Full Transcript

·YouTLDR

Philosophy of Mathematics & Frege - Michael Dummett (1994)

1:32:331,634 summary words · ~8 min readEnglishBy Philosophy OverdoseTranscribed Jun 30, 2026
Analyze another video with Pro30-day money-back guarantee
Summary

Gottlob Frege's foundational project attempted to derive arithmetic from pure logic by treating numbers as logical objects, but the failure of his system under Russell's Paradox reveals that infinite mathematical totalities cannot be treated as determinate under classical logic, necessitating an intuitionistic approach.

Understanding the collapse of Frege's logicism exposes the deep-seated tension between classical logic and the infinite, reshaping our grasp of how abstract concepts can be applied to the physical world without relying on ungrounded metaphysical assumptions.

Section summaries

0:00-11:35

The Nature of Philosophy and A Priori Puzzlement

watch

Michael Dummett begins by defining philosophy not by its subject matter, but by its unique style of thought and conceptual 'puzzlement.' Unlike empirical sciences, both philosophy and mathematics are a priori disciplines, meaning their necessary truths are discovered by reasoning alone rather than observation. Dummett compares this conceptual confusion to soldiers in a battle who understand their immediate duties but lack a clear, global view of the field, citing the interpretation of quantum mechanics as a prime example of a successful theory whose meaning remains uncomprehended.

  • Philosophy aims at understanding our conceptual frameworks, whereas science seeks empirical knowledge.
  • Mathematics and philosophy are both a priori disciplines, yet they differ deeply in how they develop non-trivial necessary truths.
  • A clear view of a concept's function is distinct from merely being able to use it successfully in local contexts.

It establishes the methodological and metaphysical framework for why mathematical philosophy exists and how it compares to physics and logic.

11:35-18:32

Frege's Life, Neglect, and Critique of Kant

watch

Dummett introduces Gottlob Frege, a mathematician who lived an uneventful academic life in Jena but pioneered modern mathematical logic to resolve a 'scandal' in mathematics: the inability of mathematicians to explain what numbers are or why their theories are true. Frege sought to prove that arithmetic (unlike geometry) is entirely analytic and independent of Kantian intuitions of space and time. To do this, Frege followed the path of Bernard Bolzano, striving to completely purge appeals to intuitive geometric evidence from rigorous mathematical analysis.

  • Frege's life's work was spent on the boundary of logic, mathematics, and philosophy, aiming to justify mathematical belief.
  • While Frege accepted Kant's intuitive view of geometry, he fiercely rejected it for arithmetic.
  • Bernard Bolzano's work served as a historical precursor to Frege's effort to expel intuition from mathematical proofs.

Explains the core intellectual conflict between Frege and Kant's synthetic a priori theory.

18:32-25:29

The Invention of Predicate Logic and the Reduct of Induction

watch

To execute his program, Frege had to invent modern mathematical logic in his 1879 work, the Begriffsschrift. Prior logic, from Aristotle to George Boole, could not analyze complex mathematical reasoning. Frege introduced a system of quantifiers that formalized first-order and second-order logic, allowing him to define the concept of a sequence in purely logical terms. By defining the natural numbers as a sequence starting from zero, Frege successfully reduced the mathematical principle of induction to a fundamental, non-intuitive law of logic.

  • Traditional Aristotelian and Boolean logic were completely inadequate for analyzing mathematical proofs.
  • Frege's formalization of second-order logic allowed him to define sequences without appealing to temporal metaphors.
  • Mathematical induction is not a special postulate of number theory, but a direct consequence of the logical definition of a sequence.

Crucial for understanding Frege's mathematical breakthroughs and how he circumvented temporal intuition.

25:29-32:26

The Fruitfulness of Deduction and the Mystery of Application

watch

Dummett addresses a deep philosophical paradox: if mathematical truths are analytic (and thus trivially logical), how can deductive reasoning lead to massive, non-trivial expansions of our knowledge? Frege's answer relies on the creative intellectual act of discerning new structural patterns within complex propositions. Additionally, Dummett emphasizes that for Frege, the applicability of arithmetic to the empirical world is not an accidental miracle, but the very feature that elevates it from a game to a science, demanding that the general principles of application be baked directly into the definitions of numbers.

  • Deductive reasoning is a creative act that involves grasping and discovering previously unperceived structural patterns in sentences.
  • Mathematics is defined by its applicability; a system without empirical application is merely a game, not a science.
  • Frege defined real numbers as ratios between quantities to ensure their immediate, pure application to physical magnitudes.

Directly addresses the relationship between formal mathematical objects and empirical physics/application.

32:26-39:23

Numbers as Objects, Nominalism, and the Context Principle

watch

This section explores Frege's controversial stance that numbers must be treated as genuine, abstract objects rather than mere properties of properties. To avoid the nominalist critique that imperceptible, abstract entities are useless fictions, Frege relied on his famous Context Principle: words only refer to things in the context of a whole sentence. Just as we can talk about the 'equator' without needing to physically see or touch a line wrapped around the Earth, we can reference numbers because we have clear, non-circular rules for evaluating the truth of sentences containing them.

  • Frege held that numbers are 'logical objects' whose existence is guaranteed by logic itself.
  • Nominalists argue that abstract objects cannot exist because their absence wouldn't change our physical observations.
  • The Context Principle solves the nominalist challenge by anchoring reference in semantic truth-conditions rather than physical causation.

This is the philosophical heart of Frege's semantic theory of reference and abstract objects.

39:23-48:39

The Fundamental Equivalence and the Threat of Circularity

optional

Frege attempted to prove the infinity of natural numbers using what is now known as Hume's Principle (the fundamental equivalence): the number of Fs equals the number of Gs if and only if there is a one-to-one correspondence between them. However, Frege rejected this principle as a self-sufficient definition because it failed to resolve identity statements involving non-numerical objects (the Julius Caesar problem) and couldn't even determine identities between different kinds of numbers. This created a troubling circularity where defining numbers depends on defining the domain of all objects, yet that domain's composition itself depends on which numbers exist.

  • Hume's Principle asserts that cardinality is defined through one-to-one mathematical mappings.
  • Frege rejected Hume's Principle as a complete foundation because it failed to define identity statements comprehensively.
  • Defining fundamental mathematical totalities simultaneously with their domains leads to a persistent problem of circularity.

Highly technical discussion of the Julius Caesar problem and the limits of Hume's Principle.

48:39-57:55

The Catastrophe of Axiom V and the Limits of the Continuum

watch

To bypass the circularity of the fundamental equivalence, Frege defined numbers as classes of equivalent classes, which shifted the burden of proof to his theory of classes. This led directly to his reliance on Axiom V, which notoriously rendered his entire system inconsistent by allowing the construction of Russell's Paradox. Dummett argues that this catastrophe was inevitable because Frege assumed a 'robust realism' that classical bivalent logic applies to uncircumscribed, infinite totalities like the continuum. Dummett asserts that we cannot construct the mathematical continuum without circularity, requiring us to view the real numbers as an open-ended, indeterminate totality governed by intuitionistic, rather than classical, logic.

  • Frege's logicist program collapsed because Axiom V permitted the formation of self-referential paradoxes.
  • Classical logic incorrectly assumes that every statement about an infinite domain has a determinate truth value.
  • The real numbers constitute an indeterminate totality that can only be legitimately analyzed using intuitionistic (constructive) logic.

It delivers the ultimate philosophical and mathematical conclusion of Dummett's lecture regarding the limits of classical logic.

57:55-1:30:21

Q&A Session: Wittgenstein, Husserl, and Ordinal Numbers

optional

The lecture concludes with a Q&A session where Dummett compares his philosophical approach to Ludwig Wittgenstein's, emphasizing his preference for detailed, systematic, and fully articulated accounts over Wittgenstein's aphoristic style. He addresses questions on the consistency of Hume's Principle, the applicability of advanced mathematical structures like group theory to physics, and the historical debates between Frege and Peano. Finally, Dummett critiques Edmund Husserl's early 'Philosophy of Arithmetic' as psychologistic and discusses why defining numbers as ordinals rather than cardinals might be more fundamental but would lead to the Burali-Forti paradox much faster.

  • Dummett distinguishes his philosophical methodology from Wittgenstein's by focusing on highly detailed, systematic elaboration.
  • The applicability of complex math (like group theory) in physics must be explained by mathematical transformation itself.
  • Husserl's early work was heavily psychologistic, diverging sharply from Frege's rigorous, anti-psychologistic realism.
  • Defining numbers as ordinals rather than cardinals introduces severe logical paradoxes (like Burali-Forti) even more rapidly.

Contains interesting historical and comparative context (Husserl, Wittgenstein, Peano) but is non-essential to the primary lecture arguments.

Key points

  • Frege's Rejection of Kantian Mathematical Intuition — Unlike Immanuel Kant, who argued that arithmetic relies on the synthetic a priori intuition of time, Frege contended that arithmetic is entirely analytic and derivable from pure logic alone, necessitating the invention of modern predicate logic to prove it.
  • The Context Principle as a Vindication of Abstract Objects — Frege's Context Principle posits that words only have meaning within the context of a complete sentence, meaning that abstract, non-actual objects (like numbers or the equator) are philosophically justified as long as we can define the truth-conditions of sentences containing them without circularity.
  • The Circularity of Infinite Totalities — Attempting to define a fundamental mathematical domain (like natural or real numbers) simultaneously with the operators that interpret its members creates a vicious circularity, as we cannot specify the domain's boundaries prior to understanding the theory's operations.
  • The Imperative for Intuitionistic Logic — Because we cannot cleanly circumscribe the infinite totality of real numbers without circularity, we are not justified in assuming that every statement about the continuum is determinately true or false, which invalidates classical bivalence and mandates the use of intuitionistic logic.
a philosopher is afflicted by an urge to subject any proposed answer to just such scrutiny and to arrive at an answer that will stand up to scrutiny because until he does he's conscious that he does not understand Michael Dummett
It is applicability alone... that raises arithmetic from the rank of the game to that of a science Michael Dummett (quoting Frege)

AI-generated from the transcript. May contain errors.

