Singapore School Mathematics Operating Manual · Chapter 42
Some mathematical moves can be undone exactly. Others cannot.
Adding 5 to both sides of an equation is reversible because subtracting 5 restores the original state. Squaring both sides is not always reversible because sign information can be lost. Rounding a decimal cannot usually recover the original exact value. Dividing by an expression can close the branch where that expression equals zero.
This chapter develops reversibility and optionality: understanding which mathematical decisions preserve every future route and which decisions collapse information, remove branches or force commitment.
1. Reversible transformations preserve information
x+4=9 and x=5 are equivalent because subtracting 4 and adding 4 are inverse operations.
No solution information is lost.
2. One-way transformations can enlarge the candidate set
From √(x+2)=x, squaring gives x+2=x².
Every original solution satisfies the squared equation, but not every squared-equation solution must satisfy the original.
Squaring loses sign restrictions.
3. One-way transformations can also shrink the candidate set
From x(x−3)=0, dividing by x gives x−3=0 and removes x=0.
The division silently assumes x≠0.
4. Optionality means keeping multiple valid future routes open
If a result can stay exact as √3, delaying decimal conversion preserves both exact symbolic manipulation and later numerical approximation.
Rounding immediately closes the exact route.
5. Exact form preserves option value
√3×√3=3 exactly.
Using 1.732×1.732 gives an approximation instead.
Keeping exact form costs little and preserves more future capability.
6. Factored form and expanded form offer different options
(x−2)(x−3) exposes roots.
x²−5x+6 exposes coefficients.
Converting between them is reversible algebraically, so both options remain available.
7. Canonicalisation can preserve meaning while reducing clutter
12/18→2/3 removes redundant scale but not rational value.
This is a safe compression because the transformation is exactly reversible at the level of value, even though the original numerator and denominator are not preserved.
8. Rounding is irreversible compression
Many exact values round to 2.7.
Knowing only 2.7 cannot recover the original value uniquely.
The inverse image is an interval.
9. Sampling is often irreversible
If a continuous curve is observed only at five points, behaviour between them is lost.
No later calculation can reconstruct the exact curve without added assumptions.
10. Aggregation closes routes to individual detail
A mean preserves average level but discards individual observations.
Once only the mean remains, questions about maximum, order or exact distribution may become impossible.
11. Variable substitution can preserve optionality when invertible
Let u=x−3.
Then x=u+3, so the substitution is fully reversible.
Temporary coordinates change representation without losing the original state.
12. Non-injective substitutions need branch restoration
Let u=x².
Recovering x requires x=±√u where allowed.
The substitution compresses sign information and creates a branch obligation on return.
13. Domain restriction is sometimes a deliberate irreversible decision
Restricting f(x)=x² to x≥0 creates a one-to-one function and permits an inverse.
The negative branch is intentionally excluded.
This is legitimate when the problem’s model or definition supports the restriction.
14. Invalid restriction manufactures certainty
Choosing x>0 merely because the positive root is convenient is not a legitimate use of optionality.
Closing a branch requires mathematical authority.
15. Inequality transformations can reverse order
Multiplying by a negative number is reversible, but the inequality sign must reverse.
The operation preserves information only when its directional effect is carried correctly.
16. Case splitting preserves optionality explicitly
For |x|, split into x≥0 and x<0.
Each branch is solved under its own condition.
The workflow keeps both routes visible until recomposition.
17. Premature case selection destroys optionality
If the problem admits both signs and one is discarded without evidence, later steps may appear simpler but the final answer becomes incomplete.
18. Approximation creates commitment
Replacing π by 3.14 commits to a finite approximation.
Further exact cancellation involving π is no longer available.
19. Calculator mode is a hidden state that changes interpretation
sin(30) means different things in degree and radian modes.
A calculation may be numerically reversible yet semantically wrong because the state was misidentified.
20. A reversible method often supports stronger checking
If every algebraic step is reversible, the final equation is equivalent to the original.
This gives stronger completeness confidence than a chain containing one-way steps.
21. One-way steps create verification debt
After squaring, cancelling or approximating, extra checking becomes necessary.
The lost information must be recovered by original-domain verification.
22. Optionality can be valuable in examination strategy
If two methods are available, delaying commitment until the structure is clearer can avoid a long dead end.
For example, factorisation may be preferable if obvious; otherwise the quadratic formula remains available.
23. Optionality has a cost
Keeping every possible route alive forever creates clutter.
Strong problem solving preserves options until enough evidence supports commitment, then closes branches decisively.
24. Branch-and-bound thinking balances options and pruning
Generate plausible branches, then eliminate those violating bounds or constraints.
This principle appears in integer search, case analysis and optimisation.
25. Irreversible decisions should be delayed near uncertainty
If a measured value lies close to a threshold, rounding or categorical classification too early can flip the decision.
Preserve the interval or higher precision until the boundary is resolved.
26. Reversibility supports debugging
When a result looks wrong, reversible steps can be walked backward.
Irreversible steps require comparison with the original source because the prior state may no longer be recoverable.
27. A practical reversibility audit
Ask:
Can this step be undone uniquely? Does it lose sign, domain, precision or individual detail? If I commit now, what future methods become unavailable? Is the branch closure justified? Will this step create a verification obligation later?
28. Independent practice
1. Is adding 7 to both sides reversible?
2. Why can squaring create extra candidates?
3. Why can dividing by x lose a solution?
4. Explain why √2 is often better than 1.414 in intermediate work.
5. Is reducing 12/18 to 2/3 value-preserving?
6. Why is rounding irreversible?
7. What branch must be restored after u=x²?
8. What is verification debt?
29. Worked answers
1. Yes; subtract 7.
2. It removes sign information, so values not satisfying the original equation may satisfy the squared version.
3. Division assumes x≠0, so the x=0 branch can disappear.
4. Exact form preserves symbolic cancellation and avoids early rounding error.
5. Yes. The rational value is unchanged.
6. Many exact inputs map to the same rounded output.
7. x=±√u where allowed.
8. Extra checking required because a one-way transformation lost information.
30. Continue through Batch 11
Use Bottlenecks and Limiting Constraints to decide which option matters most, Coupling and Interaction Effects when one decision changes another variable, and Local and Global Optimisation when preserving options can prevent a locally attractive but globally poor choice.
Return to the BTT Mathematics Hub for Batch 11.