Andrei Rodin (Ecole Normale Superieure)

On Categorical Theory-Building: Beyond the Formal

1. Introduction: Languages, Foundations and Reification of Concepts

The term "language" is colloquially used in mathematics to refer to a theory, which
grasps common features of a large range (or even all) of other mathematical theories
and so can serve as a unifying conceptual framework for these theories. "Set-
theoretic language" is a case in point. The systematic work of translation of the whole
of mathematics into the set-theoretic language has been endeavoured in 20-th
century by a group of mathematicians under the collective name of Nicolas Bourbaki.
Bourbakist mathematics proved successful both in research and higher education
(albeit not in the school math education) and is practiced until today at mathematical
departments worldwide.
The project of Bourbaki as well as other attempts to do mathematics "set-
theoretically" should be definitely distinguished from the project of reduction of
mathematics to set theory defended by Quine and some other philosophers. This latter
project is based on the claim that all true mathematical propositions are deducible
from axioms of Zermelo-Frenkael set theory with Choice (ZFC) or another appropriate
system of axioms for sets, so basically all the mathematics is set theory. A working
mathematician usually sees this claim as an example of philosophical absurdity on a
par with Zeno's claim that there is no motion, and Bourbaki never tried to put it
forward. Actually Bourbaki used set theory (more precisely their own version of
axiomatic theory of sets) for doing mathematics in very much the same way in which
classical first-order logic is used for doing axiomatic theories like ZFC. In other words
Bourbaki made set theory a "part of their logic", and then developed specific
mathematical theories stipulating new axioms expressed in this extended language.
According to the philosophical view just mentioned all Bourbaki's proofs are
nevertheless translatable into deductions from axioms of set theory. I shall not go for
pros and contras about this controversial claim here but put instead these two
questions: (1)Which features of sets make set theory a reasonable candidate for
foundations of mathematics? and (2)Which features of sets allow set theory to be an
effective mathematical language? Although the two questions are mutually related
they are not the same and require different answers.
In order to answer the first question remind Hilbert's Grundlagen der Geometrie of
1899 which provides the notion of foundation relevant to the question. Hilbert
suggests to think of geometrical points and straight lines as of abstract "things" (of
two different types) holding certain relations with required formal properties; one is
left free to imagine then these things in any way one likes or not imagine them in any
particular way at all. However abstract and unspecific might be the notion of thing
involved here one cannot avoid making certain assumptions about it. In order to clarify
these assumptions one needs an appropriate "theory of things". Set theory proves
appropriate for this purpose: sets provide the standard (Tarski) semantic for classical
first-order logic and for theories axiomatised with this logic. Hence the idea to use set
theory on a par with formal axiomatic method and the claim that mathematics is
ultimately "about" sets. Building of axiomatic set theories becomes then a rather
tricky business since any such theory involves an infinite regress: in order to build an
axiomatic theory of sets one needs to assume some (usually different) notion of set
in advance for semantic purposes. This and other relevant problems about set theory
and logic have been scrutinised by mathematicians and philosophers throughout 20th
century. I shall not explore this vast issue here but I want to stress the intimate link
between sets and formal axiomatic method just explained. True, this method allows
for building not only theories of sets but also theories of lines and points (plane
geometries), of parts and wholes (mereologies) and of whatnot. However some notion
of set (or class) is anyway required by all such theories. Stronger technical notions of
set corresponding to ZFC and other axiomatic set theories come about when this
general requirement is further strengthened for specific mathematical needs. So when
the notion of foundations is understood in the sense of Hilbert's Grundlagen or
similarly the choice of set theory (rather than mereology or anything else) as
foundations of mathematics is natural.
Second question requires, in my view, a very different answer. There is more than one
reason why set theory is helpful for mathematics but perhaps the following one is the
most important. Mathematics doesn't deal with "pure" concepts - whatever this might
mean - but deals with concepts embodied into mathematical objects. There are
different ways of such embodiment or reification of concepts. The most traditional
one is exemplification. When Euclid wants to prove that all triangles have certain
property P he always proceeds in the following way. He takes "for example" just one
triangle ABC , proves that ABC has property P and then concludes that all triangles
have this property. The conclusion is justified when the proof doesn't rely on any
specific property S of ABC, which some other triangle A'B'C' might not have. This
guarantees that the same proof can be applied to A'B'C' or any other triangle. Such
"reification through instantiation" is also widely used in today's mathematics as well. It
seems to be fairly fundamental for mathematics as we know it through its history.
However set theory provides another way of reifying concept. In the pre-set-theoretic
mathematics people had among available objects particular triangles like ABC and
particular (natural) numbers like 5 but they had no special mathematical object
corresponding to the general concept of triangle or to the general concept of number
(over and above the aforementioned particulars). Entities of this latter kind could be
believed to exist somewhere on Platonic heavens but certainly not among
mathematical objects. However set theory allowed for making such things up and
treating them mathematically. Consider set T of all triangles (on a given plane) and set
N of natural numbers. These sets are extensions of their corresponding concepts.
They are genuine mathematical objects having certain properties (e.g. cardinality) and
allowing for certain operations with them. In particular, N can be squared (the square
of N is the set of all ordered number pairs) and factorised by some equivalence
relation. Importantly T is not a triangle and N is not a natural number: the extension of
a given concept is not an instance of this concept. But like an instance
the extension is an embodiment of a given concept: it is an object one can play with,
i.e. make further constructions. Obviously this second way of reification of concepts
was not available for mathematicians until the notion of infinite set was approved by
G. Cantor and his followers.
One may remind Occam's Razor and wonder why having more objects is an advantage.
The answer is this: it is an advantage as far as it leads to new non-trivial
mathematics. This is indeed the case as far as set-theoretic mathematics is
concerned. Mathematicians can be interested in a conceptual parsimony but hardly in
an ontological parsimony. I cannot see any profit mathematics might get by preventing
certain concepts from reification. In particular, since mathematical reasoning involves
the notion of infinity anyway it is quite appropriate for this science to develop a
calculus of infinities rather than keep the notion of infinity somewhere at the limit of
the scope of the discipline without a properly mathematical treatment. The history of
mathematics of last two centuries provides numerous examples of successful
reification of "ideal elements" of different kinds. Think of ideal points in projective
geometry, for example.

