A Logical Zoo: Interesting Fallacious Mathematical

Arguments

Aubrey Weitzenkamp

In partial fulfillment of the requirements for the Master of Arts in Teaching with a Specialization
in the Teaching of Middle Level Mathematics in the Department of Mathematics.
Gordon Woodward, Advisor

July 2008

Weitzenkamp 2
Introduction
The following arguments and proofs are more advanced errors in thinking. These are a
collection of false arguments in mathematics. For each fallacy, I have attempted to show the
error and either give a correct argument or explain why the original statement could not be true.
When proving a statement mathematically, it is important to use clear thinking and careful
definition of terms.
Consider the following fallacious mathematical arguments from the University of Toronto
Mathematics Network Homepage. These proofs seem deceptively correct, yet there is an error in
the reasoning somewhere. Some mistakes are obvious, but others are harder to spot.
Fallacy #1: 1=2: A Proof using Beginning Algebra
1.
2.
3.
4.
5.
6.
7.
8.

Let a=b.
Then a2=ab
a2 + a2= a2+ab
2a2= a2+ab
2a2-2ab= a2+ab-2ab
So 2a2-2ab= a2-ab
This simplifies to 2(a2-ab)=1(a2-ab)
Cancel (a2-ab) from both sides, leaving 2=1, or 1=2.

Explanation:
I was first introduced to this fallacy when I was enrolled in the Modern Algebra course here at
the University of Nebraska. Looking at these fallacious proofs, it is intriguing to try and figure
out the error. The fallacy in this proof lies in the last step. By canceling the term (a2-ab), we are
really dividing. Since a=b was stated in the first step, this term (a2-ab) is equal to zero.
Division by zero is not possible because 0 is not invertible in the real numbers. Therefore
division by zero is undefined. In the next to last step we simplify to 2(0) = 1(0). The fallacy
here is that if a number is multiplied by zero, the product is zero. Yet, it does not imply that
these two numbers are equal. Any number multiplied by zero yields zero. (Spencer, 1998)

Weitzenkamp 3
Fallacy #2: 1=2: A Proof using Complex Numbers
1. -1/1 = 1/-1
11 = 1 1
2.
3.
1/ 1= 1/ 1
4. i/1 = 1/i
5. i/2 = 1/(2i)
6. i/2 + 3/(2i) = 1/(2i) + 3/(2i)
7. i(i/2 + 3/(2i)) = i(1/(2i) + 3/(2i))
8. (i2)/2 + (3i)/2i = i/(2i) + (3i)/2i
9. (-1)/2 + 3/2 = + 3/2
10. 1 = 2.
(Spencer, 1998)

Square root both sides.

Simplify.
Convert
1 to imaginary number i.
Multiply both sides by .
Add 3/(2i) to both sides.
Multiply both sides by i.
Simplify.
Convert i2 term to -1 and add fractions.

Explanation:
Complex numbers can be a challenge to work with. It is tempting to apply all of the same rules
from the context of real numbers to complex numbers. This example, however, shows how these
rules do not transfer from one to the other. The fallacy lies in the third step. In simplifying from
step two to three, we try to make two things equal that are not. We have sqrt(-1/1) = sqrt(1/-1)
in step two. The left-hand side (LHS) does simplify to sqrt(-1)/sqrt(1), which is the LHS in step
three. The right-hand side in step two sqrt(1/-1) does not simplify to sqrt(1)/sqrt(-1), which is
what we see as the RHS in step three. In fact, it is -sqrt(1)/sqrt(-1). A more careful examination
of negative numbers will help explain the fallacy clearly. The product of (-1)(-1) is one. So
sqrt(-1*-1) = sqrt(1) = 1. Yet sqrt(-1)sqrt(-1) = i*i = i2 = -1. So these two products are not the
same. Hence the rule of sqrt(ab) = sqrt(a)sqrt(b) holds true as a rule when a and b are both
positive numbers. For negative and complex numbers, however, this rule fails to hold true.
(Spencer, 1998)

Fallacy #3: A Proof that all people in Canada are the same age.
This proof uses induction to show that all people in Canada are the same age by showing the
following statement is true for all natural numbers n:

