Home - Wittgenstein's Logic of Language | Bibliography | Site Search | Site Map

Knowing and Saying

Does 'know' in philosophy mean 'be able to put into words'? That was the meaning of 'know' Socrates chose for his philosophy, and why Socrates made that choice can be compared to why Wittgenstein chose the meaning of 'meaning' he chose for his philosophy. Socratic philosophy and the philosophy of Wittgenstein are limited views of things (What is your aim in philosophy?)

Topics on this page ...

Background: the question of the "logic of language" (in my jargon: in what way is language-with-meaning distinguished from language-without-meaning in philosophical problems?) revised from time to time.

Words that follow "Query" are Internet search results that were directed, or more often misdirected, to this site, to which I have responded in the discussions below.


Must we be able to "give an account of what we know"?

Socrates held that if a man knew anything, he could give an account of it -- i.e. explain what he knew -- to others (Xenophon, Memorabilia iv, 6, 1, tr. Guthrie) ... an account that could be put to the test in discussion with his companions or himself alone, to be agreed to or refuted by the tests of reason and common experience. "And that which we know we must surely be able to tell?" (Plato, Laches 190c, tr. Jowett)

Compare knowing and saying ... how a clarinet sounds. (PI § 78)

Are we able to tell (put into words) how a clarinet sounds? And yet isn't this a use we make of the word 'know'? But is saying that you know how a clarinet sounds [or how an orange smells] the same as saying that you know how a toothache feels? What does 'know' mean in these cases? With the clarinet [and with the orange], (but not with the toothache), there are tests, e.g. sounds are played -- now, can you identify the clarinet sound among them?

Defining 'knowing how a clarinet sounds'

In Socratic dialectic, or in my view of it, there are two kinds of tests, the first being the tests of reasoning (i.e. of sense and nonsense, and of consistency and contradiction), and the second being the test of verifiable experience (truth and falsity). For example, the criterion 'being able to mimic the sound of a clarinet with one's voice' or 'being able to describe the sound of a clarinet in words' will not stand up to the tests either of reasoning or of experience if either is offered as the essential definition of 'knowing how a clarinet sounds', because in fact that mimicking or describing is something that very few human beings can do, despite their being able to distinguish the sound of the clarinet from other musical instrument sounds when music is played for them.

If there is to be philosophical knowledge (Socrates), and if language meaning is to be objective (Wittgenstein)

So, then, was Socrates wrong? Not if he was talking about philosophers and the claims that philosophers make, I would say. Philosophy is discourse of reason: critical discussion. Knowledge is objective; it is open to the public; testable, verifiable [These are grammatical rules, what we mean by the word 'knowledge']. Socrates set a standard that makes knowledge possible in philosophy, namely public criticism.

What Socrates did is like what Wittgenstein did when he demanded that an objective distinction between sense and nonsense be made in philosophy. That is, just as Wittgenstein selected one out of many meanings of our word 'meaning', Socrates selected a particular sense [one out of many meanings] of the word 'know'. If Socrates had not done this, he would not have been able to philosophize -- indeed, he could never have invented what we call 'Socratic philosophy'. Because philosophy without the definition of 'know' that Socrates chose -- like logic of language without the sense of 'meaning' that Wittgenstein chose -- would not allow claims to knowledge [or to meaningfulness] to be challenged -- i.e. it would not be objective: knowledge would be like meaning was for W.E. Johnson: "If I say that a sentence has meaning for me, no one has the right to say it is meaningless." -- "If I say that I know something, no one has the right to say that I don't." And then what would become of philosophy?

Rather as if Johnson had said, "I consider it a disaster that Socrates has returned to Athens. A man who is quite incapable of philosophical discussion. If I say that I know something, no one has the right to say that I don't know it." But then as with what Johnson did say about Wittgenstein, if the meaning of language is not verifiable, if whether someone knows something or does not know it is not verifiable, then there is no philosophy.

Query: knowing without language. Knowing how to.

We would certainly say -- i.e. in our normal way of talking -- that a bird knows how to build a nest, but would we also say that a bird can give an account of what it knows to others? And if it cannot explain, does it not know? There are many possible criteria for saying that someone 'knows' something.

A bird knows how to build a nest, but it cannot say, and philosophical cross-questioning cannot make a bird say what it knows. And in that sense of 'know' -- i.e. Socrates' selected sense of that word -- the bird does not know how to build a nest.

In this context, note this about Socrates' daimon: prescience [premonition, prevision] does not decide arguments. Socrates' method or tool in philosophy is rational, a thoroughgoing use of reason and public experience. It is step-by-step discussion, consisting of thesis, cross-questioning, agreement (acceptance of a clear and true thesis) or refutation (by a hidden contradiction in meaning).


To know = To have thought the thing through

We call many different things 'knowledge'. For example, 'to know' = 'to be able to state the reasons for and against'.

That is another criterion for saying that someone 'knows' or 'does not know' something. It is similar to the Socratic 'to know' = 'to be able to give an account (of what you think) you know', and 'to give an account' means 'to explain and defend when cross-questioned'.

