Home - Wittgenstein's Logic of Language | Bibliography

If everything speaks for, nothing against, the truth of a proposition

Do we not know whether the sun will rise tomorrow -- what does Wittgenstein mean by the word 'know' in his Tractatus?

Outline of this page ...

Related page: Remarks about Wittgenstein's Tractatus Logico-Philosophicus, the meaning of the title of Wittgenstein's book, and its ideas about language-meaning, and ethics.

"Everything speaks in its favor, nothing against it"

Everything speaks in its favor, nothing against it. (OC § 4; cf. ibid. § 338: "...merely extremely probable")

Rational doubt requires -- grammatically requires -- rational grounds -- i.e. objective grounds: a doubt and a method to remove that doubt. "What we can ask is if it can make sense to doubt it" (ibid. § 2) -- i.e. whether in a particular type of case a doubt would be sense or nonsense. "... rules of caution only make sense if they come to an end somewhere. A doubt without an end is not even a doubt" (ibid. § 625). But does that remark, which is a grammatical remark, fly in the face of quantum theory -- i.e. does quantum theory require a different grammatical-conceptual scheme, i.e. different rules of grammar? "Because it is impossible to trace a strict-causality mechanism -- i.e. to delimit real possibility -- it is impossible to determine whether the water in the teapot will freeze or boil. About which will happen, there is instead a doubt without an end." (The contrast between language use in science versus in philosophy: justification versus description.) Well, but is a possibility which no reasonable man takes seriously what we call a doubt?

Do we call the occurrence of an event that has never been known to happen merely extremely improbable? (In this case, that characterization is justified by a scientific theory, namely quantum theory. If that theory is replaced by another, e.g. of tracing mechanisms, strict causality, it may become nonsense to characterize such an event as improbable rather than as a real impossibility, although not a logical impossibility.)

For whether or not it is theoretically possible for the water in a teapot to freeze when placed over a fire, it will always be logically possible -- i.e. describable, which is the lowest standard of possibility: language that is not nonsense (i.e. undefined combinations of words).

Then is it correct to say that in the language of everyday, it is not possible (real possibility) for the water to boil, whereas in the language of quantum theory it is (a real possibility)? That in the language of everyday we know that the water will not freeze, whereas in the language of quantum theory we do not?

(What if those two languages used different signs rather than both using the English language? Would there be an equivalent sign for the English word 'know' in the quantum language? And for 'real impossibility'?)

Are all events accidental - Is observed regularity mere illusion?

[Comment: I suspect that I've quite misunderstood Wittgenstein's meaning in the following discussion.]

... the investigation of all regularity. And outside logic all is accident.... A necessity for one thing to happen because another has happened does not exist. There is only logical necessity. (TLP 6.3, 6.37, tr. Ogden)

Does the TLP mean that observed regularity is not really regularity, as if everything that happened were accidental, or, happened at random [ibid. 1.21] -- as if there were no laws of nature? But, of course, we don't use the word 'accidental' that way. Of course it is logically possible for something else to happen -- but that only means we can describe what it would be like if ...

But logical possibility does not imply, much less entail, real possibility (Russell's discussion of 'The golden mountain does not exist'). Eddington, on the other hand, seems to say that it does -- that (according to quantum theory) a logical possibility does imply a real possibility, e.g. water in a teapot placed over a fire may freeze or do most anything else as well . But no one lives his life (OC § 7) as if the water in the teapot's not freezing were "merely extremely probable".

(And, note, that is not the way we normally use the word 'probable' -- i.e. the antithesis of 'probable' is not 'practically impossible': the antithesis of 'practically impossible' is 'certain', not 'improbable'.)

And so, I say, there is a difficulty here: one can't say as Wittgenstein does that "outside logic, everything is accidental" (TLP 6.3). "In one sense, you can say that ..." Yes, of course, in the sense that 'random' = 'not logically necessary'; but that is not our normal way of using the word 'random', which is that 'random' = 'unpredictable', 'irregular'.

