論理哲学論考|対訳・比較

Tractatus Logico-Philosophicus|German–English

Read the German original and the English UX translation. Swipe each proposition to switch languages; the Preface switches paragraph by paragraph.

Dedicated to the memory of my friend
DAVID H. PINSENT

Motto: . . . and everything one knows, and has not merely heard rustling and roaring, can be said in three words.

Kürnberger.

Vorwort

This book will perhaps be understood only by someone who has already, at some point, thought for themselves the thoughts expressed in it—or at least similar thoughts. — So it is not a textbook. — Its purpose would be achieved if it gave pleasure to someone who reads it with understanding.

This book deals with philosophical problems and shows—as I believe—that the way these problems are posed rests on a misunderstanding of the logic of our language. One could put the whole sense of the book roughly like this: What can be said at all can be said clearly; and what we cannot talk about, we must be silent about.

So the book aims to draw a boundary for thinking, or rather—not for thinking, but for expressing thoughts: for to draw a boundary for thinking, we would have to be able to think both sides of that boundary (so we would have to be able to think what cannot be thought).

So the boundary can only be drawn within language, and what lies beyond the boundary will simply be nonsense.

How far my efforts coincide with those of other philosophers, I do not want to judge. In fact, none of the individual things I have written here claims to be new; and that is why I give no sources, for it makes no difference to me whether someone else thought what I have thought before I did.

I will mention only this: I owe much of the stimulus for my thoughts to Frege’s magnificent works and to the work of my friend Mr Bertrand Russell.

If this work has any value, it lies in two things. First, in the fact that thoughts are expressed in it, and this value will be all the greater the better those thoughts are expressed. The more squarely the nail has been hit on the head. — Here I am aware of having fallen far short of what was possible. Simply because my strength is too limited to carry out the task. — May others come and do it better.

By contrast, the truth of the thoughts set out here seems to me beyond challenge and final. So I think I have, in their essentials, solved the problems for good. And if I am not mistaken about this, then the second thing that gives this work value is that it shows how little has been accomplished by solving these problems.

L. W.

Vienna, 1918.

1

The world is everything that is the case.1

1.1

The world is the totality of facts, not of things.

1.11

The world is determined by the facts, and by their being all the facts.

1.12

For the totality of facts determines what is the case, and also everything that is not the case.

1.13

The facts in logical space are the world.

1.2

The world divides into facts.

1.21

One thing can be the case or not be the case, with everything else remaining the same.

2

What is the case, the fact, is the existence of states of affairs.

2.01

A state of affairs is a connection of objects. (Items, things.)

2.011

It is essential to a thing that it can be part of a state of affairs.

2.012

In logic nothing is accidental: if a thing can occur in a state of affairs, the possibility of that state of affairs must already be settled within the thing.

2.0121

It would seem, as it were, a coincidence if a situation were only later to fit a thing that could exist entirely on its own.

If things can occur in states of affairs, this must already lie in them.

(Something logical cannot be merely possible. Logic deals with every possibility, and all possibilities are its facts.)

Just as we cannot think of spatial objects outside space at all, or temporal objects outside time, so we cannot think of any object outside the possibility of its connection with others.

If I can think of the object as connected in a state of affairs, I cannot think of it outside the possibility of that connection.

2.0122

The thing is independent insofar as it can occur in all possible situations, but this form of independence is a form of connection with a state of affairs, a form of dependence. (It is impossible for words to occur in two different ways, on their own and in a sentence.)

2.0123

If I know an object, then I also know all the possibilities of its occurring in states of affairs.

(Every such possibility must lie in the nature of the object.)

No new possibility can be found afterwards.

2.01231

To know an object, I need not know its external properties—but I must know all its internal properties.

2.0124

If all objects are given, all possible states of affairs are given with them.

2.013

Each thing is, as it were, in a space of possible states of affairs. I can think of this space as empty, but not of the thing without the space.

2.0131

A spatial object must lie in infinite space. (A point in space is a place for an argument.)

A patch in the visual field need not be red, but it must have a colour: it has, so to speak, the space of colours around it. A tone must have a pitch, an object of touch a hardness, and so on.

2.014

Objects contain the possibility of all situations.

2.0141

The possibility of its occurring in states of affairs is the form of the object.

2.02

The object is simple.

2.0201

Every statement about complexes can be broken down into a statement about their parts and the sentences that fully describe the complexes.

2.021

Objects form the substance of the world. That is why they cannot be made up of parts.

2.0211

If the world had no substance, whether a sentence had a sense would depend on whether another sentence was true.

2.0212

It would then be impossible to draw a picture of the world (true or false).

2.022

It is obvious that a world imagined to be different from the actual world, no matter how different, must have something—a form—in common with the actual world.

2.023

This fixed form consists precisely of the objects.

2.0231

The substance of the world can determine only a form, not any material properties. For it is only through sentences that these are represented—only through the configuration of objects that they are formed.

2.0232

Incidentally: objects are colourless.

2.0233

Two objects with the same logical form are—apart from their external properties—distinguished from one another only by being different.

2.02331

Either a thing has properties that no other thing has, in which case it can readily be singled out from the others by a description and pointed to; or there are several things that share all their properties, in which case it is quite impossible to point to one of them.

For if nothing singles the thing out, I cannot single it out, for otherwise it is singled out after all.

2.024

Substance is what exists independently of what is the case.

2.025

It is form and content.

2.0251

Space, time and colour (having colour) are forms of objects.

2.026

Only if there are objects can there be a fixed form of the world.

2.027

What is fixed, what exists, and the object are one.

2.0271

The object is what is fixed, what exists; the configuration is what changes and does not remain constant.

2.0272

The configuration of objects forms the state of affairs.

2.03

In a state of affairs, objects interlock like the links of a chain.

2.031

In a state of affairs, objects relate to one another in a particular way.

2.032

The way objects are connected in a state of affairs is the structure of that state of affairs.

2.033

Form is the possibility of structure.

2.034

The structure of a fact consists of the structures of states of affairs.

2.04

The totality of existing states of affairs is the world.

2.05

The totality of existing states of affairs also determines which states of affairs do not exist.

2.06

The existence and non-existence of states of affairs is reality.

(We also call the existence of states of affairs a positive fact, and their non-existence a negative fact.)

2.061

States of affairs are independent of one another.

2.062

From the existence or non-existence of one state of affairs, we cannot infer the existence or non-existence of another.

2.063

Reality as a whole is the world.

2.1

We make pictures of the facts for ourselves.

2.11

The picture presents the situation in logical space, the existence and non-existence of states of affairs.

2.12

The picture is a model of reality.

2.13

In the picture, the elements of the picture correspond to the objects.

2.131

In the picture, the elements of the picture stand in for the objects.

2.14

The picture consists in its elements relating to one another in a particular way.

2.141

The picture is a fact.

2.15

That the elements of the picture relate to one another in a particular way presents the things as relating to one another in that way.

Let us call this connection between the elements of the picture its structure, and the possibility of this structure its form of depiction.

2.151

The form of depiction is the possibility of things relating to one another as the elements of the picture do.

2.1511

The picture is linked with reality in this way; it reaches right up to it.

2.1512

It is laid against reality like a ruler.

2.15121

Only the outermost points of the scale marks touch the object to be measured.

2.1513

On this view, then, the relation of depiction that makes it a picture also belongs to the picture.

2.1514

The relation of depiction consists of the pairings between the elements of the picture and the things.

2.1515

These pairings are, as it were, the feelers of the picture’s elements, with which the picture touches reality.

2.16

For a fact to be a picture, it must have something in common with what is depicted.

2.161

Something must be identical in the picture and what is depicted for one to be a picture of the other at all.

2.17

What a picture must have in common with reality in order to depict it in its own way—correctly or incorrectly—is its form of depiction.

2.171

A picture can depict any reality whose form it has. A spatial picture, everything spatial; a coloured one, everything coloured; and so on.

2.172

But the picture cannot depict its form of depiction; it shows it.

2.173

The picture represents its object from outside (its standpoint is its form of representation), which is why it represents its object correctly or incorrectly.

2.174

But the picture cannot place itself outside its form of representation.

2.18

What every picture, whatever its form, must have in common with reality in order to depict it at all—correctly or incorrectly—is logical form, that is, the form of reality.

2.181

If the form of depiction is logical form, then the picture is called a logical picture.

2.182

Every picture is also a logical one. (By contrast, not every picture is, for example, a spatial one.)

2.19

The logical picture can depict the world.

2.2

The picture has the logical form of depiction in common with what is depicted.

2.201

The picture depicts reality by representing a possibility of the existence and non-existence of states of affairs.

2.202

The picture represents a possible situation in logical space.

2.203

The picture contains the possibility of the situation it represents.

2.21

The picture matches reality or it does not; it is correct or incorrect, true or false.

2.22

The picture represents what it represents, independently of its truth or falsity, through its form of depiction.

2.221

What the picture represents is its sense.

2.222

Its truth or falsity consists in its sense matching or not matching reality.

2.223

To tell whether the picture is true or false, we must compare it with reality.

2.224

From the picture alone, we cannot tell whether it is true or false.

2.225

There is no picture that is true a priori.

3

The logical picture of the facts is the thought.

3.001

“A state of affairs is thinkable” means: we can make ourselves a picture of it.

3.01

The totality of true thoughts is a picture of the world.

3.02

The thought contains the possibility of the situation it thinks. What is thinkable is also possible.

3.03

We cannot think anything illogical, because otherwise we would have to think illogically.

3.031

It was once said that God could create anything, except anything that went against the laws of logic. — For we could not say what an “illogical” world would look like.

3.032

We can no more represent something “contrary to logic” in language than we can, in geometry, represent by its coordinates a figure that goes against the laws of space; or give the coordinates of a point that does not exist.

3.0321

We can indeed represent spatially a state of affairs that would go against the laws of physics, but none that would go against the laws of geometry.

3.04

A thought that was correct a priori would be one whose possibility determined its truth.

3.05

We could know a priori that a thought was true only if its truth could be recognized from the thought itself (without an object of comparison).

3.1

In the sentence, the thought is expressed in a way perceptible to the senses.

3.11

We use the sentence’s perceptible sign (a spoken or written sign, etc.) as a projection of the possible situation.

The method of projection is thinking the sense of the sentence.

3.12

I call the sign through which we express the thought the sentence sign. And the sentence is the sentence sign in its projective relation to the world.

3.13

Everything that belongs to the projection belongs to the sentence; but not what is projected.

So the possibility of what is projected, but not that itself.

So the sentence does not yet contain its sense, but it does contain the possibility of expressing it.

(“The content of the sentence” means the content of the sentence that has a sense.)

The sentence contains the form of its sense, but not its content.

3.14

The sentence sign consists in its elements, the words, relating to one another within it in a particular way.

The sentence sign is a fact.

3.141

The sentence is not a mixture of words. — (Just as a musical theme is not a mixture of tones.)

The sentence is articulated.

3.142

Only facts can express a sense; a class of names cannot.

3.143

That the sentence sign is a fact is concealed by the usual way of expressing things in writing or print.

For in a printed sentence, for example, the sentence sign does not look essentially different from a word.

(This made it possible for Frege to call the sentence a composite name.)

3.1431

The nature of the sentence sign becomes very clear if we think of it as made up of spatial objects (say tables, chairs, books) instead of written signs.

The spatial positions of these things in relation to one another then express the sense of the sentence.

3.1432

Not: “The complex sign ‘aRb’ says that a stands in the relation R to b”, but: That “a” stands in a certain relation to “b” says that aRb.

3.144

Situations can be described, not named.

(Names are like points, sentences like arrows; they have a sense.)

3.2

In the sentence, the thought can be expressed in such a way that elements of the sentence sign correspond to the objects of the thought.

3.201

I call these elements “simple signs”, and the sentence “fully analysed”.

3.202

The simple signs used in the sentence are called names.