By this criterion one must be able to say, "On the one hand, but on the other hand" in order to show that one has considered the reasons for and the reasons against, that one has thought through the objections to an action or opinion, that one's position is "the reflection of a thoughtful philosophy".


Absolute Preconception

Plato's preconception, that "knowledge is not possible of what is in flux but only of what is not in flux, i.e. only of what is unchanging" (Heraclitus effect on Plato). This is why Plato expanded Socrates' search for essences in ethics to invent his own Theory of Forms (Archetypes); an essence is unchanging -- it cannot be reduced or augmented (it is absolute).

For Plato, to 'know' means to 'be able to state the essence of a class of things'. If one cannot do this, then one is without knowledge; that is Platonic ignorance.

Would that be Plato's definition of the word 'knowledge', one selected out of many possible meanings of that word? Plato did not think that he was defining a word (i.e. making a rule for using a word), but instead that he was stating "what the essence of knowledge is"; to say otherwise would be to commit the Fallacy of Anachronism, which is a fallacy in historical research (Historiography).

But our interest in Plato's thought is philosophical rather than historical, and, from the point of view of logic of language, Plato's statement about "what the essence of knowledge is", namely that knowledge can only be of what is absolute, is not a statement of fact -- but only a definition of a word. We in fact use the word 'know' in a variety of ways: we call many different kinds of things knowledge (as Plato partly says in the Philebus).

The question a philosopher may ask is: which of the definitions of 'knowledge' is more useful to his project in philosophy than others. (A definition, a tool, that did not allow him to get any work done would not be useful.)


Is there a particular definition of the word 'know' that should be used in philosophy?

Is this a question of what anyone wants from philosophy? Someone with a quite different conception of philosophy -- i.e. view of the end-in-mind of philosophy -- e.g. someone who seeks to build up a picture of reality or worldview (Russell: "my fundamental aim has been to understand the world as well as may be") -- might say that a philosophy that has no place for knowledge in the sense of "knowing how a clarinet sounds" and "knowing how an orange tastes" -- is a philosophy that has as limited a view of reality as physics has (reducing reality only to what can be measured), that a philosophy that has no place for this kind of knowledge is "barbarous metaphysics" (F.H. Bradley). But what is neither logical nor factual is not part of Socratic philosophy.

It is impossible for me to say in my book one word about all that music has meant in my life. (DW p. xiv)

Or is it that there are limits to what philosophy can do (as Wittgenstein thought to show in his Tractatus Logico-Philosophicus and in all his later work), that this is not a question of denying the reality of what philosophy cannot deal with? But if there are limits to philosophy, what -- or who -- sets those limits?

What we can do is to say what we are calling 'philosophy', that Socratic philosophy does not deal with "what cannot be" -- i.e. what we do not -- "put into words". Philosophy as discourse of reason is a limited project in philosophy, not the only possible project. Its standard is set by language -- what cannot be put into words and tested in discussion is unintelligible (irrational), and everything non-rational is set aside

Is there a particular definition of the word 'meaning' that should be used in philosophy?

"... a philosophy that has no place for knowledge in the sense of knowing how a clarinet sounds". Socratic philosophy limits itself to criticism of propositions (theses), statements of fact in ethics, logic, and metaphysics. It does not aim to construct a picture of reality; it does not aim to have a place for every aspect of experience. Other projects in philosophy are possible. Compare Wittgenstein's words:

Now if for an expression to convey a meaning means for it to be accompanied by or to produce certain experiences, our expression may have all sorts of meanings, and I don't wish to say anything about them. (BB p. 65)

There are many meanings of the word 'meaning'; Wittgenstein chose the one he did in order to make an objective distinction between sense and nonsense in language, not because his choice was the only meaning of 'meaning' there is.

In Wittgenstein's philosophy, the master question in philosophy would be -- "What is the meaning of a word?" (ibid. p. 1) If it strikes anyone as absurd that this could be the master question of philosophy, then perhaps he will not understand the importance of Wittgenstein's work to philosophy.

Suppose someone said, "Come, surely the master question in philosophy is the one posed by the Greeks, or as Plato says in Gorgias 500c: "we are discussing no small matter, but how to live" (ethics). But Socrates needed to develop logic tools to use in his study of ethics, if ethics is to be rational rather than irrational, which in Socratic philosophy it is.

Rather than "master question", then, we could say: What is the first question in philosophy? What is the distinction between sense and nonsense in language, because isn't philosophy necessarily a use of language -- i.e. discourse of reason? And we wouldn't want "sound without sense" for a philosophy.


Knowing by Touch rather than by Sight

Query: how do the blind recognize English words and their meanings?

Defining words by the sense of touch versus by the sense of sight. Hearing a word and knowing its meaning (use in the language). For the human beings in the Fable of The Born-Blind-People ("Look at the word 'know' as a tool that has a place, a use in their life") the word 'know' is defined in the story by the sense of touch rather than by the sense of sight. [Cf. "Knowing what a tree is", knowing how to use the word 'tree': knowledge as knowing how to use language, as mastery of a technique (PI § 199).]