"From a practical point of view" -- i.e. from the point of view of how we actually live our life (OC § 7) -- "it is as close to being a logical impossibility as if it really were one; we would not send people off in airplanes otherwise." (I thought earlier that the question isn't whether we are certain, but whether the truth of the proposition is certain. -- But that is a grammatical distinction -- i.e. a distinction we can define -- but which normally we don't make.)

In philosophy we make distinctions for our particular purposes, to make something clearer to ourselves. And distinctions have a similar role in our thinking to categories and concepts (PI § 570): they "are the expression of our interest, and direct our interest".

"I am asking a question about logic (of language), not about physics (quantum theory) or metaphysics ("reality itself")." But maybe here there is more than one language to discuss the logic of.

A simple necessary-accidental distinction seems too poor in categories, too poor in propositional types. For what does the word 'law' mean in 'Laws of Nature' or 'Laws of Physics'? It doesn't mean 'Logical necessities of nature' or 'Logical necessities of physics' -- much less 'Accidents of nature' or 'Accidents of physics' ...

How do we normally use the expression 'Law of Nature'? Outside science class, most of us don't use it at all. For it is possible to just say: "You can count on this happening, that in the normal case water flows downwards, and so if you open a bottle of water upside down the water will splash out." We don't call that a "law" but instead "what always happens" or "what we have learned from experience", as we have learned that fire will burn us if we put our hand in it (cf. PI § 477).

As if e.g. a necessary-certain-random distinction were needed. But by 'certain' we don't mean merely 'extremely probable'. Although, couldn't mankind be mistaken in its fundamental certainties? What would it be like (ibid. II, xi, p. 226) if we suddenly discovered that contrary to all our experience of the world, commonplace certainties such as water flowing downwards, fire burning us, were false. Aren't these certainties "regularities" (TLP 6.3) or even "laws of nature"?

The kind of certainty is the kind of language-game.... For the question here is not one of an approximation to logical inference. (PI II, xi, p. 224, § 481)

By saying that the only kind of necessity = certainty is logical certainty -- Note that the TLP does not make that equation -- is that what is being denied: that the type of certainty is the type of language-game? Well, yes, because according to TLP there is only one type of language-game, the proposition type natural-science.

'All men are mortal.' 'No one returns from the grave.' No, these are not logical necessities. They are not necessities at all. And yet "as if they were" is their place in our life and thought.

Induction and Prediction

The principle of prediction as based on induction: 'Predictions can be made about future events based on past events.' Sometimes a mechanism is traced (Quantum theory is not a mechanism), but just as often not, and when not, that is what we call 'induction'. The TLP says that proposition can be significantly negated ("The so-called law of induction ... is obviously a significant proposition" (6.31, tr. Ogden)) ... but can it be "significantly negated" -- i.e. can its negation not be nonsense (i.e. an undefined combination of words) -- if its negation is 'No predictions can [ever] be made about future events based on past events'? Because if that were the case, then we would have to assume that the meaning of our language is also unpredictable ...

Can we describe the life of man in such a world, a world where God played god, where if we threw a ball into the sky what happened next depended on the caprice of a god or demon rather than on laws of nature -- i.e. rather than on the regularities we observe in our world in the occurrence of natural events? Is the following a rejection "of sorts" of TLP 6.31?

If anyone said that information about the past could not convince him that something would happen in the future, I would not understand him.... For note: here grounds are not propositions which logically imply what is believed. (PI § 481)

'There are compelling grounds to believe it will happen' ≠ 'It must happen'.

"It doesn't have to happen" (TLP 6.3) -- But it will! Something is wrong here, something missing, something not adequate -- i.e. there really does seem to be poverty in the categories needed to characterize this!

"The only kind of necessity is logical necessity" (TLP 6.37) -- Does that statement belong to logic (of language) -- or to metaphysics? From the point of view of the connection between grammar and sense and nonsense, is this correct: "The kind of necessity is the kind of language-game"? How do we normally use the word 'necessary' (Is it always logical necessity we mean; cf. the word 'must': sometimes we mean obligation: "You must present your passport!"; but otherwise it seems we always mean deduction), and do we use that word differently in philosophy?