3.203

The name means the object. The object is its meaning. (“A” is the same sign as “A”.)

3.21

The configuration of the simple signs in the sentence sign corresponds to the configuration of the objects in the situation.

3.22

In the sentence, the name stands in for the object.

3.221

I can only name objects. Signs stand in for them. I can only speak about them; I cannot say them. A sentence can only say how a thing is, not what it is.

3.23

The requirement that simple signs be possible is the requirement that sense be determinate.

3.24

The sentence that deals with a complex stands in an internal relation to the sentence that deals with a constituent of it.

The complex can be given only through its description, and this description will fit or not fit. If the complex does not exist, the sentence in which it is spoken of will not be nonsensical, but simply false.

That an element of a sentence designates a complex can be seen from an indeterminacy in the sentences in which it occurs. We know that not everything is yet determined by this sentence. (The notation for generality does contain a prototype.)

Condensing the symbol of a complex into a simple symbol can be expressed by a definition.

3.25

There is one and only one complete analysis of the sentence.

3.251

The sentence expresses what it expresses in a determinate way that can be clearly stated: the sentence is articulated.

3.26

The name cannot be broken down any further by any definition: it is a primitive sign.

3.261

Every defined sign designates through the signs by which it was defined; and the definitions show the way.

Two signs, a primitive sign and one defined by primitive signs, cannot designate in the same way. Names cannot be taken apart by definitions. (Nor any sign that has a meaning on its own, independently.)

3.262

What does not find expression in the signs is shown by their use. What the signs swallow, their use utters.

3.263

The meanings of primitive signs can be explained through clarifications. Clarifications are sentences that contain the primitive signs. So they can be understood only if the meanings of these signs are already known.

3.3

Only the sentence has a sense; only in the context of the sentence does a name have meaning.

3.31

I call every part of the sentence that characterizes its sense an expression (a symbol).

(The sentence itself is an expression.)

An expression is everything essential to the sense of the sentence that sentences can have in common.

The expression marks out a form and a content.

3.311

The expression presupposes the forms of all the sentences in which it can occur. It is the shared characteristic of a class of sentences.

3.312

So it is represented by the general form of the sentences it characterizes.

And in this form, the expression will be constant and everything else variable.

3.313

So the expression is represented by a variable whose values are the sentences that contain the expression.

(In the limiting case, the variable becomes a constant, the expression a sentence.)

I call such a variable a “sentence variable”.

3.314

The expression has meaning only in the sentence. Every variable can be regarded as a sentence variable.

(The variable name too.)

3.315

If we turn a part of a sentence into a variable, there is a class of sentences that are all values of the variable sentence thus formed. In general, this class still depends on what we mean, by arbitrary agreement, by parts of that sentence. But if we turn all the signs whose meaning was arbitrarily determined into variables, there is still such a class. This class, however, no longer depends on any agreement, but only on the nature of the sentence. It corresponds to a logical form—a logical prototype.

3.316

Which values the sentence variable may take is specified.

The specification of the values is the variable.

3.317

Specifying the values of the sentence variable is giving the sentences whose shared characteristic is the variable.

The specification is a description of these sentences.

So the specification will deal only with symbols, not with their meanings.

And only this is essential to the specification: that it is only a description of symbols and says nothing about what is designated.

How the sentences are described is inessential.

3.318

Like Frege and Russell, I regard the sentence as a function of the expressions it contains.

3.32

The sign is what is perceptible to the senses in the symbol.

3.321

So two different symbols can share the sign (a written or spoken sign, etc.)—they then designate in different ways.

3.322

Our designating two objects by the same sign, but by two different ways of designating, can never indicate a shared characteristic of the two objects. For the sign is arbitrary. We could just as well choose two different signs, and then where would the common element in the designation be?

3.323

In everyday language it happens extremely often that the same word designates in different ways—and so belongs to different symbols—or that two words that designate in different ways are outwardly used in the same way in the sentence.

Thus the word “is” appears as a copula, as a sign of equality and as an expression of existence; “exist” as an intransitive verb like “go”; “identical” as an adjective; we speak of something, but also of something happening.

(In the sentence “Green is green”—where the first word is a person’s name, the last an adjective—these words do not simply have different meanings; they are different symbols.)

3.324

This easily gives rise to the most fundamental confusions (of which the whole of philosophy is full).

3.325

To avoid these errors, we must use a language of signs that excludes them by not using the same sign in different symbols, and by not using signs that designate in different ways in outwardly the same way. A language of signs, then, that obeys logical grammar—logical syntax.

(Frege’s and Russell’s conceptual notation is such a language, though it does not yet exclude all errors.)

3.326

To recognize the symbol in the sign, we must attend to its meaningful use.

3.327

Only together with its use in logical syntax does a sign determine a logical form.

3.328

If a sign is not used, it has no meaning. That is the sense of Occam’s maxim.

(If everything behaves as though a sign had meaning, then it does have meaning.)

3.33

In logical syntax, the meaning of a sign must never play a part; it must be possible to lay down the syntax without speaking of the meaning of a sign; it may presuppose only the description of the expressions.

3.331

From this remark we look across to Russell’s “Theory of types”: Russell’s error shows in his having had to speak of the meanings of signs when laying down the rules for signs.

3.332

No sentence can say anything about itself, because the sentence sign cannot be contained in itself (that is the whole “Theory of types”).

3.333

A function cannot be its own argument because the function sign already contains the prototype of its argument, and cannot contain itself.

For suppose the function F(fx) could be its own argument; there would then be a sentence “F(F(fx))”, and in this the outer function F and the inner function F must have different meanings, for the inner has the form ϕ(fx), the outer the form ψ(ϕ(fx)). All the two functions have in common is the letter “F”, which on its own designates nothing.

This becomes clear at once if, instead of “F(F(u))”, we write “(∃ϕ) : F(ϕu) . ϕu = Fu”.

This disposes of Russell’s paradox.

3.334

The rules of logical syntax must be clear on their own if we just know how each sign designates.

3.34

The sentence has essential and accidental features.

The accidental features are those that arise from the particular way the sentence sign is produced. The essential ones are those that alone enable the sentence to express its sense.

3.341

So what is essential to the sentence is what all sentences that can express the same sense have in common.

And likewise, in general, what is essential to the symbol is what all symbols that can serve the same purpose have in common.

3.3411

So we could say: the name proper is what all symbols that designate the object have in common. It would thus emerge, step by step, that no composition of any kind is essential to the name.

3.342

There is indeed something arbitrary in our notations, but this is not arbitrary: that if we have determined something arbitrarily, something else must then be the case. (This depends on the nature of the notation.)

3.3421

A particular way of designating may be unimportant, but that it is a possible way of designating is always important. And this is how things stand in philosophy generally: the particular thing proves unimportant again and again, but the possibility of each particular thing tells us something about the nature of the world.

3.343

Definitions are rules for translation from one language into another. Every correct language of signs must be translatable into every other by such rules: this is what they all have in common.

3.344

What does the designating in a symbol is what is shared by all those symbols that can replace the first in accordance with the rules of logical syntax.

3.3441

For example, what all notations for truth functions have in common can be expressed like this: what they share is that all of them can be replaced, for example, by the notation of “∼p” (“not p”) and “p ∨ q” (“p or q”).

(This marks out how a particular possible notation can tell us something general.)

3.3442

Even in analysis, the sign of a complex does not break up arbitrarily—as it would, for instance, if it broke up differently in every sentence structure.

3.4

The sentence determines a place in logical space. The existence of this logical place is guaranteed by the existence of the constituents alone, by the existence of the sentence that has a sense.

3.41

The sentence sign and the logical coordinates: that is the logical place.

3.411

The geometrical and the logical place agree in that both are the possibility of an existence.

3.42

Although the sentence may determine only one place in logical space, the whole of logical space must already be given through it.

(Otherwise negation, logical sum, logical product, etc. would keep introducing new elements—in coordination.)

(The logical scaffolding around the picture determines logical space. The sentence reaches through the whole of logical space.)

3.5

The sentence sign, used and thought, is the thought.

4

A thought is a sentence with sense.

4.001

The totality of sentences is language.

4.002

Human beings have the ability to build languages in which every sense can be expressed, without having any idea how or what each word means. — Just as we speak without knowing how the individual sounds are produced.

Everyday language is a part of the human organism and no less complicated than it is.

It is humanly impossible to draw the logic of language directly from it.

Language disguises thought. In such a way, indeed, that the outward form of the clothing does not let us infer the form of the thought it clothes; because the outward form of the clothing is shaped for quite different purposes from that of letting us recognize the form of the body.

The tacit agreements through which everyday language is understood are enormously complicated.

4.003

Most of the sentences and questions that have been written about philosophical matters are not false but nonsensical. So we cannot answer questions of this kind at all, but can only establish that they are nonsensical. Most of the questions and sentences of philosophers rest on our not understanding the logic of our language.

(They are of the same kind as the question whether the good is more or less identical than the beautiful.)

And it is not surprising that the deepest problems are really no problems.

4.0031

All philosophy is “critique of language”. (Though not in Mauthner’s sense.) Russell’s achievement is to have shown that the apparent logical form of a sentence need not be its actual form.

4.01

A sentence is a picture of reality.

A sentence is a model of reality as we think of it.

4.011

At first glance a sentence—as it stands printed on paper, for example—does not seem to be a picture of the reality it deals with. But at first glance musical notation does not seem to be a picture of music either, nor our writing in sound signs (letters) a picture of our spoken language.

And yet these languages of signs turn out to be pictures of what they represent, even in the ordinary sense.

4.012

It is clear that we feel a sentence of the form “aRb” to be a picture. Here the sign is clearly a likeness of what it designates.

4.013

And if we get to the essence of this pictorial character, we see that it is not disturbed by apparent irregularities (such as the use of ♯ and ♭ in musical notation).

For these irregularities too depict what they are meant to express; only in a different way.

4.014

The gramophone record, the musical thought, the musical notation, the sound waves all stand to one another in the same internal relation of depiction that holds between language and world.

They all have logical construction in common.

(As do the two young men, their two horses and their lilies in the fairy tale. They are all, in a certain sense, one.)

4.0141

There is a general rule by which a musician can get the symphony from the score, by which the symphony can be derived from the line on the gramophone record, and, by the first rule, the score again: this is precisely what the inner similarity of these seemingly so very different formations consists in. And that rule is the law of projection that projects the symphony into the language of musical notation. It is the rule for translating the language of musical notation into the language of the gramophone record.

4.015

The possibility of all likenesses, of the whole pictorial character of our way of expressing ourselves, rests on the logic of depiction.

4.016

To understand the essence of a sentence, let us think of hieroglyphic writing, which depicts the facts it describes.

And alphabetic writing developed from it without losing what is essential to depiction.

4.02

We see this from the fact that we understand the sense of a sentence sign without its having been explained to us.

4.021

A sentence is a picture of reality: for I know the situation it represents if I understand the sentence. And I understand the sentence without its sense having been explained to me.

4.022

A sentence shows its sense.

A sentence shows how things stand if it is true. And it says that things stand that way.

4.023

Reality must be pinned down by the sentence to a yes or no.

For this, reality must be completely described by the sentence. A sentence is the description of a state of affairs.

As a description describes an object by its external properties, so a sentence describes reality by its internal properties.

A sentence constructs a world with the help of a logical framework, and so we can also see from the sentence how everything logical stands if it is true. We can draw conclusions from a false sentence.

4.024

To understand a sentence means to know what is the case if it is true.

(So we can understand it without knowing whether it is true.)

We understand it if we understand its components.

4.025

Translation from one language into another does not proceed by translating each sentence of the one into a sentence of the other; only the components of sentences are translated.

(And a dictionary translates not only nouns but also verbs, adjectives, conjunctions, etc.; and it treats them all alike.)

4.026

The meanings of simple signs (words) must be explained to us for us to understand them.

But with sentences we make ourselves understood to one another.

4.027

It belongs to the essence of a sentence that it can communicate a new sense to us.

