Literal Equations and Formulas

Intuition

The previous lesson reduced every linear equation to $ax = b$ and read its answer straight off the two numbers $a$ and $b$ . There is one solution when $a \neq 0$ , none when $a = 0$ but $b \neq 0$ , and every real number when both are zero. That worked because $a$ and $b$ were numbers you could look at. This lesson asks the same question when $a$ and $b$ are letters, and one quiet fact changes everything. Which letter you are solving for is a choice you make, not something the equation tells you. In $ax = b$ you might be solving for $x$ with $a$ and $b$ handed to you, or solving for $a$ with $x$ and $b$ handed to you. Those are the same three symbols posing two genuinely different problems. And the moment you divide by the letter $a$ to write $x = \frac{b}{a}$ , you have quietly assumed $a$ is not zero. Unlike a number, a letter will never confirm that for you. So the honest answer to a literal equation is rarely a single formula. It is a short list of cases.

Learning goals

Choose which letter is the unknown and which are parameters
Split $ax = b$ into its three parameter cases
Carry the nonzero assumption whenever you divide by a letter
Attach a new condition where a divisor can vanish
Factor the target out and condition the whole factor

The unknown and the parameters

A literal equation is an equation with more than one letter in it, and you already know the mechanics of rearranging one. Isolate the letter you want by undoing the operations around it, treating the other letters as fixed numbers. What that earlier work could only gesture at is the idea this lesson is built on. In any literal equation, one letter is singled out to be solved for, and every other letter stands for a fixed but unstated number. The letter you solve for is the unknown. Each letter held fixed while you solve is a parameter.

Here is the point that is easy to miss: nothing about the symbol $a$ makes it a parameter, and nothing about $x$ makes it the unknown. The roles are assigned by the question, not carried by the letters. Take the single equation $ax = b$ . Solve it for $x$ , treating $a$ and $b$ as parameters, and you divide by $a$ to reach $x = \frac{b}{a}$ . Solve the very same equation for $a$ instead, treating $x$ and $b$ as parameters, and you divide by $x$ to reach $a = \frac{b}{x}$ . Same three symbols, two different problems, because a different letter was chosen to be the unknown. Deciding which letter is the unknown is the first move you make, before any algebra happens at all.

By long convention, letters near the start of the alphabet ( $a$ , $b$ , $c$ , and often $k$ ) are used for parameters. By the same convention, letters near the end ( $x$ , $y$ , $z$ ) are used for the unknown, which is why $ax = b$ reads so naturally as an instruction to solve for $x$ . But that convention is only a courtesy to the reader. The mathematics is identical whichever letter you decide to isolate. A good part of this lesson is learning to see the choice rather than let the alphabet make it for you.

Dividing by a letter you cannot inspect

When the coefficient was a number, classifying $ax = b$ took a glance. When the coefficient is a parameter, the same classification is still true, but you can no longer see which case you are in. That single change is what turns solving a literal equation into a case analysis, so it is worth stating the classification carefully with $a$ and $b$ understood as parameters.

The three cases of $ax = b$ when $a$ and $b$ are parameters#

Solve $ax = b$ for the unknown $x$ , with $a$ and $b$ standing for fixed but unstated numbers. The only tool that isolates $x$ is division by $a$ , and whether that tool is available depends entirely on whether $a$ is zero.

If $a \neq 0$ , division by $a$ is legal, and it is a reversible step, so it preserves the solution set. It gives the single value $x = \frac{b}{a}$ , and the solution set is $\left\{\frac{b}{a}\right\}$ . Exactly one solution.

If $a = 0$ , division is off the table, and the equation reads $0 \cdot x = b$ . The left side is $0$ for every $x$ . When $b \neq 0$ this asks for $0 = b$ with $b$ nonzero, which no $x$ can satisfy, so the solution set is $\varnothing$ : no solution. When $b = 0$ it reads $0 = 0$ , true for every $x$ , so the solution set is all of $\mathbb{R}$ : every real number.

These three cases are exhaustive and mutually exclusive, because either $a = 0$ or it is not, and in the boundary case $a = 0$ either $b = 0$ or it is not. So the answer is not a single formula in $a$ and $b$ . It is a function of the two parameters, defined by cases, with the dividing line drawn at $a = 0$ .