Do we not know if the sun will rise tomorrow?

That the sun will rise to-morrow, is an hypothesis; and that means that we do not know whether it will rise. (TLP 6.36311, tr. Ogden)

And is it an hypothesis that All men are mortal -- i.e. something that we don't know? (Is it a prediction about the future? We treat it as being of the greatest possible certainty, as a "law of nature".) Don't we know that we are going -- indeed, that we "must" -- die (that, as the monsignor said, Siamo nati, e dovete morire)?

(How do I know that I am going to die -- for, after all, I only know this life; I have lived no other than I know of? -- Does the solipsist know that he is going to die? How does he know or not know?)

Is 'I know that I am going to die someday' a false statement -- or is it an undefined combination of words?

'I know that I am going to die' ≠ 'My death is a logical necessity' (PI § 481). But I would say, "If I don't know that (I am going to die), then what do I know?" What am I more certain of? Well, nothing. But 'I am certain' ≠ 'I know'. That a proposition stands firm (immovable) for me ≠ I know that proposition to be true.

(Are we fussing about the mere word 'know' -- or is there an important logic-of-language distinction here? Only the distinction between logical-proposition and proposition-of-experience?)

In some cases we could say, We treat this as though it were of logical necessity -- although it is not. It has the same foundational role (place) in our life as a logical necessity (deduction).

The concepts 'I believe' and 'I know' are different: knowing requires sufficient grounds, but believing does not require any grounds. The question -- the grammatical question -- is, whether with respect to future events, there are ever grounds for saying 'I know' rather than 'I believe'.)

The proposition 'I know' doesn't imply 'I can't be mistaken' (and this alone shows that 'I know' ≠ 'It is logically necessary') -- but I would say that were the proposition 'All men are mortal' to be overthrown, that couldn't be a mistake, that it would be too big for a mistake (cf. LC p. 61-62). The overthrow of the foundations of mankind's entire life could not be a mistake. "Unless I am greatly mistaken ..." You can't say that in here; it would not be the normal use of that expression.

'I am going to die someday.' -- We wouldn't normally call that an hypothesis ('a proposition to test the truth or falsity of'); but we would not call it a tautology either. It has the status of "what everyone knows"; even a court of law knows it. (What do I mean by 'status' -- "civil status" (PI § 125): what "anyone knows and must admit" (Z § 211) --? Logical-grammatical status?)

But is there such a thing as "verification before the event", a priori knowledge that an event will occur? (Experiment is not the only way to verify the truth of a proposition.)

There are compelling reasons to believe that there is a brain in your head, not sawdust. There are compelling reasons to believe that your body will stop working (that you will "give us the ghost" (Phaedo 64c)) someday. We don't believe just apropos of nothing in this particular case. (Of course there are other possibilities, even real possibilities as to what might happen. It is not logically necessary for anyone to die. It is just "extremely improbable" that they won't.)

How shall we classify this type or proposition -- is there a proposition type knowledge-of-future-events-proposition?

The word 'know' in the TLP

It is an hypothesis that the sun will rise tomorrow: and this means that we do not know whether it will rise. (TLP 6.36311, tr. Pears, McGuinness)

Is Wittgenstein using the word 'know' strangely here? -- With that remark is he redefining the word 'know' (logic) -- or saying "what knowledge really is" (metaphysics)?

Is the only knowledge, then, knowledge of the propositions of logic? But according to the TLP, the propositions of logic are "nonsense" -- i.e. they are not propositions of natural science; they do not state knowledge of the world: "Tautologies and contradictions are without sense.... e.g. I know nothing about the weather when I know that it is either raining or not raining" (ibid. 4.461): the propositions of logic do not say which facts are the case (ibid. 1.0).

As we normally use the word 'know', don't we know that the sun will rise tomorrow? And as we normally use the word 'know', don't we know that a teapot of water set over a fire won't freeze?

You could not claim quantum theory based ignorance as a defense for scalding. In a court of law: "I didn't know what would happen to the water in the teapot." -- "You knew perfectly well that if you put the teapot over a flame --!" Does the accuser have the grammatical right to say that?

