💬 NLP

On Ambiguity: The case of fraction, its meanings and roles

This paper addresses the ambiguity of the term "fraction" in elementary arithmetic by introducing precise distinctions such as "fracterm," "fracvalue," and "fracsign" to reclassify it as a collective category of concepts rather than a single mathematical concept, while also exploring its implications for the definition of number systems and structuralism.

Original authors: Jan A Bergstra, John V Tucker

Published 2026-04-07

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Jan A Bergstra, John V Tucker

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to explain the word "Fraction" to a friend. You might say, "It's a part of a whole," or "It's a number like 1/2," or "It's the way we write division on paper."

But here's the problem: You are actually talking about four different things at once.

This paper by Jan Bergstra and John Tucker is like a detective story. The detectives are trying to solve a mystery: Why is the word "fraction" so confusing in math class? They conclude that "fraction" isn't actually a single, clear mathematical concept. Instead, it's a messy umbrella term (or a "category") that covers several distinct ideas that we usually lump together.

To fix this confusion, the authors invent a new set of tools to sort the mess out. Here is the story of their discovery, explained with simple analogies.

The Problem: The "Swiss Army Knife" of Math

Think of the word "fraction" like a Swiss Army Knife.

Sometimes you use the blade (the actual number, like 0.5).
Sometimes you use the screwdriver (the written shape, like $\frac{1}{2}$ ).
Sometimes you use the corkscrew (the text on the page).

If you just say, "Pass me the Swiss Army Knife," your friend doesn't know which tool you want. In math, this causes big problems. For example, if you say "The numerator of 1/2 is 1," that makes sense if you are looking at the shape. But if you are looking at the number (which is the same as 2/4), the "numerator" isn't a fixed property of the number itself. This leads to logical paradoxes where $1 = 2$ !

The Solution: Sorting the Tools

The authors say, "Stop using the word 'fraction' as a catch-all. Let's give each tool a specific name." They introduce four new terms to act as landmarks on a map:

Fracterm (The Blueprint):
- Analogy: This is the recipe or the blueprint. It is the written expression $\frac{a}{b}$ .
- Why it matters: A blueprint has a specific numerator and denominator. You can look at the blueprint $\frac{2}{4}$ and say, "The top number is 2." But the blueprint $\frac{1}{2}$ has a top number of 1. They are different blueprints, even if they build the same house.
Fracvalue (The House):
- Analogy: This is the actual house built from the blueprint. It is the number itself.
- Why it matters: The house built from blueprint $\frac{2}{4}$ is identical to the house built from $\frac{1}{2}$ . Once the house is built, it doesn't have a "numerator" or "denominator" anymore; it's just a number.
Fracsign (The Font):
- Analogy: This is the ink or the font on the paper. It's the visual symbol $\frac{1}{2}$ regardless of what it means.
- Why it matters: It's just the shape. It could be written in red, blue, big, or small.
Fracsign Occurrence (The Specific Instance):
- Analogy: This is one specific drop of ink on a specific page.
- Why it matters: If you see the symbol $\frac{1}{2}$ written twice in a book, those are two different "occurrences," even though they look the same.

The "Fraxion": The Umbrella Term

Since we can't just delete the word "fraction" from our vocabulary (it's too useful!), the authors create a new word: Fraxion.

Think of Fraxion as a box that can contain any of the four things above.
When you say "Fraxion," you are saying, "I am talking about a fraction, but I haven't decided yet if I mean the blueprint, the number, or the ink."
The goal is to realize that when we get confused, it's usually because we are mixing up the blueprint (Fracterm) with the house (Fracvalue).

The Deeper Mystery: What is a "Number"?

Once they sorted out fractions, they realized the same confusion happens with Numbers.

Is the number "7" the same as the word "seven"?
Is it the same as the Roman numeral "VII"?
Is it the same as the binary code "111"?

The authors suggest that a "Number" is like a character in a play.