2. Language of Categories and Categorical Foundations

Unlike set theory category theory has been designed as a language to begin with and
only later has been proposed as a foundations by Lawvere and his followers. The notion
of category has been first explicitly introduced by Eilenberg and MacLane in their
(1945) paper as a purely auxiliary device, and until works of Grothendieck and his
school in late fifties, which made an essential use of category theory in algebraic
geometry, nobody would consider this theory as anything more than a convenient
system of notation. In his classical (1971) MacLane writes:

Category theory starts with the observation that many properties of mathematical
systems can be unified and simplified by a presentation with diagrams of arrows.
(MacLane 1971, p.1)

Such presentation is often possible because most of mathematical concepts come

with a corresponding notion of map (otherwise called transformation or morphism)
between tokens falling under a given concept. For example, maps between sets are
functions, maps between topological spaces are continuous transformations, maps
between groups are group homomorphisms. Such maps are composable in the usual
way corresponding to the common intuition behind the notion of transformation. The
mathematical notion of category makes this common intuition explicit assuming
associativity of composition of maps and existence of identity map for each object.
Using the "language of arrows" we may think of, say, natural numbers, not just as of
"bare set" N = {0, 1, 2, 3, ...} but as a category where numbers are provided with
succession maps:

0 --> 1 --> 2 --> 3 --> ...

Identifying number n with isomorphism classes of sets having exactly n elements and
considering classes of maps between these sets as morphisms between numbers one
gets a richer category comprising mexpn different morphisms from any number n to
any number m , in particular nexpn morphisms to each number n to itself n! of which
are isomorphisms. All these notions can be rather easily cooked set-theoretically; in
more involved constructions such a set-theoretic regression is also often (but not
always) possible but the advantage of using the language of arrows can be much more
important. I shall not talk about specific mathematical matters here but remark that
category theory like set theory allowed for reification of certain concepts which
earlier could not be reified. Consider the concept of set for example. The extension S
of this concept is the set of all sets. That S turns to be a contradictory notion is a
part of the problem but not yet the whole problem (notice that the notion of infinite
set before Cantor was believed to be contradicory too). Another part of the problem
is that S doesn't have interesting properties to be studied and apparently doesn't
allow for further non-trivial constructions. However as far as all sets are taken
together with all maps between them the situation changes. The category of sets Set
does have interesting specific properties distinguishing it from other categories and
also allows for non-trivial constructions (like that of topos). This clearly shows that
Set is a better embodiment of the general concept of set than S. The situation is
similar in the case of the concept of group, topological space and many others (see
paragraph 6 below). So category theory allows for reification of concepts when set
theory doesn't.
However useful category theory might be what has been said so far has no bearing on
the problem of foundations. One may assume standard set-theoretic foundations and
then construe the language of categories upon it. But why not to think (in particular to
think about sets) category-theoretically to start with? The first systematic attempts
of this kind has been made by Lawvere in his thesis of 1963 and papers of 1964 and
1966 based on this thesis. In these works Lawvere introduced categories using formal
axiomatic method just like Zermelo and Fraenkel did this with sets. This amounts to
the following: objects and morphisms are taken as primitives objects holding three
primitive relations with intended meaning "domain of", "codomain of", and "composition
of" plus the identity relation. Categorical objects and categorical morphisms are
treated as belonging to the same type since every categorical object is formally
identified with its identity morphism. Lawvere himself avoids speaking about objects
and relations in this context taking first purely syntactical viewpoint and after listing
the appropriate axioms saying :

By a category we of course understand (intuitively) any structure which is an

interpretation of the elementary theory of abstract categories ... (1966, p.4)

In his (1964) Lawvere adjoins to his axiomatic category theory a number of additional
axioms making an abstract category "into" the category of sets so that

There is essentially only one category which satisfies these ... axioms ... , namely the
category S of sets and mappings. (1964, p. 1506)

Remind that in order to use set theory as foundations one needs "abstract" sets
rather than "concrete" sets like sets of points, numbers, etc. What these abstract
sets are sets of? Cantor's answer is the following: abstract sets are sets of "pure
units" ("lauter Einsen"). Another answer has been later given by Zermelo: abstract
sets are sets of sets. This latter answer is obviously more economical conceptually.
For a similar reason Lawvere put forward a theory of categories of categories (but
not just a general theory of "concrete" categories like categories of sets, groups,
etc...) and suggested it as foundations of mathematics (Lawvere 1966).
While in these early papers Lawvere sticks to formal axiomatic method and the
corresponding notion of foundations in his more recent paper of 2003 this author
takes a different approach and opts for a different notion of foundations understood

... in a common-sense way rather than in the speculative way of the Bolzano-Frege-
Peano-Russell tradition. (2003 p. ??)

This change of Lawvere's view seems me remarkable. The intimate link between sets
and formal axiomatic method stressed in the previous paragraph suggests that sets
cannot be replaced in their foundational role by categories or anything else unless one
continue to use this method and applies the corresponding notion of foundation. This,
in my view, explains why the idea of making categories into foundations finally led
Lawvere to the refusal from the formal method. But Lawvere's reference to common
sense hardly solve the problem either. To get rid of "speculative foundations" one
needs a new method of theory-building. In his (2003) Lawvere doesn't aim at general
solution of this problem but gives a concrete example of how categorical foundations
may look like. The principle aim of my paper is to describe a general method of
theory-building suggested by category theory. I shall call this new method categorical
and distinguish it from formal method. Formal views on mathematics and science are
usually opposed to more traditional views according to which mathematics and
sciences always assume certain "substances" like "number" or "magnitude" as their
subject-matters. In today's philosophy of mathematics formal view is associated with
mathematical structuralism. I say that categorical method goes beyond formal method
(and beyond structuralism) in order to stress that my proposal has nothing to do with
the traditional substantialism. As we shall see the mathematical notion of category
suggests something genuinely new with respect to the traditional concepts of form
and substance.
The rest of the paper is organised as follows. In the next section I stress a distinctive
feature of formal axiomatic method (seen against traditional axiomatic method), which
concerns the notion of interpretation relevant to mathematics. Then I argue that
formal axiomatic method doesn't provide an adequate treatment of mathematical
interpretation and introduce the notion of categorical method which does this. Then I
analyse some logical aspects of categorical method and conclude with general
epistemological arguments in its favour.