0:00

people occasionally express puzzlement

0:03

that there being such a thing as

0:05

philosophy of mathematics such

0:08

puzzlement

0:09

arises from a failure to understand that

0:11

philosophy like history

0:13

is characterized not by its subject

0:15

matter but

0:16

by its style of thought by the kind of

0:19

questions it asks

0:20

and the way in which it goes about

0:22

trying to answer them

0:24

when this is understood phrases

0:26

beginning the philosophy of

0:28

will cause no more surprise than once

0:30

beginning the history of

0:33

philosophers attempt to answer questions

0:35

we are prompted to ask

0:37

by a quite special kind of puzzlement

0:40

this puzzlement arises from our

0:42

imperfect

0:43

mastery of the concepts we employ

0:47

these questions occur to everyone but

0:49

people are often content to brush them

0:51

aside

0:52

with answers that will not withstand

0:54

scrutiny

0:55

scrutiny not accorded them by those

0:58

lacking in philosophical curiosity

1:01

a philosopher is afflicted by an urge to

1:04

subject any proposed answer

1:06

to just such scrutiny and to arrive at

1:09

an answer that will stand up to scrutiny

1:12

because until he does he's conscious

1:14

that he does not understand

1:16

and what he wants above all is not so

1:19

much to know

1:20

as to understand characteristic

1:24

philosophical question is why can we not

1:27

affect the past

1:28

although we can affect the future

1:31

someone without philosophical curiosity

1:33

may answer impatiently

1:35

because the past has already happened or

1:38

because the previous event has either

1:39

occurred or not occurred

1:41

and will experience only irritation when

1:44

it's pointed out that the first answer

1:46

merely repeats the problem without

1:48

solving it and the second

1:50

may be counted by observing that a

1:52

subsequent event

1:53

either will or will not occur

1:57

the philosopher is irked by the

1:58

inadequacies

2:00

the inadequacy of these answers that

2:02

first come to mind

2:04

he's driven to seek one that will

2:06

satisfy him

2:08

well why because when we are faced with

2:11

this question

2:12

a question that asks why we cannot do

2:15

something that seems on the face of it

2:17

nonsensical

2:18

and find that we cannot clearly explain

2:20

what makes it nonsensical

2:23

we become aware that we do not really

2:25

know what past and future

2:27

are we have no

2:30

firm grasp upon the concepts of past and

2:34

future

2:35

now these are concepts we constantly

2:37

employ in the current of everyday life

2:39

and everyday conversation we recall what

2:42

happened a week ago or ten years ago

2:44

we avow what we intend to do tomorrow or

2:46

speculate

2:47

on what will happen six months from now

2:50

we use these concepts all the time of

2:53

course we understand them

2:55

and yet by our inability to answer the

2:58

question

2:59

why we cannot affect the past we show

3:02

that we have only a superficial

3:04

a superficial grasp of them we are like

3:07

soldiers in a battle

3:09

who know enough to be able to do what

3:11

they are meant to do

3:12

but have no conception of what is

3:14

happening on a larger scale

3:17

we can operate with our concepts in the

3:19

situations in which we find ourselves in

3:21

everyday life in the laboratory

3:23

on the stock exchange in the operating

3:25

theater but in wittgenstein's phrase

3:29

we do not command a clear view of them

3:32

or attain a general understanding of

3:34

them but can grasp

3:35

only how they function in particular

3:38

familiar

3:38

con contexts this does not feel good

3:42

only if the concepts we all employ in

3:45

the life of every day

3:47

it applies equally in highly technical

3:49

regions

3:50

quantum mechanics supplies a familiar

3:53

example

3:54

it's commonplace to remark that it's a

3:56

highly successful theory

3:58

physicists know how to use it to predict

4:01

observations and measurements and yet

4:03

frequent conferences

4:05

are held to discuss its interpretation

4:08

it's well understood how the theory is

4:10

to be used

4:11

it's not understood what it means that

4:13

is what it tells us

4:15

about the character of reality

4:19

mathematics generates a number of

4:21

questions causing just

4:23

this kind of puzzlement and has

4:25

fascinated and perplexed philosophers

4:27

from plato onwards for two reasons in

4:30

particular

4:31

first it's difficult to say what its

4:33

subject matter is

4:34

what it's about it's reasonably clear

4:38

what physics or geology or biology

4:40

investigates but what exactly is it that

4:43

mathematics investigates

4:46

a standard answer corresponding to the

4:48

old-fashioned division of mathematics

4:50

into arithmetic and geometry

4:52

used to be that it investigated quantity

4:56

and space but this unhelpful answer

4:59

will no longer suffice since there's so

5:01

much mathematics

5:03

that will not fit comfortably under

5:05

either head

5:07

the problem is aggravated by the second

5:09

puzzling feature of mathematics

5:11

the manner in which the mathematician

5:14

sets about

5:15

attempting to solve his problems he uses

5:18

no telescope or microscope he does not

5:21

observe

5:21

anything at all rather he reasons

5:25

he carries out complex deductive

5:27

inferences

5:29

the first question must be answered in a

5:31

way that accords with this

5:33

whatever it is that mathematics is about

5:36

must be something that can be found out

5:38

just by reasoning philosophy

5:42

resembles mathematics in this respect

5:44

the philosopher makes no observations

5:47

and requires no instruments insofar as

5:51

these disciplines can be said to involve

5:53

the making of experiments

5:55

these are thought experiments to imagine

5:57

the experiment made

5:59

is quite as good as actually to make it

6:03

the two subjects thus appear to be our

6:05

priori

6:06

their results do not require us to

6:08

observe how the world happens to be

6:11

but can be arrived at by thought alone

6:14

they are thus independent of how the

6:16

world happens to be

6:17

but would hold good whatever it was like

6:21

they are therefore not merely true but

6:23

necessarily true

6:24

and yet philosophy and mathematics

6:26

differ in all other respects

6:29

it's a puzzle how there can be even one

6:32

subject

6:33

that can be investigated our priori that

6:35

there should be two

6:37

so different from one another appears

6:39

baffling

6:42

necessary truth and our priori knowledge

6:44

are topics

6:45

that engage the attention of all

6:47

philosophers

6:49

it seems straightforward to understand

6:51

how there can be contingent truths

6:53

things that are so but might have been

6:55

different we can get to know contingent

6:57

truth

6:58

only through our experiences of the

7:00

world by observing

7:02

that they're so or deducing from what we

7:04

observe that they must be so

7:07

but how does it come about that there

7:09

should also be

7:10

necessary truths truth that we can know

7:13

independently of our experience of the

7:15

world

7:16

how is it that if we knew all contingent

7:19

truths

7:19

there would be some truths left left

7:22

over

7:23

this is easy enough to understand

7:25

concerning trivial necessary truths

7:28

such as that there are seven days in a

7:30

week or that

7:31

every widow was once married to

7:34

recognize statements of this sort as

7:36

true

7:37

we need know nothing other than the

7:39

meanings of the words

7:41

more we couldn't claim to know the

7:43

meanings of the words if we fail to

7:45

perceive

7:46

that the statements are true but

7:48

mathematical theorems are seldom

7:50

trivial in this sense we may not need to

7:53

start with any initial knowledge

7:55

other than the meanings of the words if

7:57

we are to come to recognize them as true

8:00

but the converse certainly doesn't

8:01

appear to hold

8:03

we may surely know the meanings of the

8:05

words without realizing

8:07

that the theorem so good mathematics

8:10

presents itself

8:11

as by far the most capacious repository

8:14

of non-trivial necessary truths and the

8:17

knowledge of it

8:18

is by far the most extensive body of our

8:21

priori knowledge

8:23

and this fact alone suffice is to make

8:26

it of intense interest

8:27

to philosophers admittedly certain

8:31

philosophers

8:32

of whom john stuart mill is the best

8:34

known example

8:35

have challenged the our priori character

8:38

of mathematics

8:39

claiming that mathematical theories rest

8:41

upon certain

8:42

highly general contingent facts

8:45

recognizable

8:46

by gross observation but this challenge

8:51

does not alter the situation greatly for

8:54

the most

8:55

that such empiricists can argue is that

8:58

the starting point

8:59

of this or that mathematical theory the

9:02

axioms of the theory

9:03

is a collection of readily observable

9:06

contingent facts

9:09

they cannot explain why mathematicians

9:12

fail to set about gathering by devising

9:15

experiments by closer observation or

9:18

proved observational techniques other

9:20

facts of the same kind

9:22

as other scientists do why instead they

9:25

content themselves with the meager

9:27

supply of contingent facts

9:29

with which allegedly they begin and

9:32

proceed to draw out their consequences

9:34

by means of ever lengthening chains of

9:37

deductive argument

9:38

mathematics is still a science quite

9:41

unlike

9:42

any other the only upshot

9:45

of the empiricist's contention is what

9:47

is sometimes called

9:48

if if-then-ism which restricts the

9:50

necessary truths

9:52

discovered by mathematicians to those

9:54

expressed by statements of the form

9:57

if the axioms of the theory hold good

9:59

then such and such a theorem

10:01

holds also and this leaves the problem

10:04

of necessary truth

10:06

untouched

10:10

that there should be and such an a

10:13

priori subject as philosophy

10:15

is comparatively intelligible because

10:18

the philosopher's task

10:19

consists principally of disentangling

10:22

our concepts

10:23

it does not aim so much as arriving so

10:26

much at arriving at new truths

10:28

as it coming to understand better those

10:31

which we've already arrived

10:34

disentanglement sometimes plays a

10:36

critical role in mathematics

10:38

as indeed it does in every subject to

10:41

attain the right definition of

10:43

continuous or of dimension

10:45

was a step of the highest importance

10:48

nevertheless

10:49

hitting on the correct definition of a

10:51

concept though often an essential

10:53

contribution to progress

10:55

remains a preliminary to the discovery

10:58

of mathematical truths not a means of

11:00

discovering them

11:01

it is not the characteristic activity of

11:04

ma

11:05

of the mathematician to explain the

11:08

existence of mathematics

11:10

is a greater challenge to the

11:12

philosopher

11:13

than to explain that of philosophy

11:15

itself

11:16

the capacity for wonder is a

11:18

prerequisite for the activity of

11:20

philosophizing

11:21

and anyone who retains this capacity

11:24

must marvel at the vastness of the body

11:27

of our priori knowledge

11:29

amassed by mathematicians by pure

11:31

deductive reasoning

11:34

now having tried to convey an impression

11:36

of what the philosophy of mathematics is

11:38

about

11:39

i will now tell you something of god rob

11:41

frager

11:43

superficially it might be said that he

11:45

lived an uneventful life

11:47

born in 1848 his entire professional

11:50

career from 1874 to 1918

11:53

was spent teaching at the university of

11:55

guinea and he died in retirement in

11:57

1925.

