The Knowledge Effect
How curiosity, context, and connected texts fuel reading comprehension
To become strong readers, children need more than phonics. They need to know what words mean and how they’re used in different contexts. Knowing the meaning of a word supports reading the word. For example, take these two sentences:
I suspected the suspect.
Let’s record the record.
Chances are, you read each version of the repeated word differently. Why?
It’s because of context and the role (or part-of-speech) of the word in the sentence. In English, when a noun becomes a verb, we shift the syllabic emphasis to the end of the word, so “RE-cord” becomes “re-CORD.” You have likely internalized this rule so that you understand how to read each sentence without really thinking about it.
Vocabulary knowledge facilitates the connection between the written letters and their corresponding sounds, making decoding (or reading) the word smoother. Your internalized understanding of a word and its role in a sentence helps you recognize it automatically in print and identify the correct use, meaning, and pronunciation.
Having a strong foundation of knowledge of the world also matters in how well you understand what you’re reading overall. Take this sentence for example:
"He played a straight bat for the cameras, but anyone listening closely could tell he’d been caught behind long ago."
If you’re unfamiliar with cricket (the sport), this sentence is likely confusing even if you know what each word means.
To make sense of it, you need to know:
A “straight bat” is a position in which the cricket bat is held. It is considered a safe swing or stroke, and “playing with a straight bat” is viewed by many cricket players as behaving fairly and honestly, being a good sportsman.
“Caught behind” is when a batter hits the ball and it is caught before it hits the ground, but metaphorically, it can signal a situation where someone is in a difficult or vulnerable position.
Even with these definitions, the full meaning of the sentence is layered and hard to comprehend without relevant background knowledge. As Marilyn Adams puts it:
“Research repeatedly shows prior domain knowledge to be a far stronger predictor of students’ ability to comprehend or learn from advanced texts. [. . .] No reasoning strategy can rival the speed, power, or clarity of knowledge-driven understanding; nor can it compensate for an absence of sufficient information.”
In this case, knowing about cricket is critical to understanding the meaning of this sentence. It is more important than your ability to determine the meaning of the individual words.
And, in case you want to confirm the meaning of this sentence, the subject (“he”) puts on a good show for the cameras and appears to be straightforward and honest, but that is all a show because some more observant people can tell he’s in a difficult situation even though he seems to be trying to keep that hidden.
How Does Knowledge Grow?
Knowledge is sticky and builds on itself over time. The more you know, the easier it is to gain more. In research they call it the Matthew Effect: the “knowledge rich” gain knowledge quickly and easily and the “knowledge poor” fall further behind.
It works like compound interest. Early financial investments grow faster over time because the interest, as it is earned, is reinvested and added to the principal investment, so the interest also earns interest. This compounds or snowballs the growth over time. Knowledge building operates similarly. The earlier children start building knowledge, the stronger and bigger their foundation becomes because each new idea connects to an existing idea and compounds the knowledge built. If they get a later start, it is almost impossible to catch up.
How do Kids Learn Words and Build Knowledge?
Adams explains:
“The meaning of any word consists of bundles of features and associations that are the cumulative product of the reader’s experience with both the word in context and the concepts to which it refers.”
We build the meaning of words in our brains by seeing words in context, connecting the new words to words and ideas that we have already stored, and noticing the various relationships that exist among them.
In the 1970s, researchers developed the Latent Semantic Analysis (LSA) computer model to simulate how humans learn vocabulary. They wanted to understand how we can be taught a finite number of words directly and yet go on to understand an almost infinite number of words. LSA laid the groundwork for today’s large language models or LLMs, a kind of AI. They also showed how language learning works at scale.
Here’s what they found:
Our brains organize knowledge through networks of word meaning and relationships. Each time we encounter a new word, our brain connects it to other words and concepts we’ve already learned. As those connections grow stronger, understanding accelerates.
For example, the word hypothesis has a strong connection to science. Other words like method, experiments, and research also have a strong connection to science, and so our brains will also link them to hypothesis.
Or imagine reading about plants. You may see words like stem, leaf, soil, ground, chlorophyll, sunlight, photosynthesis, growth, or flowering. Those related words and concepts are clustered together in your brain, creating a dense, connected internal mental model of “plants.” As you learn more about plants, every new encounter with one of those words reinforces and refines the connections among those ideas and builds a bigger, stickier network of knowledge.