4.03

A sentence must communicate a new sense with old expressions. A sentence communicates a situation to us, so it must be essentially connected with that situation.

And the connection is precisely that it is the situation’s logical picture.

A sentence says something only insofar as it is a picture.

4.031

In a sentence a situation is, as it were, put together as a trial.

We can simply say, instead of: this sentence has such-and-such a sense: this sentence represents such-and-such a situation.

4.0311

One name stands for one thing, another for another thing, and they are connected with one another; in this way the whole—like a living picture—presents the state of affairs.

4.0312

The possibility of a sentence rests on the principle that signs stand in for objects.

My basic thought is that the “logical constants” do not stand in for anything. That the logic of facts cannot be stood in for.

4.032

A sentence is a picture of a situation only insofar as it is logically articulated.

(Even the sentence “ambulo” is composite, for its stem with a different ending, and its ending with a different stem, yield a different sense.)

4.04

There must be exactly as much to distinguish in a sentence as in the situation it represents.

The two must have the same logical (mathematical) multiplicity. (Compare Hertz’s Mechanics, on dynamical models.)

4.041

This mathematical multiplicity itself, of course, cannot in turn be depicted. We cannot get outside it while depicting.

4.0411

Suppose we wanted to express what we express by “(x).fx” by putting an index before “fx”—say, like this: “Alg. fx”. This would not be enough—we would not know what had been generalized. Suppose we wanted to indicate it by an index “a”—say, like this: “f (xa)”. This would not be enough either—we would not know the scope of the generality sign.

Suppose we tried putting a mark in the argument places—say, like this: “(A, A) . F (A, A)”. This would not be enough—we could not establish the identity of the variables. And so on.

All these ways of designating are insufficient because they do not have the necessary mathematical multiplicity.

4.0412

For the same reason, the idealist explanation of our seeing spatial relations through “spatial spectacles” is insufficient, because it cannot explain the multiplicity of these relations.

4.05

Reality is compared with the sentence.

4.06

A sentence can be true or false only by being a picture of reality.

4.061

If we overlook the fact that a sentence has a sense independent of the facts, we can easily believe that true and false are relations of equal standing between sign and what it designates.

We could then say, for example, that “p” designates in the true way what “∼p” designates in the false way, etc.

4.062

Can we not make ourselves understood with false sentences just as we have done with true ones? As long as we know that they are meant as false. No! For a sentence is true if things stand as we say they do through it; and if by “p” we mean ∼p, and things stand as we mean, then “p” is true in the new way of taking it, not false.

4.0621

But it is important that the signs “p” and “∼p” can say the same thing. For this shows that nothing in reality corresponds to the sign “∼”.

The occurrence of negation in a sentence is not yet a feature of its sense (∼∼p = p).

The sentences “p” and “∼p” have opposite senses, but one and the same reality corresponds to them.

4.063

A picture to explain the concept of truth: a black patch on white paper; we can describe the shape of the patch by specifying, for every point of the surface, whether it is white or black. The fact that a point is black corresponds to a positive fact—the fact that a point is white (not black), to a negative one. If I designate a point of the surface (a truth value in Frege’s sense), this corresponds to the assumption put forward for judgment, etc., etc.

But to be able to say that a point is black or white, I must first know when a point is called black and when it is called white; to be able to say: “p” is true (or false), I must have determined under what circumstances I call “p” true, and in doing so I determine the sense of the sentence.

The point where the likeness breaks down is this: we can point to a point on the paper even without knowing what white and black are; but nothing at all corresponds to a sentence without sense, for it does not designate a thing (a truth value) whose properties might be called “false” or “true”; the verb of a sentence is not “is true” or “is false”—as Frege believed—but what “is true” must already contain the verb.

4.064

Every sentence must already have a sense; affirmation cannot give it one, for it is precisely the sense that affirmation affirms. And the same holds for negation, etc.

4.0641

We could say: negation already relates to the logical place that the negated sentence determines.

The negating sentence determines a different logical place from the negated one.

The negating sentence determines a logical place with the help of the logical place of the negated sentence, by describing the former as lying outside the latter.

The very fact that the negated sentence can be negated again shows that what is negated is already a sentence, and not merely a preparation for a sentence.

4.1

A sentence represents the existence and nonexistence of states of affairs.

4.11

The totality of true sentences is the whole of natural science (or the totality of the natural sciences).

4.111

Philosophy is not one of the natural sciences.

(The word “philosophy” must mean something that stands above or below the natural sciences, but not beside them.)

4.112

The purpose of philosophy is to make thoughts logically clear.

Philosophy is not a doctrine but an activity.

A philosophical work consists essentially of clarifications.

The result of philosophy is not “philosophical sentences”, but sentences becoming clear.

Philosophy should make thoughts clear and draw sharp boundaries around them, where otherwise they are, as it were, cloudy and blurred.

4.1121

Psychology is no more closely related to philosophy than any other natural science.

Epistemology is the philosophy of psychology.

Does not my study of the language of signs correspond to the study of thought processes that philosophers held to be so essential to the philosophy of logic? Only, they mostly got entangled in inessential psychological investigations, and there is a similar danger in my method too.

4.1122

Darwin’s theory has no more to do with philosophy than any other hypothesis of natural science.

4.113

Philosophy sets limits to the disputable domain of natural science.

4.114

It should draw a boundary around what can be thought, and thereby around what cannot be thought.

It should set limits to what cannot be thought from within, through what can be thought.

4.115

It will indicate what cannot be said by clearly representing what can be said.

4.116

Everything that can be thought at all can be thought clearly. Everything that can be expressed can be expressed clearly.

4.12

A sentence can represent the whole of reality, but it cannot represent what it must have in common with reality in order to be able to represent it—logical form.

To be able to represent logical form, we would have to be able to place ourselves with the sentence outside logic, that is, outside the world.

4.121

A sentence cannot represent logical form; logical form is reflected in it.

What is reflected in language, language cannot represent.

What expresses itself in language, we cannot express through language.

A sentence shows the logical form of reality.

It shows it.

4.1211

Thus a sentence “fa” shows that the object a occurs in its sense; two sentences “fa” and “ga” show that both speak of the same object.

If two sentences contradict one another, their structure shows this; likewise, if one follows from the other. And so on.

4.1212

What can be shown cannot be said.

4.1213

Now we also understand our feeling: that we have a correct logical way of taking things, if only everything in our language of signs fits.

4.122

In a certain sense we can speak of formal properties of objects and states of affairs, or of properties of the structure of facts, and in the same sense of formal relations and relations of structures.

(Instead of property of a structure I also say “internal property”; instead of relation of structures, “internal relation”.

I introduce these expressions to show the source of the confusion, so widespread among philosophers, between internal relations and relations proper (external ones).)

But the existence of such internal properties and relations cannot be asserted by sentences; it shows itself in the sentences that represent those states of affairs and deal with those objects.

4.1221

We can also call an internal property of a fact a feature of that fact. (In the sense in which we speak, for example, of facial features.)

4.123

A property is internal if it is unthinkable that its object should not have it.

(This blue colour and that one stand in the internal relation of lighter and darker eo ipso. It is unthinkable that these two objects should not stand in this relation.)

(Here the shifting use of the words “property” and “relation” corresponds to the shifting use of the word “object”.)

4.124

The existence of an internal property of a possible situation is not expressed by a sentence; it expresses itself in the sentence representing the situation, through an internal property of that sentence.

It would be just as nonsensical to ascribe a formal property to a sentence as to deny it that property.

4.1241

We cannot distinguish forms from one another by saying that one has this property and the other that property; for this presupposes that it makes sense to assert both properties of both forms.

4.125

The existence of an internal relation between possible situations expresses itself in language through an internal relation between the sentences representing them.

4.1251

Here the disputed question “whether all relations are internal or external” is settled.

4.1252

I call series that are ordered by internal relations formal series.

The number series is ordered not by an external relation but by an internal one.

So too is the series of sentences “aRb”,

“(∃x) : aRx . xRb”,

“(∃x, y) : aRx . xRy . yRb”, and so on.

(If b stands in one of these relations to a, I call b a successor of a.)

4.126

In the sense in which we speak of formal properties, we can now also speak of formal concepts.

(I introduce this expression to make clear the source of the confusion between formal concepts and concepts proper that runs through the whole of the old logic.)

That something falls under a formal concept as its object cannot be expressed by a sentence. Instead, it shows itself in the sign of that object itself. (A name shows that it designates an object, a numeral that it designates a number, etc.)

For formal concepts cannot be represented by a function in the way that concepts proper can.

For their features, formal properties, are not expressed by functions.

The expression of a formal property is a feature of certain symbols.

So the sign for the features of a formal concept is a characteristic feature of all symbols whose meanings fall under the concept.

The expression of a formal concept is therefore a sentence variable in which only this characteristic feature is constant.

4.127

The sentence variable designates the formal concept, and its values designate the objects that fall under that concept.

4.1271

Every variable is the sign of a formal concept.

For every variable represents a constant form that all its values possess, and that can be taken as a formal property of those values.

4.1272

Thus the variable name “x” is the proper sign for the pseudo-concept object.

Wherever the word “object” (“thing”, “item”, etc.) is used correctly, it is expressed in conceptual notation by the variable name.

For example, in the sentence “there are 2 objects that . . .” by “(∃x, y) . . .”.

Wherever it is used otherwise, that is, as a word for a concept proper, nonsensical pseudo-sentences arise.

Thus we cannot say, for example, “There are objects” in the way we say “There are books”. Nor can we say “There are 100 objects”, or “There are ℵ₀ objects”.

And it is nonsensical to speak of the number of all objects.

The same holds for the words “complex”, “fact”, “function”, “number”, etc.

They all designate formal concepts and are represented in conceptual notation by variables, not by functions or classes. (As Frege and Russell believed.)

Expressions such as “1 is a number”, “there is only one zero”, and all similar ones are nonsensical.

(It is just as nonsensical to say “there is only one 1” as it would be to say: 2 + 2 equals 4 at 3 o’clock.)

4.12721

A formal concept is already given with an object that falls under it. So we cannot introduce objects of a formal concept and the formal concept itself as primitive concepts. Thus we cannot, for example, introduce the concept of a function and also particular functions (as Russell does) as primitive concepts; or the concept of number and particular numbers.

4.1273

If we want to express the general sentence: “b is a successor of a” in conceptual notation, we need an expression for the general term of the formal series: aRb, (∃x) : aRx.xRb, (∃x, y) : aRx.xRy.yRb, . . . The general term of a formal series can be expressed only by a variable, for the concept: term of this formal series, is a formal concept. (Frege and Russell overlooked this; so the way they try to express general sentences such as the one above is wrong; it contains a circulus vitiosus.)

We can determine the general term of the formal series by specifying its first term and the general form of the operation that generates the next term from the preceding sentence.

4.1274

The question of the existence of a formal concept is nonsensical. For no sentence can answer such a question.

(So we cannot ask, for example: “Are there unanalyzable subject-predicate sentences?”)

4.128

Logical forms are numberless.

That is why there are no special numbers in logic, and why there is no philosophical monism or dualism, etc.

4.2

The sense of a sentence is its agreement and disagreement with the possibilities of states of affairs existing and not existing.

4.21

The simplest sentence, the elementary sentence, asserts the existence of a state of affairs.

4.211

One sign of an elementary sentence is that no elementary sentence can contradict it.

4.22

An elementary sentence consists of names. It is a connection, a chain, of names.

4.221

It is clear that in analysing sentences we must arrive at elementary sentences consisting of names in immediate connection.

The question here is how the sentence’s connection comes about.

4.2211

Even if the world is infinitely complex, so that every fact consists of infinitely many states of affairs and every state of affairs is composed of infinitely many objects, even then there would have to be objects and states of affairs.

4.23

A name occurs in a sentence only in the context of an elementary sentence.