11:59

although he published three interesting

12:01

articles towards the end of his life

12:03

almost all his work was completed by

12:06

1906

12:08

but his history during his lifetime and

12:11

that his reputation

12:12

up to the present are both extraordinary

12:16

while he lived very few among them

12:19

russell and wittgenstein

12:20

paid any attention to his work a state

12:23

of affairs that continued

12:25

until after the second world war

12:28

now he's universally recognized as the

12:31

grandfather if not the founder

12:33

of analytical philosophy and wherever

12:36

that school of philosophy flourishes

12:38

his works as indispensable reading for

12:41

any student of the subject

12:43

yet he was not a professional

12:45

philosopher

12:46

he was a professor of mathematics still

12:50

largely as as neglected by the

12:52

mathematicians

12:53

as he once was by the philosophers

12:58

this neglect was in part due to his

13:00

originality

13:02

and in part to his having chosen to

13:04

devote his life to an enterprise

13:06

lying on the borderline between

13:08

philosophy and mathematics

13:10

with very few exceptions mostly in the

13:13

early stages of his career

13:15

everything that he wrote was directed

13:18

towards

13:18

or ancillary to this enterprise

13:22

he determined to rectify a situation

13:24

that greatly distressed him

13:26

the inability of either mathematicians

13:29

or philosophers

13:30

to explain the basis of our acceptance

13:32

of mathematical theories

13:35

he wanted to show what justified our

13:37

belief in these theories

13:39

with what right we assumed the theories

13:41

to be correct

13:42

and their theorems to be true his

13:44

attempt to do this

13:46

made him the first modern philosopher of

13:49

mathematics

13:50

now this description of the task fraga

13:53

set himself to accomplish is somewhat

13:55

too wide

13:56

he had the traditional view of

13:58

mathematics

13:59

as subdivided into arithmetic and

14:02

geometry

14:02

and he believed that these two parts of

14:05

mathematics

14:06

demanded different accounts of the

14:08

grounds of our belief in them

14:10

his general remarks about geometry were

14:13

almost all for the purpose of

14:15

contrasting it

14:16

in this regard with arithmetic his

14:18

positive work

14:19

related entirely to the latter by

14:22

arithmetic

14:23

fragment number theory and analysis

14:26

the theory of natural numbers and the

14:28

theory of real numbers

14:30

he did not view set theory as a separate

14:33

mathematical theory but rather as a part

14:35

of logic

14:36

of which he believed that he could also

14:38

give a satisfactory account

14:41

and indeed if if the problem can be

14:44

solved for number theory

14:45

analysis and set theory there should be

14:48

little fear

14:49

that any other branch of mathematics

14:51

will give rise to further difficulties

14:54

fragra thought that the inability of

14:56

mathematicians to provide

14:59

a justification for our acceptance

15:02

even of number theory was a scandal to

15:05

the science

15:06

they could not even give an intelligible

15:08

account

15:09

of what the natural numbers are and

15:12

hence they could not so much as say what

15:14

number theory is about

15:15

let alone explain how we know that it is

15:18

true

15:19

his first step in carrying out his

15:22

self-appointed task

15:24

was to invent modern mathematical logic

15:29

i said earlier that fraga was the first

15:31

modern philosopher of mathematics

15:33

the title might be claimed by some for

15:36

but fragra differed from kant and from

15:39

all his predecessors

15:40

is in his determination to treat of the

15:43

mathematical

15:44

theories with which he concerned himself

15:47

in detail

15:48

rather than contenting himself with

15:50

observations which even if sound

15:52

had not been demonstrated to hold good

15:55

for all propositions of those theories

15:58

can't have maintained the a priori

16:01

character of mathematics

16:02

but to distinguish two varieties of our

16:04

priori truths

16:06

the analytic and the synthetic analytic

16:09

truths were those guaranteed by logic

16:11

alone

16:12

and where on can't view all trivial

16:15

they must be recognized immediately by

16:18

anyone who understood the words in which

16:20

they were expressed

16:21

and hence they could not extend our

16:23

knowledge

16:25

the truths of mathematics by contrast

16:27

were synthetic

16:28

and hence substantial the recognition of

16:31

geometrical truths

16:32

depended on our our intuition of space

16:36

and the various medical truths on our

16:38

intuition of time

16:40

our mathematical knowledge was

16:41

nevertheless our priori

16:43

because can't help that we possess

16:45

conceptions of space and time

16:48

independently of any particular

16:50

experience of

16:52

our priori intuitions freyja

16:55

accepted consternation of our our priori

16:58

intuition of space and with it

17:00

his view of geometry he thought that

17:03

while non-euclidean geometries are

17:05

logically consistent and tense

17:07

intelligible

17:08

we know our priori that the the geometry

17:12

of physical space

17:13

is euclidean but he utterly opposed the

17:16

kantian view of arithmetic

17:18

believing that neither temporal nor

17:21

spatial intuition

17:22

played any essential role in our

17:25

recognition of its truths

17:28

in this he was following the footsteps

17:30

of the

17:31

great czech mathematician and

17:33

philosopher boltzano

17:34

who died in the year of fragrance birth

17:38

bolsano had initiated the process of

17:41

rigorizing analysis

17:43

explicitly arguing the propositions

17:45

concerning real numbers

17:47

such as the mean value theorem ought not

17:50

to be accepted

17:51

in virtue of their apparent obviousness

17:53

to geometrical intuition

17:55

but could and therefore should be proved

17:58

in a purely arithmetical manner

18:02

as both sauna had striven to expel

18:04

appeals to intuition from analysis

18:07

fraga wished to expel them even from

18:09

number theory

18:11

since proof is the principal instrument

18:13

for establishing mathematical truths

18:16

he considered the process of

18:18

mathematical proof

18:20

must be subjected to scrutiny

18:23

do we know that all principles of proof

18:26

that we use in number theory

18:28

or that we need in order to establish

18:30

the basic number theoretic propositions

18:33

that we frequently take for granted

18:35

are of a purely logical character

18:38

as long as we reason without paying

18:41

conscious attention to the steps we take

18:43

in reasoning

18:44

we do not we may then mistake for

18:47

logical transitions

18:49

ones which in fact rely upon intuition

18:52

conversely we may suppose an appeal to

18:55

intuition to be demanded

18:57

by what in fact can be accomplished by

18:59

purely logical deduction

19:02

it was therefore necessary in fragrance

19:04

eyes to attain

19:05

an explicit systematization of the

19:08

process of mathematical

19:10

proof this he did in his first book the

19:13

gristrift of 1879

19:16

a completely original work that did not

19:18

build

19:19

on the recent advances in logic by bull

19:22

and his successors but adopted a wholly

19:25

new approach

19:27

not only have they extended the scope of

19:29

logic by only a small degree

19:32

beyond what aristotle had achieved but

19:35

they had provided no more than a means

19:37

of

19:37

encoding any given argument that fell

19:41

within the scope of their theories fraga

19:44

desired something different a language

19:46

in which mathematical theorems could be

19:48

expressed and their proofs carried out

19:51

in accordance with strictly formal

19:53

principles of inference

19:56

it is with hindsight astonishing that

19:59

despite an

20:00

intense study of logic over many

20:02

centuries indeed millennia

20:04

it failed to register the glaring fact

20:07

that logical theory was incapable of

20:09

analyzing

20:11

even the simplest piece of mathematical

20:13

reasoning

20:15

fraga's little book completely rectified

20:17

this

20:18

using a notation quite different from

20:21

but isomorphic with

20:22

that used nowadays or by any of his

20:25

successes

20:26

parts one and two of the book presented

20:29

a complete formalization

20:31

of what we now call first order logic

20:36

but part three

20:40

explored the realm of second order logic

20:42

also that is the logic

20:43

governing statements that generalize not

20:46

only over

20:47

objects for example the natural numbers

20:49

but over properties

20:50

of and relations between these objects

20:53

and here fragrance scored his first

20:56

success

20:57

in the war against intuition a logical

20:59

analysis

21:00

of the notion of a sequence rather

21:03

naturally it was common to think of

21:05

sequences in temporal terms

21:08

but frag observed that the notion was a

21:11

far

21:11

greater generality a sequence can be

21:14

generated by any relation

21:16

the term generated by i've just used

21:19

itself relies on a way of picturing

21:21

sequences

21:22

as proceeding in time and i should have

21:24

said something like

21:26

characterized by reference to

21:29

the relation r say r is that which any

21:32

term in the sequence

21:34

has to the next fragra defined the

21:36

expression

21:37

b follows a object b follows the object

21:40

a in the

21:41

r sequence to mean b has every property

21:45

f possessed by every object which either

21:48

a

21:49

or an object having the property f

21:51

stands in the relation

21:53

r um

21:58

this definition and the theorems

22:01

concerning sequences that fragra

22:04

proved by appeal to it were not

22:06

important to him

22:08

merely as samples of how one could

22:10

dispense

22:11

means with what was often thought to

22:14

rely upon intuition

22:16

it was important because the notion of a

22:18

sequence

22:19

is fundamental to the theory of numbers

22:22

since the natural numbers form precisely

22:24

such a sequence

22:26

given the number naught and the relation

22:28

that holds between any natural number n

22:31

and its successor n plus one the natural

22:33

numbers could be defined

22:35

to be those things that followed naught

22:37

in the successor sequence

22:39

together with naught itself moreover

22:42

such a definition

22:44

would render the principle of induction

22:46

its most immediate consequence

22:48

the principle allows something to be

22:50

proved to hold good of all natural

22:52

numbers

22:53

by showing the totals of naught and of

22:56

the successor of any number of which it

22:58

holds

22:59

and had frequently been said to be a

23:01

form of inference

23:02

peculiar to number theory but fragrance

23:05

definition of natural number

23:07

reduced it to pure logic

23:10

plainly investigation of what justifies

23:13

our accepting

23:14

mathematical theories other than

23:16

geometry must begin with the most basic

23:19

theory

23:20

number theory and the first step might

23:23

be expected to be

23:24

to axiomatize it it's odd that despite

23:28

the praise conventionally lavished on

23:30

euclid for

23:31

his axiomatization of geometry no one

23:34

attempted to do this for any

23:36

branch of arithmetic until the 19th

23:38

century

23:40

we can extract an axiomatization from

23:43

frager's work

23:44

more precisely an abstract

23:46

characterization

23:47

about which he proved any structuring

23:50

exemplifying it

23:51

to be isomorphic to the natural numbers

23:55

but he did not proceed in this way after

23:58

what he had achieved in his first book

24:01

it's unsurprising that he came to

24:03

believe that the whole of number theory

24:05

could be derived from logic alone in

24:08

1884 he published the foundations of

24:11

arithmetic

24:12

in which having subjected all existing

24:15

accounts to deadly criticism

24:17

he sketched his own demonstration of the

24:19

logical character of number theory