Two concepts: 'knowledge' versus 'certainty'. In these cases 'knowing' = 'being objectively certain', not being certain that an empirical proposition is logically necessary, but being certain that the proposition is true beyond a reasonable doubt. That is the closest thing there is to empirical necessity [-- The importance of the word 'necessity' in our vocabulary; cf. 'knowledge', 'truth' --]. Except that it is not necessary because "with respect to future events we can always be mistaken", here 'can' signifying logical (although maybe not real [i.e. 'not real' = 'unprecedented']) possibility. 'There is no necessity for it to happen' means only that the combination of words 'real necessity' is undefined.

So in the court of law: 'You knew what would happen' = 'You were objectively certain what would happen to the water'. And 'What grounds had you for your certainty?' = 'How did you know?' The most compelling ground here is of course past experience (PI § 481).

As we normally use the word 'know', is that word not applied to future events, just as there is no first-person present progressive of 'to sleep' [as in 'I am asleep']? And why don't I know? And what would I need to know in order to know? Both those questions are about rules of grammar: the first notes the apparent absence of a rule, while the second asks for a rule -- or what the basis of a rule would be. Are we talking about grammar or about general facts of nature? If I don't know -- is that ignorance because of the nature of things determining concept-formation -- or because of a defective (ill-formed) concept -- i.e. one inadequate in rules?

[What is the point of these questions that make nothing clearer! Precisely that they make nothing clearer? Maybe the answer is that there is no general answer: the meaning of 'know' -- i.e. what the person employing that word means -- is shown in the particular case only. (Concept fluidity.) Well, I don't think that's how we use the word 'know' -- i.e. as individuals rather than as a community; but, on the other hand, I also don't think there is a rule that says whether it is correct to say that we know, or only believe, that the sun will rise tomorrow.]

'I know that I am going to die' -- does the addition of 'I know' imply that there is no other real possibility? Would we say that for all practical purposes, "Yes, it does imply that"?

'I am going to die' does not imply that nothing else is logically possible, and indeed, it cannot imply that. Because for a proposition to be "significantly negated" its negation must describe a logical possibility -- i.e. its negation cannot be undefined language.

When do we say that there is no other possibility? 'If today is Saturday, then tomorrow is Sunday' is not a prediction (hypothesis) but a logical proposition (Knowing that proposition to be true means knowing a rule of English grammar). If today is Saturday, it is not logically possible for tomorrow to be Monday (not without redefining the language involved). And of course if it is logically impossible, it is also not a real possibility.

In philosophy, the imagination is ordered to run wild, to describe this, that or the other logical possibility. But where 'possible' means no more than 'can be described' or 'can be put into words' -- i.e. where logical possibility has no corresponding real possibility -- then that possibility doesn't affect our everyday view of the world -- and then isn't it the normal use of 'know' to say 'I know A will happen'?

Maybe it should be 'We know' rather than 'I know'. Or rather, in this type of case, remember that 'I' is a subset of 'we' -- i.e. not only do I know that A will happen but everyone knows that A will happen.

For note: here grounds are not propositions which logically imply what is believed. (PI § 481)

'I believe A will happen' has a different grammar from 'I know A will happen' -- i.e. those propositions are used differently in the language.

Doesn't knowing -- I mean the concept 'knowing' -- require sufficient grounds for saying 'I know'? Well, but haven't I sufficient grounds? What do we count as sufficient grounds in this type of case (for this proposition type)? -- If A has always happened (e.g. if touching a fiery oven has always burned my hand), then haven't I sufficient grounds -- as we normally use the word 'know' [Well, but how we normally use the word 'know' is the very question we want answered, and so we can't assume this to be its normal use] -- to say that I know that A will happen?