4.24

Names are the simple symbols; I indicate them by single letters (“x”, “y”, “z”).

I write an elementary sentence as a function of names in the form: “fx”, “ϕ(x, y)”, etc.

Or I indicate it by the letters p, q, r.

4.241

If I use two signs with one and the same meaning, I express this by putting the sign “=” between them.

So “a = b” means: the sign “a” can be replaced by the sign “b”.

(If I introduce a new sign “b” by an equation, stipulating that it is to replace an already familiar sign “a”, I write the equation—definition—in the form “a = b Def.” (as Russell does). A definition is a rule for signs.)

4.242

So expressions of the form “a = b” are only aids to representation; they say nothing about the meaning of the signs “a”, “b”.

4.243

Can we understand two names without knowing whether they designate the same thing or two different things? — Can we understand a sentence in which two names occur without knowing whether they mean the same thing or different things?

If, for example, I know the meaning of an English word and of a German word with the same meaning, it is impossible for me not to know that the two mean the same; it is impossible for me not to be able to translate them into one another.

Expressions such as “a = a”, or ones derived from them, are neither elementary sentences nor any other signs with sense. (This will become clear later.)

4.25

If the elementary sentence is true, the state of affairs exists; if the elementary sentence is false, the state of affairs does not exist.

4.26

Specifying all true elementary sentences describes the world completely. The world is completely described by specifying all elementary sentences plus specifying which of them are true and which false.

4.27

Concerning the existence and nonexistence of n states of affairs, there are Kn=∑ν=0n(nν) possibilities.

Any combination of the states of affairs can exist, while the others do not exist.

4.28

These combinations correspond to just as many possibilities of the truth—and falsity—of n elementary sentences.

4.3

The truth possibilities of elementary sentences mean the possibilities of states of affairs existing and not existing.

4.31

We can represent the truth possibilities by tables of the following kind (“W” means “true”, “F”, “false”. The rows of “W” and “F” below the row of elementary sentences mean their truth possibilities in an easily understood symbolism):

pqr
WWW
FWW
WFW
WWF
FFW
FWF
WFF
FFF
pq
WW
FW
WF
FF
p
W
F
4.4

A sentence is the expression of agreement and disagreement with the truth possibilities of elementary sentences.

4.41

The truth possibilities of elementary sentences are the conditions of the truth and falsity of sentences.

4.411

It is probable from the outset that introducing elementary sentences is fundamental to understanding all other kinds of sentence. Indeed, the understanding of general sentences palpably depends on that of elementary sentences.

4.42

Concerning the agreement and disagreement of a sentence with the truth possibilities of n elementary sentences, there are ∑κ=0Kn(Knκ)=Ln possibilities.

4.43

We can express agreement with the truth possibilities by assigning them, for example, the mark “W” (true) in the table.

The absence of this mark means disagreement.

4.431

The expression of agreement and disagreement with the truth possibilities of elementary sentences expresses the truth conditions of the sentence.

A sentence is the expression of its truth conditions.

(So Frege was quite right to put them first as an explanation of the signs of his conceptual notation. Only, Frege’s explanation of the concept of truth is wrong: if “the True” and “the False” really were objects and the arguments in ∼p, etc., then according to Frege’s determination the sense of “∼p” would by no means be determined.)

4.44

The sign that results from assigning those “W” marks to the truth possibilities is a sentence sign.

4.441

It is clear that no object (or complex of objects) corresponds to the complex of signs “F” and “W”; no more than to the horizontal and vertical lines or the brackets. — There are no “logical objects”.

The same naturally holds for all signs that express the same thing as the “W” and “F” tables.

4.442

For example:

pq
WWW
FWW
WF
FFW

is a sentence sign.

Frege’s “judgment stroke” “⊢” has no logical meaning at all; in Frege (and Russell) it only indicates that these authors take the sentences so designated to be true. So “⊢” no more belongs to the structure of a sentence than, say, the sentence’s number does. A sentence cannot possibly assert of itself that it is true.)

If the order of the truth possibilities in the table has been fixed once and for all by a combinatorial rule, the last column alone is already an expression of the truth conditions. If we write this column out as a row, the sentence sign becomes:

“(WW–W)(p, q)” or, more clearly, “(WWFW)(p, q)”.

(The number of places in the left pair of brackets is determined by the number of terms in the right pair.)

4.45

For n elementary sentences there are Ln possible groups of truth conditions.

The groups of truth conditions belonging to the truth possibilities of a number of elementary sentences can be arranged in a series.

4.46

Among the possible groups of truth conditions there are two extreme cases.

In the one case, the sentence is true for all truth possibilities of the elementary sentences. We say that the truth conditions are tautological.

In the second case, the sentence is false for all truth possibilities: the truth conditions are contradictory.

In the first case we call the sentence a tautology, in the second a contradiction.

4.461

A sentence shows what it says; tautology and contradiction show that they say nothing.

A tautology has no truth conditions, for it is unconditionally true; and a contradiction is true under no condition.

Tautology and contradiction are without sense.

(Like the point from which two arrows diverge in opposite directions.)

(I know nothing about the weather, for example, if I know that it is raining or not raining.)

4.4611

But tautology and contradiction are not nonsensical; they belong to the symbolism, much as “0” belongs to the symbolism of arithmetic.

4.462

Tautology and contradiction are not pictures of reality. They represent no possible situation. For the former allows every possible situation, the latter none.

In a tautology the conditions of agreement with the world—the representing relations—cancel one another out, so that it stands in no representing relation to reality.

4.463

The truth conditions determine the room that a sentence leaves to the facts.

(A sentence, a picture, a model are, in the negative sense, like a solid body that restricts the freedom of movement of the others; in the positive sense, like the space bounded by solid material in which there is room for a body.)

A tautology leaves reality the whole—infinite—logical space; a contradiction fills the whole logical space and leaves reality no point. So neither can determine reality in any way.

4.464

The truth of a tautology is certain, of a sentence possible, of a contradiction impossible.

(Certain, possible, impossible: here we have the sign of that gradation we need in probability theory.)

4.465

The logical product of a tautology and a sentence says the same thing as the sentence. So that product is identical with the sentence. For we cannot change what is essential to a symbol without changing its sense.

4.466

A particular logical connection of signs corresponds to a particular logical connection of their meanings; any arbitrary connection corresponds only to the unconnected signs.

That is, sentences that are true for every situation cannot be connections of signs at all, for otherwise only particular connections of objects could correspond to them.

(And to no logical connection corresponds no connection of objects.)

Tautology and contradiction are the limiting cases of the connection of signs, namely its dissolution.

4.4661

Of course, even in tautology and contradiction the signs are still connected with one another, that is, they stand in relations to one another; but these relations have no meaning, are inessential to the symbol.

4.5

Now it seems possible to specify the most general sentence form: that is, to give a description of the sentences of any language of signs, such that every possible sense can be expressed by a symbol that fits the description, and every symbol that fits the description can express a sense if the meanings of the names are chosen accordingly.

It is clear that in describing the most general sentence form, only what is essential to it may be described—for otherwise it would not be the most general.

That there is a general sentence form is proved by the fact that there must be no sentence whose form could not have been foreseen (that is, constructed). The general form of a sentence is: Things stand in such-and-such a way.

4.51

Suppose all elementary sentences were given to me: then we can simply ask: what sentences can I form from them? And these are all sentences, and this is how they are bounded.

4.52

Sentences are everything that follows from the totality of all elementary sentences (and naturally also from its being the totality of all of them). (So in a certain sense we could say that all sentences are generalizations of elementary sentences.)

4.53

The general sentence form is a variable.

5

The sentence is a truth function of elementary sentences.

(An elementary sentence is a truth function of itself.)

5.01

The elementary sentences are the truth arguments of the sentence.

5.02

It is easy to confuse the arguments of functions with the indices of names. For both the argument and the index let me recognise the meaning of the sign containing them.

In Russell’s “+c”, for example, “c” is an index indicating that the whole sign is the addition sign for cardinal numbers. But this notation rests on an arbitrary agreement, and a simple sign could be chosen instead of “+c”; in “∼p”, however, “p” is not an index but an argument: the sense of “∼p” cannot be understood unless the sense of “p” has been understood beforehand. (In the name Julius Caesar, “Julius” is an index. An index is always part of a description of the object to whose name we attach it. For example: the Caesar of the Julian family.)

If I am not mistaken, the confusion of argument and index underlies Frege’s theory of the meaning of sentences and functions. For Frege, the sentences of logic were names, and their arguments the indices of those names.

5.1

Truth functions can be arranged in series. This is the basis of probability theory.

5.101

The truth functions of any number of elementary sentences can be written out in a scheme of the following kind:

(WWWW)(p, q)Tautology(If p, then p; and if q, then q.) (p ⊃ p . q ⊃ q)
(FWWW)(p, q)In words:Not both p and q. (∼(p . q))
(WFWW)(p, q)″ ″If q, then p. (q ⊃ p)
(WWFW)(p, q)″ ″If p, then q. (p ⊃ q)
(WWWF)(p, q)″ ″p or q. (p ∨ q)
(FFWW)(p, q)″ ″Not q. ∼q
(FWFW)(p, q)″ ″Not p. ∼p
(FWWF)(p, q)″ ″p or q, but not both. (p . ∼q : ∨ : q . ∼p)
(WFFW)(p, q)″ ″If p, then q; and if q, then p. (p ≡ q)
(WFWF)(p, q)″ ″p
(WWFF)(p, q)″ ″q
(FFFW)(p, q)″ ″Neither p nor q. (∼p . ∼q) or (p | q)
(FFWF)(p, q)″ ″p and not q. (p . ∼q)
(FWFF)(p, q)″ ″q and not p. (q . ∼p)
(WFFF)(p, q)″ ″q and p. (q . p)
(FFFF)(p, q)Contradiction(p and not p; and q and not q.) (p . ∼p . q . ∼q)

Those truth possibilities of its truth arguments that make the sentence true I will call its grounds for truth.

5.11

If the grounds for truth shared by a number of sentences are all also grounds for truth of a particular sentence, we say that the truth of this sentence follows from the truth of those sentences.

5.12

In particular, the truth of a sentence “p” follows from the truth of another, “q”, if all the grounds for truth of the second are grounds for truth of the first.

5.121

The grounds for truth of the one are contained in those of the other; p follows from q.

5.122

If p follows from q, the sense of “p” is contained in the sense of “q”.

5.123

If a god creates a world in which certain sentences are true, he thereby already creates a world in which all the sentences that follow from them hold. Similarly, he could not create a world in which the sentence “p” is true without creating all its objects.

5.124

A sentence affirms every sentence that follows from it.

5.1241

“p . q” is one of the sentences that affirm “p” and at the same time one of the sentences that affirm “q”.

Two sentences are opposed to one another if there is no sentence with sense that affirms them both.

Every sentence that contradicts another negates it.

5.13

We see from the structure of the sentences that the truth of one sentence follows from the truth of others.

5.131

If the truth of a sentence follows from the truth of others, this is expressed by relations in which the forms of those sentences stand to one another; and we do not first need to put them into those relations by joining them together in a sentence: these relations are internal and exist as soon as, and because, those sentences exist.

5.1311

When we infer q from p ∨ q and ∼p, the notation here conceals the relation between the sentence forms “p ∨ q” and “∼p”. But if, for example, we write “p | q . | . p | q” instead of “p ∨ q” and “p | p” instead of “∼p” (p | q = neither p nor q), the internal connection becomes apparent.

(That we can infer fa from (x) . fx shows that generality is also present in the symbol “(x) . fx”.)

5.132

If p follows from q, I can infer p from q; derive p from q.

The manner of the inference is to be gathered from the two sentences alone.

Only they themselves can justify the inference.

“Laws of inference” that are supposed—as with Frege and Russell—to justify inferences are without sense, and would be superfluous.

5.133

All inference takes place a priori.

5.134

No other elementary sentence can be derived from an elementary sentence.

