Aristotle On Number Theory - Peter Atkinson
Aristotle On Number Theory - Peter Atkinson
Aristotle On Number Theory - Peter Atkinson
Can we make an assumption that there is a unified theory of number in Aristotle? Annas thinks so: All these passages
show a consistent attitude to the question of the existence of numbers. (Annas, 99)
1 | Peter Atkinson
One element of a things existence is its unity - its one-ness (which Aquinas in Questiones de
Veritate takes to be a transcendental). But Aristotle says that Unity is not itself a number:
We should generally say by any number except unity but a unit here is excluded from the meaning of number.
According to Aristotle, the unit is not a number: cf. Metaph. N. I. 1088a4-8: The term one means that it is a
measure of some multitude, and number that it is a measured multitude, or a multitude of measures. It is only
reasonable therefore that one is not a number, for neither is a measure measures, but the measure and the one
are the beginning, i.e. the unit if the beginning of number but not a number. (Heath p.84)
To count, we need to pick something out as our unit. This unity we pick is what numbers measure. One is
the measure of number. (Annas, 99) Counting is then the notation of unified beings.
One, unity, and unit all indicate the same element: a measure picked out by the mind to
measure a multitude. We focus on a particular unity and make that the measure by which we count: If we
have 10 sheep and 10 men, we have two species of the genus 10. Counting then, is the measuring of
ones and all numbers are derivative from counting.2
In sense (1) then, numbers are measures of the unities of a group of objects. Numbers are
derived from counting and are identical with measure and measuring: Aristotle is quite happy to
interchange number and counting with measure and measuring. (Annas 98) Since counting is - in
every instance - concerning with a unit, and every unit is a definite thing, then counting will always be of
specific things:3 When we have 10 dogs and 10 sheep we have different units and therefore different tens
(because they are tens of different kinds of thing) but we still have the same number. (Annas 107)
Therefore definite numbers have species, depending upon what you are counting.
Exempli gratia: 1000 is the measure of the body of students. 1000 is impossible to be had apart
from counting the number of students who go here. 1000 is not a thing itself, but a measure of the unit
we have chosen: student, which is itself recognizable by the fact that the substance man unifies each
student. Therefore our genus is 1000 and our species is students.
Is there any difference between numbers and counting? Yes, since numbers are accidents of things
and counting is an activity of the mind. Numbers, though derived from counting, can be considered apart
and their various aspects may be considered (such as equal and unequal). Their ontological status is
questionable, since Aristotle does not want to grant them independent status as numbers and since they are
2
Measure is that by which quantity is known and quantity qua quantity is known either by a one or by a number, and all
number is known by a one. 1052b20 (Ross translation)
A number is several ones or a certain quantity of them. Hence number must stop at the indivisible two and three are
derivative words, and so is every other number. (Phys. III. 7.207a33-b34, cf. Heath p110)
3
Numbers are derived from one, and one is - in every kind - a definite thing: In a sense, unity means the same as being.
[...] To be one is just to be a particular thing. (Metaphysics 1054a10-19)
2 | Peter Atkinson
The possible bisections of a magnitude are infinite in number this infinite is potential, not actual, but you can always
assume a number (of such bisections) exceeding any assigned number. But this number is not separable from the process of
bisection, and its infinity is not a stationary one but it is in process of coming to be, like time and the number of time. With
magnitudes the contrary is the case for the continuous magnitude is divisible ad infinitum, but in the direction of increase
there is no infinite. Whatever its size potentially, that size it can be actually hence, since there is no sensible magnitude that
is infinite, it is not possible to have an excess over every determinate magnitude if it were, there would have to be
something greater than the universe.
