‑codes information into more abstract and flexible formats — moving from:
implicit → explicit
procedural → conceptual
step‑bound → schema‑based
The most basic method for achieving the above is chunking. Recall Miller’s 7 ± 2 explains the difficulty of memorizing a random ten-digit sequence–5104284976. Chunking turns ten separate items into three–510-428-4976, the familiar phone number format. As powerful as the method of loci and chunking are, they are only the beginning.
Let’s make this explicit. Consider the following basic Algebra problem:
Solve 3(x - 4) = 18
This already comes with a fair amount of intrinsic load (Paas & Sweller, 2012): What is an equation? How do parentheses work? What does distribution mean? Step‑bound students (freezers) compound this with extrinsic load when they treat the problem as a memorized sequence, cluttering their working memory with a host of template-driven operations.
Distribute
Combine like terms
Move constants
Divide
If they forget any step, or if the problem deviates from the template, they stalll. Their inner monologue might sound something like this:
“First I distribute… I think?”
“Then I… move something?”
“Wait, do I divide now?”
“What if I do it in the wrong order?”
Because they’re juggling isolated steps, not meaning, working memory overloads, and they freeze. This is Miller’s 7±2 bottleneck in action.
Conversely, schema‑based, flexible students have a mental model of what an equation is: “An equation is a balance scale. Whatever I do to one side, I do to the other.” They don’t rely on a memorized sequence. They rely not on steps but on a conceptual structure.
Their approach:
“The 3 is multiplying the whole group. I can undo that first.”
Divide both sides by 3 → (x - 4 = 6)
“Now it’s just a shift.”
Add 4 → (x = 10)
They’re not recalling steps.
They’re applying a schema:
isolate the variable
undo operations in reverse
maintain balance
This uses far less working memory because the schema chunks the entire process into one idea, not four separate steps. In short, step‑bound students remember procedures; schema‑based students understand structures.
Put another way, frozen students start from Step 1; flexible students start from the goal.This is not just a stylistic difference. It’s a representational difference, a thinking difference.
Let me make it concrete. The step‑bound student starts at the surface, thinking “What’s the first step I’m supposed to do?” They scan for:
the template
the memorized procedure
the teacher’s last example
the “right” opening move
This forward‑chaining from the surface form of the problem is brittle:
if Step 1 isn’t obvious, they freeze
if the problem deviates from the template, they freeze
if they forget a step, they freeze
They’re navigating horizontally across the surface of the problem, not vertically through its structure. This is exactly what Karmiloff‑Smith calls procedural, implicit, step‑bound representation. Representational redescribers start from the goal, thinking “What does the answer need to look like, and how do I get there?”
They reverse‑engineer:
the target form
the structure of the expression
the operation that isolates the variable
the conceptual “shape” of the solution
This is backward‑chaining from the goal state, and it’s robust:
the goal constrains the path
the schema chunks the steps
the student can improvise
the student can adapt when the surface form changes
They’re navigating vertically, moving up and down the ladder of abstraction–what Karmiloff‑Smith describes as explicit, flexible, schema‑based representation.
The difference isn’t intelligence — it’s representation. One student is following a recipe; the other understands the dish. The first approach maximizes load; the second minimizes it.
These differences explain why guided practice is better than unguided problem sets (Gupta & Zheng, 2020). Astute teachers give students the “view from above” and the means to get from the messy details of the problem to the abstract representation of the problem. This is why your teachers told you to “show your work.” It wasn't to police your procedure but to force you to redescribe the information. After all, 3(x - 4) = 18 and x = 10 are the same information, just represented differently.
And now for some good news: A recent review confirms that overload → shutdown is still one of the most robust findings in learning science (Ouwehand et al., 2025). Why good? Because the alternative is much worse–that some of us just aren’t math people. Students aren't failing math tests because their lazy or genetically ill-equipped; they’re failing because they have not been taught memory-minimizing ways to think about math. They’ve been taught to memorize steps, but not to think mathematically.
If your child is freezing, the problem isn’t genetics — it’s representation. Change the representation, and you change what’s possible.
References
Chinnappan, M. (2010). Cognitive load and the role of schema-based instruction in algebra learning. Mathematics Education Research Journal, 22(2), 139–163.
Gupta, U., & Zheng, R. Z. (2020). Cognitive load in solving mathematics problems: Validating the role of motivation and the interaction among prior knowledge, worked examples, and task difficulty. European Journal of STEM Education, 5(1), 05.
Karmiloff‑Smith, A. (1992). Beyond modularity: A developmental perspective on cognitive science. MIT Press.
Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63(2), 81–97.
Ouwehand, K., Lespiau, F., Tricot, A., & Paas, F. (2025). Cognitive Load Theory: Emerging Trends and Innovations. Education Sciences, 15(4), 458.
Paas, F., & Sweller, J. (2012). An evolutionary upgrade of cognitive load theory: Using the human motor system and collaboration to support the learning of complex cognitive tasks. Educational Psychology Review, 24(1), 27–45.
Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257–285.
TeachFind. (n.d.). Why students struggle with word problems. Retrieved from https://www.teachfind.com