Nothing in that argument is new; it is the trichotomy the previous lesson proved. What is new is that you cannot carry it out by inspection. With a number in front of $x$ you simply look and see which case holds. With a parameter you cannot look, because $a$ is a letter whose value you were never told. Writing $x = \frac{b}{a}$ in one confident line is really a bet that you are in the first case. And the whole discipline of this lesson is refusing to place that bet silently. This is also what the earlier advice to check your result after dividing by a letter was quietly protecting against. Division by $a$ is safe only in the first case, and the check is what catches the times you had wandered into one of the other two.

Check your understanding

You solve $ax = b$ for $x$ , and a classmate writes $x = \frac{b}{a}$ as the complete answer. For which values of the parameters is that single line the whole story?

A When

a \neq 0

B When

b \neq 0

C For all

a

and

b

D When

a = 0

For which values of the parameter?

The most revealing use of the classification runs the logic backward. Instead of being handed fixed numbers $a$ and $b$ , you are handed a whole family of equations that depends on a parameter $k$ . The question is this: for which values of $k$ does the equation have one solution, no solution, or infinitely many? The recipe is always the same. Reduce the equation to the form $(\text{coefficient})\,x = (\text{constant})$ , where the coefficient and the constant now depend on $k$ , and then apply the trichotomy to those two expressions. The following pair looks almost identical, yet the pair is chosen to show that the coefficient alone does not settle the answer.

Worked example 1 For which $k$ does $kx + 3 = 2x + k$ have one, none, or infinitely many solutions?

Treat $k$ as the parameter and $x$ as the unknown. Collect the $x$ terms on the left and the constants on the right:

kx - 2x = k - 3, \qquad (k - 2)x = k - 3.

The coefficient of $x$ is $k - 2$ and the constant is $k - 3$ . Apply the classification to them.

If $k \neq 2$ , the coefficient $k - 2$ is nonzero, so there is exactly one solution:

x = \frac{k - 3}{k - 2}.

If $k = 2$ , the coefficient is zero and the constant is $k - 3 = -1$ , so the equation becomes $0 \cdot x = -1$ . No value of $x$ works, and the solution set is $\varnothing$ .

Could any $k$ give infinitely many solutions? That would need the coefficient and the constant to vanish together, $k - 2 = 0$ and $k - 3 = 0$ at once, which asks $k$ to be both $2$ and $3$ . Impossible. So this family has one solution for every $k \neq 2$ and no solution at $k = 2$ , and infinitely many solutions simply never occur.

Worked example 2 For which $k$ does $kx + 6 = 2x + 3k$ have one, none, or infinitely many solutions?

Again treat $k$ as the parameter. Collect the $x$ terms and the constants:

kx - 2x = 3k - 6, \qquad (k - 2)x = 3(k - 2).

The coefficient is the same $k - 2$ as before, but the constant is now $3(k - 2)$ , which carries that same factor.

If $k \neq 2$ , the coefficient is nonzero, and dividing gives

x = \frac{3(k - 2)}{k - 2} = 3.

The factor $k - 2$ cancels, so the solution is $x = 3$ for every $k \neq 2$ , the identical value no matter which such $k$ you were handed.

If $k = 2$ , the coefficient is zero and the constant is $3(k - 2) = 0$ as well, so the equation becomes $0 \cdot x = 0$ , true for every $x$ . The solution set is all of $\mathbb{R}$ . This time no value of $k$ produces “no solution,” because the instant the coefficient $k - 2$ vanishes, the constant $3(k - 2)$ vanishes with it.

Put the two families side by side. Their coefficient of $x$ is the same expression, $k - 2$ . So the coefficient by itself cannot be what decides the outcome, since one family loses its solution at $k = 2$ and the other gains a whole line of them. The tiebreaker is the constant at the exceptional value. In the first family the constant is $-1$ at $k = 2$ , nonzero, so $0 \cdot x = -1$ is a contradiction. In the second the constant is $0$ at $k = 2$ , so $0 \cdot x = 0$ is an identity. That is exactly the fork from the previous lesson: once the coefficient $a$ is zero, it is whether $b \neq 0$ or $b = 0$ that separates no solution from every real number. The coefficient tells you when you are standing on the boundary $k = 2$ ; the constant at that boundary tells you which way you fall off it.