24:21

without using his logical notation

24:24

in 1893 and 1903 he published the first

24:28

and second volumes

24:29

of his basic laws of arithmetic which

24:32

set out

24:34

fully formalized proofs in his logical

24:36

system

24:37

both for number theory and for analysis

24:41

given fragra's view that all

24:44

arithmetical notions can be defined

24:46

in logical terms and all our

24:47

arithmetical propositions prove

24:50

from logical first principles the task

24:52

of

24:54

justifying arithmetic necessarily

24:56

involved for him

24:57

much detailed mathematical work of which

25:00

his books are full

25:01

largely neglected if only because very

25:04

few

25:04

were prepared to learn to read his

25:06

symbolism

25:08

but the task involved philosophical

25:10

argument also

25:11

if the mathematical proofs were to be

25:13

shown to establish what was claimed

25:16

and in the process of carrying it out

25:18

fraga turned himself into a philosopher

25:20

and one of genius as well as a

25:22

mathematician

25:24

although his philosophical work was

25:26

restricted in scope

25:28

he plainly became interested in the

25:30

topics he tackled

25:31

for their own sake so that his

25:33

discussions

25:34

particularly in the ancillary articles

25:37

that he wrote

25:38

extend well beyond what was strictly

25:41

necessary

25:42

for his primary objective it's this that

25:45

has made him have

25:46

made them of such vivid interest the

25:48

modern philosophers

25:50

not especially concerned with the

25:52

philosophy of mathematics

25:53

but that's not our pr our present

25:56

concern

25:57

what to my mind makes him of primary

26:00

importance to present day philosophy of

26:02

mathematics

26:03

is the clarity with which he presented

26:06

its problems

26:07

even when he did not attain more than

26:10

partial

26:10

solutions to them for these are often

26:13

problems

26:14

which later writers have scarcely

26:16

attempted to tackle

26:18

given his conclusion that analytic

26:21

truths cannot extend our knowledge

26:24

can't could could not regard

26:27

mathematical truths as analytic

26:29

since on the face of it mathematics

26:32

massively extends our knowledge

26:35

freyja having decided that arithmetical

26:38

truths are logical in character and

26:39

therefore analytic

26:41

was faced with explaining how it is that

26:44

analytic truth

26:45

can extend our love knowledge and this

26:48

is the same problem

26:50

as how deductive reasoning can lead to

26:53

new knowledge

26:54

for such reasoning to be valid a single

26:57

inferential step

26:59

must be both recognizable and compelling

27:03

anyone who understands the statements

27:05

figuring as premises and conclusion

27:08

must thereby grasp that acknowledging

27:10

the former is true

27:12

requires acceptance of the latter but if

27:15

this

27:16

is so how can a sequence of such steps

27:19

take you any distance from where you

27:22

started

27:23

the problem has been addressed by few

27:25

philosophers other than fragra notably

27:28

by mill

27:28

and none has offered a satisfactory

27:31

solution

27:33

if fragrance solution is not correct in

27:35

detail

27:36

i'm not going to give the detail it must

27:38

i believe be correct in outline

27:42

for fragrance not merely to discover a

27:46

deductive argument but even to

27:48

follow it is to engage in a creative act

27:52

that is because it involves more than

27:54

merely

27:55

understanding the statements involved

27:58

to understand the statement to grasp the

28:00

thought it expresses

28:02

in his terminology it's necessary to

28:05

apprehend its structure how it's put

28:07

together

28:08

out of its paths a sentence is not just

28:11

a string of words

28:13

but to recognize the validity of an

28:15

argument in which it figures

28:17

we need to do more than this sentences

28:20

exhibit patterns to use a term fragrant

28:23

did not

28:24

imply that we do not need to discern

28:28

in order merely to know what they say

28:31

in it's on the relationship between

28:33

these patterns

28:35

that the validity of an inference

28:36

depends

28:38

a pattern is not imposed but is there to

28:41

be discerned

28:42

but discerning it is an intellectual act

28:45

and not

28:45

mere passive reception

28:50

fraga is sometimes supposed to have

28:52

thought that arithmetic

28:54

describes the plutonic realm of abstract

28:57

objects

28:58

having no con connection with contingent

29:01

reality

29:02

such a view would provide no room for

29:05

the application of arithmetic

29:07

though in fact even plato's ideas bear a

29:10

direct relation

29:11

to the empirical world but the

29:13

misrepresentation of freyja

29:15

is nearly total the real fragra

29:19

gave a more central role to application

29:22

than almost any other philosopher of

29:25

mathematics

29:26

the applicability of mathematics is

29:29

nowadays often described as miraculous

29:32

fragra would have denied not merely that

29:34

it was a miracle

29:36

but that it was even a surprise it is

29:39

applicability alone he wrote that raises

29:42

arithmetic from the rank of the game

29:45

to that of a science his view of

29:48

application was a subtle one

29:51

arithmetic must remain pure any

29:54

importation of notions relating to

29:57

particular empirical applications would

29:59

sully that purity

30:01

and he criticized both mill and

30:03

helmholtz

30:04

on this ground but the general principle

30:08

exemplified by every application of a

30:11

fundamental

30:12

mathematical notion and in particular by

30:14

the notions of the natural numbers and

30:16

of the real numbers

30:18

must be discerned formulated and made

30:21

central to its definition not tagged on

30:24

as an appendage

30:25

as dedicated did in his treatise only in

30:28

his treaties

30:29

on the natural numbers rather

30:32

the general principle in accordance with

30:34

which all applications are made

30:36

must be incorporated in the definition

30:40

that's why fraga held that since the

30:43

fundamental use

30:45

of the natural numbers is to specify in

30:48

finite cases

30:49

how many objects there are that satisfy

30:52

some given condition

30:53

they must be defined as cardinal numbers

30:56

not indeed objects of any specific kind

31:00

but of objects of any kind whatever

31:02

since all objects can be counted

31:05

and it's also why frago regarded the

31:08

definitions of the real numbers

31:10

given by kanto and dedekind as

31:13

unsatisfactory

31:15

though their definitions were different

31:17

both assumed the rationals

31:19

as already given whereas on fragra's

31:22

view

31:23

the term number is used in just the same

31:26

sentence in rational number as in real

31:29

number

31:30

so that the rationals ought not to be

31:32

defined separately

31:33

in advance of the real numbers but

31:36

simply as a particular type of real

31:38

number

31:39

by contrast the term number has a

31:42

different sense

31:44

in natural number making it necessary

31:47

to define the natural numbers separately

31:50

the natural numbers serve to answer the

31:53

question

31:53

how many so they together with the

31:56

trans-finite cardinals

31:58

are numbers in one sense of the word the

32:00

real numbers

32:01

including the rational ones serve to

32:03

give the ratio of a quantity

32:05

a mass a temporal duration an electric

32:08

charge etc

32:10

to a unit quantity of the same kind and

32:13

are therefore numbers in a different

32:14

sense they must therefore be defined as

32:18

such ratios not between quantities of

32:20

any specific kind

32:22

but between any two quantities of the

32:25

same kind

32:25

that of course requires a prior

32:28

analysis of the notion of a quantity or

32:32

a range of quantities i've been

32:34

describing a number of near-misses

32:38

fragrance account of the fruitfulness of

32:40

deductive reasoning

32:42

must be close to the truth but i don't

32:44

think

32:45

it was the precise truth his emphasis

32:48

on the applicability of mathematics was

32:52

surely correct

32:53

but his demand on the way mathematical

32:55

concepts are defined

32:57

is difficult to satisfy he was clearly

33:00

right

33:01

to maintain the necessity of

33:03

arithmetical truth

33:05

and to deny that this would derive from

33:07

intuition in can't sense

33:09

but from features of concepts sharing

33:12

with logical ones their applicability to

33:15

any domain of discourse

33:17

even if he was wrong to claim them to be

33:20

strictly logical in character

33:23

the most interesting of all the

33:25

components of

33:26

his philosophy of arithmetic is however

33:30

the most contentious it attempts to

33:32

resolve a problem

33:34

that hardly any of his successes has

33:37

faced

33:38

it's not a single thesis but a complex

33:40

of ideas

33:42

that can with difficulty be disentangled

33:44

from one another

33:46

and it's quite certainly wrong as a

33:49

whole

33:50

because it led fraga into the

33:51

catastrophe that caused him to

33:53

acknowledge

33:54

that his life's work had been a failure

33:59

someone who knew nothing about fragra

34:02

but learned that he believed

34:03

arithmetical theorems to be expressible

34:06

in purely logical terms and provable

34:08

from purely logical principles

34:11

would expect on reflection that he did

34:13

not take statements of arithmetic

34:15

at face value the statement

34:19

pi is transcendental is on the face of

34:21

it of the same

34:22

form as clinton is unsuccessful

34:26

and says that a particular object of a

34:29

certain kind

34:30

has a specific property well there's a

34:33

prime number between 44 and 52 appears

34:36

to be of the same form as

34:38

there's a small country between france

34:40

and spain

34:41

and says that some object of a certain

34:43

kind has a particular property

34:45

and stands in a specific relation to two

34:48

named objects of the same kind

34:51

but statements of forms such as these

34:54

are hardly candidates for being

34:55

propositions of logic

34:57

for the principles of logic must hold

35:00

independently of which

35:02

particular object or even of how many

35:04

objects

35:05

there may happen to be so it appears to

35:08

follow

35:09

that fraga did not think that

35:11

arithmetical statements are truly

35:13

of the form they appear to be he must

35:15

rather have construed them as disguised

35:18

versions of statements of very different

35:20

forms

35:21

for instance reinterpreting any

35:23

proposition referring

35:25

to a natural number n as involving the

35:28

statement

35:28

that there are endings of some kind

35:32

well this conclusion is highly

35:34

reasonable but it is

35:36

wrong fraga did understand arithmetical

35:40

statements at face value

35:42

numbers were in his few objects

35:45

it followed that logic must after all

35:48

guarantee the existence of certain

35:50

abstract objects

35:51

indeed of infinitely many of them

35:54

numbers

35:54

are what he called logical objects

35:58

now this didn't to any significant

36:00

extent

36:01

hamper his current account of the

36:03

application of arithmetic

36:05

mathematical theorems were for him

36:07

encapsulated results of complex chains

36:10

of

36:10

deductive reasoning enabling us by

36:13

specializing them to less general cases

36:16

to pass from contingent premises to

36:18

contingent

36:19

conclusions without having to carry out

36:22

the reasoning process afresh

36:25

the existence of the objects

36:27

constituting the elements of the

36:29

structure which each particular theory

36:31

number theory or analysis treated was

36:34

guaranteed by logic

36:35

and in this sense the purity of

36:37

arithmetic was secured

36:39

but they were not pure in the sense in

36:42

which set theorists

36:43

nowadays speak of pure sense a pure set

36:47

in this sense is one whose

36:49

transitive closure contains nothing but

36:51

sex

36:52

