Questions: Vocabulary Growth and Semantic Development

Question 1

A 20-month-old calls all four-legged animals 'dog.' Over the next few months, as she learns 'cat,' 'horse,' and 'rabbit,' her use of 'dog' becomes restricted to actual dogs. What best explains this self-correction?

• A. Her parents explicitly teach her the correct boundaries of each animal category
• B. Her phonological development improves, allowing her to hear differences between animal names
• C. Adding new words forces finer distinctions, which simultaneously sharpens the meaning of existing words
• D. She passes through a phase of underextension of 'dog' before reaching correct boundaries

Answer: (C)

Overextension (using 'dog' for all four-legged animals) reflects a category whose semantic boundaries are drawn too broadly. When the child acquires 'cat,' 'horse,' and 'rabbit,' those words compete for semantic space and force finer distinctions — the meaning of 'dog' gets sharpened by what it is *not*. This is the semantic network in action: vocabulary breadth and depth are not separate goals. Option A plays a role but doesn't explain the general self-correcting mechanism. Option B is irrelevant — overextension is a semantic problem, not phonological. Option D is wrong: she doesn't pass through underextension on the way to correct use.

Question 2

Why does explicitly teaching semantic relationships ('this is the opposite of X,' 'this belongs to the same category as Y') accelerate vocabulary acquisition more effectively than simply repeating new words?

• A. Repetition fails to activate the phonological loop needed for word retention
• B. Semantic relationships tap into the thematic organizational system children prefer at all ages
• C. It builds the network infrastructure of connections that makes words accessible and usable
• D. It reduces cognitive load by limiting the number of new words introduced per session

Answer: (C)

The key insight is that words are stored in semantic networks, and the density of connections determines how accessible and usable a word is. Teaching semantic relationships (synonymy, antonymy, category membership) directly builds those connections. A richly connected word is easy to retrieve, deploy correctly, and distinguish from related words. Option B is partially backwards — the developmental shift is *from* thematic to taxonomic organization, and taxonomic (category-based) teaching supports the academically important later style, not an always-preferred thematic one.

Question 3

Fast mapping gives children a complete and stable word meaning after one or a few exposures, which is why the vocabulary explosion is possible.

• T. True
• F. False

Answer: False

Fast mapping creates a *rough initial representation* — a placeholder — not a complete definition. The placeholder is refined through subsequent encounters in different contexts. What fast mapping achieves is speed of *initial binding*: children can form a working representation after very few exposures and begin using the word, without requiring extended repetition. The vocabulary explosion is possible because of this speed, not because children achieve full word knowledge immediately. Word meaning deepens over many subsequent encounters.

Question 4

As children acquire more vocabulary, breadth and depth of lexical knowledge are separate goals — knowing more words does not inherently sharpen the meaning of words already known.

• T. True
• F. False

Answer: False

The semantic network model predicts the opposite: learning more words *simultaneously* sharpens existing words. When a child learns 'cat,' the meaning of 'dog' becomes more precise because it now has a closer semantic competitor. Vocabulary breadth and depth are mutually reinforcing within the network structure. This is why rich vocabulary environments accelerate the quality of word knowledge, not just the quantity — and why overextension errors self-correct automatically as the network densifies.

Question 5

Why do young children who know fewer words show more overextension and underextension errors, and what does this reveal about how word meanings are stored?

Model answer: Because word meanings are defined partly by their relationships to other words in the semantic network. A child who knows only 'dog' has no neighboring concepts to mark its boundaries — the category spreads until it encounters some other stored concept that blocks it. As more words are learned, each carves out semantic space, forcing neighboring words to contract to correct boundaries. This reveals that word meaning is not a fixed private definition but an emergent property of how a word is positioned relative to other words in the network.

Meaning is relational. A word stored in isolation is like a point on a map with no other points to provide scale — its boundaries remain undefined. Words become precise when the network around them densifies. Overextension and underextension are symptoms of an underdense network, not of incorrect learning, which is why they self-correct automatically as more words are acquired and the network fills in.