Weitzenkamp 4
S(n): In any group of n people, everyone in that group has the same age.
The conclusion follows from that statement by letting n be the number of people in Canada.
In any group that consists of just one person, everybody in that group has the same age, because
it is just one person.
Hence statement S(1) is true.
Now we must prove that whenever S(n) is true for one number (n=k), it is also true for the next
number (n=k+1). By doing this, we will be proving this statement true for all natural numbers
by induction.
So if we assume that, in every group of k people, everyone has the same age, then we can deduce
from it that in every group of k+1 people, everyone has the same age.
Let G be a random group of k+1 people; we just need to show that every member of this group
has the same age.
To do this, we just need to show that, if P and Q are any members of G, then they have the same
age.
Take into account everyone in G except P. These individuals form a group of k people, so they
must all have the same age. Remember that we are assuming in any group of k people, everyone
has the same age.
Now consider everyone in G except Q.
Let R be someone else in G other than P or Q.
Since Q and R are both elements in the group considered before, they are the same age.
Likewise, since P and R are both elements in the group considered before, they are the same age.
Since Q and R are the same age, and P and R are the same age, it follows that P and Q are the
same age.
We have now seen that, if we consider any two people P and Q in G, they have the same age. It
follows that everyone in G has the same age.
The proof is now complete: we have shown that the statement is true for n=1, and we have
shown that whenever it is true for n=k it is also true for n=k+1, so by induction it is true for all
n. (Spencer, 1998)
Explanation:
There might not be anyone else in G other than P or Q! Hence the fallacy lies in the step Let R
be someone else in G other than P or Q. Previously it was established that G was a random
group of k+1 people that we needed to show was comprised of people the same age. We are
letting S(1) imply S(2), which is not necessarily true. Therefore we fail to prove that P and Q are
the same age, and cannot relate them to R, so the proof fails.
Fallacy #4: All natural numbers can be described in fourteen words or less.
Suppose there is some natural number which cannot be unambiguously described in fourteen
words or less.
Then there must be a smallest such number n.

Weitzenkamp 5
But now n is the smallest number that cannot be unambiguously described in fourteen words or
less.
This is a complete and unambiguous description of n in fourteen words, contradicting the fact
that n was supposed to not have such a description.
Since the assumption from the first step of the existence of a natural number that cannot be
unambiguously described in fourteen words or less led to a contradiction, it must be an incorrect
assumption.
Therefore, all natural numbers can be unambiguously described in fourteen words or less!
(Spencer, 1998)
Explanation:
The fallacy is found in step four. By saying that there was such a number, it was actually
including itself in that set of numbers. Similarly, there is error is saying something like S is the
set of all sets because the set S would be included, or would it not? Therefore it is a
contradiction in saying that.
Fallacy #5: Chessboard Paradox

This fallacious argument is quite interesting. The area of the first figure appears to have one less
square unit of area than the second figure with 64 square units compared to 65 square units on
the right. The second figure is simply the first one with the defined figures rearranged. How
does this happen? Euclidean geometry tells us that the area should not decrease by
rearrangement of the figure. The answer lies in the diagonal of the rectangle.

Weitzenkamp 6
Looking at the rectangle, by rearranging the figure, the diagonal line will not be perfectly
straight. The difference is so slight that it is hard to notice. In fact the slope in one triangle is 3/8
= .375 and the other 2/5 = .4, a difference of only .025. When the figures are switched, then the
diagonal overlaps in the rectangle. Then the missing square unit is hidden in the diagonal
overlap. (Sillke, 2004)
Fallacy #6: Confusion in statistical arguments
When [the projection of past trends] used in conjunction with much knowledge about the
variables playing upon a series of past numbers, trend fitting and projecting can be most helpful;
the use of the tool is not, per se, a statistical fallacy. (Campbell, 1974)
For example, if the United States had a population of 90 million in 1910 and 180 million in 1960,
the arithmanticist would say that population doubles every 50 years and that we will have 360
million people in the year 2010, 720 million by 2060, and just under a billion and a half by 2110.
Arithmancy is a derogatory term suggesting that somehow the manipulation of past numbers has
imbued the resulting estimates of future numbers with a kind of magic. Another example could
be the finishing times of races. Since 100-yard dash times have decreased over time, we infer
that at some point a sprinter will run the 100-yard dash in literally no time at all, or even finish
the race before it begins. (Campbell, 1974)
How can we know if the statistical information we are reading is carefully prepared or hastily
thrown together? Campbell suggests five questions a person could ask to use as a screening
device for poor estimates. First, what kind of reputation does the source enjoy as a supplier of
statistical estimates or as an authority on the relevant subject? Also, does the supplier of the data
have an axe to grind? In other words, does the supplier have intentional or unintentional bias
on the material. A third question is what supportive evidence is offered? Does the underlying