that's to say every element of the set

36:54

is a set and every element of an element

36:56

of the set is a set

36:58

every element an element of an element

37:00

of the set is a set and so on

37:02

on the contrary for fragrant natural

37:04

numbers are cardinal

37:07

and the cardinal number n is the class

37:09

of all classes

37:11

with just n members since that includes

37:14

classes whose

37:15

members are empirical objects there's no

37:18

problem

37:18

how theorems about natural numbers can

37:21

be applied

37:22

to empirical circumstances likewise real

37:25

numbers where for him

37:26

ratios between quantities and such

37:29

quantities include

37:30

physical magnitude such as mass and

37:32

length

37:33

the theory was designed to be directly

37:35

applied and the fact

37:37

that the numbers with which it deals are

37:39

taken to be objects

37:40

in no way hinders its applicability

37:43

what is problematic is how the existence

37:46

of logical objects can be justified

37:50

logical objects form a special class of

37:53

abstract objects

37:54

called by fragra non-actual objects

37:57

because they don't

37:58

act on other objects or bring about

38:00

effects in them

38:01

as physical objects do and are therefore

38:04

not perceptible by the senses

38:07

his account of what we do when we refer

38:10

to non-actual objects

38:12

was based on his celebrated context

38:14

principle

38:16

this principle says that it's only in

38:18

the context of a sentence

38:20

that we can refer to anything otherwise

38:23

expressed

38:23

we can't mention anything without saying

38:27

save as part of a process of saying

38:30

something about it

38:34

expressing a thought about it not

38:36

necessarily asserting anything about it

38:38

this principle is the most profound the

38:40

most difficult

38:41

and the most contentious ingredient in

38:44

all fragrance philosophy

38:47

philosophers who deny the existence of

38:49

abstract objects are labeled

38:51

nominalists they usually characterize

38:54

abstract objects precisely

38:56

by their not being actual in fragrance

38:58

sense that is by their

38:59

lack of causal powers and the standard

39:02

nominalist argument against their

39:04

existence

39:05

is that since they cannot affect

39:07

anything everything must appear exactly

39:10

the same if they do not exist

39:12

as if they do and hence that we can have

39:15

no reason

39:16

to suppose them to exist

39:19

well the equator is cited by fraga

39:22

as a non-actual object so suppose you're

39:25

in an airplane and you remark to your

39:27

neighbour

39:28

that the plane has just crossed the

39:29

equator

39:31

to your surprise he doesn't know what

39:32

the term equator

39:35

he doesn't know the term equator and

39:36

asks you what it means

39:38

you try to explain to him and he asks

39:40

whether you can see the equator

39:42

or feel the equator when you tell him

39:44

that it's not that sort of thing

39:46

he asks what reason you have to suppose

39:49

that there's any such object

39:51

seeing that everything will be the same

39:52

exactly the same if they were not

39:56

you can do no more than patiently

39:58

explain to him

39:59

how sentences containing the term the

40:02

equator

40:03

are used and in particular how we judge

40:06

of their truth and falsity

40:07

and according to fraga in order to

40:10

justify the use of the term

40:12

the equator you need to know more than

40:15

that

40:16

this is what is meant by saying that

40:18

it's only in the context of a sentence

40:20

that we can refer to an object or at

40:22

least to an abstract object

40:24

and we may add what's intended to be

40:27

part of the content

40:28

of the context principle that when we

40:31

know

40:32

how sentences mentioning such an object

40:34

are used

40:35

we thereby know what it is to refer to

40:38

that object

40:40

well in my view this is a wholly

40:42

satisfactory vindication

40:44

of the general vindication of the use of

40:46

terms

40:47

referring to abstract objects and there

40:49

can

40:50

therefore doubt that the context

40:52

principle

40:53

is in broad outline sound but we should

40:56

not

40:57

jump to the conclusion that abstract

40:59

objects present no further difficulties

41:01

and fraga did not jump to that

41:03

conclusion

41:04

to justify the use of abstract terms of

41:08

any

41:08

given kind in the light of the context

41:10

principle

41:11

we must be able to do what that

41:13

principle demands that is

41:15

to explain without circularity the use

41:18

of sentences containing such terms

41:20

and the conditions of their truth and

41:22

falsity and this is by no means always a

41:25

simple task

41:28

why did fraga require that numbers be

41:30

recognized as objects

41:32

to use the for example the sorry

41:36

pi is transcendental should be treated

41:38

as being of the same form as

41:40

clinton is unsuccessful his primary

41:43

reason was to guarantee the existence of

41:46

sufficiently many

41:47

elements of each theory for all possible

41:50

applications of it while preserving the

41:53

purity of arithmetic

42:05

it's easy to imitate fragrance

42:07

constructions at a higher level for

42:09

example take

42:10

cardinal numbers as properties of

42:12

properties of objects so

42:13

they're no longer themselves construed

42:17

as objects and that's essentially what

42:19

russell whitehead did

42:20

in principia mathematics matica taking

42:23

properties

42:24

for this purpose extensionally the

42:27

drawback of doing this is that one can't

42:29

guarantee

42:31

then guarantee that there are infinitely

42:33

many natural numbers

42:35

the solution adopted by russell

42:37

whitehead

42:38

was to assume an axiom stating that

42:41

there are infinitely many individuals

42:43

a proposition certainly not a logical

42:46

truth and

42:46

dubiously drew true at all and thereby

42:50

the entire project

42:51

of deriving arithmetic from logic was

42:53

abandoned

42:54

similar difficulty arose with the real

42:56

numbers

42:58

it was by taking numbers to be objects

43:00

that frag was able to circumvent this

43:03

problem

43:04

he insisted on the generality of the

43:07

notion of cardinal number

43:08

objects of all kinds can be counted and

43:11

among things that can be counted

43:13

are numbers themselves as when we speak

43:17

of the number of roots of an equation

43:20

or of prime numbers less than or equal

43:22

to a given number

43:24

so numbers must be objects since a

43:27

cardinal number is always the odd number

43:29

of

43:29

objects satisfying some given condition

43:33

and this allowed trigger to prove the

43:35

existence

43:36

of infinitely many natural numbers

43:39

independently

43:40

the existence of objects of any other

43:43

kind

43:45

we can show a number n to exist by

43:48

producing a predicate true of just

43:50

n objects so the number not exists

43:54

since is different from itself is true

43:56

of not objects

43:58

and so is the number naught is true of

44:00

just one object

44:02

and hence the number one exists and

44:05

is the term of the sequence not one is

44:08

therefore true

44:09

of just two objects and so the number

44:11

two also exists

44:12

and in general given the existence of

44:15

the numbers from not to n

44:17

the number n plus one must exist since

44:20

it's the number of

44:21

terms in that sequence this is a

44:24

informal sketch of the theorem fraga

44:26

proved really

44:28

rigorously from his assumptions the

44:31

assumptions are easily stated

44:33

natural numbers were to be treated as

44:35

cardinal numbers

44:37

and the basic notion for the theory of

44:39

cardinality

44:40

is that expressed in natural language by

44:42

saying that there are just as many

44:44

objects of one kind

44:45

as of another frago adopted as a

44:48

definition of this notion

44:50

one that had recently been accepted by

44:53

other mathematicians of his day

44:55

namely the existence of a relation

44:58

mapping

44:58

the objects of the one kind one to one

45:01

onto those of the other

45:03

he tacitly assumed that every term

45:06

standing for a cardinal number

45:08

could be framed by means of the operator

45:11

the number of objects which

45:13

so given that by appending

45:16

any well-defined predicate to this

45:18

operator one would obtain a term

45:20

standing for an object and given the

45:23

definition of justice many

45:25

fragra formulated a fundamental

45:27

equivalence

45:28

namely the number of objects of one kind

45:31

is the

45:32

same as the number of objects of another

45:34

if and only if there are just as many

45:36

objects of the one kind

45:38

as of the other this fundamental

45:41

equivalence

45:42

looks at first sight tautologous but it

45:44

isn't it's a principle

45:46

governing the introduction of terms

45:48

standing for numbers

45:49

the right hand side says nothing about

45:52

any such objects as numbers

45:54

the left hand side says that the numbers

45:56

denoted by

45:57

two basic terms for them coincide

46:01

now against the that background fragra

46:04

proved

46:05

by means of suitable definitions that

46:08

all the basic principles of number

46:10

theory

46:11

could be derived by means of second

46:13

order logic

46:14

from this fundamental equivalence

46:18

so a justification for the fundamental

46:20

equivalence is required

46:23

if the introduction of terms for

46:28

cardinal numbers is to be defended by

46:30

appeal to the context principle

46:32

we need to show that we succeeded in

46:34

specifying

46:35

the condition for the truth of any

46:37

sentence containing such

46:38

terms and fragra discussed whether the

46:41

fundamental equivalence

46:43

could itself be regarded as affecting

46:45

this

46:46

well some present day philosophers and

46:49

mathematics

46:50

enthusiastically answer yes to this

46:53

question or

46:54

for some rather similar question but

46:56

fraga's own answer

46:57

was no so he resorted

47:01

to his definition in terms of classes

47:03

the number of objects of a given kind

47:06

is the class of classes a such that just

47:09

as many objects

47:10

of that kind as members of a

47:13

the only use he made of this definition

47:16

was to derive the fundamental

47:18

equivalence from it

47:19

its purpose was simply to introduce

47:22

terms for numbers in a way

47:24

he considered unexceptionable

47:28

his ground for denying that the

47:30

fundamental equivalence

47:31

served to do what the context principle

47:35

required

47:36

was that it failed to determine the

47:38

condition for the truth or falsity

47:41

of a statement of identity between a

47:43

number and an

47:44

object denoted by a term not given as a

47:47

number not

47:48

formed by means the operator the number

47:50

of objects

47:52

the objection is sound but he overlooked

47:54

a far more basic one

47:56

namely that the fundamental equivalence

47:58

does not even determine

48:00

the truth or falsity of every statement

48:03

of identity between numbers

48:06

it doesn't for instance give us any

48:09

means of deciding

48:10

whether the number of natural numbers is

48:12

or is not the same

48:14

as the number of all cardinal numbers or

48:16

of all objects whatever

48:19

now consider what frag was about he was

48:22

trying to justify the introduction of

48:24

terms for cardinal numbers

48:27

satisfying the fundamental equivalence

48:30

the numbers denoted by those terms were

48:33

then to be treated as belonging to the

48:35

same domain of generality

48:37

as all other objects those to be covered

48:39

by the quantifiers

48:41

for every object x or there is an object

48:43

x

48:44

and in accordance with this the terms

48:46

being introduced

48:47

were to include ones for the number of

48:50

cardinal numbers of some

48:52

given kind as specified by means of the

48:55

some predicate numbers it's essential

48:58

for the proof of the infinity of the

48:59

natural numbers and all this was to be