(Phys. III. 7.207a33-b34, cf. Heath p110)
5
Even the definition of an odd number sounds physical. In the Topics he speaks of a contemporary definition - which he
rejects - of an odd number as a number having a middle. Unclear in what way odd could mean having a middle
except as a physical object. (Heath p.91)
3 | Peter Atkinson
All numbers above one are derivative of one, and - it seems to me - have only res cogitans status (a mental
idea that we have in organizing the world).6 The meaning of statements about numbers is exhausted by
statements in which no such apparent reference [to number] occurs, which involve reference only to
counting and measuring. (Annas, 99) Number is not counting. Counting is the measuring (and therefore
creation/abstraction) of number based on unity in being. (Annas 103) It seems that Aristotle must - if he
were to systematically write on number - place numbers ontology in thought. But thought is not in one of
his categories so where does it belong?
Which leaves us with a question: where/how do we get our knowledge of the number sequence?
Numbers are derivative of multitudes of unities and dont have real existence outside of counting, but it is
very unclear how we get our idea of the number sequence. Aristotle would presumably call it by
abstraction, but how can the number sequence be abstracted from a group of physical objects? How do we
know that 2 is after 1, and 3 after 2? Where does this come from in Aristotle?
Say that number is a res cogitans. It remains that number is itself objective (and not arbitrary) in
some degree. Yes, we can have different base systems (binary vs. Hindu-Arabic ten-base-system) but there
is a consistent order to it. No culture counts randomly. This is what I take Aristotle to mean by a
certain order () if not the certain position towards other parts (). One comes after
another, but it doesnt lie anywhere, so that it cant have position. The ontological status of the number
sequence is unclear.
Aristotle thinks that number is discrete (i.e. parts without a common boundary, discontinuous,
definite). [For not one of the parts of number is a common boundary. (Cat 4b25)] Numbers, as discrete
quantities, have an order (), but not position () towards each other.
But, at any rate, concerning number, one cant observe its parts holding some position
towards each other or pick out where it lies. Nor can you pick out which of the parts join together
with others. (Cat 5a23-26)
Similarly with a number also, in that one is counted before two and two before three in this
way they may have a certain order, but you would certainly not find position. (Akrill)
A number line - insofar as it shows the - is accurate, but insofar as it seems to show that numbers
border each other is a misrepresentation. As a number sequence, it is accurate.
What then would Aristotle do with decimals?
I am unsure as to whether Aristotle would have had the concept of res cogitans. It is certainly anachronistic to read that
back into Aristotle (it is Latin, after all), but it seems that it is the correct category for where Aristotle would place it had he
had access to this concept.
4 | Peter Atkinson
Ultimately, Annas sees these two sense of number as a confusion: This distinction of two sense of
number seems to rest on a confusion of number with numbered group. (Annas 111) Whether or not she
is correct, Aristotles understanding of the relationship between number and numbered group is clear:
number (sense 2) is derived from the activity of counting a numbered group (sense 1). Heath writes:
Mathematica have no separate existence apart from things in nature, physical objects. (Heath 224)
Aristotles View:
If, he says, mathematical objects exist, we have to consider the alternatives: (a) they may exist in
sensible objects as some maintain, (b) they may be separate from existing things (a view which also has its
supporters), or (c) they must exist in some other way. [...] You may have arguments about moving objects
without reference to what the moving objects are or what attributes they possess, but treating them merely
as moving objects, it is not necessary to assume that there exist any movable entities separate from sensible
ones, or that these have in them some natural character separate from their matter. In like manner you may
have a science dealing with these same sensible moving objects not qua moving but qua bodies, and again,
qua planes only, qua lengths only, qua divisible [...] This science is mathematics. (Heath 225)
Perhaps an anachronistic reading, but Aristotle seems to think of mathematics as the science of
manipulating the res cogitans/mental realities of number, which are derived from - and ultimately have to
existence separate from - the counted objects. Heath makes an interesting note about the incarnate nature
of Aristotelian number theory:
But algebra in our sense was impossible for Greeks in Aristotles time, because no symbols has been invented,
and such problems as are equivalent to the algebraic solution of quadratic or cubic equations the Greeks could
only solve by geometry. (Heath 223-4)
5 | Peter Atkinson