5.135

There is no way to infer from the existence of any situation the existence of a situation entirely different from it.

5.136

There is no causal connection that would justify such an inference.

5.1361

We cannot infer future events from present ones.

Belief in a causal connection is superstition.

5.1362

Free will consists in the fact that future actions cannot be known now. We could know them only if causality were an internal necessity, like that of logical inference. — The connection between knowing and what is known is that of logical necessity.

(“A knows that p is the case” is without sense if p is a tautology.)

5.1363

If it does not follow from a sentence’s seeming obvious to us that it is true, then its seeming obvious is no justification for our belief in its truth either.

5.14

If one sentence follows from another, the latter says more than the former, the former less than the latter.

5.141

If p follows from q and q from p, they are one and the same sentence.

5.142

A tautology follows from all sentences: it says nothing.

5.143

Contradiction is what sentences have in common that no sentence has in common with another. Tautology is what all sentences that have nothing in common with one another have in common.

Contradiction vanishes, so to speak, outside all sentences, tautology inside them.

Contradiction is the outer boundary of sentences, tautology their centre without substance.

5.15

If Wr is the number of grounds for truth of the sentence “r”, and Wrs the number of those grounds for truth of the sentence “s” that are also grounds for truth of “r”, then we call the ratio Wrs : Wr the measure of the probability that sentence “r” gives to sentence “s”.

5.151

In a scheme like the one above in 5.101, let Wr be the number of “W”s in sentence r; let Wrs be the number of those “W”s in sentence s that stand in the same columns as “W”s in sentence r. Sentence r then gives sentence s the probability Wrs : Wr.

5.1511

There is no special object peculiar to sentences about probability.

5.152

We call sentences that share no truth arguments independent of one another.

Sentences independent of one another (for example, any two elementary sentences) give one another the probability ½.

If p follows from q, sentence “q” gives sentence “p” the probability 1. The certainty of logical inference is a limiting case of probability.

(Application to tautology and contradiction.)

5.153

A sentence in itself is neither probable nor improbable. An event occurs or it does not occur; there is nothing in between.

5.154

Let an urn contain equal numbers of white and black balls (and no others). I draw one ball after another, putting each back in the urn. I can then establish by experiment that, as I keep drawing, the numbers of black and white balls drawn approach one another.

This, then, is not a mathematical fact.

If I now say that I am as likely to draw a white ball as a black one, this means: all the circumstances known to me (including the laws of nature assumed hypothetically) give the occurrence of one event no more probability than the occurrence of the other. That is, they give each the probability ½—as can easily be gathered from the explanations above.

What I confirm by experiment is that the occurrence of the two events is independent of the circumstances I do not know in detail.

5.155

The unit of a sentence about probability is: the circumstances—about which I otherwise know nothing further—give the occurrence of a particular event such-and-such a degree of probability.

5.156

So probability is a generalisation.

It involves a general description of a sentence form.

We use probability only when certainty is lacking. — When we do not know a fact completely, but do know something about its form.

(A sentence can indeed be an incomplete picture of a particular situation, but it is always a complete picture.)

A sentence about probability is, as it were, an extract from other sentences.

5.2

The structures of sentences stand in internal relations to one another.

5.21

We can bring out these internal relations in our way of expressing ourselves by representing a sentence as the result of an operation that produces it from other sentences (the bases of the operation).

5.22

The operation is the expression of a relation between the structures of its result and its bases.

5.23

The operation is what must happen to one sentence to make the other out of it.

5.231

And this will of course depend on their formal properties, on the internal similarity of their forms.

5.232

The internal relation that orders a series is equivalent to the operation by which one term arises from another.

5.233

An operation can occur only where one sentence arises from another in a logically significant way. That is, where the logical construction of the sentence begins.

5.234

The truth functions of elementary sentences are results of operations that have elementary sentences as their bases. (I call these operations truth operations.)

5.2341

The sense of a truth function of p is a function of the sense of p.

Negation, logical addition, logical multiplication, etc., etc. are operations.

(Negation reverses the sense of the sentence.)

5.24

The operation shows itself in a variable; it shows how to get from one form of sentences to another.

It expresses the difference between the forms. (And what the bases and the result of the operation have in common is precisely the bases.)

5.241

The operation marks out no form, only the difference between forms.

5.242

The same operation that makes “q” out of “p” makes “r” out of “q”, and so on. This can be expressed only in “p”, “q”, “r”, etc. being variables that give general expression to certain formal relations.

5.25

The occurrence of the operation does not characterise the sense of the sentence.

The operation says nothing, after all; only its result does, and this depends on the bases of the operation.

(Operation and function must not be confused with one another.)

5.251

A function cannot be its own argument, but the result of an operation can become that operation’s own base.

5.252

Only in this way is it possible to advance from term to term in a series of forms (from type to type in Russell’s and Whitehead’s hierarchies). (Russell and Whitehead did not admit the possibility of this advance, but made use of it again and again.)

5.2521

I call the continued application of an operation to its own result its successive application (“O′O′O′a” is the result of applying “O′ξ” successively to “a” three times).

In a similar sense I speak of the successive application of several operations to a number of sentences.

5.2522

I therefore write the general term of a series of forms a, O′a, O′O′a, . . . . like this: “[a, x, O′x]”. This expression in brackets is a variable. The first term in the brackets is the beginning of the series of forms, the second the form of any term x of the series, and the third the form of that term of the series which immediately follows x.

5.2523

The concept of the successive application of an operation is equivalent to the concept “and so on”.

5.253

An operation can undo the effect of another. Operations can cancel one another out.

5.254

The operation can disappear (for example, negation in “∼∼p”, ∼∼p = p).

5.3

All sentences are results of truth operations on elementary sentences.

The truth operation is the way in which the truth function arises from elementary sentences.

By the nature of the truth operation, a new truth function arises from truth functions in the same way as their truth function arises from elementary sentences. Every truth operation produces from truth functions of elementary sentences another truth function of elementary sentences, a sentence. The result of every truth operation on the results of truth operations on elementary sentences is again the result of one truth operation on elementary sentences.

Every sentence is the result of truth operations on elementary sentences.

5.31

The schemes in 4.31 also have a meaning when “p”, “q”, “r”, etc. are not elementary sentences.

And it is easy to see that the sentence sign in 4.442 expresses one truth function of elementary sentences even when “p” and “q” are truth functions of elementary sentences.

5.32

All truth functions are results of successively applying a finite number of truth operations to elementary sentences.

5.4

Here it becomes apparent that there are no “logical objects”, “logical constants” (in Frege’s and Russell’s sense).

5.41

For: results of truth operations on truth functions are all identical when they are one and the same truth function of elementary sentences.

5.42

It is obvious that ∨, ⊃, etc. are not relations in the sense of right and left, etc.

The possibility of defining Frege’s and Russell’s logical “primitive signs” in terms of one another already shows that they are not primitive signs, and all the more that they do not designate relations.

And it is apparent that the “⊃” we define by means of “∼” and “∨” is identical with the one by means of which, together with “∼”, we define “∨”, and that this “∨” is identical with the first. And so on.

5.43

It is surely hard to believe from the outset that infinitely many other facts should follow from one fact p, namely ∼∼p, ∼∼∼∼p, etc. And it is no less strange that the infinitely many sentences of logic (of mathematics) follow from half a dozen “basic laws”.

But all the sentences of logic say the same thing. Namely, nothing.

5.44

Truth functions are not material functions.

If, for example, an affirmation can be produced by double negation, is negation then—in any sense—contained in affirmation? Does “∼∼p” negate ∼p, or affirm p; or both?

The sentence “∼∼p” does not deal with negation as an object; but the possibility of negation is already settled in affirmation.

And if there were an object called “∼”, then “∼∼p” would have to say something different from “p”. For one sentence would then be about ∼, the other not.

5.441

This disappearance of the apparent logical constants also occurs when “∼(∃x) . ∼fx” says the same thing as “(x) . fx”, or “(∃x) . fx . x = a” the same thing as “fa”.

5.442

If a sentence is given to us, then with it the results of all truth operations that have it as their base are already given too.

5.45

If there are logical primitive signs, a correct logic must make their positions relative to one another clear and justify their existence. The construction of logic from its primitive signs must become clear.

5.451

If logic has basic concepts, they must be independent of one another. Once a basic concept has been introduced, it must have been introduced in every combination in which it occurs at all. So it cannot be introduced first for one combination, then again for another. For example: once negation has been introduced, we must now understand it in sentences of the form “∼p” just as in sentences such as “∼(p ∨ q)”, “(∃x) . ∼fx”, and others. We must not introduce it first for one class of cases and then for the other, for it would then remain doubtful whether its meaning was the same in both cases, and there would be no reason to use the same kind of combination of signs in both cases.

(In brief, what Frege said in “Grundgesetze der Arithmetik” about introducing signs by definitions applies, mutatis mutandis, to the introduction of primitive signs.)

5.452

Introducing a new device into the symbolism of logic must always be an event with serious consequences. No new device may be introduced into logic—so to speak, with a perfectly innocent face—in brackets or in a footnote.

(Thus, in Russell and Whitehead’s “Principia Mathematica”, definitions and basic laws occur in words. Why suddenly words here? This would need a justification. It is lacking and must be lacking, since the procedure is in fact inadmissible.)

But if introducing a new device has proved necessary at one point, one must immediately ask: where must this device now always be applied? Its place in logic must now be explained.

5.453

All numbers in logic must be capable of being justified.

Or rather: it must turn out that there are no numbers in logic.

There are no privileged numbers.

5.454

In logic nothing stands side by side; there can be no classification.

In logic there cannot be anything more general and anything more specific.

5.4541

The solutions to logical problems must be simple, for they set the standard of simplicity.

People have always sensed that there must be a field of questions whose answers lie—a priori—symmetrically, united in a closed, regular formation.

A field in which the sentence holds: simplex sigillum veri.

5.46

If logical signs were introduced correctly, the sense of all their combinations would thereby already have been introduced too; so not only “p ∨ q” but already “∼(p ∨ ∼q)”, etc., etc. The effect of every possible combination of brackets would thereby already have been introduced too. And it would then have become clear that the actual general primitive signs are not “p ∨ q”, “(∃x) . fx”, etc., but the most general form of their combinations.

5.461

The apparently unimportant fact that apparent logical relations such as ∨ and ⊃ need brackets—unlike actual relations—is significant.

The use of brackets with those apparent primitive signs already points to their not being the actual primitive signs. And surely no one will believe that the brackets have an independent meaning.

5.4611

The signs of logical operations are punctuation marks.

5.47

It is clear that everything that can be said in advance about the form of all sentences must be capable of being said at once.

For all logical operations are already contained in the elementary sentence. “fa” says the same thing as “(∃x) . fx . x = a”.

Where there is composition, there are argument and function, and where these are, all logical constants are already there.

One could say: the one logical constant is what all sentences, by their nature, have in common with one another.

But this is the general sentence form.

5.471

The general sentence form is the essence of the sentence.

5.4711

To give the essence of the sentence is to give the essence of all description, and so the essence of the world.

5.472

The description of the most general sentence form is the description of the one and only general primitive sign of logic.

5.473

Logic must look after itself.

A possible sign must also be able to designate. Everything possible in logic is also permissible. (“Socrates is identical” means nothing because there is no property called “identical”. The sentence is nonsensical because we have not made an arbitrary stipulation, not because the symbol in itself is inadmissible.)

In a certain sense, we cannot make mistakes in logic.

5.4731

The seeming obviousness Russell spoke of so much can be dispensed with in logic only if language itself prevents every logical mistake. — That logic is a priori consists in the fact that one cannot think illogically.

5.4732

We cannot give a sign the wrong sense.

5.47321

Occam’s maxim is of course neither an arbitrary rule nor a rule justified by its practical success: it says that unnecessary sign units mean nothing.

Signs that serve one and the same purpose are logically equivalent; signs that serve no purpose are logically without meaning.

