Is Calculus Obsolete?

Author(s): Ed Dubinsky
Source: The Mathematics Teacher, Vol. 88, No. 2 (FEBRUARY 1995), pp. 146-148
Published by: National Council of Teachers of Mathematics
Stable URL: https://www.jstor.org/stable/27969236
Accessed: 13-10-2018 00:03 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

National Council of Teachers of Mathematics is collaborating with JSTOR to digitize,

preserve and extend access to The Mathematics Teacher

This content downloaded from 200.10.131.129 on Sat, 13 Oct 2018 00:03:43 UTC
All use subject to https://about.jstor.org/terms
 Is Calculus Obsolete?
So why do we need calculus if we have all this
technology? I mean, if you want me to find all the
relative maximum values of a function, 1 just
draw the graph, zoom in on the humps, and get
as close an approximation as you want.
with such a problem only by using a technological
tool to draw its graph. It is assumed that the tool
will draw the graph on some default domain, usual
ly [-10,10], but the user may specify any interval
for the domain. Consider two examples, fandg.

I
The definitions of the functions will remain a secret
f a teacher would answer this question by com
until later. All graphs have been redrawn for
plaining that the graphical approach gives only an readability.
approximate answer, a student might then ask why
Figures 1 and 2 illustrate the graphs on a
an "exact" answer like
default domain [-10,10]. The reader will notice

( /2+ 3)
'. ?? .' A ?:' '.'?'

is any better, more accurate, or more insightful " ?. } '

than 0.002634 or 0.002633823309767369.
Of course, the question has many answers, which
can be found by studying partial derivatives, solv
ing optimization problems for functions of several
variables, or applying calculus to such higher math
ematics as real analysis or differential geometry.
However, most students in our classes are, unfortu
nately, thinking about fewer answers, not more. We
-lis: i
would like our students to be interested in?and
appreciate?the power, beauty, and subtlety of a

If we set of ideas like finding the shape of a curve by

studying derivatives. We cannot get there by point
have all this ing to items that they are not likely to care about
technology, and that might include so-called real-world prob
why lems that many students seem to dread. We need to
look for more effective justification for studying
do we need powerful, but not immediately accessible, concepts.
calculus? One item in our present educational culture
about which many students care is getting the right
answer. So why not try to argue for the inadequacy
of studying a graph with only a computer or graph
ing calculator by suggesting to students that they
might not get the right answer?
Consider, given a specific function, the problem
of drawing its complete graph and showing, in par
ticular, all turning points, changes of concavity,
zeros, and asymptotes. This article illustrates what
might happen if a student was capable of dealing
In this month's column, Ed Dubinsky gives an
example of one danger of using graphing soft
ware without an accompanying mathematical
analysis.?Eds.
3:*:?
Edited by Albert A Cuoco
E. Paul Goldenberg
June Mark
Education Development Center (EDC)
55 Chapel Street
Newton, MA 02158-1060
Ed Dubinsky is a mathematician and mathematics edu
cator at Purdue University, West Lafayette, IN 47907,
who works in research and curriculum development in
postsecondary mathematics. He coedits Research in Colle
giate Mathematics Education and UME Trends.
This department is a forum where teachers can discuss
issues related to the use of technology in mathematics
class. Tips for using technology, classroom activities, and
ideas related to teaching and learning mathematics with
technology are all appropriate. Ideas that may not fill an
entire column are encouraged. We particularly welcome
tips involving calculators. Submissions may be sent to the
editors at the Education Development Center, 55 Chapel
Street, Newton, MA 02160, or to the Mathematics
Teacher. Electronic mail can be sent to techtips@edc.org.

146 THE MATHEMATICS TEACHER

This content downloaded from 200.10.131.129 on Sat, 13 Oct 2018 00:03:43 UTC
All use subject to https://about.jstor.org/terms
that this particular tool intentionally plots only a
discrete set of points and does not try to connect
them with a continuous line. The arguments in this
article are just as valid, maybe even more so, if the
tool being used connects the points somehow.
A fairly reasonable first guess is that these two
fimctions are the same but with different scales; they
are probably the same expression but with very dif
ferent coefficients. The x-axis is quite clearly a hori
zontal asymptote, but more needs to be done to be
sure what is happening near they-axis. If one zooms
in and looks at the two graphs in the domain [-1,1],
they still look very much the same, and it appears
as though the y-axis is a vertical asymptote.
At this point, a student might concentrate on f
One
and keep looking at smaller and smaller intervals might not
around 0. Later, the student might conclude that the
s-axis is a horizontal asymptote, they-axis is a ver
get the
tical asymptote, no turning points occur, and the con right
cavity is always down. These conclusions are actu answer
ally correct because the function fis indeed given by

0. Many teachers would be happy if a student

made such an investigation and gave this answer.