The solution set as a function of the parameter k. Away from the single exceptional value k = 2, both equations of the contrasting pair have exactly one solution. At k = 2 they split: equation (i) reduces to 0 = -1 and has no solution, while equation (ii) reduces to 0 = 0 and is solved by every real number. The coefficient k - 2 is identical in both, so it is the constant on the right at k = 2 that decides which outcome you meet.

Check your understanding

For which value of $k$ does $kx - 8 = 4x - 2k$ have infinitely many solutions?

k = 4

k = 2

k = -4

D no such

k

When a formula hides a condition

Solving a real formula for one of its variables can produce a restriction the original never visibly had. That restriction is not arbitrary, and it is not a ritual to attach to every answer. It appears at exactly the step where you divide by an expression that could be zero, the same division whose danger the classification spelled out. Whenever a rearrangement carries such a step, name the condition out loud.

Worked example 3 Solve the thin-lens equation for the image distance $v$

The thin-lens equation from optics ties the focal length $f$ of a lens to the object distance $u$ and the image distance $v$ :

\frac{1}{f} = \frac{1}{u} + \frac{1}{v}.

Solve it for $v$ . Isolate the $\frac{1}{v}$ term by subtracting $\frac{1}{u}$ from both sides, then combine the right side over a common denominator:

\frac{1}{v} = \frac{1}{f} - \frac{1}{u} = \frac{u - f}{fu}.

Take the reciprocal of both sides:

v = \frac{fu}{u - f}.

The original already carries hidden restrictions, since $f$ , $u$ , and $v$ all sit in denominators and so must be nonzero. On top of those, the rearranged formula divides by $u - f$ , so it adds the new demand $u \neq f$ . That extra condition is telling the truth about lenses. An object placed exactly at the focal length, $u = f$ , produces no image at any finite distance, because the rays leave the lens parallel, so there is no $v$ to report. The algebra discovered the physics: at $u = f$ the denominator $u - f$ is zero and $v$ does not exist.

Worked example 4 Solve the parallel-resistor formula for $R_2$

Two resistors $R_1$ and $R_2$ wired in parallel act like a single resistor $R$ given by

R = \frac{R_1 R_2}{R_1 + R_2}.

Solve for $R_2$ . Clear the fraction by multiplying both sides by $R_1 + R_2$ , then gather every term containing the target $R_2$ on one side:

R(R_1 + R_2) = R_1 R_2, \qquad R\,R_1 = R_1 R_2 - R\,R_2.

The target now sits in two terms on the right, so factor it out and divide, the collect-and-factor move you already know:

R\,R_1 = R_2(R_1 - R), \qquad R_2 = \frac{R\,R_1}{R_1 - R}.

The rearranged formula divides by $R_1 - R$ , so it requires $R_1 \neq R$ . Once more the condition is physically honest: the combined resistance $R$ of a parallel pair is always smaller than either resistor alone. So $R$ can never equal $R_1$ , and asking for a fixed $R$ equal to $R_1$ describes an impossible circuit.

Worked example 5 Solve $C = \frac{5}{9}(F - 32)$ for $F$

It is worth seeing a rearrangement that produces no condition at all, so the restriction never looks like a reflex. The Celsius-to-Fahrenheit relationship is

C = \frac{5}{9}(F - 32).

Solve for $F$ . Multiply both sides by the reciprocal $\frac{9}{5}$ , then add $32$ :

\frac{9}{5}C = F - 32, \qquad F = \frac{9}{5}C + 32.

Every step here either multiplied by the nonzero constant $\frac{9}{5}$ or added the constant $32$ . Nowhere did we divide by an expression that could be zero, so no value of $C$ is excluded and the formula works for every temperature. In a rearrangement built only from adding constants and multiplying or dividing, a new condition appears exactly where you divide by an expression that can vanish. Divide by a genuine constant and nothing is excluded.

Check your understanding

The sum of an infinite geometric series is $S = \dfrac{a}{1 - r}$ . Solving it for the ratio $r$ gives $r = \dfrac{S - a}{S}$ . Which nonzero condition does this rearrangement introduce?

S \neq 0

a \neq 0

r \neq 1

D no condition is needed

The condition belongs to the factor

You already know that when the target appears in several terms you collect those terms, factor the target out, and only then divide. The one thing to get right here is where the nonzero condition lands. It lands on the whole factor you divide by, not on the individual letters inside it.

Worked example 6 Solve $ax + b = cx + d$ for $x$

The unknown $x$ starts on both sides. Collect its terms on the left and the constants on the right:

ax - cx = d - b.

