Logarithm Challenges: Spot the Structure, Check Every Step

A squared equation can admit an extra answer

Squaring both sides is useful when an equation contains a square root. It can also hide a sign difference. The statements $A=B$ and $A^2=B^2$ are not equivalent without extra information: both $A=B$ and $A=-B$ give equal squares.

A principal square root is nonnegative. Therefore, in $\sqrt{5-t}=3-t$, you need two restrictions before squaring: the radicand requires $t\le5$, and the right side must be nonnegative, requiring $t\le3$. The stronger combined condition is $t\le3$.

If $t=\log_2 x$, also retain the original logarithm condition $x>0$. A temporary variable has its own allowed values; carrying them alongside the equation makes it much easier to reject an unsuitable quadratic root.

After solving the squared equation, check each candidate in the unsquared one. A candidate that gives 1 on the left and−1 on the right will pass the squared equation, but fail the original. Do not report its corresponding $x$-value as a solution.

Use the activity to compare the two sides at the candidate values. First predict which candidate survives. Then reveal the actual unsquared outputs. This check is short, exact and more informative than simply being told that a root is “extraneous”.

The general routine is: identify the radical’s domain, require the other side to have the same nonnegative sign, square, solve, and substitute back. The sign condition and original check explain why the final answer may contain fewer roots than the quadratic.

The square root returns a nonnegative value

Recognizing a hidden square can turn a nested radical into a few clean steps. But one small sign matters throughout:

It is not always $z$. If $z=-3$, then $\sqrt{z^2}=\sqrt9=3$, not−3. The vertical bars mean absolute value (परम मान), which makes the result nonnegative.

To denest an expression such as $\sqrt{A+B-2\sqrt{AB}}$, recognize $(\sqrt A-\sqrt B)^2$, with $A,B\ge0$. Its principal square root is $|\sqrt A-\sqrt B|$. Compare the two terms before dropping the bars.

For instance, $9-4\sqrt5=(\sqrt5-2)^2$. Since $\sqrt5>2$, its square root is $\sqrt5-2$. You could also write the radicand as $(2-\sqrt5)^2$, but then the expression inside the absolute value is negative. The final answer must still be $\sqrt5-2$.

For a longer nested expression, start at the innermost root. Look for two squares whose sum gives the constant term and whose double product gives the surd term. Simplify that layer, substitute it into the next, and check signs again.

The activity separates a number, its square and its principal square root. Use an exact surd preset after trying ordinary positive and negative numbers. The notation is doing the same job in both situations: a square forgets the original sign, and the principal square root chooses the nonnegative result.

A long sum with just two surviving endpoints

The symbol $\sum$, read “sum”, tells you to add a family of terms over an index range. Before simplifying a long sum, write its first few terms and its final term. This reveals whether the pattern really links together and protects you from an off-by-one error.

For the family $\log_b((r+1)/r)$, the first term at $r=m$ is $\log_b((m+1)/m)$, while the last at $r=n$ is $\log_b((n+1)/n)$. There are $n-m+1$ terms. Assume positive integers $1\le m\le n$, and a valid fixed base $b$.

Because the bases agree, adding the logs forms one logarithm of a product:

Now the common factors cancel. Everything in the middle appears once above and once below, leaving

This is telescoping. The cancellation happens inside the product after using the log law; you are not cancelling separate terms across addition.

Use the activity with a small range so every factor is visible. Then try the larger worked example. The first denominator and final numerator follow exactly the same rule, however many middle terms there are. Keep the index endpoint $n$ distinct from the final numerator $n+1$; confusing them changes the answer.

A product of logs needs a different first move

Now change the operation. The capital product symbol $\prod$ tells you to multiply, not add. In a product such as $\prod_{r=2}^{n}\log_r(r+1)$, both the base and the argument change from factor to factor. The same-base sum rule from the previous topic is not the appropriate first step.

Use change of base instead:

For integer indices starting at 2, every base is positive and different from 1, so every denominator $\ln r$ is nonzero. Multiplying the fractions creates a chain in which all the intermediate logarithms cancel.

The starting index and ending index determine what remains. More generally, for integers $2\le m\le n$,

There are $n-m+1$ factors. The first denominator is $\ln m$; the last numerator is $\ln(n+1)$. This looks like the preceding telescoping sum, but the objects being cancelled are logarithm values rather than consecutive integers.

Starting at 1 would break the original product: its first factor would use base 1. You cannot simplify an expression containing an undefined factor and then claim that the final formula supplies its value.

Before solving, say which operation connects the terms. A sum of same-base logs combines first; a product of linked changing-base logs uses change of base first. Recognizing that difference is the main skill. A long string of symbols becomes manageable once its structure is clear.

A factorial can be recognized without being calculated

A factorial such as $2001!$ represents the product $1\cdot2\cdot3\cdots2001$. Its numerical size is enormous, but a logarithm problem may only need its factor structure. Avoid calculating a huge integer when the definition itself is enough.

Let $P=2001!$, and consider terms of the form $1/\log_r P$, where the integer base $r$ runs from 2 to 2001. The original bases are all valid. Also $P>1$, so none of these logarithms is zero.

Reciprocity turns each term into $\log_P r$. Now every term has the same base, allowing a sum to combine into a logarithm of a product. The product of the arguments is $2\cdot3\cdots2001$, which is exactly $2001!$: the missing factor 1 would not change it.

This gives a useful general pattern. For an integer $n\ge2$, set $P=n!$. Then

There are $n-1$ terms. The original sum must start at 2. Although $\log_P1=0$ would be legal in the transformed form, the corresponding original expression $1/\log_1P$ is undefined. Rewriting does not authorize adding a forbidden index.

Try a smaller factorial first, such as $6!=720$, and list all its terms. Then omit the final term and predict the change. The full sum is 1 because the whole factorial is present; a partial product generally gives a different logarithm.