5.4733

Frege says: every properly formed sentence must have a sense; and I say: every possible sentence is properly formed, and if it has no sense, this can only be because we have given some of its parts no meaning.

(Even if we believe we have done so.)

Thus “Socrates is identical” says nothing because we have given the word “identical” no meaning as a word for a property. For when it occurs as an equality sign, it symbolises in quite a different way—the designating relation is different—and so the symbol too is quite different in the two cases; the two symbols happen to share only the sign.

5.474

The number of basic operations needed depends only on our notation.

5.475

All that matters is to form a system of signs with a particular number of dimensions—with a particular mathematical multiplicity.

5.476

It is clear that what is at issue here is not a number of basic concepts that must be designated, but the expression of a rule.

5.5

Every truth function is a result of successively applying the operation (– – – – –W)(ξ, . . . .) to elementary sentences.

This operation negates all the sentences in the right-hand brackets, and I call it the negation of those sentences.

5.501

I indicate an expression in brackets whose terms are sentences—when the order of the terms in the brackets does not matter—by a sign of the form “(ξ¯)”. “ξ” is a variable whose values are the terms of the expression in brackets; and the bar over the variable indicates that it stands for all its values in the brackets.

(So if, for example, ξ has the three values P, Q, R, then (ξ¯)=(P,Q,R).)

The values of the variable are fixed.

Fixing them is describing the sentences that the variable stands for.

How the terms of the expression in brackets are described is inessential.

We can distinguish three kinds of description: 1. Direct enumeration. In this case we can simply put the variable’s constant values in its place. 2. Giving a function fx whose values for all values of x are the sentences to be described. 3. Giving a formal law according to which those sentences are formed. In this case the terms of the expression in brackets are all the terms of a series of forms.

5.502

So instead of “(– – – – –W)(ξ, . . . .)” I write “N(ξ¯)”.

N(ξ¯) is the negation of all the values of the sentence variable ξ.

5.503

Since it is evidently easy to express how sentences can be formed with this operation and how sentences cannot be formed with it, it must also be possible to give this an exact expression.

5.51

If ξ has only one value, then N(ξ¯) = ∼p (not p); if it has two values, then N(ξ¯) = ∼p . ∼q (neither p nor q).

5.511

How can logic, which encompasses everything and mirrors the world, use such special hooks and manipulations? Only by all of these linking together into an infinitely fine network, into the great mirror.

5.512

“∼p” is true if “p” is false. So in the true sentence “∼p”, “p” is a false sentence. How, then, can the stroke “∼” make it agree with reality?

But what negates in “∼p” is not the “∼”, but what is shared by all the signs of this notation that negate p.

So, the common rule according to which “∼p”, “∼∼∼p”, “∼p ∨ ∼p”, “∼p . ∼p”, etc., etc. (ad inf.) are formed. And what is common mirrors negation.

5.513

One could say: what all symbols that affirm both p and q have in common is the sentence “p . q”. What all symbols that affirm either p or q have in common is the sentence “p ∨ q”.

And so one can say: two sentences are opposed to one another if they have nothing in common, and: every sentence has only one negative, because there is only one sentence that lies entirely outside it.

In this way it also shows itself in Russell’s notation that “q : p ∨ ∼p” says the same thing as “q”; that “p ∨ ∼p” says nothing.

5.514

Once a notation is fixed, there is in it a rule according to which all sentences negating p are formed, a rule according to which all sentences affirming p are formed, a rule according to which all sentences affirming p or q are formed, and so on. These rules are equivalent to the symbols, and their sense is mirrored in them.

5.515

It must show itself in our symbols that what is joined together by “∨”, “.”, etc. must be sentences.

And this is indeed the case, for the symbols “p” and “q” themselves presuppose “∨”, “∼”, etc. If the sign “p” in “p ∨ q” does not stand for a complex sign, it cannot have a sense on its own; but then the signs “p ∨ p”, “p . p”, etc., which have the same sense as “p”, cannot have a sense either. But if “p ∨ p” has no sense, then “p ∨ q” cannot have a sense either.

5.5151

Must the sign of the negative sentence be formed using the sign of the positive one? Why should it not be possible to express the negative sentence by a negative fact? (For example: if “a” does not stand in a particular relation to “b”, this could express that aRb is not the case.)

But here too the negative sentence is formed indirectly by the positive one.

The positive sentence must presuppose the existence of the negative sentence, and vice versa.

5.52

If the values of ξ are all the values of a function fx for all values of x, then N(ξ¯) = ∼(∃x) . fx.

5.521

I separate the concept all from the truth function.

Frege and Russell introduced generality in connection with the logical product or the logical sum. This made it difficult to understand the sentences “(∃x) . fx” and “(x) . fx”, in which both ideas are contained.

5.522

What is distinctive about the notation for generality is, first, that it points to a logical prototype, and second, that it brings out constants.

5.523

The notation for generality occurs as an argument.

5.524

If the objects are given, then all objects are thereby already given to us too.

If the elementary sentences are given, then all elementary sentences are thereby given too.

5.525

It is incorrect to render the sentence “(∃x) . fx” in words—as Russell does—by “fx is possible”.

The certainty, possibility or impossibility of a situation is not expressed by a sentence, but by an expression’s being a tautology, a sentence with sense, or a contradiction.

That precedent to which one always wants to appeal must already lie in the symbol itself.

5.526

The world can be described completely by fully generalised sentences, that is, without assigning any name in advance to a particular object.

To arrive at the ordinary way of expressing ourselves, we then need simply say, after an expression “there is one and only one x which . . . .”: and this x is a.

5.5261

A fully generalised sentence is, like every other sentence, composed of parts. (This shows itself in our having to mention “ϕ” and “x” separately in “(∃x, ϕ).ϕx”. Both stand independently in designating relations to the world, as in the ungeneralised sentence.)

Mark of a composite symbol: it has something in common with other symbols.

5.5262

For the truth or falsity of every sentence changes something in the general structure of the world. And the room left in its structure by the totality of elementary sentences is exactly the room bounded by the completely general sentences.

(If an elementary sentence is true, then in any case there is one more elementary sentence that is true.)

5.53

I express sameness of the object through sameness of the sign, not with the help of an equals sign. Difference between objects through difference between signs.

5.5301

That identity is not a relation between objects is clear. This becomes quite clear when one considers, for example, the sentence “(x) : fx. ⊃ .x = a”. What this sentence says is simply that only a satisfies the function f, not that only those things satisfy the function f which have a certain relation to a.

Of course, one could now say that only a has this relation to a, but to express this we would need the equals sign itself.

5.5302

Russell’s definition of “=” is not enough; for according to it one cannot say that two objects have all their properties in common. (Even if this sentence is never correct, it still has sense.)

5.5303

In passing: to say of two things that they are identical is nonsense, and to say of one thing that it is identical with itself says nothing at all.

5.531

So I write not “f (a, b) . a = b”, but “f (a, a)” (or “f (b, b)”). And not “f (a, b) . ∼a = b”, but “f (a, b)”.

5.532

And similarly: not “(∃x, y) . f (x, y) . x = y”, but “(∃x) . f (x, x)”; and not “(∃x, y) . f (x, y) . ∼x = y”, but “(∃x, y) . f (x, y)”.

(So in place of Russell’s “(∃x, y) . f (x, y)”: “(∃x, y) . f (x, y) . ∨ . (∃x) . f (x, x)”.)

5.5321

So instead of “(x) : fx ⊃ x = a” we write, for example, “(∃x) . fx. ⊃ .fa : ∼(∃x, y) . fx . fy”.

And the sentence “only one x satisfies f ()” reads: “(∃x) . fx : ∼(∃x, y) . fx . fy”.

5.533

So the equals sign is not an essential component of conceptual notation.

5.534

And now we see that pseudo-sentences such as “a = a”, “a = b . b = c. ⊃ a = c”, “(x) . x = x”, “(∃x) . x = a”, etc., cannot be written down at all in a correct conceptual notation.

5.535

This also settles all the problems that were bound up with such pseudo-sentences.

All the problems that Russell’s “Axiom of infinity” brings with it are to be solved here already.

What the axiom of infinity is supposed to say would be expressed in language by there being infinitely many names with different meanings.

5.5351

There are certain cases in which one is tempted to use expressions of the form “a = a” or “p ⊃ p” and the like. This happens when one wants to speak about the prototype: sentence, thing, etc. Thus in the “Principles of Mathematics” Russell rendered the nonsense “p is a sentence” in symbols as “p ⊃ p” and put it as a hypothesis before certain sentences, so that their argument places could be occupied only by sentences.

(It is already nonsense to put the hypothesis p ⊃ p before a sentence to secure arguments of the right form for it, because with something other than a sentence as argument the hypothesis becomes not false but nonsensical, and because the sentence itself becomes nonsensical through arguments of the wrong kind, and so guards itself just as well, or just as poorly, against the wrong arguments as does the hypothesis without sense attached for this purpose.)

5.5352

In the same way, one wanted to express “There are no things” by “∼(∃x) . x = x”. But even if this were a sentence—would it not also be true if “there were things”, but these were not identical with themselves?

5.54

In the general form of a sentence, a sentence occurs in a sentence only as a base of truth operations.

5.541

At first sight it seems as if a sentence could also occur in another in a different way.

Especially in certain forms of sentence in psychology, such as “A believes that p is the case”, or “A thinks p”, etc.

For on the surface it seems here as if the sentence p stood in some kind of relation to an object A.

(And in modern epistemology (Russell, Moore, etc.) these sentences have also been understood in this way.)

5.542

But it is clear that “A believes that p”, “A thinks p”, “A says p” are of the form “‘p’ says p”: and here it is a matter not of pairing a fact and an object, but of pairing facts by pairing their objects.

5.5421

This also shows that the soul—the subject, etc.—as it is understood in today’s superficial psychology, is an impossibility.

For a composite soul would no longer be a soul.

5.5422

The correct explanation of the form of the sentence “A judges p” must show that it is impossible to make a judgment that is nonsense. (Russell’s theory does not meet this condition.)

5.5423

To perceive a complex is to perceive that its component parts are related to one another in such-and-such a way.

This probably also explains why the figure

A cube with its eight vertices marked a and b, which can be seen in two ways.

can be seen in two ways as a cube; and all similar phenomena. For we really do see two different facts.

(If I first look at the corners a and only glance at b, then a appears in front; and conversely.)

5.55

We must now answer a priori the question of all the possible forms of elementary sentences.

The elementary sentence consists of names. But since we cannot give the number of names with different meanings, we also cannot say how the elementary sentence is composed.

5.551

Our principle is that every question that can be decided by logic at all must be decidable without further ado.

(And if we find ourselves having to answer such a problem by looking at the world, this shows that we are on a fundamentally wrong track.)

5.552

The “experience” we need in order to understand logic is not that something is so-and-so, but that something is: but this is precisely no experience.

Logic is before every experience—that something is so. It is before the How, not before the What.

5.5521

And if this were not so, how could we apply logic? One could say: if there were logic even if there were no world, how could there then be logic, given that there is a world.

5.553

Russell said that there were simple relations between different numbers of things (individuals). But between which numbers? And how is this to be decided?—By experience?

(There is no privileged number.)

5.554

Giving any particular form would be completely arbitrary.

5.5541

It should be possible to say a priori whether, for example, I could find myself having to designate something with the sign for a 27-place relation.

5.5542

But may we ask in this way at all? Can we set up a form of sign and not know whether anything could correspond to it?

Does the question have a sense: What must exist so that something can be the case?

5.555

It is clear that we have a concept of the elementary sentence apart from its particular logical form.

But where symbols can be formed according to a system, it is this system that is logically important, not the individual symbols.

And how could it be possible for me, in logic, to deal with forms that I can invent; I must instead deal with what makes it possible for me to invent them.

5.556

There can be no hierarchy of the forms of elementary sentences. We can foresee only what we construct ourselves.