Both terms on the left contain $x$ , so factor it out. The factor left behind is $a - c$ :

(a - c)x = d - b.

Now $x$ appears once, multiplied by $a - c$ , so divide by that whole factor:

x = \frac{d - b}{a - c}, \qquad a \neq c.

The condition is $a \neq c$ , a single statement about the factor $a - c$ . It is not ” $a \neq 0$ and $c \neq 0$ .” You are dividing by $a - c$ , not by $a$ and not by $c$ , so the only quantity that must stay nonzero is the difference. Coefficients $a = 5$ and $c = 5$ break the formula even though neither is zero, while $a = 0$ and $c = 3$ are perfectly fine, because there $a - c = -3$ is nonzero.

And when the forbidden case does happen, when $a = c$ , the equation has not disappeared; it has returned to the classification. Setting $a = c$ turns $(a - c)x = d - b$ into $0 \cdot x = d - b$ , the familiar fork one more time: no solution when $d \neq b$ , and every real number when $d = b$ . The excluded condition on a formula and the boundary case of the case analysis are one fact seen from two sides. Dividing by $a - c$ writes the answer as a formula valid on the region $a \neq c$ . And the trichotomy takes over precisely on the line $a = c$ , where the formula is not allowed to go.

Common mistakes

Thinking a letter has a fixed role. Which letter is the unknown and which are parameters is decided by the question, not by the alphabet. The same $ax = b$ can be solved for $x$ (giving $x = \frac{b}{a}$ , needing $a \neq 0$ ) or for $a$ (giving $a = \frac{b}{x}$ , needing $x \neq 0$ ).
Dividing by a letter without splitting into cases. Writing $x = \frac{b}{a}$ as the whole answer to $ax = b$ silently assumes $a \neq 0$ . The honest answer also covers $a = 0$ , which gives no solution when $b \neq 0$ and every real number when $b = 0$ .
Expecting all three outcomes to be reachable. In a parameter family, only the outcomes the constant permits can occur: $kx + 3 = 2x + k$ never has infinitely many solutions, and $kx + 6 = 2x + 3k$ never has none.
Reading the outcome off the coefficient alone. The coefficient $k - 2$ is identical in both families above; it is the constant on the right at $k = 2$ that separates no solution from every real number.
Putting the condition on the wrong thing. Solving $ax + b = cx + d$ needs $a \neq c$ , one condition on the factor $a - c$ , not the two separate conditions $a \neq 0$ and $c \neq 0$ .
Attaching a condition to every rearrangement. A restriction appears only where you divide by something that can be zero. The formula $F = \frac{9}{5}C + 32$ excludes no values of $C$ , because nothing in it can vanish.

Key takeaways

In a literal equation you choose one letter to be the unknown; the rest are parameters, fixed but unstated numbers. The choice is yours, not a property of the symbols.
Solving $ax = b$ for $x$ has no single answer: one solution $x = \frac{b}{a}$ when $a \neq 0$ , no solution when $a = 0$ and $b \neq 0$ , and every real number when $a = 0$ and $b = 0$ . The answer is a function of the parameters, defined by cases.
Every time you divide by a letter you assume you are in the first case. With a number you can check by looking; with a parameter you cannot, so the cases must be carried explicitly.
To find which parameter values give one, none, or infinitely many solutions, reduce to $(\text{coefficient})\,x = (\text{constant})$ and apply the classification. The coefficient marks the boundary, and the constant at the boundary decides the fork.
In a rearrangement built only from adding constants and multiplying or dividing, a new nonzero condition appears exactly where you divide by an expression that can vanish. The rearrangement $v = \frac{fu}{u - f}$ picks up the condition $u \neq f$ that way, while a rearrangement that divides only by constants, like $F = \frac{9}{5}C + 32$ , excludes nothing.
When the target sits in several terms, factor it out and put the condition on the whole factor. For example, $ax + b = cx + d$ gives $x = \frac{d - b}{a - c}$ with $a \neq c$ , and on the line $a = c$ the trichotomy takes over.

Literal Equations and Formulas

Learning goals

The unknown and the parameters

Dividing by a letter you cannot inspect

For which values of the parameter?

When a formula hides a condition

The condition belongs to the factor

Common mistakes

Practice

Multiple Choice Questions (MCQ)

Free Response Questions (FRQ)