3. Formal Axiomatic Method in The Nutshell

A today's mathematical student can read in various textbooks that formal axiomatic
method invented by Hilbert is nothing but a perfectioned version of the traditional
axiomatic method known since Euclid. True, Hilbert certainly had Euclid's Elements in
mind writing his Grundlagen, so his method can be rightly seen as a modification of
Euclid's. However I don't think that the description of this modification as
perfectioning sheds a lot of light on it. To see clearly what is specific for formal
axiomatic method as distinguished from more traditional versions of axiomatic method
a historical regression seems me helpful. Soon after the publication of Hilbert's
Grundlagen in 1899 Frege sent Hilbert a letter (the precise date is missing) containing
a severe criticism of Hilbert's approach. Frege's had the following traditional
understanding of axiomatic method in mind. A given theory starts with axioms, which
are truths taken for granted. These non-demonstrable truths are truths about certain
objects. The theory proceeds with inferences from the axioms made according rules
of inference, which must be also assumed. As a matter of course for any given theory
meanings of all terms used in its axioms and further inferences must be unequivocally
fixed once and for all. This general epistemological view dating back to Aristotle has
been recently called classical model of science (Jong&Betti, forthcoming).
Frege pointed to Hilbert that his Grundlagen falls short of meeting the requirements
just mentioned and in particular the unequivocality requirement. Here is a quotation
from Hilbert's reply to Frege where Hilbert explains his new method in the nutshell:

"You say that my concepts, e.g. "point", "between", are not unequivocally fixed <...>.
But surely it is self-evident that every theory is merely a framework or schema of
concepts together with their necessary relations to one another, and that basic
elements can be construed as one pleases. If I think of my points as some system or
other of things, e.g. the system of love, of law, or of chimney sweeps <...> and then
conceive of all my axioms as relations between these things, then my theorems, e.g.
the Pythagorean one, will hold of these things as well. In other words, each and every
theory can always be applied to infinitely many systems of basic elements. For one
merely has to apply a univocal and reversible one-to-one transformation and stipulate
that the axioms for the transformed things be correspondingly similar. Indeed this is
frequently applied, for example in the principle of duality, etc." (cit. by Frege 1971,
p.13, italic mine).

Since a point is allowed to "be" (or to "be thought of as") a "system of love and
chimney sweeps" (or a beer mug according to another popular Hilbert's saying) - and all
this within one and the same theory - Frege's notion of axiomatic method is certainly
no longer relevant. But let's look for a serious mathematical reason behind Hilbert's
colourful rhetoric. In the end of the quotation Hilbert refers to the duality principle in
projective geometry. Given a true proposition of this theory, which involves straight
lines and points, one may formally exchange terms "line" and "point" and get another
true proposition. (This doesn't reduce to the trivial remark that one may call lines
"points" and call points "lines" without changing the given theoretical structure since
the original proposition remains true as it stands.) This suggests the following idea: a
given mathematical object can be occasionally "thought of" or "interpreted" as another
mathematical object. In particular in projective geometry one may "think of lines as
points and think of points as lines". Such a liberal treatment of mathematical objects
is common in today's mathematics. In the end of 19th century it was not yet common
but a number of important examples were already around (I elaborate on one such
example in the next section). Hilbert's Grundlagen provides a justification for this
apparently careless conceptual game. The problem Hilbert addressed can be
formulated as follows: How to construe a mathematical concept, which can be
occasionally "interpreted" as another concept?; How to formulate a theory in which
basic concepts are defined only "up to interpretation"? Hilbert's answer is roughly this.
One should first conceive of mathematical objects as bare "things" (possibly of
different types) standing in certain relations to each other, and then describe these
relations stipulating their formal properties as axioms. Any "system of things", which
hold relations satisfying the axioms would be a model of the theory.
As Hilbert makes it explicit in the quoted passage he thinks about re-interpretations
of theories as "reversible one-to-one transformations", i.e. as isomorphisms. Hilbert's
clearness often absent in later expositions of formal method allows one to see its
limits: mutual interpretations of mathematical theories are, generally, not reversible.
To build a theory "up to isomorphism" is not the same thing as to build a theory "up to
interpretation". For interpretations are, generally, not isomorphisms. Let me now
demonstrate this fact using a historical example, which was already available to
Hilbert.

4. Irreversible Interpretations
Non-Euclidean geometries emerged in mathematics of 19th century as a result of at
least two different developments. The first started in Ancient times and culminated
with (Bolyai 1832) and (Lobachevsky 1837). These mathematicians like their
predecessors tried to prove the 5-th Postulate of Euclid's Elements (the "Axiom of
Parallels") by getting a contradiction from its negations but at certain point changed
their attitude and came to a conviction that they were exploring a new vast territory
rather than approaching the desired dead end. The second line of development, which I
associate with the names of Gauss and Riemann, was relatively recent. Gauss had a
genuinely new insight on the old "problem of parallels" guessing a link with geometry of
curve surfaces. This allowed his pupil Riemann to build a new generalised concept of
geometrical space (Riemanean manifold), which still serves us as the best
mathematical description of the physical space-time (in General Relativity).
The two lines of development were brought together by Beltrami in his two prominent
papers (1868) and (1868-69). Anachronistically speaking, in his Saggio of 1868
Beltrami gave a 3D Euclidean model of plane Lobachevskian geometry. More precisely it
was only a partical model where finite segments of geodesics of a surface named by
Beltrami pseudo-sphere represented straight line segments of Lobachevskian plane.
But Beltrami didn't have the notions of formal theory and model in mind. He first
thought he discovered what the Lobachevskian plane was indeed: he believed it was the
pseudo-sphere. However this conclusion was not quite satisfactory even in Beltrami's
own eyes. He didn't notice that his model for plane Lobachevskian gometry was only
partial (this was first noticed by Helmholz in 1870, see Kline 1972) but he saw that
Lobachevskian 3D geometry couldn't be treated in the same way. So he looked for a
better solution. He found it after reading Riemann's (1854) and presented in his Teoria
of 1869: Lobachevskian space is a Riemanean manifold of constant negative
curvature. This holds for spaces of any number of dimensions.
The latter solution apparently makes the talk of interpretation no longer necessary.
Let's however see how the result of Saggio looks from the point of view of Teoria: a 2-
manifold of constant negative curvature is partially embeddable into 3D Euclidean
space (which is another Riemanean manifold). So we have here two manifolds and a
map, which can still be tought of as interpretation as suggested by Saggio. The point I
want to stress is that this map is not an isomorphism, it is not reversible. It restricts
to an isomorphism (a part of Lobachevskian plane is isomorphic to a part of Euclidean
space) but the whole construction cannot be conceived on this restricted basis alone:
the map in question is a map between two spaces (manifolds) but not between their
"parts". As a surface in Euclidean space the pseudo-sphere cannot be "carved out" of
this space. One may remark that we are talking about a map between spaces
(manifolds) but not about an interpretation between theories, and so this example is
not quite relevant to the issue discussed in the previous section. But it is obvious that
however the notion of theory is construed in this case the situation remains
asymmetric: while Lobachevskian plane geometry can (modulo needed reservations) be
explained in or "translated into" terms of Euclidean 3D geometry the converse is not
the case. Observe that the mere existence of interpretation f :A-->B of theory A in
terms of another theory B and a backward interpretation g:B-->A is not sufficient for
considering f as reversible: f and g should "cancel" each other for it. To give a
precise definition one needs to stipulate appropriate "identical interpretations" idA , idB
(which leave A,B correspondingly "as they are") and then require fg=id A and gf=id B ,
where fg and gf denote composition interpretations (written here in the direct
geometrical order). This standard categorical definition of isomorphism often allows
for but doesn't require thinking of it in terms of set-theoretic "correspondences"
between elements. Hereafter talking about isomorphisms I shall understand this notion
in the sense of the above definition.
A similar point can be made about arithmetical models of plane Euclidean and other
geometries used by Hilbert in Grundlagen. Perhaps one can indeed imagine geometrical
points as usual dots, "systems of loves" or beer mugs indiscriminately. But
representation of points by pairs of real numbers (or pairs of elements of another
appropriate algebraic field) is a different matter. Unlike dots and beer mugs numbers
are mathematical objects on their own rights belonging to a different mathematical
theory, namely arithmetic. "Translations" of geometrical theories into arithmetic used
by Hilbert are obviously non-reversible: they allow for translation of geometrical
theories into arithmetic terms but not the other way round. Hilbert certainly saw this.
He didn't mean to say that geometry and arithmetic seen from a higher viewpoint turn
to be the same theory; actually he considered a possibility of reduction of geometry to
arithmetic. Nevertheless he thought about this translation as an isomorphism, namely
an isomorphism between basic geometrical objects and relations, one the one hand,
and specially prepared arithmetical constructions, on the other hand. But such
constructions obviously cannot be made outside an appropriate arithmetical theory.
As far as this "target" arithmetical theory is wholly taken into consideration (as it
must be!) the translation in question doesn't look like like an isomorphism any longer.
We see that the example of projective duality mentioned by Hilbert (to leave alone
"systems of love, of law or of chimney sweeps") is special and cannot be used as a
model for treating the notion of interpretation in mathematics in the general case.
One may perhaps remark that the point I am making is obvious but trivial or at best
merely technical: the categorical notion of morphism allows indeed for a better
treatment of interpretations between theories but doesn't essentially change
anything. Let me now show that in fact it does. What might look like a minor technical
amendment suggests a revision of the whole idea of "formal" mathematics and formal
theory-building in general. (Endnote 1)