Having sufficient grounds for predictions isn't always only a question of correlations (or "constant concomitance"), of course. Sometimes there is a mechanism to trace (-- pace quantum theory: the notion of a change in grammatical rules being justified by a scientific theory, of a scientific theory justifying new rules of language use (grammar) --), such that we have the right to say 'We know that A will happen, given the right set of circumstances'. (The grammatical, not as it were the ontological [i.e. knowledge of real possibility], right.) Weather forecast: "If current trends hold ..."

It is as if Hume wants to claim that all talk of causality commits the post hoc fallacy ('A happened before B; therefore, A was the cause of B'). But what must apply to all must also apply to none -- i.e. it is nonsense. We call neither all nor no cases of constant concomitance cases of causality. (The colliding billiard balls were used to define 'cause'; cf. Eddington's table was used to define 'solid' (BB p. 45-46). "Something must be taught us as a foundation" (OC § 449); no language-game could [or, can (logical possibility)] get off the ground [be put in motion] without one.)

Is there a difference between saying 'We have every reason to believe, and no reason not to believe, that the sun will rise tomorrow' and saying 'We know that the sun will rise tomorrow'? It is not as if the word 'know' had magical power, as if it could force events to happen. It is not as if it tied God's hands. ("The reasonable man says 'I know' here.")

I am asking about logic (of language), not about the metaphysics of knowledge ("epistemology"). For example, "... there are things that I know. God himself can't say anything to me about them" (OC § 554) is metaphysics. As is the question, "Is God bound by our knowledge? Are a lot of our statements incapable of falsehood? For that is what we want to say" (ibid. § 436).

Because what a puzzled someone "wants to say" in this case does not belong to grammar-logic, but to "metaphysics" -- the distinction between What must the reality be? and How is this language used?

Wittgenstein: "the hardness of the logical must" (PI § 437). We could speak of the hardness of the word 'know' -- of the importance we assign to that word. Knowledge is very serious, belief less so. (The emphasis given the word 'know' -- cf. the importance we attach to the word 'fact' is the importance we attach to the words 'truth' and 'knowledge'. A word is assigned to a concept, and the importance of the concept is then attached to the word itself.)

Is the question here about language, or about what we really know versus what we only fancy we know but do not? G.E. Moore and "absolute certainty" -- is that certainty the condition only of a madman? For isn't absolute certainty what Moore is talking about: "I cannot be mistaken about these facts -- not logically cannot, but factually cannot." No, only a madman can be mistaken about these facts, e.g. whether or not he has two hands. But that is a grammatical remark: "you play the game wrong" or, in the case of the madman, not at all (OC § 446).

"Because we have experience of people being mistaken." "It is a fact of experience [PI II, xii, p. 230] that human beings make mistakes; that is why our language does not allow for absolute certainty except in logic. That is why we make a distinction between knowing and believing, why we have these two concepts, this conceptual distinction. And why a question such as "Do we know that the sun will rise tomorrow? or only believe that?" seems to put everything in its place." "Say whichever you like. Your answer won't change the facts themselves, although it may change our attitude to the facts." But our attitude -- or point of view, our understanding/interpretation of things -- is important to us.

'Just because A has never happened, doesn't mean that it won't happen.' The question is: which type of possibility is being asserted by this proposition: 'won't' (real possibility) or 'can't' (logical possibility)?

The relation between language and the world -- the limits of convention, where facts (-- "If the formation of concepts can be explained by facts of nature, should we not be interested, not in grammar, but rather in that in nature which is the basis of grammar?" (PI II, xii, p. 230) --) of nature take command: e.g. the concepts 'object' and 'possible' -- the place where rules for using language [i.e. the mandate/compulsion to use some particular concept] looks like obedience to natural law.

We are interested in the concept and its place among the concepts of experience. (PI II, xi, p. 193)

"It is not a convention that 'I know I am going to die' is true." It is the definition of the proposition that is conventional, not its truth or falsity. ("The mind-body interchange." We are always contrasting language with the world, things versus our ways of talking about things. On the one side language, and on the other side the world. How to bridge the gap. Why does it matter what we say? -- Because words name concepts, and concepts = thought. And our thoughts in some way = we ourselves, what we are.)