These knowledge networks in our brains are vast and complex and nearly impossible to map, but we can visualize them like a graph (see the image below). That knowledge graph is constantly growing and shifting as more words are encountered and the ideas associated with those words develop into knowledge. As we read more words in different contexts, our network grows in size and the associations among words grow stronger, allowing us to know the meaning of more words from reading than we can be taught.
Licensed CC-BY-SA by Mark Müller-Linow; Claus C Hilgetag; Marc-Thorsten Hütt - https://www.researchgate.net/figure/Graph-representation-of-the-modular-scale-free-network-The-nodes-are-colored-according_fig6_23283823
In a graph representation, the words, ideas, and their meaning would be the nodes or entities on the graph. These are called “vertices” in graphs. The various relationships among all the words and ideas are represented by the connecting lines, which are called “edges” in graphs.
So Why Don’t We Just Teach More Words?
If knowledge matters so much and words represent bundles of knowledge, then wouldn’t it stand to reason that if we taught more words, children would gain more knowledge?
Yes and no.
One study found that to understand the vocabulary in children’s literature, kids would need exposure to over 600,000 words. But if a child goes to school 180 days a year for 13 years and learns 20 words a day (which is generous), that still only totals 46,800 words.
We can’t directly teach every word a child will need to know. Does this mean we shouldn’t teach vocabulary directly? No. Direct vocabulary teaching is necessary.
Direct vocabulary teaching increases knowledge of words that are taught.
Direct vocabulary teaching promotes learning of words beyond those that have been explicitly taught (e.g., being taught a word like thermometer, helps with indirectly learning words like thermal, hypothermia, geothermal).
Direct vocabulary teaching is also insufficient.
This is why the LSA findings are so important.
We are able to comprehend a text more deeply when we have an internal mental model of the context and its related words and ideas. We can build up our internal mental models more quickly by encountering sets of related words and concepts in context, not by studying words in isolation.
What Does This Mean for Children?
One of the best ways to build knowledge and understanding is to engage children in reading a series of texts about the same topic or idea, a text set. In fact, the LSA computer model “learned” around three times more about words not in the text when reading a series of texts.
For example, imagine children reading a text set about cats. They learn about:
Domesticated cats and big cats (like lions)
Cat anatomy, size, behavior, and habitats
Related ideas like carnivores and food chains
With every new text, the network in their brain starts to shift. They refine and develop their understanding of cats and big cats with facts as well as the relationships between each idea (e.g., lions are bigger than domesticated cats, lions live in the wild while domesticated cats live with people, lions are more dangerous to humans than domesticated cats).
But what’s even more remarkable is that even though the texts are about cats, the networks in the children’s brains also refine and adjust for dogs, carnivores, and living animals and non-living things. While not directly mentioned, our brains process words and ideas for concepts that may, on the surface, seem either distantly related or completely unrelated to cats. For example, cats are about the size of river otters, so assuming you have some knowledge of river otters, your brain may develop a connection between those ideas to reinforce your understanding of both cats and river otters when you’re reading a series of texts about cats. These indirect connections make comprehension faster and richer. The more we read about cats, the more complex these ideas and their associations become.
Text sets and our brains are made for each other!
How Does Wonderwood Help?
If we want children to read well, think critically, and learn with ease, we need to help them build knowledge. That happens best when children:
Listen to texts read aloud that are written above what they can read on their own.
Read a large volume of texts, driven by curiosity and interest.
Read a series of texts on the same topic.
Wonderwood offers all three.
As Marilyn Adams writes:
“To grow, [children] must read lots. More specifically, they must read [. . .] texts that offer them new language, new knowledge, and new modes of thought.”
— Marilyn Adams
When children open Wonderwood, they can explore rockets, rainforests, or rabbits and almost any other topic imaginable in a series of engaging, connected texts. Each time they read or listen in Wonderwood, they strengthen the knowledge networks in their brains.
Each visit to Wonderwood provides new words, new ideas, new connections, and new opportunities for building knowledge. The more they read, the more they understand. And because knowledge is sticky, the more they use the app, the more knowledge they gain both in and outside of Wonderwood.
When kids know more, they read better. They think more critically, more independently, and they build a foundation for learning that lasts a lifetime.