Progress is being made. After looking around, the

student might conclude that the graph resembles
figure 5. The description might then state again
that the x-axis is a horizontal asymptote. The y-axis
is also a vertical asymptote; however, / was on the
negative side and g was on the positive side. More
over, two more vertical asymptotes occur at = ? a
where a is some fairly small positive number.
Again, no turning points are possible, but the con

What about g? To save space in this article, as

sume that for some time, as the domain interval gets
smaller, the same graph results. If a student made
such an analysis and concluded that the graph of g
was pretty much the same as that of /but with dif
ferent scaling, the teacher would have to be satisfied;
same analysis, same answer. But a really persistent
student might go as far as the interval [-0.1,0.1]
and get a picture like the one in figure 3. This
image is not terribly informative. The problem is
that the graph has a point at around (0,1042), but
most of the points are much closer to the jc-axis and
so little is seen. It does show that g is not very much
like f It also suggests that something is occurring
around = 0 by the clustering of points, so perhaps
the student might try the domain [-0.02,0.02] as in
figure 4.

Vol. 88, No. 2 ? February 1995 147

This content downloaded from 200.10.131.129 on Sat, 13 Oct 2018 00:03:43 UTC
All use subject to https://about.jstor.org/terms
cavity does change from up to down at the = -a
asymptote and from down to up at = a.
This result is certainly a much richer and more
thoughtful use of the graphing tool. However, it has
two serious drawbacks.
then put them together. Only by using concepts
from calculus, possibly with the help of graphing
software, will a student be able to arrive at the
solution to the original problem, given in figure 7.

No graphing tool will create a graph such as the

one just described, which was drawn by hand. To
show the various features, very different scales
must be used. A scale that is appropriate to show
one feature, such as large positive values close to
= 0, will cause squeezing in other places, such
as near = 0.0004, which may conceal important
features. Also, one cannot show on the same
graph the behavior near 4 10"1 and near 4 IO"4.
The result is wrong.

Next we can reveal the definition of g. It is given

A little calculus will very quickly show that a criti

cal point arises at = 0.001, but care is needed
when making a graph. If an interval such as
[-0.0015,0.0015] is used, then the large values
As a final comment, the reader might think this
near zero will squeeze out all the action near 0.001.
problem was created to illustrate the author's
But if realizing from the expression that this graph
point. Not so. The coefficients were created, with
will be symmetric about the y-axis, one can look at
some help from Andy Gleason, but the basic expres
only one side and move away from 0.
sion for g describes a chemical reaction between
The reader might think that it was enough to
two molecules in a certain substance; it came from
notice the symmetry and use the graphing tool to
look at domains of the form [6, c] where 6 is a small
a chemistry book. ^
positive number. Try it; if not hitting on an interval
in which c is very close to 0.001, little will be seen.
The contribution from calculus is to suggest such Summer Math
The third annual Canada/USA camp for mathematically
values for c. Some experimentation will lead to a
choice of such a domain interval as [0.0009, talented high school students will be held in 1995 in
Canada: July 3 to July 28 in Vancouver (Univ. of British
0.0011], which gives the graph in figure 6.
Columbia) and July 31 to August 25 in Toronto (Univ.
Again the vastly different scales prevent graph of Toronto). The 4-week camp, known as Mathcamp, is
ing software from presenting a single picture of the a re-education of the mathematically precocious by
curve, and it is necessary to analyze pieces and university mathematicians who are also experts in
communicating concepts to these teenagers. Fee: $550
US ($690 CDN). Meals and supervised university
^MW!0WSBm__. dormitory room $30US/day extra. Application deadline
is May 15. For camp's brochure which also contains the
Application Form, call (519) 672-7990.

"?s?asicasvtgets"
THE PLATONIC SOLIDS
StedtModets of lfe Regular ofyhcdra
?tidtfi?ir contiitutt?ofis
Durable Classroom Manipulatives
?? SEND FOR FREE BROCHURE
M2 SOLIDS
t?mufacturersofSo?d
24351 Condon, Oak Park, MI 48237
_(810) 399-9571_

THE MATHEMATICS TEACHER

This content downloaded from 200.10.131.129 on Sat, 13 Oct 2018 00:03:43 UTC
All use subject to https://about.jstor.org/terms