The Label is the character's name (e.g., "The Hero").
The Shape is the actor playing the role.
- Actor A might wear a tuxedo (Decimal shape).
- Actor B might wear a suit (Binary shape).
- Actor C might wear a robe (Roman numeral shape).

They are all the same "Hero" (the abstract concept), but they look different and act slightly differently depending on the "Shape" (the system) they are in. The paper argues that we shouldn't try to find one perfect definition of "7" that covers all actors. Instead, we should just pick a "Shape" (like the decimal system) and stick with it for school math.

Why Does This Matter?

You might ask, "Why do we need all these fancy words?"

To Stop Logic Errors: If you mix up the blueprint and the house, you might think $\frac{2}{4}$ is different from $\frac{1}{2}$ in a way that breaks math rules. By separating them, we stop these errors.
To Help Teachers: Teachers often struggle to explain why $\frac{2}{4}$ can be "simplified" but the number $0.5$ cannot be "simplified." The answer is: You are simplifying the blueprint, not the house.
To Handle "Divide by Zero": In computer science, what happens if you divide by zero?
- If you treat it as a blueprint (Fracterm), you can say "This is an error."
- If you treat it as a number (Fracvalue), you might say "It's undefined."
- The authors suggest a middle ground: treat it as a special "error number" (like a warning light) that spreads through your calculation, so the computer knows something went wrong without crashing.

The Bottom Line

The paper concludes that "Fraction" is not a single concept; it is a "Category" of related ideas.

Just as "Vehicle" is a category that includes cars, bikes, and boats (which are all different things), "Fraction" is a category that includes written expressions, numbers, and symbols. By using the new terms (Fracterm, Fracvalue, Fraxion), we can talk about math with much more precision, avoiding the confusion that has plagued students and teachers for centuries.

In short: Stop trying to force the blueprint, the house, and the ink to be the same thing. Give them their own names, and the math will make much more sense.

1. Problem Statement

The paper addresses the fundamental ambiguity of the term "fraction" in elementary arithmetic and mathematical discourse. Despite the high value placed on precision in mathematics, the word "fraction" lacks a single, well-defined mathematical concept. Instead, it functions as a collective term for a plurality of interpretations ranging from:

Syntactic constructions: Expressions with a division operator (e.g., $1/2$ , $4/6$ , $1/0$ ).
Semantic values: Rational numbers or quotients.
Textual representations: Signs or symbols on a page.

The authors argue that this ambiguity is not merely a linguistic nuisance but a structural issue that leads to logical paradoxes (e.g., proving $1=2$ ) and pedagogical confusion. They posit that "fraction" does not qualify as a mathematical concept (which requires a unique, precise definition) but rather as a category (a bundle of related concepts).

2. Methodology

The authors employ a method of ambiguity resolution based on the following steps:

Identification of Landmarks: Instead of forcing a single definition, they introduce precise "landmark notions" that approximate the various meanings of the ambiguous term.
Disambiguation via New Terminology: They propose a new vocabulary to distinguish between the syntactic, semantic, and textual layers of fractions.
Level of Abstraction Analysis: They analyze the hierarchy of abstraction, moving from concrete text to abstract values.
Application to "Fractalk": They test these new terms by reconstructing common arithmetic arguments and paradoxes to see if the ambiguity causes logical failures.
Extension to Number Theory: They generalize the findings to the concept of "number," introducing the framework of Labels and Shapes to analyze how different number systems are constructed.

3. Key Contributions and Definitions

A. The Four Landmark Notions

To resolve the ambiguity of "fraction," the authors define four distinct concepts:

Fracterm: A syntactic expression where division is the leading function symbol (e.g., $a \div b$ ). It is a structural entity, not necessarily a number. It allows for undefined operations like $1/0$ .
Fracvalue: The semantic value denoted by a fracterm. These are typically rational numbers but can include "peripheral numbers" (e.g., $\bot$ for error, $\infty$ , NaN) depending on the algebraic structure (e.g., Common Meadows).
Fracsign: The abstract symbol or sign representing a fraction (e.g., the glyph " $\div$ " or the layout " $\frac{a}{b}$ ").
Fracsign Occurrence: A specific instance of a fracsign in a text (e.g., a specific ink mark on a blackboard), which carries context-dependent meaning.