49:02

done

49:03

without first specifying of what

49:06

objects the domain of generality was to

49:09

consist

49:10

the procedure has a troubling

49:12

circularity

49:13

which objects there are depends on which

49:16

numbers there are

49:18

but which numbers there are depends on

49:20

which objects

49:21

the domain contains

49:24

well what do we say about this situation

49:28

should we say as the philosophers i

49:30

mentioned believe that the conditions

49:32

required for appeal to the context

49:34

principle were too stringent

49:37

must need not determine the truth value

49:39

of every statement

49:41

if so just how could they be weakened

49:44

without destroying the plausibility of

49:47

the principle

49:48

or should we say that the whole

49:50

procedure is misconceived

49:53

the usual conception of how an

49:56

interpretation of a formal theory

49:58

should be laid down is that one should

50:00

begin

50:01

by specifying the intended domain over

50:04

which the variables that arrange

50:06

and then interpret with reference to

50:09

that

50:09

the expression special to the theory

50:12

saying for example of which elements of

50:14

the domain some given predicate

50:16

is to be true so according to this

50:19

conception it's impossible

50:21

simultaneously to determine the domain

50:24

and the interpretation of the symbols

50:27

intended to denote

50:28

elements of that domain but if that's

50:31

the only way in which one can go about

50:34

laying down how a mathematical theory is

50:36

to be understood

50:37

we can never explain how a fundamental

50:41

theory

50:41

such as number we can never explain a

50:44

fundamental theory

50:45

such as number theory or as analysis by

50:48

a fundamental

50:49

mathematical theory i mean one whose

50:51

elements cannot be defined as resulting

50:54

from some simple operation on those of

50:57

another theory

50:58

and this amounts in practice to one with

51:00

a greater number of elements

51:02

than any prior theory the natural

51:05

numbers form the prototype

51:07

but innumerable totality one all whose

51:10

elements can be

51:11

generated as the terms of an infinite

51:14

sequence

51:15

it's by grasping the conception of the

51:18

totality of

51:19

natural numbers that we first come by an

51:21

understanding of the phrase

51:23

infinitely many likewise the

51:26

real numbers form the prototype of a

51:28

larger infinite totality that is a

51:32

non-innumerable one with too many

51:34

elements for them to be generated

51:36

as the terms of a sequence our problem

51:39

is to explain

51:40

how we first attain a conception of a

51:42

totality

51:43

the one or the other size cardinality

51:47

so long as these theories are presented

51:49

to us before any others with domains of

51:52

these sizes

51:53

we have no means of specifying their

51:56

domains

51:56

in advance of expounding the

51:58

interpretation of the theory as a whole

52:01

and yet we do come to understand

52:04

it seems that fraga must have been right

52:07

in thinking that for these fundamental

52:10

theories

52:10

we have to acquire a conception of the

52:13

domain

52:14

simultaneously with that of the meanings

52:16

of the basic notions defined

52:18

over it and yet such a process

52:22

appears doomed to vicious circularity

52:26

it's one of the great merits of

52:28

fragrance philosophy of arithmetic

52:30

that it faces this difficulty squarely

52:33

even if he's attempted to solution up to

52:36

it failed

52:37

he addressed the problem that's usually

52:39

ignored

52:40

but cannot be evaded his solution did

52:43

indeed fail

52:44

catastrophically i've discussed the

52:47

point is it applies to the cardinal

52:49

numbers taken as introduced by

52:51

means of the fundamental equivalence to

52:54

make clear that it doesn't relate either

52:56

to the consistency

52:57

or to the power of the method of

52:59

introduction but fraga rejected that

53:02

method and gave instead his definition

53:04

in terms of classes

53:06

but that just shifted the problem to the

53:09

justification for introducing classes

53:12

this he affected by a method precisely

53:15

analogous to the use of the

53:17

fundamental equivalence he treated the

53:20

operator the class of objects which

53:22

as a primitive symbol forming a term

53:25

standing for an

53:26

object whenever supplemented by a

53:28

predicate of objects

53:30

lay down an axiom stating that the class

53:32

of objects

53:33

satisfying any given condition is the

53:36

same as the class of those satisfying

53:37

some other conditions

53:39

just in case every object that satisfies

53:41

other condition

53:42

also satisfies the other the only

53:45

difference

53:46

was that in this case he had a

53:48

supplementary stipulation

53:50

to handle the case he believed to give

53:53

rise to the only problem

53:54

that of a statement equating a class

53:57

with some object

53:58

not given as a class and notoriously

54:01

this axiom rendered his system

54:04

inconsistent

54:05

it yielded the celebrated paradoxes of

54:08

said theory

54:09

and after struggling to escape this

54:11

calamity

54:12

fraga accepted that his entire life's

54:15

work

54:15

had failed

54:18

in setting out the problem i've tacitly

54:21

relied on an

54:22

attitude towards mathematical objects

54:25

that we naturally have

54:26

but seldom remark on fragra took for

54:30

granted that the logic

54:32

appropriate to mathematical theories is

54:34

that we now call

54:35

classical one that assumes every

54:38

statement

54:38

to be determinantly either true or false

54:41

and considers the truth or falsity of a

54:43

complex statement

54:45

to depend only on the truth or falsity

54:47

of its constituent sub-statements

54:49

or in the case of a universal

54:51

generalization or existential statement

54:54

of its instances this assumes

54:58

that the operations of generalization or

55:00

existential quantification

55:02

will preserve determinateness of truth

55:05

failure

55:06

that is we shall always obtain a

55:08

statement determinantly true or false

55:11

by attaching say every number or there

55:13

is a number which

55:15

to a predicate definitely true or false

55:18

of any specific number

55:20

well in just consider the empirical case

55:24

what then do we require of a domain of

55:27

generality

55:28

if this is to be so if that's to be

55:30

enough

55:31

to guarantee determinateness of truth

55:33

value we

55:34

normally assume it's sufficient that the

55:37

concept by means of which we specify the

55:39

domain

55:40

should have determinate application and

55:43

determinant conditions for identity

55:46

to guarantee a definite value true or

55:48

false

55:49

for every statement about all stars or

55:51

all books

55:53

and every statement to the effect that

55:55

there is a star or a book of a certain

55:57

kind

55:58

we take two conditions two things to

56:01

suffice

56:02

that the concept star or book should

56:04

have a quite precise application

56:06

no borderline cases and that it should

56:09

be definite what counts as the same

56:11

style

56:11

or book we don't need in addition

56:16

to lay down what stars or books there

56:19

are

56:20

reality does reality does that

56:23

for us now this assumption may be

56:26

challenged i'm not concerned with that

56:28

i'm concerned with the contrast to how

56:30

we normally think about mathematical

56:32

objects

56:34

to endow every statement about all real

56:37

numbers

56:37

or asserting the existence of a real

56:39

number of a given kind

56:41

with a definite truth value we don't

56:43

normally think it

56:44

enough to lay down what's the count as a

56:47

real number

56:49

following dedicant we might do that by

56:51

requiring

56:52

requiring it to have a determinant

56:54

relation of magnitude to every rational

56:57

but that would

56:58

merely tell us how to recognize a real

57:01

number

57:01

when presented with one it doesn't tell

57:04

us

57:05

what real numbers there are and it would

57:07

need a very robust realism about

57:10

mathematics

57:11

to think that that could be left the

57:13

mathematical reality

57:15

to determine normally we think

57:18

that that's something that we have

57:22

by some means or other to circumscribe

57:28

fraga might be accused of having

57:32

believed that it was

57:35

her classes or cardinal numbers could in

57:37

general be specified

57:38

and the conditions under which two such

57:40

specifications

57:42

determined the same class or number but

57:44

that would be unfair

57:45

he asserted a need to supplement any

57:48

such account by

57:49

certifying that it did confer on every

57:51

statement of the theory

57:53

a definite value true or false but

57:55

unhappily his attempted proof of this

57:58

for the theory with classes was

58:00

fallacious

58:02

and because of this he left unresolved

58:04

the problem

58:05

to which he thought he had found the

58:07

solution

58:09

my own view is that as it stands it

58:12

cannot be solved

58:14

fragrant sought a justification of our

58:16

arithmetical theories satisfying three

58:18

criteria

58:20

it must accord to arithmetic the status

58:23

of a science that is a body of truths

58:26

it must exhibit it as apt for

58:28

application to empirical reality

58:31

while not itself invoking any empirical

58:34

notions

58:35

or once derived from spatial or temporal

58:37

intuition

58:39

and it must leave intact the classical

58:41

theories

58:42

including the classical canons of

58:44

mathematical reasoning

58:46

it's probable that no account of what

58:49

justifies arithmetic

58:50

can satisfy all three criteria

58:54

no means exists to circumscribe the

58:56

totality of real numbers

58:58

without circularity and in so definite a

59:01

manner

59:02

as to warrant confidence that every

59:04

general statement about them

59:06

has a definite truth value it wouldn't

59:09

even be any use to relinquish the

59:11

requirement of the purity of arithmetic

59:14

the classical continuum cannot be

59:16

derived from physical reality

59:18

as we experience it but is rather a

59:20

concept

59:21

generated within mathematics and imposed

59:24

by us

59:24

in thought upon physical reality

59:28

rather we should see the totalities real

59:30

numbers

59:31

as an immediately indeterminate one

59:34

we can prove some statements about all

59:36

real numbers on the basis of the

59:38

knowledge

59:39

of what must hold good of anything for

59:41

it to be a real number

59:43

and we can prove some statements that a

59:45

real number of a certain kind exists

59:47

by finding a way to construct one but we

59:50

do not have

59:51

so sharp a conception of the totality as

59:54

to justify

59:55

our assuming every statement of either

59:57

kind to be true or false independently

1:00:01

of our being able to prove or refute it

1:00:04

and if that's right we are not entitled

1:00:06

to reason

1:00:07

in accordance with the canons of

1:00:08

classical logic

1:00:10

a weaker logic that we should be

1:00:12

justified in using has long been in

1:00:14

existence

1:00:15

so-called intuitionistic logic used by

1:00:18

constructive mathematicians most

1:00:21

mathematicians are reluctant to restrict

1:00:24

the allowable

1:00:25

methods of mathematical proof in this

1:00:27

way because they value the

1:00:29

power of classical reasoning but there's

1:00:32

no real

1:00:33

merit in presenting mathematical results

1:00:36

in a form more course than a careful

1:00:39

attention to the meanings

1:00:40

that can be legitimately attached to

1:00:43

them would warrant

1:00:44

it may well be that a version of

1:00:47

analysis

1:00:48

purified in this way would prove better

1:00:50

adapted to what fraga prized

1:00:53

as the basis for accounting it with

1:00:55

science

1:00:56

its application to the physical world

1:01:12

thank you michael very much um when i

1:01:14

was um

1:01:15

[Applause]