The concepts 'knowledge' and 'justification' are interconnected. -- "How do you know? What is the justification for saying you know?" -- "I am touching the tree." (Unlike belief, knowledge belongs to the community (OC § 555); that is a grammatical, not a sociological note, about the concepts 'knowledge' and 'belief'.)


Abstraction, Clarinets and Games

Query: Philosophical Investigations, what the word 'game' means, what a clarinet sounds like.

This juxtaposition is very suggestive of exactly what Wittgenstein did not want to suggest -- "abstraction", a way to smuggle it in through the back door. The query's suggestion is that, just as one can know "what a clarinet sounds like" (PI § 78) without being able to say, so too one can know what "the essence of games" (ibid. § 66) is without being able to say.

Aural definition

However, is there no sense in which one can say what a clarinet sounds like? If one can select the sound of the clarinet from among a sample of sounds of various musical instruments, has not one by this selection answered the question of "what a clarinet sounds like"? Wouldn't that be what 'know how a clarinet sounds' mean in this case -- or what else would it mean? That is, wouldn't one have given an ostensive definition of 'sound of a clarinet'?

Not every explanation of meaning is given by means of [being put into] words. Compare: "knowing how the color red looks". The question of essence doesn't arise here: one points to examples of red objects and adds the words "and similar things".

Definition by enumeration

Query: 'philosopher' means person as Socrates, Plato, Aristotle or Kant. What type of definition technique?

It seems to be "definition by enumeration", i.e. by giving examples of representative members of a class, as in Euthyphro 5d-e: "I say that the holy is what I am now doing, prosecuting the wrongdoer who commits a murder or a sacrilegious robbery, or sins in any point like that .... And not to prosecute would be unholy" (tr. Cooper).

If someone who thinks as Plato does, in response to our giving examples of games to him, told us that there were after all many, many games, but that what he wanted instead was for us to tell him what all games have in common and distinguishes them from all other things -- i.e. what the essence of games is -- (Euthyphro 6d), we could only reply that games do not, as a public matter of fact, have a common nature but instead that there are many "family" resemblances -- i.e. various similarities -- among the many things that are called by the common name 'game'.

Which members must be included in a definition by enumeration of the members of a class, e.g. of the members of the class (or category) 'games'? Are some members representative of games, others not so much? (Leading models or paradigms.)

Knowing in Socrates in contrast to Plato

And when I give examples of games in order to define the word 'game', I do not tell you less than I know about games myself: every explanation of meaning I can give myself I can give you too. (cf. PI §§ 210, 208)

According to Plato's requirement -- i.e. the meaning of the word 'know' he chose from among the many uses we make of that word -- Wittgenstein does not "know" what games are, because he cannot say what Plato's postulated essence (common nature) of games is.

In Socrates' sense of the word 'know', only if I can explain in words how a clarinet sounds do I know how a clarinet sounds. So if the Socratic standard were broadened to "If a man knows anything he can explain or demonstrate what he knows to others" that would be closer to how we normally use the word 'know'.

Query: if we can pose the question within the language we can answer it as well [TLP 6.5]. Wittgenstein.

"Then me, then, how does a clarinet sound?" -- "I will play you a recording; I will mimic it with my voice." -- "No, I want the answer within the language -- i.e. put into words -- if that is what the TLP claims is possible." Is a musical score a picture of the music? Well, if proposition 'The book is on the table' is a picture, then why wouldn't a musical score also be. But is the score a picture of the form: "This is how things stand"? The score of the music is a "proposition" if there is a one-to-one correspondence between the written notes and the sounds they transcribe, but a musical score does not "put into words" how a clarinet sounds even if the music was written to be played by a clarinet, and we do not call a musical score true or false.

The essence is hidden from us: this is the form our problem now assumes. (PI § 92)

The "theory of abstraction" is precisely that: a metaphysical theory (speculation) about a conjured-up unseen reality (an occult mental process of "abstracting the essence of common names"), despite the evidence in plain view that most common names do not name essences. For the theory of abstraction, to 'know' means to 'be able' -- e.g. if you can use the word 'dog' correctly, then you "know" what the essence (common nature) of dogs is. [A "theory of language-meaning is not what I want from philosophy. An objective logic of language, on the other hand, is what I am looking for, and that is why, despite its limitations, I will go on using Wittgenstein's logic until I find something better.]

Can we say that the method of verification shows you the meaning of the combination of words 'know how a clarinet sounds'?

Pace Samuel Johnson, there are countless problems "that the mind of man can set that the mind of man cannot solve".


"Is the meaning of a word justifiable?"