B. The Concept of "Fraxion"

The authors introduce Fraxion as a disjunctive concept:
$\text{Fraxion} = \text{Fracvalue} \oplus \text{Fracterm} \oplus \text{Fracsign}$
A fraxion is an unambiguous category that explicitly allows for the choice of interpretation based on context. It is designed to be a "near notion" to the ambiguous term "fraction," allowing for semantic freedom while maintaining logical rigor.

C. Labels and Shapes (Number Theory)

Extending the analysis to numbers, the authors propose:

Label: A type identifier (e.g., Natural, Integer, Rational).
Shape: A specific construction or presentation method for a label (e.g., Decimal naturals, Von Neumann naturals, Dedekind cuts).
Normal vs. Subnormal Shapes: A shape is normal if syntactic equality implies semantic equality ( $i =_S j \iff i \equiv_L j$ ). A shape is subnormal if distinct syntactic forms represent the same value (e.g., $007 \equiv 7$ in decimal).

D. Vertical vs. Horizontal Cross-Cutting Notions

Vertical Cross-Cutting: A notion that spans different levels of abstraction (e.g., "fraction" spans fracterm $\to$ fracvalue). The authors argue "fraction" is a vertical cross-cutting notion.
Horizontal Cross-Cutting: A notion spanning different shapes at the same abstraction level (e.g., "number 7" across decimal, binary, and Roman shapes).

4. Results and Analysis

Resolution of Paradoxes

The authors demonstrate how their terminology resolves classic arithmetic paradoxes:

The $1=2$ Paradox: In the argument " $1/2$ has numerator 1, $2/4$ has numerator 2, $1/2 = 2/4$ , therefore $1=2$ ," the error arises from confusing fracterms (where numerators are unique) with fracvalues (where $1/2$ and $2/4$ are the same value). By distinguishing levels, the contradiction vanishes.
Division by Zero: The framework allows for a rigorous treatment of $1/0$ . In Common Meadows (algebraic structures), $1/0$ is assigned a peripheral value $\bot$ (error), which propagates through calculations. This distinguishes the fracterm $1/0$ (which exists syntactically) from its fracvalue ( $\bot$ ).

The Nature of "Fraction"

The paper concludes that "fraction" is a productively ambiguous category. It is not a single concept but a collection of concepts that reinforce each other's usage in education and practice. Attempting to force a single definition is counter-productive; instead, one should use the precise landmark terms (fracterm, fracvalue) when precision is required.

Implications for Education and Computing

Pedagogy: Elementary arithmetic is essentially the teaching of shapes (e.g., decimal fractions) and rules for their transformation. The ambiguity of "fraction" in schools often stems from mixing syntactic rules (simplifying $4/6$ to $2/3$ ) with semantic values.
Computing: In computer arithmetic, the distinction is vital. "Prospective division" (computing $P/Q$ without knowing $Q \neq 0$ ) requires handling $Q=0$ via peripheral values ( $\bot$ ), whereas "retrospective division" (mathematical proofs) assumes the quotient exists.

5. Significance

Philosophical: The paper challenges the structuralist view that mathematical objects are defined solely by their axiomatic properties. It argues that the textual and syntactic presentation (the "shape") is intrinsic to the concept in elementary contexts.
Technical: It provides a formal calculus (Fracterm Calculus) for handling division, including division by zero, which is essential for computer algebra systems and theorem provers.
Methodological: It offers a general method for analyzing ambiguous terms in mathematics by decomposing them into "landmark notions" and "cross-cutting categories," applicable beyond fractions to other ambiguous terms like "probability" or "number."

In summary, Bergstra and Tucker argue that the ambiguity of "fraction" is not a flaw to be eliminated but a feature of mathematical language that must be managed by distinguishing between syntax (fracterm), semantics (fracvalue), and text (fracsign). This distinction is crucial for rigorous logic, computer science, and effective mathematics education.