Lexical ambiguity in educational contexts refers to the condition where a student carries an everyday-English semantic model for a technical term that directly conflicts with its domain-specific meaning. The conflict is invisible to the student — they believe they understand the term — and produces errors that appear to be conceptual or computational failures when they are in fact linguistic failures upstream of the mathematics.
The Lacefield framework treats lexical ambiguity as a first-class architectural problem, not a side note. Language is not background noise in mathematics. It is one of the primary places where wrong schemas hide. Every tier map node explicitly flags high-ambiguity terms. The intake diagnostic includes a dedicated lexical ambiguity assessment. The DLP stores lexical ambiguity flags as a separate data structure within the language ceiling object.
A student who doesn't know what "product" means mathematically will fail multiplication word problems — not because they can't multiply, but because they don't know what operation is being requested. This is a lexical failure masquerading as a math failure.
These terms carry everyday-English definitions that directly conflict with their mathematical meanings. Each requires explicit disambiguation in the diagnostic and explicit flagging in the DLP when a failure is observed.
| Term | Everyday meaning (failure mode) | Mathematical meaning (required) |
|---|---|---|
| product | Something that gets made or manufactured | The result of multiplication |
| factor | An element that influences something; "a contributing factor" | An integer that divides evenly into another |
| mean | Unkind; or "to signify" | The arithmetic average of a data set |
| range | A span or variety; "a range of options" | The difference between the maximum and minimum values |
| rational | Logical, sensible | Expressible as a ratio of two integers |
| variable | Something that changes or varies | A symbol representing an unknown quantity in an expression |
| root | The base of a plant; origin | A value that makes an equation true; or a radical operation |
| term | A word; a period of time | A single component of an algebraic expression |
| expression | A facial gesture; a saying | A mathematical phrase containing numbers and/or variables without an equals sign |
| value | Worth; importance | The numerical quantity a variable or expression represents |
| difference | The way things are unlike each other | The result of subtraction |
| even | Level, flat, equal | Divisible by 2 with no remainder |
Lexical ambiguity failures are stored in the DLP under the language_ceiling object, separate from schema floor and wrong schema flags. The separation is architecturally intentional — a lexical ambiguity failure is not the same as a conceptual schema failure and requires a different intervention.
general_comprehension: integer 1-5. Overall prose comprehension level. Determines which word problems the student can read precisely enough to set up correctly. Assessed during intake via word problem presentation and read-aloud test.
lexical_ambiguity_flags: array of objects. Each entry: term, observed_failure (what the student did that revealed the ambiguity), everyday_definition_used (what they believed the term meant), session_first_observed, and resolution_status (active | addressed | resolved).
Detection rule: When a student uses a technical term, ask them to define it. If the definition is a correct everyday-English definition that conflicts with the mathematical meaning, log a lexical ambiguity flag for that term. One flag is not enough — watch for it across multiple sessions before marking it resolved.
Every concept node in the tier maps carries a lexical_ambiguity_risks field listing the high-ambiguity terms that appear in problems targeting that concept. The problem generation prompt receives this list and is instructed to include a definitional probe for any term flagged as active in the student's DLP.
This means lexical ambiguity detection is not a separate assessment layer — it is embedded in regular practice. A student working on word problems involving multiplication automatically encounters probes for "product" if that term is flagged in their DLP. The system surfaces the conflict in context, not in isolation.
Lexical ambiguity is categorized within the schema-tracing framework as a specific type of wrong schema — one where the incorrect model lives in semantic memory rather than mathematical schema. The student has a model for the term. The model is wrong (from a domain-specific perspective). It produces predictable failures. And it feels internally consistent to the student, who is using the word correctly by everyday-English standards.
The intervention is identical to the wrong-schema protocol: surface the conflict explicitly before proceeding. Simply providing correct instruction without addressing the semantic conflict will not resolve it — the student will integrate the new information using their existing (incorrect) definition of the term.