In the earliest version of "Socrates' Logic of Language" the second paragraph was:

For Socrates in Plato (i.e. Plato's literary character Socrates) the questions 'How is a word used?' and 'What justifies the use of the word?' are different questions: what justifies the use of a common name is our being able to "define" the name -- not by describing the name's use in the language -- i.e. its grammar -- but by saying what the common nature the common name presumably names is.

For Wittgenstein, by contrast, they are the same question: the justification for using a common name is our having learned to what to apply the name (What we have learned is shown by what we do: understanding is a public event).

Was I correct when I wrote that for Wittgenstein (1) a description of how a word is used -- i.e. what our common practice is -- and (2) what justifies the use of a word are the same? Wittgenstein wants to know the rules of the game: for example if our investigation of games shows that all the things that we call 'game' have no common quality that defines them as games, then that just is how we play this particular game -- i.e. language game with the word 'game'.

Wittgenstein's way of looking at language is anthropological (ethnological): an anthropologist does not seek to justify what the tribe he is observing does; he simply tries to describe it (CV p. 38). His account may be true or false, but it is only a description, not a justification. "Wittgenstein, why do you call badminton a game?" What can he say to justify myself beyond that, "I have learnt English" (PI § 381)? Plato: "You could say that such-and-such quality is what all games have in common and that badminton shares that quality." Wittgenstein: "But our investigation shows that games do not have any such common quality." Plato: "Then you do not know what games are, and therefore your use of the word 'game' is not justified."

Grammar, in Wittgenstein's revision of the concept 'grammar', consists of more or less arbitrary rules, i.e. conventions. But for Socrates these rules cannot be in any way arbitrary (in Wittgenstein's family likeness sense): they must identify an essence -- i.e. a unique defining characteristic.

Suppose someone said: "Socrates was not an anthropologist; he was a philosopher: he wanted to know what justifies the community's way of life (Remember that it was Socrates who introduced ethics into philosophy), including our language practices. And Plato thought that the only thing that could justify our calling various things 'game' would be that all games have a quality "in which they do not differ but are all alike" (Meno 72c).

In the original paragraph either I used the word 'use' to mean 'utterance' (to 'utter' is simply to 'make a sound') -- which is what I should have done -- or I used that word in Wittgenstein's sense of "in a large class of cases ... the meaning of a word is its use in the language" (PI § 43), with the similes that we use words like we use a tool (e.g. a hammer or glue pot, a fountain pen, a magnifying glass, sticky tape or a ruler) or like we use a game piece (e.g. a knight or a net, a racket or a ball, a playing card or an arrow). And if I used the word 'use' in Wittgenstein's way, then the question 'What justifies the use of a word?' is equivalent to 'What justifies the meaning of a word?' And now, does it mean anything to say that 'the meaning of a word is justified'? I cannot think of anything to do with that combination of words. You cannot say that a meaning that is not a common quality is nonsense. -- You cannot call a meaning meaningless, a sense senseless (or what would you mean by that?) (PI § 500).

Eccentric meaning of 'meaningless'

Is the definition This, this, and the like, are called 'games' (PI § 69) nonsense, whereas A 'simile' is: a comparison using the words 'like' or 'as' is not nonsense? If we say that then we are defining the word 'nonsense' in an eccentric way. Are we uttering meaningless signs when we use the word 'game' or the word 'shape' (Meno 74d) or the countless other words of our language that do not name an essential meaning? Is the question here: Do you want to say that? At the very least it might be remarked that you would then be using the word 'meaning' in an extremely restricted way ... although if one says that about Plato, one must remember that it is also the case with Wittgenstein's selection of a meaning for 'meaning': If a sign is not governed by rules, it hasn't got what I call a 'meaning'.

Antithesis and 'hidden'

The essence is hidden from us: this is the form our problem now assumes. (PI § 92)

To what kinds of cases do we correctly apply the word 'hidden'? To cases where 'hidden' contrasts with words such as 'found', 'uncovered', 'discovered, 'disclosed'. But in the case of words such as 'game' -- "What is the common nature of games?" -- the word 'hidden' is being used [i.e. that sign is being uttered] without any such antithesis. In the absence of any antithesis the word 'hidden' is nonsense.

Hide and Seek (apropos of antitheses)

"Is reality confined to what is perceptible to the senses?" I called that a question without answer, an eternal question. Is it possible -- logically possible -- that anything is essentially imperceptible, essentially hidden?

Yes, the words are there, the combination of words 'essentially hidden', but have they a meaning, a use in the language? They have no normal use in language, but have they -- is there such a thing as -- a "metaphysical use" of language? ("Metaphysical analogies": from absolute pitch to tone deafness, from four to five to six senses, from blind to sighted to -- but without that sense it is hidden from you. A ghost world. Analogies created, create metaphysics.)

"There is a wall on the other side of which there is nothing; nothing whatever -- not even a place for anything to be!" (Cf. Russell's There is a hippopotamus in the room, but it has no legs, no tail, no trunk, no ears, no head, no body ...." and "grammar stripping") You are playing with language -- playing off syntactic analogies, trying to make abnormal use of normal usage.