5.5561

Empirical reality is bounded by the totality of objects. The boundary shows itself again in the totality of elementary sentences.

The hierarchies are, and must be, independent of reality.

5.5562

If we know on purely logical grounds that there must be elementary sentences, then everyone who understands sentences in their unanalysed form must know it.

5.5563

All the sentences of our ordinary language are in fact, just as they are, in perfect logical order.—That simplest thing which we are to give here is not a likeness of the truth, but the full truth itself.

(Our problems are not abstract, but perhaps the most concrete there are.)

5.557

The application of logic decides which elementary sentences there are.

Logic cannot anticipate what lies in its application.

This is clear: logic must not collide with its application.

But logic must touch its application.

So logic and its application must not encroach on one another.

5.5571

If I cannot give the elementary sentences a priori, then wanting to give them must lead to obvious nonsense.

5.6

The limits of my language mean the limits of my world.

5.61

Logic fills the world; the limits of the world are also its limits.

So in logic we cannot say: this and this exist in the world, that does not.

For this would seem to presuppose that we rule out certain possibilities, and this cannot be the case, since logic would otherwise have to go beyond the limits of the world; if, that is, it could view these limits from the other side too.

What we cannot think, we cannot think; so we also cannot say what we cannot think.

5.62

This remark gives the key to deciding the question of how far solipsism is a truth.

For what solipsism means is quite right, only it cannot be said, but shows itself.

That the world is my world shows itself in the fact that the limits of the language (the language which I alone understand) mean the limits of my world.

5.621

The world and life are one.

5.63

I am my world. (The microcosm.)

5.631

The thinking, imagining subject does not exist.

If I wrote a book, “The world as I found it”, I would also have to report in it on my body and say which limbs are subject to my will and which are not, etc.; for this is a method of isolating the subject, or rather of showing that in an important sense there is no subject: for it alone could not be spoken of in this book.—

5.632

The subject does not belong to the world, but is a boundary of the world.

5.633

Where in the world is a metaphysical subject to be noticed?

You say that it is just as with the eye and the visual field. But you really do not see the eye.

And nothing about the visual field allows one to infer that it is seen by an eye.

5.6331

For the visual field does not, for example, have a shape like this:

Diagram labelled Auge (eye), showing an eye at the pointed end of an enclosed visual field.
5.634

This is connected with the fact that no part of our experience is also a priori.

Everything we see could also be different.

Everything we can describe at all could also be different.

There is no a priori order of things.

5.64

Here one sees that solipsism, strictly carried through, coincides with pure realism. The I of solipsism shrinks to a point without extension, and the reality coordinated with it remains.

5.641

So there really is a sense in which one can speak in philosophy about the I in a non-psychological way.

The I enters philosophy through the fact that “the world is my world”.

The philosophical I is not the human being, not the human body or the human soul that psychology deals with, but the metaphysical subject, the boundary—not a part of the world.

6

The general form of a truth-function is: [p¯,ξ¯,N(ξ¯)].

This is the general form of a sentence.

6.001

This says nothing other than that every sentence is a result of successively applying the operation N′(ξ¯) to the elementary sentences.

6.002

If the general form of how a sentence is built is given, then the general form of how another sentence can be produced from a sentence by an operation is already given with it.

6.01

So the general form of the operation Ω′(η¯) is:
[ξ¯,N(ξ¯)]′(η¯)(=[η¯,ξ¯,N(ξ¯)]).

This is the most general form of the transition from one sentence to another.

6.02

And so we come to numbers: I define

x=Ω0′xDef. and
Ω′Ων′x=Ων+1′xDef.

Following these rules for signs, we therefore write the series

x,Ω′x,Ω′Ω′x,Ω′Ω′Ω′x,…
as: Ω0′x,Ω0+1′x,Ω0+1+1′x,Ω0+1+1+1′x,…

So instead of “[x,ξ,Ω′ξ]” I write:

“[Ω0′x,Ων′x,Ων+1′x]”.

And define:

0+1=1Def.
0+1+1=2Def.
0+1+1+1=3Def.

(And so on.)

6.021

A number is the exponent of an operation.

6.022

The concept of number is nothing other than what all numbers have in common, the general form of number.

The concept of number is the variable number.

And the concept of numerical equality is the general form of all particular numerical equalities.

6.03

The general form of a whole number is: [0,ξ,ξ+1].

6.031

The theory of classes is entirely unnecessary in mathematics.

This is connected with the fact that the generality we need in mathematics is not accidental generality.

6.1

The sentences of logic are tautologies.

6.11

So the sentences of logic say nothing. (They are the analytic sentences.)

6.111

Theories that make a sentence of logic seem to have content are always false. One might believe, for example, that the words “true” and “false” designate two properties among other properties, and then it would seem a curious fact that every sentence has one of these properties. This now seems anything but self-evident, no more self-evident than, say, the sentence “All roses are either yellow or red” would sound, even if it were true. Indeed, that sentence now takes on the full character of a sentence of natural science, and this is the sure sign that it has been taken wrongly.

6.112

The correct explanation of logical sentences must give them a unique position among all sentences.

6.113

It is the distinctive mark of logical sentences that one can recognize from the symbol alone that they are true, and this fact contains the whole philosophy of logic. And so it is also one of the most important facts that the truth or falsity of non-logical sentences cannot be recognized from the sentence alone.

6.12

That the sentences of logic are tautologies shows the formal—logical—properties of language, of the world.

That their components, connected in this way, yield a tautology characterizes the logic of their components.

For sentences, connected in a certain way, to yield a tautology, they must have certain structural properties. That they yield a tautology when connected in this way therefore shows that they have these structural properties.

6.1201

That, for example, the sentences “p” and “∼p” yield a tautology in the combination “∼(p.∼p)” shows that they contradict one another. That the sentences “p ⊃ q”, “p”, and “q”, connected in the form “(p ⊃ q).(p) : ⊃ : (q)”, yield a tautology shows that q follows from p and p ⊃ q. That “(x) . fx : ⊃ : fa” is a tautology shows that fa follows from (x) . fx. Etc., etc.

6.1202

It is clear that one could also use contradictions instead of tautologies for the same purpose.

6.1203

To recognize a tautology as such, in cases where no notation for generality occurs in it, one can use the following visual method: instead of “p”, “q”, “r”, etc., I write “WpF”, “WqF”, “WrF”, etc. I express the truth-combinations by brackets. For example:

Brackets joining the truth-combinations of p and q; W and F are the original labels.

and the assignment of the truth or falsity of the whole sentence to the truth-combinations of the truth-arguments by lines, as follows:

Lines assigning W or F to the truth-combinations of p and q.

So this sign would represent, for example, the sentence p ⊃ q. Now I want to examine, for example, the sentence ∼(p . ∼p) (the law of contradiction), to see whether it is a tautology. In our notation the form “∼ξ” is written like this:

The diagram for negation, with the original W, F, and xi labels.

and the form “ξ . η” like this:

The diagram for conjunction, with the original W, F, xi, and eta labels.

So the sentence ∼(p . ∼q) is written like this:

The combined diagram for the sentence not (p and not q), with the original labels.

If we put “p” in place of “q” here and examine how the outermost W and F are connected with the innermost, we find that the truth of the whole sentence is assigned to all the truth-combinations of its argument, its falsity to none of the truth-combinations.

6.121

The sentences of logic demonstrate the logical properties of sentences by connecting them into sentences that say nothing.

One could also call this method a zero method. In a logical sentence, sentences are brought into balance with one another, and the state of balance then indicates how these sentences must be constituted logically.

6.122

It follows that we can also do without logical sentences, since in a suitable notation we can recognize the formal properties of sentences simply by looking at these sentences.

6.1221

If, for example, two sentences “p” and “q” yield a tautology in the combination “p ⊃ q”, then it is clear that q follows from p.

That, for example, “q” follows from “p ⊃ q . p” can be seen from these two sentences themselves, but we can also show it this way: by connecting them into “p ⊃ q . p : ⊃ : q” and then showing that this is a tautology.

6.1222

This throws light on the question of why logical sentences cannot be confirmed by experience any more than they can be refuted by experience. A sentence of logic must not only be incapable of being refuted by any possible experience; it must also be incapable of being confirmed by such an experience.

6.1223

Now it becomes clear why people often felt as if we ought to “demand” the “logical truths”: we can indeed demand them insofar as we can demand an adequate notation.

6.1224

It now also becomes clear why logic was called the study of forms and of inference.

6.123

It is clear: the laws of logic must not themselves be subject in turn to logical laws.

(There is not, as Russell thought, a separate law of contradiction for each “type”; one is enough, since it is not applied to itself.)

6.1231

The mark of a logical sentence is not universal validity.

Being general means only: happening to hold for all things. An ungeneralized sentence can be just as tautological as a generalized one.

6.1232

One could call logical universal validity essential, in contrast to the accidental kind, as in the sentence “All human beings are mortal”. Sentences such as Russell’s “Axiom of reducibility” are not logical sentences, and this explains our feeling that, if true, they could still be true only by a fortunate accident.

6.1233

One can think of a world in which the axiom of reducibility does not hold. But it is clear that logic has nothing to do with the question of whether our world is actually like that or not.

6.124

Logical sentences describe the framework of the world, or rather, they represent it. They are not “about” anything. They presuppose that names have meaning and elementary sentences have sense: and this is their connection with the world. It is clear that it must indicate something about the world that certain combinations of symbols—which essentially have a particular character—are tautologies. This is the decisive point. We said that some things about the symbols we use were arbitrary, some not. In logic only the latter expresses anything: but this means that, in logic, it is not we who use signs to express what we want; in logic the nature of the inherently necessary signs itself speaks: once we know the logical syntax of any sign-language, all the sentences of logic are already given.

6.125

It is possible, even on the old view of logic, to give a description in advance of all “true” logical sentences.

6.1251

That is why there can never be surprises in logic.

6.126

Whether a sentence belongs to logic can be calculated by calculating the logical properties of the symbol. And this is what we do when we “prove” a logical sentence. For, without concerning ourselves with sense or meaning, we form the logical sentence from others according to mere rules for signs.

The proof of logical sentences consists in making them arise from other logical sentences by successively applying certain operations that produce tautologies from the initial ones over and over again. (And only tautologies follow from a tautology.)

Of course, this way of showing that its sentences are tautologies is entirely inessential to logic. If only because the sentences from which the proof starts must show without proof that they are tautologies.

6.1261

In logic, process and result are equivalent. (Hence no surprise.)

6.1262

Proof in logic is only a mechanical aid for making a tautology easier to recognize when it is complicated.

6.1263

It would indeed be too strange if one could prove a sentence that has sense logically from others, and a logical sentence as well. It is clear from the start that the logical proof of a sentence that has sense and proof in logic must be two quite different things.

6.1264

A sentence that has sense says something, and its proof shows that it is so; in logic every sentence is the form of a proof.

Every sentence of logic is a modus ponens represented in signs. (And modus ponens cannot be expressed by a sentence.)

6.1265

Logic can always be understood in such a way that every sentence is its own proof.

6.127

All sentences of logic stand on an equal footing; there is no essential distinction among them between basic laws and derived sentences.

Every tautology shows for itself that it is a tautology.

6.1271

It is clear that the number of “basic laws of logic” is arbitrary, since one could derive logic from a single basic law simply by forming, for example, the logical product of Frege’s basic laws. (Frege would perhaps say that this basic law was no longer immediately evident. But it is strange that so exact a thinker as Frege appealed to how evident something is as a criterion for a logical sentence.)

6.13

Logic is not a doctrine but a mirror image of the world. Logic is transcendental.

6.2

Mathematics is a logical method.

The sentences of mathematics are equations, and so are pseudo-sentences.

6.21

A mathematical sentence expresses no thought.

6.211

In life, it is never the mathematical sentence that we need; we use the mathematical sentence only to infer, from sentences that do not belong to mathematics, others that likewise do not belong to mathematics.

