42 Lehoux
retical understanding is not a measure of its intrinsic inadequacy—it was
very adequate for a very long time—so much as a function of historical
and cultural distance.
We know the universe to be fi nite, said Aristotle. But he offered dif-
ferent (if compatible) arguments to support this claim in different works.
In On the Heavens, he began with the observation that we see the heavens
revolving around us each day, and argued that if the heavens were infi nite
then they could not complete a revolution in any fi nite period of time.^9
In the Physics, he began with the fact that we know from observation that
all bodies have proper places to which they tend: heavy bodies fall down-
ward, and fi re and air tend upward.^10 This claim builds on the Aristotelian
arguments for a four- element theory, which sees all physical objects as
composed of some combination of four basic elements, called earth, air,
water, and fi re, which have two pairs of opposing qualities through which
they act (hot and cold, wet and dry).^11 A consequence of this theory is that
there must be a center to the cosmos as a whole, which turns out to be
the center of Earth, toward which heavy bodies can fall and away from
which light bodies can rise. If the cosmos were infi nite, the center could
not be defi ned, and so bodies would have nowhere or everywhere to fall
toward, which is contradicted by experience. That the fi nite cosmos must
be spherical emerged as a result of Aristotle’s analysis of motion and his ar-
gument that the universe must be both eternal and uncreated in time. He
divided all motion into three types: circular, rectilinear, and some mixture
of the two.^12 Any motion that is along a straight line cannot be continu-
ous, since in a fi nite universe all straight lines must be fi nite, and so the
only possible eternal motion would involve eventually getting to the end
of the line and then moving back again (thus coming to a brief moment of
standstill in between the two directions, and suffering periods of accelera-
tion and deceleration toward these rest points). But if eternal rectilinear
motion is thus discontinuous, then any motion composed even partly of
rectilinear motion (that is, the third type of motion, mixed motion) must
also be discontinuous, and so the only type of continuous and eternal mo-
tion possible is circular.^13 Circular motion also makes sense in this context,
Aristotle was quick to point out, insofar as it is both the simplest eternal
motion, and is “complete” in itself, with any one point on the circumfer-
ence serving simultaneously as a direction- toward- which motion is occur-
ring and a direction- away- from- which.^14 Furthermore, unlike an eternal
rectilinear motion back and forth along a fi nite line, circular motion does
not require an extra cause to account for the changes of direction. Once
it’s going, it just keeps going. That the cosmos as a whole rotates in place
is derived from a number of considerations, and Aristotle returned to the