"You play the game wrong or not at all" (OC § 446). Hide and seek -- but if there is no way, no defined method for how, to seek, are you -- i.e. can you -- seek at all? "What you are to seek can't be perceived." Now, that is outright nonsense.

"Metaphysics is a speculative position." Metaphysics is "playing the game wrong or not at all"? "There is a hidden mechanism", where 'hidden' = 'impossible to perceive by any means'. But is that what we mean by 'impossible', because here that word is not the antithesis of 'possible'?The word 'impossible' has no antithesis in this case. (There's no such thing as an impossible possibility.)

"There must be a hidden mechanism!" Cause and effect: The imperceptible cause is deduced from its perceptible presumed effect. "You play the game wrong." There is something wrong here, but I'm not seeing it: "calloused by doing philosophy" (PI § 348). I keep wanting to say: "But the meaning of 'essentially hidden' is obvious!" And that's because I'm not seeing why it isn't.

What I am looking for is the grammatical difference. (PI II, viii, p. 185)

And not finding it.


Does Augustine know what time is?

What is time? If no one asks me, I know. (Confessions xi, 17, quoted in PI § 89)

In which sense of the word 'know' does Augustine know what time is? Wittgenstein: something we know until we are asked to explain it to others ..." -- Because if Augustine cannot tell (explain to others) what he knows, then in Socrates' sense of 'know', he does not know. Nonetheless, Augustine does know how to use the word 'time' in "the language games that are its original home" (PI § 116) -- and that is what "knowing what time is" means (or what else will it mean? or is there a "phenomenon of time" to know what that is).

To Socrates' way of looking at the meaning of words (i.e. definition), is knowing how to use a word -- without being able to state a general definition -- not "knowing" at all?

I put the word 'knowledge' in double quotes here because there is a danger of equivocation (or grammatical confusion) here, as there was with the word 'use' above. Note that there is no necessity to our classifying "able to give an account [i.e. explain what you know to others]" and "able to do, without being able to give an account" in the same category -- something which the word 'know' may lead us to overlook.

Compare Wittgenstein's rejection of "I think I mean something by it", as if that vague "something" could be the meaning of a word: No, Wittgenstein wanted an objective distinction between sense and nonsense: if you cannot express a word's meaning in rules (conventions) for its use, then it hasn't got what I call a 'meaning' (cf. BB p. 65).

"The true meaning of the words"

In no case can we say: This is the only, the true meaning of 'know', the only, the true meaning of 'meaning'. There are many meanings of 'know', many meanings of 'meaning', not only the ones Socrates, Plato, and Wittgenstein selected for their work in philosophy.


What would it be like for everyone to be mistaken?

Query: Wittgenstein's theory of a family resemblance.

"Wittgenstein's theory." What would it be like if everyone were mistaken, if e.g. games really did have a defining common characteristic (essence)? "The definition had remained unformulated (PI § 75) for thousands of years, but now it has been discovered that the essence of games is ..." This imagining is not nonsense if Wittgenstein's simile "family likeness" is a theory rather than a statement of fact: "Don't say there must be something common -- but look and see if there is!" (PI § 66), "for nothing is hidden" (ibid. § 435): logic of language is a public event. Grammar is a public event.

Is there a limit to looking -- There is looking and finding. At what point can't you (logically) look any further; at what point is 'to look further' nonsense (i.e. undefined language)?

What would it be like if it were never possible to be certain about any matter of fact -- for example, if one could never say with certainty that By the word 'simile' educated speakers of the English language mean: a comparison using the words 'like' or 'as'? Suppose someone were to say, "Maybe some day it will be discovered that similes don't have a common nature; you can't be certain it won't!" That would be to treat all report-of-normal-use definitions as if they were more or less plausible or probable, but never certain, statements of fact. Indeed, there would no longer be what we call definitions -- i.e. statements of rules that "anyone [who speaks the language] knows and must admit" (Z § 211; PI § 599).

Can the sky be discovered not to be blue -- i.e. can we have misperceived its color for millennia? Can we have been blind [so to speak] to the common nature of games?


Language, knowledge, Socrates

Query: without language we cannot know.

But it does seem natural to say that a dog knows things, e.g. that it knows its owner and the house where it lives. And yet the dog does not have language. "Percepts without concepts are blind" in Kant's view, but I don't think that by 'concept' we always mean 'language'. The dog does not seem to be "blind" (in Kant's way). The dog's knowledge can be put to the test (observed), because a dog is in many ways similar to a human being (PI § 360), like the human knowledge of "knowing what a clarinet sounds like" can be put to the test. (In what way "like" or "similar"? In this: that neither kind of knowledge involves the use of language.)

So then can knowledge always be put to the test? Is that one criterion (i.e. what is the grammar of 'know'?) we always use? If someone cannot prove that he knows -- or if his behavior does not show that he knows, then do we normally call it 'knowledge', (place it in that category)? But what of "knowing how a toothache feels"? That cannot be proved (put to the test).