(In philosophy the question “What do we actually use that word, that sentence for?” leads again and again to valuable insights.)

6.22

The logic of the world, which the sentences of logic show in tautologies, mathematics shows in equations.

6.23

When two expressions are connected by the equals sign, this means that they can replace one another. But whether this is the case must show itself in the two expressions themselves.

That they can replace one another characterizes the logical form of two expressions.

6.231

It is a property of affirmation that it can be taken as double negation.

It is a property of “1 + 1 + 1 + 1” that it can be taken as “(1 + 1) + (1 + 1)”.

6.232

Frege says that the two expressions have the same meaning but different senses.

But what is essential about an equation is that it is not needed to show that the two expressions joined by the equals sign have the same meaning, since this can be seen from the two expressions themselves.

6.2321

And that mathematical sentences can be proved means nothing other than that their correctness can be seen without what they express itself having to be compared with the facts for correctness.

6.2322

The identity of the meaning of two expressions cannot be asserted. For to be able to assert anything about their meaning, I must know their meaning: and in knowing their meaning, I know whether they mean the same or different things.

6.2323

The equation marks only the standpoint from which I look at the two expressions, namely the standpoint of their sameness of meaning.

6.233

The question of whether intuition is needed to solve mathematical problems must be answered by saying that here language itself supplies the necessary intuition.

6.2331

The process of calculating provides precisely this intuition.

Calculation is not an experiment.

6.234

Mathematics is a method of logic.

6.2341

What is essential to the mathematical method is working with equations. For it is because of this method that every mathematical sentence must be clear on its own.

6.24

The method by which mathematics arrives at its equations is the method of substitution.

For equations express that two expressions can replace one another, and we move from a number of equations to new equations by replacing expressions with others in accordance with the equations.

6.241

So the proof of the sentence 2 × 2 = 4 runs as follows:

(Ων)μ′x=Ων×μ′xDef.
Ω2×2′x=(Ω2)2′x=(Ω2)1+1′x=Ω2′Ω2′x=Ω1+1′Ω1+1′x
(Ω′Ω)′(Ω′Ω)′x=Ω′Ω′Ω′Ω′x=Ω1+1+1+1′x=Ω4′x
6.3

To investigate logic is to investigate all lawfulness. And outside logic everything is chance.

6.31

The so-called law of induction cannot in any case be a logical law, since it is plainly a sentence that has sense. — And that is why it cannot be an a priori law either.

6.32

The law of causality is not a law but the form of a law.

6.321

“Law of causality” is a name for a kind. And just as in mechanics there are, say, minimum laws—for example, the law of least action—so in physics there are causal laws, laws of the causal form.

6.3211

People already had an inkling that there must be a “law of least action” before they knew exactly how it went. (Here, as always, what is certain a priori turns out to be something purely logical.)

6.33

We do not believe a priori in a law of conservation; we know a priori the possibility of a logical form.

6.34

All those sentences, such as the principle of reason, of continuity in nature, of least expenditure in nature, etc., etc.—all these are a priori insights into the possible ways of giving form to the sentences of science.

6.341

Newtonian mechanics, for example, brings the description of the world into a uniform form. Let us imagine a white surface with irregular black patches on it. We now say: whatever picture this produces, I can always come as close as I like to describing it by covering the surface with a sufficiently fine square grid and then saying of each square that it is white or black. In this way I will have brought the description of the surface into a uniform form. This form is arbitrary, for I could have used a grid of triangular or hexagonal mesh with the same success. It may be that a description using a triangular grid would have been simpler; that is, we could describe the surface more accurately with a coarser triangular grid than with a finer square one (or the other way round), and so on. Different grids correspond to different systems for describing the world. Mechanics determines a form for describing the world by saying: all sentences that describe the world must be obtained in a given way from a number of given sentences—the axioms of mechanics. In this way it supplies the building blocks for the scientific building and says: whatever building you want to put up, you must somehow put it together with these building blocks and these alone.

(Just as any number can be written down with the number system, so any sentence of physics must be writable with the system of mechanics.)

6.342

And now we see the positions of logic and mechanics in relation to one another. (The grid could also consist of different kinds of figures, say triangles and hexagons.) That a picture like the one mentioned earlier can be described by a grid of a given form says nothing about the picture. (For this holds for every picture of this kind.) But this characterizes the picture: that it can be described completely by a particular grid of a particular fineness.

Likewise, that the world can be described by Newtonian mechanics says nothing about the world; but that it can be described by that mechanics in the way it actually is described does say something. That the world can be described more simply by one mechanics than by another also says something about the world.

6.343

Mechanics is an attempt to construct, according to a single plan, all the true sentences that we need for describing the world.

6.3431

Through the whole logical apparatus, the laws of physics still speak of the objects of the world.

6.3432

We must not forget that the description of the world by mechanics is always completely general. In it there is never, for example, any talk of particular material points, but always only of some or other.

6.35

Although the patches in our picture are geometrical figures, geometry can of course say nothing at all about their actual shape and position. But the grid is purely geometrical; all its properties can be given a priori.

Laws such as the principle of reason, etc., deal with the grid, not with what the grid describes.

6.36

If there were a law of causality, it could run: “There are laws of nature”.

But of course this cannot be said: it shows itself.

6.361

In Hertz’s way of speaking, one could say: only connections governed by laws can be thought.

6.3611

We cannot compare a process with the “passage of time”—there is no such thing—but only with another process (say, the running of a chronometer).

So a course in time can be described only by relying on another process.

Exactly the same applies to space. Where, for example, it is said that neither of two events (which exclude one another) can occur because there is no cause for one to occur rather than the other, what is really at issue is that we cannot describe even one of the two events unless there is some asymmetry. And if such an asymmetry exists, we can take it as the cause of one occurring and the other not occurring.

6.36111

Kant’s problem of the right and left hands, which cannot be made to coincide, already exists in a plane, indeed in one-dimensional space, where the two congruent figures a and b likewise cannot be made to coincide without being moved out of this space. Right and left hands are in fact completely congruent. And the fact that they cannot be made to coincide has nothing to do with this.

The two congruent one-dimensional figures a and b, oriented in opposite directions.

One could put the right-hand glove on the left hand if one could turn it around in four-dimensional space.

6.362

What can be described can also happen, and what the law of causality is supposed to rule out cannot be described either.

6.363

The procedure of induction consists in accepting the simplest law that can be brought into agreement with our experience.

6.3631

But this procedure has no logical justification, only a psychological one.

It is clear that there is no reason to believe that the simplest case will now actually occur.

6.36311

That the sun will rise tomorrow is a hypothesis; and this means that we do not know whether it will rise.

6.37

There is no compulsion that makes one thing have to happen because something else has happened. There is only logical necessity.

6.371

The whole modern view of the world rests on the illusion that the so-called laws of nature are explanations of natural phenomena.

6.372

So people stop at the laws of nature as at something beyond challenge, just as earlier people stopped at God and fate.

And both are right, and wrong. The people of old are admittedly clearer insofar as they recognize a clear stopping point, whereas in the new system it is supposed to seem as if everything had been explained.

6.373

The world is independent of my will.

6.374

Even if everything we wished for happened, this would still be only, so to speak, a grace of fate, since there is no logical connection between will and world to guarantee it, and we could not in turn will the supposed physical connection itself.

6.375

Just as there is only logical necessity, so there is only logical impossibility.

6.3751

That, for example, two colours are at the same place in the visual field at the same time is impossible, and logically impossible, since it is ruled out by the logical structure of colour.

Let us think about how this contradiction appears in physics: roughly, that a particle cannot have two velocities at the same time; that is, that it cannot be in two places at the same time; that is, that particles in different places at one time cannot be identical.

(It is clear that the logical product of two elementary sentences can be neither a tautology nor a contradiction. The statement that a point in the visual field has two different colours at the same time is a contradiction.)

6.4

All sentences have equal value.

6.41

The sense of the world must lie outside it. In the world everything is as it is and everything happens as it happens; there is no value in it—and if there were, it would have no value.

If there is a value that has value, it must lie outside all happening and all being as things are. For all happening and all being as things are is accidental.

What makes it non-accidental cannot lie in the world, since then this too would be accidental.

It must lie outside the world.

6.42

That is why there can be no sentences of ethics either. Sentences can express nothing higher.

6.421

It is clear that ethics cannot be put into words. Ethics is transcendental.

(Ethics and aesthetics are one.)

6.422

The first thought when an ethical law of the form “You ought to . . . .” is laid down is: And what if I do not do it? But it is clear that ethics has nothing to do with punishment and reward in the ordinary sense. So this question about the consequences of an action must be irrelevant. — At the very least, these consequences must not be events. For there must, after all, be something right about that way of putting the question. There must indeed be a kind of ethical reward and ethical punishment, but they must lie in the action itself.

(And it is also clear that the reward must be something pleasant, the punishment something unpleasant.)

6.423

The will, as the bearer of the ethical, cannot be spoken of.

And the will as a phenomenon interests only psychology.

6.43

If good or evil willing changes the world, it can change only the boundaries of the world, not the facts; not what can be expressed through language.

In short, the world must then become altogether different as a result. It must, so to speak, shrink or grow as a whole.

The world of the happy person is different from that of the unhappy person.

6.431

Just as, in death too, the world does not change but ends.

6.4311

Death is not an event in life. We do not experience death.

If by eternity we mean not an endless duration but timelessness, then whoever lives in the present lives eternally.

Our life is as endless as our visual field is without boundaries.

6.4312

The temporal immortality of the human soul—that is, its continuing to live forever even after death—is not only in no way guaranteed; above all, this assumption does not accomplish at all what people have always wanted to accomplish with it. Is a riddle solved by my living on forever? Is this eternal life not then just as puzzling as the present one? The solution to the riddle of life in space and time lies outside space and time.

(After all, it is not problems of natural science that are to be solved.)

6.432

How the world is makes no difference at all to what is higher. God does not reveal himself in the world.

6.4321

The facts all belong only to the task, not to the solution.

6.44

It is not how the world is that is mystical, but that it is.

6.45

To view the world sub specie aeterni is to view it as a—bounded—whole.

The feeling of the world as a bounded whole is the mystical feeling.

6.5

For an answer that cannot be put into words, the question cannot be put into words either.

The riddle does not exist.

If a question can be asked at all, it can also be answered.

6.51

Scepticism is not irrefutable but plainly nonsensical when it tries to doubt where no question can be asked.

For doubt can exist only where a question exists; a question only where an answer exists, and an answer only where something can be said.

6.52

We feel that even when all possible scientific questions have been answered, the problems of our lives have still not been touched at all. Admittedly, no question remains then; and this very thing is the answer.

6.521

We notice the solution to the problem of life in the disappearance of this problem.

(Is this not the reason why people to whom the sense of life became clear after long doubts—why they then could not say what this sense consisted in.)

6.522

There is indeed what cannot be put into words. It shows itself; it is what is mystical.

6.53

The correct method in philosophy would really be this: to say nothing but what can be said, that is, sentences of natural science—and so something that has nothing to do with philosophy—and then, whenever someone else wanted to say something metaphysical, to show them that they had given no meaning to certain signs in their sentences. This method would be unsatisfying to the other person—they would not feel that we were teaching them philosophy—but it would be the only strictly correct one.

6.54

My sentences clarify in this way: anyone who understands me eventually recognizes them as nonsensical, once they have climbed through them—on them—beyond them. (They must, so to speak, throw away the ladder after they have climbed up it.)

They must get beyond these sentences; then they see the world rightly.

7

What we cannot speak about, we must be silent about.

1.

The decimal numbers used to number the individual sentences indicate their logical weight, the emphasis placed on them in my presentation. Sentences n.1, n.2, n.3, etc. are remarks on sentence no. n; sentences n.m1, n.m2, etc. are remarks on sentence no. n.m; and so on.

German original and English UX translation from UX Translation Universal (2026-11-19). Source texts unchanged.