5. Forms, Categories and Structuralism

Many traditional mathematical concepts have the following property: all items falling
under a given concept are isomorphic or, in other words, are defined "up to
isomorphism". I shall call such concepts form-concepts or simply forms. The colloquial
expression "up to isomorphism" apparently involves a systematic ambiguity between
identity and isomorphism but it doesn't matter here. Think about any Euclidean
geometrical form (shape) like that of circle. A circle allows for (Euclidean) motions
and scalings. I leave aside the tricky question of whether these transformations
preserve identities of circles or rather associate with given circles some other circles
(see Rodin, forthcoming). I only point here to the fact that all (Euclidean) circles are
isomorphic in the sense that for any given pair of circles there always exist a
reversible transformation (motion, scaling or their composition) transforming one
circle into the other. Through an appropriate modification of the class of admissible
transformations one may modify a given form-concept. Thinking about a circle up to
reversible continuous transformation one gets a more general topological notion of
circle. Klein (1872) first put forward the idea of description of geometrical spaces
through group-theoretic properties of transformations available in these spaces. This
approach requires all transformations in question to be reversible (to be
isomorphisms) for otherwise they don't form a group by composition.
For a further example consider the traditional notion of (natural) number conceived as
a "shared form" of isomorphic finite sets. (Number n is a form shared by all sets of
exactly n elements; any pair of such sets allows for a one-to-one correspondence
between their elements.) The colloquial notion of "algebraic form" provides yet another
class of elementary examples of form-concepts. Take expression like (x+y)expn. All
expressions obtained from this one through any admissible substitution of symbols
x,y, n by some other variables, constants or algebraic expressions are told to share
the "same algebraic form". Isomorphisms associated with this algebraic form are such
substitutions (notice the reversibility of substitutions). For more modern examples of
form-concepts think about any concept build up as "structured set". Think, for
example, about "the" group of plane Euclidean motions already mentioned. Obviously
one can take as many isomorphic "copies" of this group as one likes. The mathematical
notion of form (and the notion of group of symmetry associated with it) plays a major
role in physics and other sciences (for a popular account see Weyl 1952).
The above examples are so various that one might think that in fact all mathematical
concepts are form-concepts. This view has been held by Plato who, however, made a
distinction between mathematical forms and forms strictu sensu. In todays philosophy
of mathematics this view is known under the name of mathematical structuralism.
Mathematical structuralism exists in a number of different versions but that
structures are things determined up to isomorphism seems to be a common
assumption. Understandably Hilbert is often referred to as one of founding farthers of
mathematical structuralism (Hellman, forthcoming).
But the view according to which all mathematical concepts are form-concepts is
obviously wrong. Think about the general concept of group - not any particular group
like the group of Euclidean motions but the concept of group as such. There are
certainly isomorphic groups but not all groups are isomorphic. So the general concept
of group is not determined up to isomorphism. Hence it is not a form-concept.
Similarly general concepts of set, natural number, etc. are not form-concepts.
This sounds like a trivial point. However one may still argue that all mathematical
objects are instances of form-concepts. However important the general concept of
group might be one arguably doesn't need in mathematics anything like "general group"
over and above all particular groups. Since these particular groups are all form-
concepts the structuralist view remains plausible. But the situation can be seen
differently. Notice that like in the case of any "particular" (that is, specified up to
isomorphism) group there is a notion of transformation associated with the general
concept of group. I mean the notion of group homomorphism (I shall explain it shortly).
Homomorphisms are non-reversible (except when they are isomorphisms!). One might
say that the general concept of group is determined "up to homomorphism" like the
group of plane Euclidean transformations (or any other particular group) is determined
up to isomorphism. But a better way to put is this: groups with group homomorphisms
make a category. The category of groups is evidently something "over and above" all
particular groups that it comprises; it has properties, which cannot be detected
through studying particular groups. Importantly the concept of category of groups is
not a form-concept as we shall see. The same holds for the category of sets, etc.
By analogy with form-concepts I shall speak about category-concepts. General
concepts of sets, group, manifold, topological space or the general concept of
category itself are category-concepts. This means that items falling under such
concepts and their associated transformations make categories (in particular all
possible categories make the category of categories, see (Lawvere 1966)). The notion
of category-concept is more general than that of form-concept: any form-concept is
a special case of category-concept where all morphisms are isomorphisms.
(Awodey 1996) and some other people argue that the notion of category is a typical
example of structure , which appears to be a different name for (or at least a special
case of) what I call here form. Since the general concept of category is instantiated
by sets, groups, etc., etc., an abstract category can be colloquially called a "common
conceptual form" of all these things . However this is misleading and, as far as one
assumes the proposed understanding of the notion of form, simply wrong. For the
categories in question are not isomorph. They can be mapped to each other by suitable
functors but these functors are never reversible. So it is misleading to interpret the
fact that sets, groups, etc. all make categories in the sense that all these thing share
a "form of category". The erroneous thinking about categories as forms is, in my view,
responsible for the infamous description of category theory as "abstract nonsense".