The concept 'normal'

What is normal, what not, like what is and what is not knowledge belongs to the community not the individual. Philosophical investigations = grammatical investigations ... with exceptions. (The word 'normal' doesn't simply mean 'usual' here; it is not purely descriptive; it indicates a rule or norm or standard, in this case the public standard for the use of language.)

Of course to the metaphysician this is not a question of "many ways to slice a pie" (invent categories, classes), but of discovering the one true or real (as in 'reality') way of knowing.

Is there a conceptual essence (a defining common quality) of tests? (Cf. Is there an essence of measurement?) Some tests require language, but not all. But in philosophy, as we have it from Socrates: you must be able to tell -- i.e. put into words -- others what you (think you) know (Plato, Laches 190c); you must be able to explain and defend what you (think you) know to others (Xenophon, Memorabilia iv, 6, 1). Otherwise you do not know. ("Think you know", because it is presumed that before testing by cross-questioning that you do not know.)

Aristotle, imagery and thought

The active intellect abstracts forms from the images or phantasmata [a mental image or representation of a real object], which, when received in the passive intellect, are actual concepts. (Aristotle considered that the use of imagery is involved in all thinking.) (Frederick Copleston, A History of Philosophy: Greece and Rome (1947), II, xxx, 10, p. 329)

That seems to be a theory of concept-formation: abstraction. But can what is abstracted be pictured (e.g. drawn on paper) outside the intellect? (But if not, then it cannot be pictured by the intellect either.)

"Knowing and saying how a clarinet sounds" need not involve imagery, and surely Aristotle is aware that born-blind men use language, expressing themselves in words (so that images are not essential to the use of language). And so Copleston's report is difficult to interpret.


Plato's Meno and Grammatical Reminders

Query: Plato and concept-formation.

In Plato's Meno: knowledge is said to really be [This is of course a metaphysical theory] reminiscence or recollection [i.e. remembering]: we saw the Forms ["Ideas", Archetypes] before we were born (i.e. when we were not in the body): that was where our concepts were formed -- but now we must recollect them. Wittgenstein compared his method of "assembling [grammatical] reminders" (PI §§ 109, 127) to this "theory of knowledge" in the Meno, according to Norman Malcolm (Memoir, 2nd ed. (1984) p. 44). In which aspect were they alike?

Something that we know when no one asks us, but no longer know when we are asked to give an account of what we know to others, is something we need to remind ourselves of. (PI § 89)

Is the Meno a denial of our ability to learn anything new (while still "in the body")? In which sense of 'to know' does Augustine already know what time is -- because if he cannot tell (give an account, explain), then in Socrates' sense, he does not know at all? (He already knows how to use the word 'time' ... But that is not what interests Socrates.) If shapes must have an essence and we do not perceive [= cannot say] what their essence is, then how is it that we can recognize shapes? We must, therefore, have seen them before we were born into our bodies -- otherwise we would not be able to recognize that a, b, c are shapes, because apparently no one can say what the essence of shapes is. The Meno is an example of Rationalism: the "must" here is a requirement [imposed on our investigation], not the conclusion [of our investigation] (PI § 107).

[Note: there is a later discussion of Plato and concept-formation in the context of the meaning of common names. For if the meaning of a common name is not a common nature that it names, then what is its meaning?]

The Socrates with questions and not the Socrates with answers (History and Plato)

Query: Socrates, rationalism or empiricism?

Here is an example of empiricism: "I turned these facts over in my mind, endeavoring to hit upon some theory which could reconcile them all," Dr. Watson wrote ("The Adventure of the Empty House"): the hypothesis follows from the facts, not vice versa. Is that the method which Aristotle says that Socrates used? In Plato's dialogs a requirement (for it was not the result of dialog) is set for Socrates' investigations in ethics -- namely, that there must be common natures that define such common names as 'courage', 'piety', 'justice', 'temperance' (Being brave, just, pious, and self-controlled are examples of excellences that are proper to man, according to the Greeks; Plato does not prove that they are, but he might well do), as in Plato's Euthyphro. But a requirement of that type belongs to Rationalism, not empiricism.

Note that the word 'Rationalism' is capitalized (and Plato describes this method), for that is not the only thing we mean by the word 'rationalism'; we also mean 'thoroughgoing reason working on our common experience' -- which is to say that there is a double-test in Socratic "rationalism": reason and public experience, not reason in despite of any experience to the contrary of reason (The latter is "Rationalism").

Query: philosophy as questioning. Ignorance, thinking we know, but we do not know.

There are ways we can make a distinction between the historical Socrates, i.e. the Socrates of Plato's Apology, Xenophon's Memories of Socrates, and the account in Diogenes Laertius -- in so far as these are history, which is farther than we know -- and the literary character especially of Plato's post-Apology dialogues. One way is to distinguish between the Socrates who questions and the Socrates who suggests answers (Thus the theories in Plato's dialogs are Platonic, not a Socratic).