Can't we make a list of the things we know? -- I am asking a logic of language (grammatical) question ["For each one of these sentences I can imagine circumstances that turn it into a move in one of our language-games ..." (OC § 622)], not a question about epistemology (Moore's claims to know are metaphysical claims; that is the difference between our normally saying 'I know' (Z § 223) and Moore's saying 'I know' in the context of his philosophy), or a question about mental states either (ibid. § 6), as if the latter were of concern to logic-philosophy (RPP i § 212).

Encyclopedias as a list of what mankind knows. And so, haven't we the grammatical right to list the things we know -- because 'We know' ≠ 'We can't be mistaken'? But Moore is saying that in some cases we/I can't be mistaken.

"The kind of grounds is the kind of language-game" (cf. PI II, xi, p. 224). Then is it not nonsense to say 'We know that the sun will rise tomorrow'?

Possible ≠ There is reason to believe

I imagine that if the moon were to collide with the earth or if the earth's core were to explode, there wouldn't be an earth for the sun to rise over. But we have no reason to believe that either event will occur overnight, before tomorrow's sunrise.

"What can be described can happen too" (TLP 6.362), but that doesn't mean we have reasons justified by experience (PI §§ 480-481) to believe that just anything we can -- "can" either logically, or logically and really as well -- describe will happen.

If we -- i.e. knowledge belongs to the community, not to the individual (This is a grammatical remark) -- don't know that the sun will rise tomorrow, or that if we throw a stone up into the sky it will come crashing down to earth, then what do we know?

Any factual proposition can be "significantly negated" -- i.e. its negation is not nonsense but a description of a state of affairs -- that is to say it may be true or false, as an hypothesis may be verified or falsified. Thus if in order to know a proposition to be true, the proposition can have only the possibility of being true, then there is no knowledge of reality. (Tautologies are not knowledge, except maybe knowledge of grammar.)

If the only difference between saying 'I know the sun will rise tomorrow' and 'I believe the sun will rise tomorrow' is that it is possible (not only logically but also really) for the sun not to rise, then we have no knowledge of future reality but only beliefs about it.

What is important to see here is that this is a grammatical investigation, a query about rules or their absence: How do we use -- i.e. what work do we do with the word 'know'? (Where is that word's original home in the language? (PI § 116))

In its language-game ... it has no higher position than simply the human language-game. For there it has its restricted application.

But as soon as I say this sentence outside its context, it appears in a false light. (OC § 554)

"... its restricted application" = its normal use in the language, in "the language-game that is its original home" (PI § 116).

It is as if 'I know' did not tolerate a metaphysical emphasis. (OC § 82)

And a metaphysical emphasis is what we are trying to give it when we ask, 'If I don't know this, then what do I know? If I don't know this I don't know anything.' As if to say, Don't I really (The word 'really' is the hallmark of metaphysics) know that the sun will rise tomorrow?

If in our normal way of speaking, we say we are certain the sun will rise tomorrow, we are not making a metaphysical assertion. "Will it happen?" -- "You can be just as certain it will happen as that the sun will rise tomorrow." -- i.e. the sunrise is a very model ("paradigm") of certainty in the context of statements of fact about future events. (The table as a model of solidity; the table was used to define the word 'solidity'. (Cf. BB p. 45))

The kind of certainty is the kind of language-game. (PI II, xi, p. 224)

If anyone says with justification 'I know', he is saying only that he has what the community agrees are compelling, sufficient grounds. His statement means nothing more than that. That is the point of The Fable of the Born-Blind People. The word 'know' means certainty -- not that the contrary might not happen -- but that we have no reason to believe that it will.

The question is about the rules of the game, about how we normally use certain words of our language. Look at the word 'know' as a tool we use to do some work in our life.

Query: Wittgenstein, Tractatus attack induction.

The TLP aims to describe the nature of "the world" by contrasting its world with logic.