6. Transformations Instead of Relations?

The notion of transformation can be formulated in mathematics most easily in the
case when transformed objects are construed a la Bourbaki as "structured sets". For
"bare" ("non-structured") sets we have the notion of set isomorphism as one-to-one
correspondence between their elements and a more general notion of morphism
between (or "transformation of") sets, which are called functions: each element of the
domain is sent to a certain (unique) element of co-domain. Since two different
elements of the domain can be sent to the same element of the codomain morphisms
(transformations) of sets are, generally speaking, non-reversible. When a given set is
equipped with a certain structure making it into a group or something else the
permissible transformations are those transformations of underlying sets (that is,
functions), which are said to "preserve" the corresponding structure. For example the
group structure on a given set is a binary operation defined for elements of this sets
and verifying certain axioms. Given two groups G and G' construed as just explained
with corresponding group operations + and * think about function f: G-->G'. It is said
to preserve the group structure (and hence to be a group homomorphism) iff

(GS): f(x+ y)=f(x)*f(y) .

When f is reversible groups G and G' are called isomorphic or "identical up to

isomorphism". Such groups cannot be distinguished by their group-theoretic
properties. However (GS) may hold in the non-reversible case as well. Then the
colloquial talk of "preservation of structure" becomes rather misleading. Consider the
case when G' consists of only one element 1 (such that 1+1=1). In this case
homomorphism f "destroys" the structure of G rather than preserves it! A better way
to put (GS) into words is to say that in the general case homomorphisms respect the
group structure (even if they destroy it).
Things work similarly with all "structured sets" (with different conditions amounting to
"preservation" or "respect" of the corresponding structure). This points to an
important link between category theory and the notion of mathematical structure (and
hence with Mathematical Structuralism). However I see this link as historical rather
than theoretic. For the notion of transformation in fact doesn't depend of the set-
theoretic background involved in definitions of morphisms of groups, topological
spaces and other concepts construed as structured sets. This is true in the case
when transformations are conceived "naively" as well as in the case when the notion of
transformation acquires a technical definition other than set-theoretic. The notion of
geometrical transformation came about long before set theory; in particular the
notion of geometrical motion is implicit already in Euclid's Elements (think about
Euclid's notion of congruence). Klein in his Erlangen Program made geometrical
transformations into foundations and treated them algebraically using the notion of
(algebraic) group. In this latter context the notion of transformation could be hardly
called naive any longer. Finally the notion of transformation or morphism became a
basic notion of category theory. Although in the early days of category theory all
important examples of morphisms were morphisms of structured sets MacLane and
Eilenberg realised it from the very beginning that certain morphisms are not
"structural" in this sense. Two basic examples of "non-structural" categories which
immediately suggest themselves are the following: (1) group construed as a category
with just one object and such that all its morphisms are isomorphisms; (2) preoder
construed as a category having no more than one morphism between any (ordered)
pair of its objects. The categories just mentioned shouldn't be confused with
categories of "all" groups and "all" preorders correspondingly since (1) is an "individual"
group and (2) is an "individual" preorders. But category of "all" groups and category of
"all" preorders could be construed by taking, correspondingly, (1) and (2) as their
objects. Groups and preorders, of course, "are" structured sets in the sense that
they can be construed as such things. However as we have just seen they can be also
construed "purely categorically", so their morphisms are not supposed to "preserve"
or "respect" any structure. However trivial may be the translation of notions of group
and preorder from the set-theoretic into the categorical language it clearly shows
that the notion of morphism is more general than that of structure-preserving
morphism.
Let's now go back to formal axiomatic method. Its basic idea can be expressed by this
slogan: describe mathematical objects in terms of formal properties of their relations.
The slogan of categorical method reads slightly differently: describe mathematical
objects in terms of categorical properties of their transformations. The principle
assumption behind the categorical approach is that transformations of mathematical
objects (including both isomorphisms and non-reversible transformations) indeed
essentially characterise these objects.
A comparison between the two methods suggests an analogy between relations and
morphisms. One can say indeed that morphisms "relate" objects to each other in a
way. As far as this claim is taken in the general philosophical sense it sounds
reasonable. But it is in odds with the standard technical notion of relation as function
sending tuples of individuals (relata) to truth values. For morphism f:A-->B sending
object A to object B is just another mathematical object, which prima facie has nothing
to do with truth values. Further, in a given binary relation R(x,y) arguments (relata)
x,y can be, generally speaking, replaced by some other arguments x',y' so that R(x',y')
again makes sense (i.e. is true or false). This allows for thinking about R as
"structural relationships" (between its relata) appearing in Hellmann's official definition
of structuralism as
a view about the subject matter of mathematics according to which what matters are
structural relationships in abstraction from the intrinsic nature of the related objects

But for morphism f:A-->B nothing similar is possible. The fact that object A is domain
of f and object B is codomain of f characterises f essentially: there is no sense in
which f may survive replacement of A, B by some other objects A', B'.
These observations point to a gap between the general philosophical intuition behind
the concept of relation and the standard logical notion of relation just mentioned. The
intuition says that morphisms are relation-like while within the standard logical
formalism they should be treated as objects. This can be seen in Lawvere's early
papers of 1964 and 1966 already mentioned where categories are treated as classes
of morphisms with relations "domain of", "codomain of" and "composition of". Let me
now provide a further argument showing that doing category theory formally (i.e.
using formal axiomatic method) is not a good idea.