The portrait I draw of Socrates is different -- much more limited -- from W.K.C. Guthrie's view of him (other historians have their own views). I am not a scholar; for me Socrates is philosophy (as are all other philosophers), not history. But I think that even of history scholars it can be said that we make for ourselves a portrait of Socrates; what we draw is not the true face of Socrates, but only a possible selection made from the historical texts known to us.

"The Socrates with questions and not the Socrates with answers, the Socrates of the give-and-take of dialectic (dialog) who does not think he knows what he does not know, but who questions because he wants to know" -- that is the one I call by the name 'Socrates' (cf. Russell's "theory of descriptions", as in PI § 79).

Plato's Socrates (literature vs. historiography)

After the first book of the Republic, Plato has Socrates say again and again, "In my opinion". The historical Socrates, the one who does not think he knows what he does not know, is replaced by a literary character who is full of opinions, Plato's opinions. But opinions -- i.e. conjecture -- is not what I want from philosophy. Further, Plato's method of preferring his own preconceptions despite any evidence to the contrary (Phaedo 99d-100a; 114d and 91b amount to saying that the Socrates of the Phaedo is not the historical Socrates, at whose death Plato was not present) is the very opposite of what I want from philosophy.

It was only once I discovered the passage in Aristotle about Plato's inheritance from Heraclitus that I saw another way to distinguish between Socrates and Plato, one which has changed my view of Plato, whom I can now appreciate a great inventor of philosophical ideas, and no longer feel anger towards him for filching Socrates' good name.

Plato and Opinion and Knowledge (slippery slope)

Questions: does Plato think he knows what he does not know? If his method in philosophy is to hold opinions that he does not subject to refutation in dialectic (i.e. Plato's sometime method of letting his preconceptions override any evidence to their contrary), then the answer sometimes is Yes. But on the other hand, the answer is much more often No, -- because to state an opinion is to state what one believes but does not know to be true: the word 'opinion' contrasts with the word 'knowledge', and Plato of course makes that distinction. But the trouble is that the line in one's own thinking between one's opinions and one's knowledge can easily become blurred, with 'I know' taking the place of what should rightfully be 'I think, and 'I think' means I don't know'.


Is there no general definition of 'to measure'?

You have been taught to measure lengths; but this does not tell you how to measure time. -- There is no general definition of 'to measure'. Perhaps you learned to use a ruler to measure fabric lengths; do you then use that ruler to measure time? [Perhaps you learned to use a ruler; do you use a ruler to measure time?] (cf. Malcolm's Memoir, 2nd ed. (1984), p. 41)

'How is that measured?' performs the same service as 'How is that verified?' -- It makes the grammar of the language that is being asked about clear or clearer. (The meaning, or part of the meaning, is the method of measurement or the method of verification.)

"To give an account"

Query: Theaetetus' three ways to give an account.

But there are far more than three ways to give an account of what we know, as many ways as there are to measure what we know. We don't measure the temperature outside our front door the same way we measure the distance to the mail box in our courtyard, and we don't measure either the same way we measure the distance to the moon. We don't measure the circumference of the earth and the girth of a horse the same way. We don't measure a child's height the same way we measure its pulse rate, and we don't measure either the same way we measure its weight or mass. What do all these methods of measurement have in common -- that they result in a number?

So does a child's practicing "adding and take away", but we don't normally call that a method of measurement. When we measure a child's knowledge of geography by asking it the location of Russia on a world map, or its knowledge of geometry by asking it to draw a right triangle, that does not result in a number, and yet it is a method of measuring knowledge.

That is, what is "an account of what we know" differs from case to case: there is no general definition of 'to give an account'; and so, neither is there a general definition of 'to know', if by 'to know' we mean 'to be able to give an account of what you know to others' (Xenophon, Memorabilia iv, 6, 1).

Heraclitean Preconception

But that is not the way Plato thinks; he assumes that every name that is a true name (i.e. the name of an Archetype or Absolute Reality) has a general definition, and he tries, unsuccessfully, by the method of guessing, to always find one. [Heraclitus and Plato]

Suppose we were asked ... what clay is; it would be absurd to answer: potter's clay, and ovenmaker's clay, and brickmaker's clay, and dollmaker's clay.... it is absurd to imagine that our answer [which would be merely a particular example, not a general definition] conveys any meaning to the questioner [who does not know what clay is], when we use the word 'clay' ... (Theaetetus 147a-c; Plato gives another example of a general definition, this time of 'quickness', in Laches 192a)

Even if there were a general (or defining common quality or defining common nature) definition of the word 'clay', it would not follow as a matter of course that if there is a general definition of 'clay' then there must also be a general definition of 'knowledge'. Plato says that "clay is earth mixed with moisture", but there are many kinds of earth of which sand is one and moist sand is not clay (and therefore Plato's definition is not Socratic, according to Aristotle's account, because it doesn't answer the question: what differentiates the earth of clay from all other kinds of earth?)

