. 3
( 60 .)


4 1 Categories
and adjoints. (Limits and adjoints are de¬ned later in this chapter.) By the use
of this technique, category theory has made mathematically precise the unity of a
variety of concepts in di¬erent branches of mathematics, such as the many prod-
uct constructions which occur all over mathematics (described in Section 1.7) or
the ubiquitous concept of isomorphism, discussed below. Besides explicating the
unity of concepts, categorical techniques for de¬ning concepts without mention-
ing elements have enabled mathematicians to provide a useful axiomatic basis for
algebraic topology, homological algebra and other theories.
Despite the possibility of giving element-free de¬nitions of these constructions,
it remains intuitively helpful to think of them as being de¬ned with elements.
Fortunately, this can be done: In Section 1.4, we introduce a more general notion
of element of an object in a category (more general even when the category is
Set) which in many circumstances makes categorical de¬nitions resemble familiar
de¬nitions involving elements of sets, and which also provides an explication of
the old notion of variable quantity.

Isomorphisms and terminal objects

The notion of isomorphism can be given an element-free de¬nition for any
category: An arrow f : A ’ B in a category is an isomorphism if it has an

inverse, namely an arrow g: B ’ A for which f —¦ g = idB and g —¦ f = idA . In

other words, both triangles of the following diagram must commute:

id id
© c
In a group regarded as a category, every arrow is invertible, whereas in a poset
regarded as a category, the only invertible arrows are the identity arrows (which
are invertible in any category).
It is easy to check that an isomorphism in Grp is what is usually called an
isomorphism (commonly de¬ned as a bijective homomorphism, but some newer
texts give the de¬nition above). An isomorphism in Set is a bijective function,
and an isomorphism in T is a homeomorphism.
Singleton sets in Set can be characterized without mentioning elements, too.
A terminal object in a category C is an object T with the property that for
every object A of C there is exactly one arrow from A to T . It is easy to see
1.1 De¬nition of category 5
that terminal objects in Set, T and Grp are all one element sets with the only
possible structure in the case of the last two categories.


If C is a category, then we de¬ne C op to be the category with the same objects
and arrows as C , but an arrow f : A ’ B in C is regarded as an arrow from B

to A in C . In other words, for all objects A and B of C ,

HomC (A, B) = HomC op (B, A)

If f : A ’ B and g: B ’ C in C , then the composite f —¦ g in C op is by de¬nition
’ ’
the composite g —¦ f in C . The category C op is called the opposite category of
If P is a property that objects or arrows in a category may have, then the
dual of P is the property of having P in the opposite category. As an example,
consider the property of being a terminal object. If an object A of a category C
is a terminal object in C op , then HomC (B, A) has exactly one arrow for every
object B of C . Thus the dual property of being a terminal object is the property:
Hom(A, B) has exactly one arrow for each object B. An object A with this
property is called an initial object. In Set and T the empty set is the initial
object (see “Fine points” on page 7). In Grp, on the other hand, the one-element
group is both an initial and a terminal object.
Clearly if property P is dual to property Q then property Q is dual to property
P. Thus being an initial object and being a terminal object are dual properties.
Observe that being an isomorphism is a self-dual property.
Constructions may also have duals. For example, the dual to the category of
objects over A is the category of objects under A. An object is an arrow from
A and an arrow from the object f : A ’ B to the object g: A ’ C is an arrow
’ ’
h from B to C for which h —¦ f = g.
Often a property and its dual each have their own names; when they don™t
(and sometimes when they do) the dual property is named by pre¬xing “co-”.
For example, one could, and some sources do, call an initial object “coterminal”,
or a terminal object “coinitial”.

De¬nition of category by commutative diagrams

The notion of category itself can be de¬ned in an element-free way. We
describe the idea behind this alternative de¬nition here, but some of the sets we
6 1 Categories
construct are de¬ned in terms of elements. In Section 1.6, we show how to de¬ne
these sets without mentioning elements (by pullback diagrams).
Before giving the de¬nition, we mention several notational conventions that
will recur throughout the book.

1. If X and Y are sets, p1 : X —Y ’ X and p2 : X —Y ’ Y are the coordinate
’ ’
2. If X, Y and Z are sets and f : X ’ Y , g: X ’ Z are functions,
’ ’

(f, g): X ’ Y — Z

is the function whose value at a ∈ X is (f (a), g(a)).
3. If X, Y , Z, and W are sets and f : X ’ Z, g: Y ’ W are functions, then
’ ’

f — g: X — Y ’ Z — W

is the function whose value at (a, b) is (f (a), g(b)). This notation is also
used for maps de¬ned on subsets of product sets (as in 4 below).