7. Categories and Categoricity.

As the above quote from Hilbert clearly shows he thinks of a formal theory as
construed up to isomorphism of its interpretations. But what guaranties that a given
theory like that of Grundlagen is indeed formal in this sense, i.e. that all its models are
in fact isomorphic? The desired property has been called by Veblen (1904)
categoricity. This term has nothing to do with category theory.
When Hilbert was preparing his Grundlagen for publication he apparently didn't yet see
the problem. He discovered it about the time of the first publication of Grundlagen. In
his lecture Ueber den Zahlbegriff delivered in 1899 and published in 1900 Hilbert first
introduced an "axiom of completeness" (Vollstandigkeitsaxiom) requiring from any
model of a given theory (this time it was arithmetic) this maximal property: given
model M satisfying the rest of the axioms one cannot obtain another model satisfying
the same axioms by extending M with new elements. In the second and following
editions of Grundlagen Hilbert used a similar axiom under the same name. Hilbert's
Vollstandigkeitsaxiom implies categoricity (although the converse is obviously not the
case). From the point of view of today's model theory this axiom looks very dubios if
not plainly "wrong": Hilbert's account doesn't provide any reason why a model with the
desired property should exist (in any appropriate sense of "exist") but apparently
relies onto the intuition which suggests that the intended model (i.e. "usual"
geometrical space) has this property. This is a very shaky ground indeed. The
standard Tarskian model theory doesn't allow for a model with the required maximal
property (because of "upward" Skolem-Lowenheim's theorem: given that a theory has
an infinite model it has other models in higher cardinalities). So the
Vollstandigkeitsaxiom turned to be incompatible with later model theory. Since the
notion of categoricity has been formulated by Veblen the idea that this property can
be stipulated by a fiat has been largely abandoned.
For the reason just explained categoricity is commonly viewed as a desired property of
formal theories. However in 20-th century people have learnt to be tolerant to the
lack of categoricity. For Zermelo-Frenskel axiomatic set theory (ZF), Peano Arithmetic
(PA) and some other theories commonly viewed as important turned to be non-
categorical. To preclude the right of these theories to be qualified as formal on this
ground would apparently mean to go too far. To save the situation philosophers
invented the notion of "intended model", that is of model chosen among others on an
intuitive basis. Isn't this ironic that such a blunt appeal to intuition is made in the core
of formal axiomatic method? I'm agree with F. Davey who recently argued that "no-one
has ever been able to explain exactly what they mean by intended model". (FOM, 13 Oct
2006). Other people question the categoricity requirement. Asks R. Lindauer: "Why
rule out non-standard models of 1st-Order PA? What's wrong with having other
models? Why should we be making our model-world smaller and not larger?" (FOM, 27
Oct 2006).
I believe that the lack of categoricity of theories like ZF and PA is indeed a serious
flaw because the lack of categoricity undermines the very idea of formal theory. At
the same time I agree with Lindauer and other people who think that the pursuit of
categoricity is misleading. These two claims might seem to contradict each other but
they don't: instead of forcing categoricity or looking for a philosophical excuse of the
lack of categoricity of formal theories I suggest to change the method of theory-
building and the corresponding notion of theory. As far as non-reversible morphisms
are treated on the same footing with isomorphisms the pursuit of categoricity has no
sense any longer. Trying to describe a model of a given theory "up to arbitrary
morphism" rather than up to isomorphism one may get a category of models which has
"good" categorical properties making it "well-manageable". In the following paragraph I
provide some details of how this can be achieved. As we shall see the categorical
approach undermines the usual distinction between a (formal) theory and its models: in
the new context a theory can be naturally seen as one of the models having this
specific property that it "generates" all the others. This is hardly surprising given that
the notion of formal theory (as distinguished from its models) requires categoricity in
Veblen's sense.
A great advantage of formal axiomatic method is that it provides a clear idea about
the role and the place of logic of in theories built by this method (albeit details can
always be a subject of philosophical discussion). Talking about categorical method as
an alternative to formal method we cannot avoid this important issue either. In the
next paragraph I shall briefly review classical and formal notions of logic, and then
develop a notion of categorical logic and show its role in categorical method of theory-
building.

8. Formalising Logic
Traditionally logic is conceived as a general theory of reasoning independent of any
particular subject matter one may reason about. On Aristotle's account logic is closely
related to ontology. In particular, Aristotle treats the logical law of non-contradiction
as a fundamental ontological principle, and his so-called perfect syllogism reflects the
structure of entity (as Aristotle understands it). Logical truths are grounded upon
ontological truths even if the former do not coincide with the latter. There is, of
course, a sense in which Aristotle's logic could be called formal. For it captures and
studies common forms shared by various reasonings about different matters. These
forms are called logical forms; perfect syllogism is a typical example. However other
sciences like biology proceed similarly: biology captures forms shared by different
organisms and so brings about the notion of biological form (differently called "form of
life" or "living form"). However biology can be hardly called a formal science on this
ground. This shows that the notion of being formal relevant to modern logic is
different.
The Hilbertian notion of formal theory is that of "framework or schema of concepts
together with their necessary relations to one another" taken in abstraction from its
possible "basic elements", that is, technically speaking, from its possible
interpretations. Formal logic in the usual today's understanding of this expression is
formal in the same sense. This gives the idea of distinction between logical syntax and
logical semantics which is not found in the traditional logic. "Formal" means here
"syntactic". As Carnap puts this

The task of formalisation of any theory ... belongs to syntax, not to semantics.
(1947, Preface)

However there is a problem here, which make it difficult to apply the notion of formal
theory to logic. Remind that not all terms used in axioms of Hibert's Grundlagen have
variable meanings. Meanings of terms "and", "or", "exist" and of some others are fixed;
such terms form the logical vocabulary of the given theory, and the (maximal)
fragment of the theory which involves no other terms but logical can be identified with
logic. So unlike geometrical theories themselves their underlying logic is fixed and
doesn't allow for different interpretations. But this means that one cannot distinguish
here between formal and interpreted logic or between logical syntax and logical
semantics. However with this latter distinction we loose the modern notion of formal
logic.
Actually this notion is apparently absent in Hilbert. The idea behind his Grundlagen is to
base mathematics in general and geometry in particular on logical rather than intuitive
or other grounds. Discovery of non-Euclidean geometries and some other
developments in mathematics led many people to the belief that since the traditional
mathematical intuition proved unreliable logic remained the only firm foundations
available for mathematics. (Nagel 1939) provides a thorough historical analysis of
mathematics of 19th century showing where this logicist view on mathematics stems
from (Endnote 2). Hilbert's Grundlagen showed how this general approach can be
realised in practice. A formal theory in Hilbert's sense is a "logical skeleton" (or
"logical form" liberally understood) shared by a class of traditional so-called "naive"
mathematical theories. But logic itself on this account is not formal in anything like
the same sense. Applying the notion of formal theory just given to logic one would
need to speak of "logical skeleton of logic" which is at least unclear and at most
senseless.
Carnap and other promoters of the idea of formal (or "formalised" logic) largely
disregarded this philosophical difficulty and applied formal or "syntactic" method to
logic itself, introducing the nowadays standard distinction between logical syntax and
logical semantics. Instead of saying that in formal theories of the type of Grundlagen
logic is not a subject of interpretation these people (as well as many today's
professors of logic after them) would say that logical terms get interpreted together
with non-logical terms but unlike the latter they are always interpreted in the same
way (i.e. logical terms are invariant under all possible interpretations). Unless the
class of "all interpretations" is precisely determined this move seems me purely
rhetorical, and if such a class is determined then the notion of logicicity becomes
relational (dependent of the given class of interpretations). Anyway this doesn't solve
the problem, which is the following.
The initial hope that logic unlike geometry will always be rigidly fixed on the pain of
absurdity turned to be futile and logic ramified into multiple systems just like earlier
did geometry. Formalisation of logic played an important role in this development
because it allowed to treat systems of logic on equal footing with systems of
geometry or algebraic systems. However since the assumption about rigidity of logic is
given up the whole idea of formal approach (at least in Hilbert's sense) is shaken, so it
becomes rather unclear what is meant by "formal" logic except that this kind of logic
is symbolic and mathematical. What kind of new philosophy of logic is needed to
replace Hilbert's (or Frege's) logicism in order to cope with these developments
remains an open question. Most philosophers working today in logic share Hilbert's
weak logicism according to which logic has to do with foundations of mathematics and
of other sciences. At the same time only few of them if any hold the old-fashioned
view according to which there is only one "true" system of logic. As the above analysis
shows the two parts of the popular view are hardly compatible with each other.
Categorical logic suggests a solution of this problem through a revision of formal
axiomatic method and more broadly - of formal approach. In order to show this
solution I shall first develop a speculative notion of categorical logic as a generalisation
of formal logic, and then point to some technical developments supporting this
speculative notion.