A necessity for one thing to happen because another has happened does not exist. [6.37] There is only logical necessity. [6.37] And outside logic all is accident. [6.3] That the sun will rise to-morrow, is an hypothesis; and that means that we do not know whether it will rise. [6.36311] (tr. Ogden)

What would it be like if it were otherwise, if there were "necessity outside logic" -- i.e. in the world? The clockwork of a machine that cannot melt? Is there such a thing as real impossibility, or does only the combination of words 'logical impossibility' have a defined use in our language?

Metaphysical speculation is like thought experiment: it is imagination. That is the only source of the cartoon-like picture of gears that cannot melt.

Are the rules of inference really natural laws as the TLP seems to say [6.3], or is the question instead about ways of life, about our community of ideas, its idea about sound reasoning -- or rather, about a sound world-picture (in contrast to foolishness or madness)?

Does belief in the causal nexus (rail lines that cannot break) = belief in fate? Is that belief superstition [5.1361] only in our community of ideas? Our community treats logic as an absolute point of reference -- but are we free to do otherwise: are the rules of inference only rules for sound reasoning in the eyes of our community of thought? (How would an illogical community live? Not only "In this case they do not use reason" -- but nowhere do they use reason?)

The Philosophical Investigations in contrast

Does Wittgenstein contradict TLP 6.3 in PI § 481? In the latter he says only: "Here grounds are not propositions which logically imply what is believed." And so he is talking about grounds for belief, not about logical inference. -- That is, in the Philosophical Investigations Wittgenstein is only describing "our right" to say that "information about the past" is grounds for believing "that something [will] happen in the future" -- i.e. Wittgenstein is only describing our way of life ("language game"); he is not justifying that way of life (PI II, xi, p. 226). Nor is he denying that "it is an hypothesis that the sun will rise tomorrow" (We can describe events that would preclude its happening). But whatever certainty or uncertainty we have about the sun's rising belongs to our way of life (We do not pray every evening at dusk that the sun will return next day, for example).

Does the TLP "attack induction"? I don't know, but it does attack "belief in the causal nexus" [5.1361]. Although to me it seems to say no more than the obvious -- which may have been Wittgenstein's aim here: Look at this! (Z § 461, PI § 144) -- because the answer to the question 'How do I know that the only necessity is logical necessity?' is that the combination of words 'empirical necessity' is undefined except as a metaphysical picture, fantasy, fairy tale in our community of ideas.

[Words are tools, but concepts are also the expression of a world-picture.]

The way philosophers (and students of philosophy) think

Query: who discovered philosophy?

A question like that assumes that phenomena exist before they are conceptualized, so that philosophy is a phenomenon to be discovered: philosophy in itself, independently existing, like the western hemisphere waiting for Columbus to discover it (although did America exist before it was conceived to be "a new world" rather than merely an outskirt of Asia?)

The way of metaphysics

Philosophers try to use forms of expression in an abnormal way to emphasize what does not need to be emphasized -- if we understand the grammar of our language. They stop short of seeking that understanding because they imagine that we are talking -- not about rules for using words -- but about independently existing phenomena, of which even "knowledge" is an example (What is the essence of knowledge -- i.e. Absolute Knowledge -- Plato asks in the Theaetetus). They see nothing worthwhile therefore in a description of how a community (Knowledge belongs to the community, not to the individual) uses the word 'know' in particular circumstances, but instead they "let the word speak to them".

'You can only know the present and the past, not the future.' -- What kind of proposition, what kind of possibility ("can") is that? It is a rule for using a word; but is that the way we normally use the word 'know'? Wouldn't we contrast knowing that the sun will rise tomorrow with knowing what next week's weather will be? It is not a forecast that the sun will rise tomorrow; you can't unmisleadingly say 'If present trends continue, the sun will rise tomorrow'; as a prediction, that is nonsense (i.e. that's not what we normally mean by 'prediction').

Weather forecast: The sun will rise tomorrow. (That is a grammatical joke.)

Site copyright © September 1998. Please send corrections and criticism of this page to Robert [Wesley] Angelo. Last revised: 20 August 2015 : 2015-08-20

The URL of this Web page:

Back to top of page

Wittgenstein's Logic of Language - Introduction and Table of Contents | Bibliography | Site Search | Site Map