1:01:17

when i was preparing an introduction for

1:01:18

michael i ran into a puzzle

1:01:20

i wanted to try to convey uh

1:01:24

clearly very briefly totally

1:01:26

non-technically

1:01:27

the kind of work that michael did and i

1:01:29

found that

1:01:30

a lot of the things i was noting down

1:01:33

sounded very much like wittgenstein

1:01:35

so a question what's the

1:01:38

difference between michael and

1:01:41

wickenstein and michael has

1:01:42

i think himself told us

1:01:46

given us the slogan in his own lecture

1:01:49

there are of course

1:01:50

myriad differences in detail but

1:01:54

what's significant beyond that is the

1:01:56

very existence

1:01:58

of the detailed fully articulated

1:02:02

and carefully elaborated account um

1:02:06

michael has told us that what

1:02:08

distinguished

1:02:10

fragra from kant was fregger's

1:02:12

determination

1:02:13

to work through his problems in detail

1:02:17

and that is indeed as i believe this

1:02:20

lecture has given us some sense of

1:02:22

what singles out michael himself and

1:02:24

makes him the truly great philosopher

1:02:26

that he is

1:02:27

and i'm very proud to have him here for

1:02:29

this lecture thank you michael

1:02:36

now we've had a

1:02:40

we've had a bit of a gap uh we usually

1:02:42

try to leave a moment or two for people

1:02:44

who

1:02:44

aren't able to stay a little longer for

1:02:48

a question period but

1:02:51

in a moment if it's all right with you

1:02:52

michael we'll have some we'll have some

1:02:54

questions

1:02:55

uh followed then by the award uh

1:02:58

presentation of the award

1:02:59

uh later

1:03:05

oh no it was this beautiful making clear

1:03:09

for the people who don't know about it

1:03:12

yes i mean it was the

1:03:13

you did the balance i thought wonderful

1:03:16

oh

1:03:16

thank you very much because it's

1:03:18

terribly yes it's

1:03:20

it's terribly complicated and one

1:03:21

doesn't want to say what's not exactly

1:03:23

right

1:03:23

sure

1:03:36

i wondered if at some point i should

1:03:38

bring it over i thought well that's a

1:03:39

little bit

1:03:39

distracting it also looks pointy

1:03:45

on the other hand when i heard your

1:03:46

coughing yes well

1:03:54

i'm sorry it was a bit long no no we

1:03:56

didn't it's they take an hour and it

1:03:58

took an hour

1:03:59

yes yeah right now that's uh

1:04:04

just yesterday can we start please i

1:04:07

mean

1:04:07

if people would be a little quieter and

1:04:09

i think you need all this time to

1:04:12

get yourselves organized well it's

1:04:14

really these people though who are

1:04:15

someone else oh yes david papanowa

1:04:36

you said it wasn't there wasn't just a

1:04:38

problem

1:05:07

no that's the whole point of this

1:05:09

example that

1:05:10

it's certainly perfectly it's even

1:05:13

provably consistent if you

1:05:15

take the fundamental equivalent what i

1:05:17

call fundamental equivalence

1:05:20

uh as an axiom and with that definition

1:05:24

of just as many as a

1:05:25

one-on-one mapping it's certainly

1:05:28

uh in a setting of uh of second order

1:05:34

logic

1:05:36

derivation allows you to get the whole

1:05:38

arithmetic by that way and

1:05:40

the theory is provably equivalent that's

1:05:43

something

1:05:44

not equivalent that big or fun

1:05:45

consistent so the objection is not

1:05:47

consistent

1:05:49

in inconsistency the objection is

1:05:52

what the context principle demands

1:06:00

the introduction of a range of abstract

1:06:03

terms

1:06:04

is supposed to be justified providing

1:06:06

that you can lay down

1:06:08

what the truth values are to be of

1:06:12

um senses containing that type

1:06:15

the truth varies or in the case that

1:06:19

reference the empirical world is

1:06:21

necessary at least the truth conditions

1:06:23

right and i just remark

1:06:27

it the problem

1:06:31

of identity statements which are what

1:06:35

are directly dealt with

1:06:37

by the fundamental equivalent doesn't

1:06:39

just arise for once with

1:06:41

numerical terms on one side and some

1:06:43

non-numerical terms on the other

1:06:45

which is what pregame makes all the fuss

1:06:47

about but

1:06:48

even for cases when there are numerical

1:06:51

terms on both sides

1:06:53

how are we to decide i mean quite

1:06:55

obviously

1:06:57

the unless we know a lot more about what

1:07:00

there is in the domain of the

1:07:02

what objects there are we can't

1:07:05

possibly decide uh

1:07:08

identity statements of the kinds i

1:07:11

studied for example

1:07:12

the number of natural numbers equals the

1:07:15

number of cardinal numbers

1:07:18

uh so this

1:07:23

so there are there's a choice here

1:07:26

either the context principle

1:07:28

only justifies the introduction of these

1:07:31

abstract terms if it really does

1:07:33

settle i'll show you how to settle in

1:07:36

the case of

1:07:37

things involving empirical matters the

1:07:40

truth of

1:07:41

every statement in containing one of

1:07:43

these terms

1:07:45

or we've got to place much

1:07:52

less honoris requirement on what has to

1:07:55

be done

1:07:56

in order to justify the introduction of

1:07:59

abstract terms

1:08:00

by appeal to the context principle i

1:08:02

don't know exactly what that

1:08:04

ought to be or what would make it

1:08:07

compelling so fraga doesn't

1:08:11

acknowledge this fact that's all i

1:08:14

complained of

1:08:33

ability of mathematics to uh to heal

1:08:36

nature

1:08:38

and then we have a a long story about

1:08:40

numbers

1:08:41

but the in my domain where one hears

1:08:45

this

1:08:45

is about much of different kinds of

1:08:47

branches of mathematics like

1:08:49

um group theory isn't it amazing that

1:08:53

uh the elaborate mathematics of group

1:08:56

theory has any bearing whatsoever on

1:08:58

reality or hilbert's faces yeah so i

1:09:00

wonder if uh

1:09:02

is there a connection that you can trace

1:09:04

briefly for us to

1:09:06

how to think about the other branches of

1:09:09

mathematics

1:09:11

that get used in modern physics well i'm

1:09:13

not sure that i can

1:09:14

but i mean

1:09:17

what i gave is the recipe

1:09:21

for fragrance says there is

1:09:27

there is a question about

1:09:31

applicability let's say some theory

1:09:33

ought to explain

1:09:35

how mathematics gets applied

1:09:40

and the only theory that can explain

1:09:42

this is mathematics itself otherwise he

1:09:44

says the question falls

1:09:46

into the void right the philosopher the

1:09:49

physicist

1:09:50

uh not my business and the mathematician

1:09:53

says it's not his

1:09:55

uh in the case of groupthink i can't

1:09:59

answer this very well the uh

1:10:07

when you introduce the notion of groups

1:10:10

you're

1:10:12

normally introduced to it via the notion

1:10:14

of a transformation of some

1:10:16

kind of transformations closed under

1:10:19

obvious operations uh under

1:10:23

composition in inverse

1:10:27

and when that's not enough

1:10:33

the the whole idea of the vrega has

1:10:37

and i said it's very difficult to apply

1:10:40

is

1:10:40

you should look to see what is the most

1:10:44

general characteristic of the theory

1:10:46

that enables

1:10:47

one to apply and then you should define

1:10:50

the notions in those terms

1:10:54

that's not very easy to do i don't even

1:10:57

know whether it's practicable

1:10:59

but uh well you can say whether

1:11:02

whether you think that's

1:11:07

look group the very sophisticated theory

1:11:09

but it starts with this

1:11:11

notion of transformation and that

1:11:14

enables you to get

1:11:15

a grip on something

1:11:19

so i don't know that i can say any more

1:11:21

than that at the

1:11:24

moment

1:11:44

uh that was a question discussed between

1:11:48

frank and piana

1:11:52

themselves

1:11:55

yes

1:11:59

piano derived

1:12:02

a great i mean piano was

1:12:06

concerned somewhat

1:12:09

no that would be a wrong thing but piano

1:12:13

was

1:12:13

concerned to develop

1:12:17

[Music]

1:12:18

a certain kind of foundational

1:12:21

presentation of mathematical theories in

1:12:24

a

1:12:24

systematic way but he he wasn't i think

1:12:28

a lot of the what piano did was actually

1:12:32

derivative even what we call the piano

1:12:35

axiom

1:12:37

were given by dedicate in the first

1:12:40

instance that particular accentization

1:12:42

of numbered view

1:12:45

[Music]

1:12:48

and

1:12:51

well i don't know that i can say a great

1:12:53

deal more than that i think that the

1:12:56

the logical notions that piano used were

1:13:00

probably uh

1:13:06

derived from in in part from drago's own

1:13:09

work i

1:13:10

i may be wrong about that this

1:13:11

historical question as you said and

1:13:14

i'm not dead certain of the answer to it

1:13:17

the historical question wasn't what was

1:13:19

discussed between frank and piano what

1:13:21

was discussed was

1:13:23

questions of rigor and frag objected to

1:13:27

the

1:13:27

kind of definitions that piano

1:13:45

yes almost certainly they create nothing

1:13:48

to fragra

1:13:50

and

1:13:56

yes and they also as i said probably

1:14:00

still credit to piano things that was

1:14:02

due to dedicate

1:14:05

um i'm not trying to detract from

1:14:09

piano's reputation i thought things i

1:14:11

said just sound a bit like that but

1:14:13

i think look if you look at

1:14:17

russell's principles of mathematics he

1:14:20

very openly uses a lot of work done

1:14:23

principally by german mathematicians he

1:14:25

doesn't

1:14:26

attempt to to pretend that

1:14:30

uh that it's all his own work by no

1:14:32

means

1:14:33

but that's a very useful summary

1:14:37

of a lot of foundational work that

1:14:40

existed

1:14:40

right at the beginning of this century i

1:14:42

think that

1:14:44

piano did something of the same kind he

1:14:47

systematized a lot of stuff

1:14:49

but i think that a lot of it wasn't due

1:14:52

to his

1:14:53

own work

1:14:56

um no

1:15:00

do you think that's wrong

1:15:16

can you speak more lovely places i

1:15:18

couldn't catch no

1:15:22

no

1:15:34

[Music]