9. Categorical Logic
Both traditional Aristotelian logic and modern formal logic hinge on the notion of logical
form. What kind of forms are logical forms is a difficult question which I shall not now
try to answer here. Let's see instead what happens to logic when the notion of form is
upgraded to that of category. Remind that categories unlike forms, generally
speaking, don't allow for a straightforward abstraction: given a class of balls one may
think about them "up to isomorphism" and stipulate The Ball as their shared abstract
form but nothing similar works when objects of a given class make a category. So a
categorical system of logic unlike formal logic cannot be anything like a self-standing
structure occasionally applied in this or that particular context. Instead it must be
"internal" or "intrinsic" with respect to a given category playing the role of such
context. This rises the question of universality of categorical logic: Is this indeed
appropriate to give the title of logic to something, which applies to a particular
category rather than to everything? Let me make three remarks concerning this
question. First, nothing prevents one to conceive of "everything" as a category (rather
than as a class). This idea is behind (Lawvere 1966) "The Category of Categories"
approach. Personally I'm not sympathetic with this idea. Actually I consider the "local"
character of categorical logic as its advantage rather than otherwise. The second
remark is that the idea of "local" or "regional" logic as a notion of a system of logic
designed for some specific purposes has been around already during few decades, and
it better fits today's technical developments in logic than the traditional idea of the
universal logic. The third remark is that in the categorical setting the notion of
"regional" or "local" logic can better cope with the following important objection:
Ramification of logic brakes the rational thought into a number of incompatible
domains and this contradicts the whole idea of rationality. The usual response to this
problem is the suggestion to find a weak system U of universal logic such that regional
logics could be seen as specifications of U in corresponding local contexts.
Alternatively one may challenge the assumption about the unity of rationality on
philosophical grounds. Categorical logic allows for a different solution, namely it
provides means for translation between different local logics. Such translation doesn't
intervene here as a new external principle since what I call here local or regional logic is
construed in the categorical setting in terms of morphisms which can be naturally
viewed as translations. One would still need, of course, some universal principles,
namely the general principles of category theory. What makes the difference is this: in
the categorical setting universal principles of rationality are principles of translation
rather than formal principles imposing universal forms of reasoning indifferent to its
content. One may argue that general principles of translation I'm talking about are
themselves formal but this is, in my view, an abuse of the language. As far as one
tries to be precise about the meaning of "formal" it becomes clear that the argument
is wrong. Importantly categorical logic assumes the possibility of multiple local logics
to begin with, so that no counterpart of the aforementioned problem about
formalisation of logic arises in the categorical context.
Let's now see how this speculative notion of categorical logic can be realised
technically. There are several different ways to "do logic" with categories but the
most relevant in the present context is apparently the so-called topos logic. The
notion of topos is of a geometrical origin; it was a discovery of Lawvere that this
notion can be introduced axiomatically through an appropriate specification of the
abstract notion of category. It turns out that given a topos one may associate with it
a logical calculus called "internal language". Then the given topos can be viewed as a
geometric model of this calculus. However a more suggestive view on this situation is
the converse one: the given topos has certain specific "logical properties" which
determine its "internal logic". This gives the idea of "reasoning in a topos"; reasonings
in different toposes can be always compared through morphisms (functors) between
these toposes (which can be of different sorts). Remark that the view just mentioned
(colloquially known as "toposophy") is in odds with the usual (weak) logicism which
requires to "fix logic first". For according to toposophy logic is an element (or perhaps
an aspect) of the overall construction of topos, which doesn't have an epistemic
priority with respect to the rest of this construction. This rises anew the traditional
philosophical issue about first principles. But let me now turn this discussion in a more
technical mode.
Technically speaking the problem is that the general idea of categorical logic doesn't
provide by itself any clue of how it could be used for theory-building. Even if one
refuses the idea of logical foundations of theories and tries to recover logic
afterwards the problem of foundations still persists at least as a pedagogical problem.
For no theory can be grasped at once but needs certain guiding mechanisms allowing to
explore it piece by piece. In addition any theory needs an entry (or multiple entries).
Let me briefly describe the idea of functorial semantics put forward in (Lawvere
1963), which serves these purposes. It can be viewed as a simulation of formal
axiomatic method by categorical means. This view on functorial semantics is helpful
for comparing formal and categorical methods.
Instead of writing axioms with usual strings of formulas one encodes axioms into a
special "syntactic" category T which plays the role of "formal theory". Then like in the
case of standard semantics one takes for granted a "background category" B, which is
usually taken to be the category of sets but can be chosen differently. Models of T
are functors of the form F: T-->B . This construction is that it allows for different
notions of model dependent one specific properties of functors of form F. Another
remarkable fact is that under rather general conditions T can be embedded into a
category M(T,B) of its functorial models. This definitely changes the whole idea of
theory as a structure over and above all its possible models and suggests the view on
a theory as "generic model" (Lawvere 1963-2004, p.19) which generates other models
like circles and straight lines generate further constructions in Euclid's Elements. The
functorial semantics makes it clear that the requirement of categoricity (in Veblen's
sense) is as much unrealistic as unnecessary: although "good" categorical properties
of M(T,B) are much desirable there is no good reason to require that this category
reduces to a single object.