"Well, I have a general idea of it"

How many types of clay must you be familiar with before you "know what clay is"? We say things like, "Well, I have a general idea; I know, or I think I know, what to expect when someone talks about clay, e.g. that he has found clay while digging in his garden, or that he needs clay for his potter's wheel."

Can we say the same about knowledge -- i.e. that we have "a general idea; we know what kind of thing to expect when a claim to knowledge is made"? Maybe, but what do we mean by 'a general idea': can we put that idea into words, e.g., as a definition? -- or is it an "as yet unformulated definition" (PI § 75)?

If by the expression 'I have a general idea of it' someone simply means 'I think I know what to expect' or 'I think I am oriented, that I know the correct neighborhood' or 'I think I know my way about' or 'I have a sense of it, a knack', then it is harmless. But if it at all suggests the metaphysical theory of abstraction, then in philosophy it is very far from harmless (How to stay confused) indeed.

If Plato had been much interested in the question of "what clay is", would he have accepted [been satisfied with] his own general definition of 'clay'?

"I know it when I see it"

"I can't tell you what a game is, but I know one when I see one." But what do you mean you "can't tell" me, for you can certainly give me examples of games -- and in this case isn't giving examples giving an account of all that you know about "what a game is"? (But the English-speaking lion does not know humor when he hears it -- i.e. he does not know when (or how) to be amused -- despite his being able to give examples of things human beings call 'humor'.) Is the criterion for saying that someone who says "I know one when I see one" is correct that when presented with a game that is new to him (Perhaps he has never heard of backgammon) he identifies it as belonging to the category 'game'? But what if he is never confronted with a game that is new to him?

Someone who says that the meaning of the expression 'I know one when I see one' = 'I can't say what the essence of x is, but I must have abstracted it', is not describing how we use the expression 'I know one when I see one', but is instead substituting a metaphysical theory -- i.e. a speculation about a "hidden reality" that is presumed to be the explanation of "how anyone knows, but cannot say, what the essence of x is" -- for a description of the actual grammar.

This is not a case of theory versus practical application, but a description of reality versus a speculation that floats free of reality -- i.e. that is anomaly-proof: it is logically impossible for it come into conflict with experience, but it is instead consistent with any and all experience; and therefore it is an idle speculation in masquerade: an account that accounts for nothing because it, as it were, accounts for everything. (The theory of abstraction is on the same level as the Cartesian theory that God puts these essences into the human mind.)

Examples and definition

Either logic of language overstates concept fluidity (the indefiniteness of the borders of our concepts) -- or the role of examples in defining words is understated, and most common names are defined, not by essential definitions, but by examples (most clearly in the case of ostensively defined names such as 'cat' and 'cow', and by well-established lists of examples such as in the case of the common name 'game'). Words defined by essential definitions have definite (fixed) borders, but their role in our normal use of language is overstated (the preoccupation of philosophers, lexicographers, and English teachers).

Defining words by giving examples ("enumeration") is not a defective or inferior way of defining words -- it is the normal way. Essential (common nature) definitions are the exception.

As we normally use language essential definitions are for the most part unnecessary, but in Plato's Euthyphro Socrates argues that an essential definition of 'piety' is necessary if we are to always know what is and what is not pious, and not disagree and fall out with one another over that question. (It seems that such a definition of 'justice' would be even more needed.) And so Socrates makes an important point in Plato's dialog, but it doesn't follow that Plato understands the logic of our language when he presumes that phenomena can be "defined" by hypotheses -- as if phenomena defined concepts rather than concepts -- i.e. rules for the use of words -- define phenomena. In other words, we define words not things: there is a relationship between facts and concepts, but we set and revise the criteria for a correct understanding of that relationship. Essence "belongs to grammar" because essence is grammar.

Essential definitions are the exception in natural language; we do not learn to use our language by learning them. (The general definitions in dictionaries that we are taught at school are too general -- they cannot be understood without examples to explain their meaning.)

[Plato is trying to identify "the essence of the phenomenon of piety", but in this case is there a difference between that and trying to formulate a general definition of the word 'piety'? Is the only difference that the latter way of posing the question recognizes that concepts -- i.e. definitions of words (conventions) -- define phenomena? Phenomena without concepts are blind. ]


Internal Connections

Why all these cross-references (internal links)? Because of the immense background of preparation my remarks presuppose. As I once headed these appendixes: "Nothing in this appendix should be taken to represent Wittgenstein's views, but this appendix won't be understood without first understanding Wittgenstein's logic of language." I feel that if I am to explain anything, I must explain everything. [All the contents of Wittgenstein's "logical rucksack" must be opened out onto the table. (Wittgenstein: "A thinker is very much like a draughtsman whose aim it is to represent all the interrelations between things" (CV p. 12, remark from 1931). That is a "bird's-eye view".)]


Site copyright © September 1998. Send Internet mail to Robert [Wesley] Angelo. Last revised: 12 January 2022 : 2022-01-12

The URL of this Web page:
https://www.roangelo.net/logwitt/logwit31.html

Back to top of page

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