Weitzenkamp 7
assumption, theory, or methodology seem okay? Finally, do the estimates seem plausible?
(Campbell, 1974)
Fallacy #7: Improper Comparisons with Statistics
Sometime you may have heard the phrase comparing apples to oranges. It is true that this
often happens in statistics. Campbell makes an interesting point for consumers with the
following story of statistical fallacy from Flaws and Fallacies of Statistical Thinking.
The advertising claims of National Bakers Services in behalf of Hollywood Bread were taken to
task by the Federal Trade Commission on the grounds that they made a misleading statistical
comparison. The claim made was that Hollywood Bread contains fewer calories per slice than
other brands of standard commercial bread. However, the FTC maintained that Hollywood has
as high a calorie content as any other breadabout 276 calories per 100 gramsand that the
only reason a slice of Hollywood Bread contains fewer calories is that it is more thinly sliced;
the comparison, therefore, was based on unequal units of bread. Or, expressing the same point in
somewhat different words: like was not being compared with like. (Campbell, 1974)
Fallacy #8: Speeders Delight
The following problem is from the book Mathematical Fallacies, Flaws, and Flimflam by
Edward J. Barbeau. It provides a delightful example of how a false assumption of a diagram
provides an incorrect solution to the problem.
The situation. Poised 50 feet to the side of the road, a speed trap awaits 120 feet beyond the crest
of a hill. In a flash, the officer pulls me over and accuses me of going 65 in a 60 mile-per-hour
zone. I say that there has been a math error: Even though the radar gun registered 65, I was
going only 60. As everyone knows, the Pythagorean theorem enables us to compute the third
side of a right triangle. So if the radar gun indicated 65 (along its line of sight), then by similar
triangles my actual speed (in the direction of the road) was 60 (Figure 1). Please refer to the
pictures below to illustrate this problem.
50 ft
By the Pythagorean theorem, this third side is 130 feet.
120 ft

Weitzenkamp 8

This situation is given by Carl E. Crockett of the United States Air Force Academy in Colorado.
He asks his students to prepare a response for the prosecution to use in court. There are three
possible responses:
1) The velocity of the car can be expressed as the result of two components, one in the
direction of the line of sight of the radar equipment and one in the direction perpendicular
to this. These two components, equal to the speed of the car multiplied by the cosine of a
suitable angle will both not exceed the speed in magnitude. Thus, we should have drawn
the following triangle, from which we find that the speed s of the car satisfies (12/13)s =
65 or s = 70.4.
25

60

65

2) While there is a speed triangle with the dimensions of the triangle above, it does not
correspond to the given situation.
3) Pythagoras theorem works for distances. In the following triangle, x and s (distances)
are both changing as time passes: x = x(t) and s = s(t). The rate of change x(t) is the
speed of the car and is denoted x(t). Similarly s(t) is the speed indicated by the radar.
The theorem tells us that x2 + y2 = s2, but it does not say x = kx, y = ky, or s = ks. The
latter was improperly implicitly assumed when the defendant used proportions from the
distance triangle to incorrectly determine the speed. (Barbeau, 2000)
y

Fallacy #9: Every Triangle is Isosceles

Given: Triangle ABC
Prove: AB=AC, Triangle ABC is isosceles.
1. For any triangle ABC, let the angle bisector from vertex A meet the perpendicular bisector of
BC at P.

Weitzenkamp 9
2. Drop perpendiculars from P to sides AB and AC, meeting these sides at points E and F.
3. Since EAP
FAP and AP = AP, the right triangles RAP and SAP are congruent by AAS.
Thus, AE = AF and EP = FP by CPCTC.
4. BP = CP because they are corresponding parts of the two congruent base triangles BPD and
CPD. Thus, triangles BPE and CPF are congruent and EB = FC.
5. AE + EB = AF + FC, or AB = AC and triangle ABC is isosceles. (Brown, K., 2007)

Explanation:
When a picture accompanies a fallacious proof, it most likely has an error somewhere. This is
the case for the example above. If we try to accurately construct the points and lines described in
this proof we will discover that the actual configuration doesn't look like the figure above. The
point P is not inside triangle ABC. Instead it falls outside the triangle ABC. However, if we
carry out the proof on this basis, and if we now assume the points E and F also fall outside the
triangle, we still conclude that the triangle is isosceles. This too is an incorrect configuration.
The actual configuration of points given by the stated construction is for the point P to be outside
the triangle ABC, and for exactly one of the points E,F to be between the vertices of the triangle,
as shown below.

Weitzenkamp 10

We still have AE=AF, PE=PF, and PB=PC, and it still follows that BE=FC, but now we see that
even though AE=AF and BE=FC it does not follow that AB=AC, because while F is between A
and C, E is not between A and B. (Brown, K. 2007)

Weitzenkamp 11

Works Cited
Barbeau, E.J. (2000). Mathematical fallacies, flaws, and flimflam. Washington, D.C.:
Mathematical Association of America.
Brown, K. (2007). Are all triangles isosceles?. Retrieved November 16, 2007, from MathPages
Web site: http://www.mathpages.com/home/kmath392.htm
Campbell, S.K (1974). Flaws and fallacies in statistical thinking. Englewood Cliffs, N.J.:
Prentice-Hall, Inc.
Maxwell, Ph.D., E.A. (1959). Fallacies in mathematics. Cambridge, England: University Press.
Sillke, T. (2004-10-13). Geometrical paradox. Retrieved November 14, 2007, from Jigsaw
Paradox Web site: http://www.mathematik.uni-bielefeld.de/~sillke/PUZZLES/jigsawparadox.html
Spencer, P. (1998, May 26). Classic Fallacies. Retrieved November 11, 2007, from University of
Toronto Mathematics Network Home Page Web site:
http://www.math.toronto.edu/mathnet/falseProofs/fallacies.html