1:15:36

is

1:15:49

he was introducing

1:16:12

yes well i think he was much more on

1:16:15

russell's side than

1:16:17

on puerto rico so far as i know you

1:16:20

never took any notice of that

1:16:21

the particular dispute between those two

1:16:25

but certainly he was in the same sort of

1:16:28

position as

1:16:30

russell namely he did not think

1:16:33

that induction was a

1:16:36

form of reasoning special to mathematics

1:16:40

on the contrary he made a

1:16:42

very specific point of what russell

1:16:44

later

1:16:45

perhaps independently made a point of

1:16:49

this definition that i cited

1:16:52

allows us to exhibit if we define the

1:16:55

natural numbers

1:17:00

using this definition of a sequence

1:17:04

then induction falls out as immediate

1:17:07

consequence of it

1:17:09

and therefore not as depending on any

1:17:11

special

1:17:13

mathematical principle of reasoning

1:17:16

but as uh

1:17:22

just on purely logical principles given

1:17:24

this

1:17:25

definition quackers

1:17:29

position had to do with

1:17:34

the idea that you have to use induction

1:17:37

in seeing that the system is consistent

1:17:41

or the proofs or proofs from the axioms

1:17:45

again

1:17:46

going to yield true conclusions or

1:17:48

something which is

1:17:49

true in a way but it doesn't seem to me

1:17:52

to invalidate the position

1:17:54

of russell and and fragra

1:17:58

and it's like saying you can't

1:18:01

show a logical system to be sound

1:18:05

because look in your argument you've

1:18:08

used some of the principles

1:18:10

of that logical system well of course

1:18:12

any argument has to use

1:18:13

some principles but

1:18:16

there's a difference between a form of

1:18:20

argument that you

1:18:21

use and one that is the object

1:18:25

of of your reasoning

1:18:29

that is that you're talking about so i

1:18:32

don't believe that that argument of

1:18:34

pancreas is sound

1:18:36

at any rate if from the point of view of

1:18:38

your purely historical

1:18:40

question craig was certainly on

1:18:42

russell's side

1:18:44

and would have been against poincare if

1:18:46

he had read him

1:18:54

yes he said it perfectly explicit

1:18:58

well i can't get he certainly says it in

1:19:00

the foundations of arithmetic

1:19:02

um i can't give you an exact reference

1:19:06

at the moment i haven't got the book

1:19:07

with

1:19:08

me but

1:19:11

i'll send you a i'll send you a

1:19:20

reference

1:19:34

um

1:19:46

correct me if i don't exactly answer

1:19:49

your question i

1:19:50

quite got it but the thing was there was

1:19:53

this long

1:19:55

correspondence between russell

1:19:58

and uh freyja in the course of which

1:20:02

russell kept trying all sorts of uh

1:20:06

possible ways out and uh fraga

1:20:10

uh with great confidence despite this

1:20:13

disaster that has occurred to him with

1:20:15

great confidence kept shooting them

1:20:18

down and saying you can't say that and

1:20:22

and in the middle of it fraga announces

1:20:26

that she's found this

1:20:27

solution which then russell pays no

1:20:30

attention to

1:20:32

and says you're well you're probably

1:20:33

right but um

1:20:35

and actually of course the solution

1:20:37

didn't work

1:20:38

now the thing about that is it's

1:20:42

surprising it took so long

1:20:46

it took fraga till 1906 i think to

1:20:50

to to recognize that it didn't work i

1:20:53

don't think he ever knew

1:20:55

that you can still get a contradiction

1:20:58

in his system i'm not sure

1:21:00

but he certainly must have known as soon

1:21:03

as he

1:21:03

addressed himself to the question that

1:21:06

the proofs he had given

1:21:08

would break down under this weakening it

1:21:11

wasn't

1:21:11

a weakening of action five it wasn't

1:21:13

weakened enough

1:21:15

to avoid contradiction but nevertheless

1:21:18

it was weakened enough to invalidate

1:21:21

the simplest proofs in his theory you

1:21:24

couldn't even prove

1:21:26

using that that naught is not equal to

1:21:28

one that's

1:21:29

fatal for any foundations for

1:21:32

arithmetic obviously um

1:21:36

and i think you can locate the exact

1:21:39

point

1:21:39

at which he realized that uh slightly

1:21:43

indirect

1:21:44

but not very injury and namely in august

1:21:47

1906 and after that

1:21:51

he simply acknowledged that that his

1:21:54

work in the philosophy of mathematics

1:21:56

had

1:21:58

had crashed right and he thought

1:22:02

all that remained was the logic that was

1:22:05

all he had achieved but but not the

1:22:08

theory of classes not the

1:22:10

value ranges and therefore not the

1:22:13

foundations of

1:22:15

of number theory or analysis

1:22:22

i want just to say this i suppose

1:22:27

well no problems i'll wait and say

1:22:29

something more in the

1:22:36

moment

1:22:50

was

1:23:04

um

1:23:08

could you just say um what would you

1:23:11

perceive as

1:23:12

the main difference between the research

1:23:14

programs because in the outset

1:23:16

it seems that they shared a good view

1:23:27

yes this is a contentious matter

1:23:31

people disagree about this quite a lot

1:23:34

so i'll just say what uh

1:23:36

briefly what i think uh it's true that

1:23:39

was a converted

1:23:42

no they're not exactly the same fragra

1:23:47

was not a converted mathematician he

1:23:49

remained a mathematician who also

1:23:51

engaged in philosophy engaged in it

1:23:54

while being

1:23:55

while teaching mathematics and ordinary

1:23:57

mathematics counselors

1:23:58

all the time tiene uh

1:24:03

well jose really was a converting this

1:24:05

was someone who had been

1:24:07

started as a mathematician moved into

1:24:09

being a philosopher

1:24:11

um however that's not the main

1:24:13

difference

1:24:14

jose wrote only one book in fact it was

1:24:19

a book which was supposed to have a

1:24:21

second volume and never did

1:24:23

his first book on the philosophy of

1:24:25

arithmetic

1:24:28

it was written after fraker's book it

1:24:30

contained

1:24:31

criticisms of fraga and

1:24:35

it was a book of which fraga later

1:24:39

on four years afterwards wrote

1:24:42

of savage review

1:24:45

um there'd been some correspondence

1:24:48

between them

1:24:48

and i think it's certainly known or

1:24:52

reasonably guessed that jose was deeply

1:24:55

hurt by this review

1:24:57

as indeed anyone would be who read such

1:25:00

a review

1:25:01

his book

1:25:05

i think that fragrance review was

1:25:08

in part unfair in places unfair

1:25:12

i nevertheless think it by and large

1:25:15

fair

1:25:17

that is

1:25:22

the point at which fraga and jose were

1:25:24

close

1:25:26

was let us say just after

1:25:30

jose had published the prologue to his

1:25:34

logical investigation and

1:25:37

containing this attack on psychologism

1:25:41

now i don't know whether jose was

1:25:44

influenced by frago or not but what i

1:25:47

do myself think is that the book on

1:25:50

philosophy of

1:25:51

arithmetic was a thoroughly

1:25:53

psychologistic work

1:25:56

on a basis quite different from

1:25:59

his work at the time of the logical

1:26:01

investigations on the

1:26:03

basis that he had repudiated and i think

1:26:06

it's

1:26:08

a far inferior work to to

1:26:11

to uh to fragrance and

1:26:14

didn't lead in a fruitful direction i

1:26:17

entirely agree that jose the very

1:26:20

interesting philosopher

1:26:22

i think a lot of his work is very well

1:26:24

worth studying i

1:26:25

don't believe this to be true of his

1:26:28

work

1:26:29

on philosophy of mathematics that's my

1:26:32

own view

1:26:32

other people think quite differently and

1:26:35

some people who say

1:26:38

the philosophy of arithmetic isn't

1:26:40

psychologistic at all

1:26:42

there's very little difference between

1:26:43

that logical investigations

1:26:45

i i think this is a bizarre view i can't

1:26:49

do that

1:26:51

so but i will have to discuss the text

1:26:54

in detail but i just that's my reply to

1:26:57

it

1:26:59

michael i think it's time to turn to uh

1:27:01

more of a celebration

1:27:12

um you don't want the clothes by saying

1:27:15

gregor was more interesting

1:27:18

no i don't much want to close by saying

1:27:20

that but

1:27:24

stupid i should have said at that moment

1:27:27

what

1:27:28

it came to me to say um

1:27:34

and sorry i'm just trying to remember

1:27:36

what it was because it was quite an

1:27:38

interesting

1:27:38

point um but

1:27:42

oh yes just get yes i think what i

1:27:44

wanted to say was this

1:27:46

there are a lot of people who study

1:27:48

fragrance philosophy of

1:27:50

mathematics but actually concentrate

1:27:52

only

1:27:53

on the work of number theory and

1:27:55

sometimes you'll see it

1:27:57

argued that well maybe what he ought to

1:28:00

have

1:28:01

done was simply to take this

1:28:04

the numerical operator a number of

1:28:06

objects which as primitive

1:28:08

and we know that gives us a consistent

1:28:11

theory and that's all he needed for

1:28:13

for his construction of listening and so

1:28:16

on

1:28:17

that overlooks two things

1:28:21

um or let's say it overlooks his

1:28:25

attempt in uh the

1:28:28

basic laws of arithmetic which is very

1:28:31

little studied and deserves study it has

1:28:34

some very interesting mathematical

1:28:36

results

1:28:38

to to do a similar thing for

1:28:41

analysis for theory of real numbers

1:28:44

there's no way in which you could set

1:28:47

about reconstructing that as it stands

1:28:50

without using notion of of

1:28:53

of class and you would have to change it

1:28:57

very radically if you wanted to base it

1:29:00

on a

1:29:01

on a consistent century so

1:29:07

that suggestion simply overlooks half of

1:29:10

what

1:29:11

fraga wanted to do the trouble is people

1:29:14

read the foundations for arithmetic

1:29:15

which hardly discusses real numbers at

1:29:17

all

1:29:18

they think it's all about number theory

1:29:20

instead of being just as much about

1:29:23

analysis

1:29:27

that's one thing i want to say the other

1:29:28

thing i want to say is this

1:29:31

you could say fraca took

1:29:36

the fundamental application of the

1:29:39

natural numbers

1:29:40

to be as cardinal numbers say how many

1:29:44

things you

1:29:45

made this very very plain you could it's

1:29:48

perfectly reasonable to

1:29:49

argue that you ought to have taken their

1:29:53

use as ordinal numbers as when you

1:29:55

number houses in the street

1:29:57

as more fundamental because after all

1:30:01

when you count you can't find a number

1:30:03

of things or even if you enumerate

1:30:06

infinite number of things you impose an

1:30:08

order

1:30:09

it doesn't matter what the order is

1:30:11

particular order is

1:30:13

you're in the interest in the

1:30:14

cardinality but you get it the

1:30:15

cardinality by imposing an order so

1:30:18

that's one reason for saying notion of

1:30:20

ordinal number is more

1:30:22

fundamental another reason is that

1:30:27

uh without the notion of the

1:30:30

trans-financial

1:30:32

you have only one way

1:30:36

of getting bigger cardinal numbers

1:30:40

namely the power set operation start

1:30:42

with

1:30:43

the number of natural numbers number of

1:30:45

sets of natural numbers number of sets

1:30:47

sets natural numbers and so on you

1:30:50

haven't got the way that counts or used

1:30:53

by

1:30:53

considering the totality of the

1:30:55

enumerable ordinal gives you a left one

1:30:58

and then there's the interesting

1:31:00

continuum problem how that relates to

1:31:02

the number of real numbers and so on so

1:31:07

there are good reasons for saying a

1:31:09

notion of ordinal number is more basic

1:31:12

actually

1:31:12

the notion of cardinal number you

1:31:16

couldn't as this wasn't my

1:31:19

observation originally somewhere else

1:31:22

you couldn't

1:31:24

set about it in the same way or if you

1:31:26

did you would run into contradiction

1:31:28

much quicker than fraca did because

1:31:31

while the fundamental equivalence which

1:31:33

i discussed is perfectly consistent

1:31:36

if you try to do that

1:31:39

with the notion of the order type of an

1:31:41

ordering

1:31:43

all the type of r is equal to the old

1:31:46

type of best just in case there's a

1:31:48

similarity map order preserving map

1:31:52

on the field once the fields yeah if you

1:31:55

just run into the burally 40 paradox

1:31:57

straight away

1:31:58

without using classes at all

1:32:02

so that's quite that raises a big

1:32:05

question

1:32:07

how far can you push this

1:32:10

way of introducing operators yielding

1:32:13

abstract terms

1:32:15

um well without running into

1:32:19

inconsistencies to start with uh

1:32:22

well i'll just leave it at that you

Continue with YouTLDR

Analyze another video with Pro

Process a new video, search every timestamp, compare sources, and keep the result in your library.

Get Pro — $12/month30-day money-back guarantee

More transcripts

Explore other videos transcribed with YouTLDR.