10. Conclusion: Formalisation versus Categorification

Lawvere's functorial semantics has been developed for a special case of algebraic
theories and so it cannot be immediately used as a method of theory-building
applicable in all areas of mathematics. Since then a lot of technical work been done in
related fields of categorical logic and categorical model theory. For a historical
introduction and further references I refer the reader to (Bell 2005). The categorical
method of theory-building is a work in progress, so to the date it doesn't exist in any
standard form. The purpose of this non-technical paper is to provide this method with
an appropriate epistemological background, which might motivate further technical
work. So let me conclude with some general epistemological remarks.
The categorical method outlined above suggests an epistemic strategy, which differs
from that suggested by formal method. In the latter case the general epistemic
strategy is to subsume different objects under a common form and then stipulate this
form as a self-standing abstract object. This amounts to identifying of given objects
"up to isomorphism". Here "objects" may stand for various mathematical
constructions including whole mathematical theories, so the sense in which the
obtained formal object is "abstract" is relational. This leads to a traditional
hierarchical structure of organisation of mathematical knowledge, where theories and
concepts are subsumed under other theories and concepts, which are more abstract,
more general and more formal. Apparently the same pattern applies at the interface
between mathematics and the material world and is responsible for the usual
qualification of mathematics as a formal science. This epistemic strategy I call
formalisation.
The alternative strategy of categorification is different. It is more general and in a
sense more straightforward. Categorification amounts to taking into consideration all
transformations, which can be described as categorical morphisms but not only
reversible ones, i.e. not only isomorphisms. The requirement that objects and
transformations in question make a category is much weaker than the requirement
that these transformations are reversible (and so make a group). So formalisation is a
special case of categorification. Unlike formalisation categorification in the general
case doesn't subsume objects in question under a conceptual umbrella but simply links
them (by mophisms) into into a whole, namely into a category. When objects are
theories and morphisms are mutual interpretations of theories one gets a network of
theories. Although this network might have no single "centre" it may be still coherent
and well-manageable if it has "good" categorical properties.
We see that categorification like formalisation serves for integration of mathematical
knowledge. But categorification unlike formalisation doesn't bring about a hierarchical
structure. Given that the very idea of foundations seems to imply a hierarchical
structure of knowldge (which "starts with" foundations and then branches into various
specific sub-domains) one may wonder if categorification is compatible with it. I think
that at least one version of the notion of foundations remains viable in a categorical
context, namely the pedagogical one stressed by Lawvere in his (2003). I mean the
notion of foundations as an "entry" into a theoretical network. Such entry should
exist for any theoretical network but it obviously needs not to be unique.
I suggest that hierarchical structures cannot any longer serve as universal models of
organisation of knowledge just like they cannot any longer serve as models of
organisation of our societies. But the task of integration of knowledge into a
manageable whole remains pertinent as ever. I believe that categorical method of
theory building outlined in this paper can be helpful for this task.

Endnotes:
(1) One might argue that I systematically confuse here two different notions of
interpretation: (a) assignment of referents to primitive terms of a formal theory like
"point" and "straight line" (evaluation of logical variables) and (b) interpretation of one
non-formal theory in terms of another non-formal theory like in the case when
traditional geometrical points are represented by pairs of numbers. In fact in
preceding paragraphs I'm talking only about (b) as also does Hilbert in the given quote.
The argument which I develop below in the main text is this: the idea of formal theory
and of interpretation in the sense (a) assumes that all models of a given formal theory
are mutually reversibly interpretable in the sense (b). But generally they are not.

(2) This logicism about mathematics should be distinguished from stronger form of
logicism aiming at reduction of mathematics to logic.

Literuature:
Awodey, S., 1996, "Structure in Mathematics and Logic: A Categorical Perspective",
Philosophia Mathematica (3), N4, p. 209-237
Bell, J., 2005, "The Development of Categorical Logic", Handbook of Philosophical
Logic, Vol. 12. Springer
Beltrami, E., 1868, "Saggio di interpetrazione della geometria non-euclidea", Giornale di
Matematiche, v.6, p. 284-312
Beltrami, E., 1868-69, "Teoria fondamentale degli spazii di curvatura constante",
Annali di matematica pura et applicata (2), 2, p. 232-255
Bolyai, J., 1832, Scientia absoluta spatii. Appendix to Bolyai, F., Tentamen juventutem
studiosam in elementa matheseos purae elementis ac sublimioris, methodo intuitiva,
evidentiaque huic propria, introducendi, Tomus Primus.
Bourbaki, N., 1939 - , Elements de mathematique , Hermann Paris
Carnap, R., 1947 Formalisation of Logic (Studies in Semantics, vol. 2) Harvard
University Press
Eilenberg, S., MacLane, S., 1945, "General theory of natural equivalences",
Transactions of American Mathematical Society, 58, p. 231-294
FOM (Foundations of mathematics): http://www.cs.nyu.edu/pipermail/fom/
Frege, G., 1971, On the Foundations of Geometry and Formal Theories of Arithmetic,
E.H.W. Kluge (ed.), Yale University Press
Hellman, G., forthcoming, "Structuralism, mathematical", to appear in The Encyclopedia
of Philosophy, 2nd ed., MacMillan; available at G. Hellman's webpage
Hilbert, D., 1899, Grundlagen der Geometrie, Leipzig, Teubner; second ed.: 1903
Jong W.R., Betti A., "The Aristotelian Model of Science: a Millennia-Old Model of
Scientific Rationality", Synthese, forthcoming in 2007
Klein, F., 1872, Vergleichende Betrachtungen ueber neuere geometrische Forschungen
("Erlanger Programm") , Erlangen
Kline, M., 1972, Mathematical Thought from Ancient to Modern Times, Oxford
Lawvere, W., 1963-2004, "Functorial Semantics of Algebraic Theories" (Thesis1963)
and "Some algebraic problems in the context of functorial semantics of algebraic
theories " (1968) Reprint with author commentary Reprints in Theory and Applications
of Categories, No. 5, p. 1-121
Lawvere, W., 1964, "Elementary Theory of the Category of Sets", Proceedings of the
National Academy of Science, vol. 52, N6, p.1506-1511
Lawvere, W., 1966, "The Category of Categories as a Foundation for Mathematics", La
Jolla Conference on Categorical Algebra, Springer-Verlag , p.1-20
Lawvere, W., 2003, "Foundations and Applications: Axiomatization and Education", The
Bulletin of Symbolic Logic, vol. 9, No. 2, p. 213-224
Lobatcheffsky, N.I., 1837, Geometrie Imaginaire, Journal fuer die reine und
angewandte Mathematik N17
MacLane, S., 1971, Categories for Working Mathematician
Nagel, E., 1939, "The Formation of Modern Conceptions of Formal Logic in the
Development of Geometry", Osiris, vol.7, p.142-223
Riemann, B., 1854, “Ueber die Hypothesen, welche der Geometrie zu Grunde liegen”
(Habilitationsvortrag), Abhandlungen der Kgl. Gesellschaft der Wissenschaften zu
Göttingen. 13 (1867): 133-152; reprinted in Gesammelte Mathematische Werke
(Narasimhan, ed.), Berlin, Heidelberg, NewYork: Springer-Verlag, 1990
Rodin, A., , forthcoming "Identity and Categorification" , to appear in Philosophia
Scientia; available at: <http://arxiv.org/pdf/math.CT/0509596>
Veblen, O., 1904, "A system of axioms for geometry", Trans. Amer. Math. Soc. N5, p.
343-384.
Weyl, H., 1952, Symmetry, NJ : Princeton University Press