A category consists of two sets A and O and four functions d0 , d1 : A ’ O,

u: O ’ A and m: P ’ A, where P is the set
’ ’

{(f, g) | d0 (f ) = d1 (g)}

of composable pairs of arrows for which the following Diagrams 1“4 commute.
For example, the left diagram of 2 below says that d0 —¦ p1 = d0 —¦ m. We will treat
diagrams more formally in Section 1.7.
uE 'u
d0d  d1
‚  ©
This says that the source and target of idX is X.
p2 E p1 E

m m
d0 d1 (2)
c c c c
d0 d1
1.1 De¬nition of category 7
This says that the source of f —¦ g is that of g and its target is that of f .

(u —¦ d0 , 1) E ' (1, u —¦ d1 )
d   (3)
idA d   idA
‚  ©
This characterizes the left and right identity laws.
In the next diagram, Q is the set of composable triples of arrows:

Q = {(f, g, h) | d1 (h) = d0 (g) and d1 (g) = d0 (f )}


c c
P m
This is associativity of composition.
It is straightforward to check that this de¬nition is equivalent to the ¬rst one.
The diagrams just given actually describe geometric objects, namely the clas-
sifying space of the category. Indeed, the functions between O, A, P and Q
generated by u, d0 , d1 , m and the coordinate maps form a simplicial set trun-
cated in dimension three. The reader needs no knowledge of simplicial sets for
this text.

Fine points

Note that a category may be empty, that is have no objects and (of course)
no arrows. Observe that a subcategory of a monoid regarded as a category may
be empty; if it is not empty, then it is a submonoid. This should cause no more
di¬culty than the fact that a submonoid of a group may not be a subgroup. The
basic reason is that a monoid must have exactly one object, while a subcategory
need not have any.
It is important to observe that in categories such as Set, Grp and T in which
the arrows are actually functions, the de¬nition of category requires that the
function have a uniquely speci¬ed domain and codomain, so that for example
8 1 Categories
in T the continuous function from the set R of real numbers to the set R+ of
nonnegative real numbers which takes a number to its square is di¬erent from the
function from R to R which does the same thing, and both of these are di¬erent
from the squaring function from R+ to R+ .
A de¬nition of “function” in Set which ¬ts this requirement is this: A func-
tion is an ordered triple (A, G, B) where A and B are sets and G is a subset of
A — B with the property that for each x ∈ A there is exactly one y ∈ B such that
(x, y) ∈ G. This is equivalent to saying that the composite

G ‚’ A — B ’ A

is an isomorphism (the second function is projection on the ¬rst coordinate).
Then the domain of the function is the set A and the codomain is B. As a
consequence of this de¬nition, A is empty if and only if G is empty, but B may
or may not be empty. Thus there is exactly one function, namely (…, …, B), from
the empty set to each set B, so that the empty set is the initial object in Set,
as claimed previously. (Note also that if (A, G, B) is a function then G uniquely
determines A but not B. This asymmetry is reversed in the next paragraph.)
An equivalent de¬nition of function is a triple (A, G— , B) where G— is the
quotient of the disjoint union A + B by an equivalence relation for which each
element of B is contained in exactly one equivalence class. In other words, the
B ’ A + B ’’ G—
’ ’
is an isomorphism, where the ¬rst arrow is the inclusion into the sum and the
second is the quotient mapping. This notion actually corresponds to the intu-
itive picture of function frequently drawn for elementary calculus students which
illustrates the squaring function from {’2, ’1, 0, 1, 2} to {0, 1, 2, 3, 4} this way:

’2 4

’1 1

0 0

The set G is called the graph and G— the cograph of the function. We will see
in Section 1.8 that the graph and cograph are dual to each other.
1.1 De¬nition of category 9
Exercises 1.1

1. Show that the following de¬nition of category is equivalent to the de¬nition
given in this section. In this de¬nition, to say that an element e has the identity
property means that for all f and g, e —¦ f = f whenever e —¦ f is de¬ned and
g —¦ e = g whenever g —¦ e is de¬ned.
This is the alternative de¬nition: A category is a set with a partially de¬ned
binary operation denoted “—¦” with the following properties:
(a) the following statements are equivalent:
(i) f —¦ g and g —¦ h are both de¬ned;
(ii) f —¦ (g —¦ h) is de¬ned;
(iii) (f —¦ g) —¦ h is de¬ned;
(b) if (f —¦ g) —¦ h is de¬ned, then (f —¦ g) —¦ h = f —¦ (g —¦ h);
(c) for any f , there are elements e and e with the identity property for which
e —¦ f and f —¦ e are de¬ned.

2. Verify that the following constructions produce categories.
(a) For any category C , the arrow category Ar(C ) of arrows of C has as
objects the arrows of C , and an arrow from f : A ’ B to g: A ’ B is a pair
’ ’
of arrows h: A ’ A and k: B ’ B making the following diagram commute:
’ ’



. 3
( 60 .)