16:31 “We’re striving to eliminate error … we’re not aspiring to be able to utter true statements. We’re aspiring to eliminate false statements from the things that we say, which always leaves more false statements and we can’t be sure we haven’t eliminated a true statement” 17:56 “We guess that our statement is some kind of indicator of the abstraction” 21:00 “My problem is: if truth is correspondence with the facts, how can a statement - which is incapable of being true or false - correspond to a fact, which is either there or is not?”
A statement like "This statement is false" does not represent a proposition, because when you try to find the proposition that it represents there is nothing. Statements are imperfect tools for pointing at propositions, just like your finger is an imperfect tool for pointing at physical objects. They're usually imperfect for their finite ambiguity, but also they can fail to represent any proposition at all, not unlike how your finger can be pointing toward nothing in particular.
Statements represent propositions and propositions correspond to reality. A statement is true when the proposition it represents corresponds to reality. Statements are always insufficient to correspond to propositions, so they can only be representations of a proposition. We can improve our representations, but they're always just stand-ins. Statements and models must be constructed from a finite physical substrate, while the propositions they represent are infinite in their logical content. Propositions are abstract objects that correspond to other objects, including physical and abstract objects.
Once again, a fascinating musing from David. The only problem with the theory that comes to mind is how do statements exist, ontologically speaking, if they are not themselves abstractions (propositions)? David mentions that statements are physical entities (like audible vibrations or pen marks) that get interpreted into abstractions, but the concept of "physical" is itself an abstraction that we think describes/corresponds to reality. I just wanted clarity on this issue, as the idea currently seems internally inconsistent. Perhaps, statements are abstractions that exists at a different level of emergence? the realm of "meaning"?
Maybe the solution lies somewhere in the fact that statements are imprecise guesses at abstractions which themselves imprecisely correspond to reality. After all, everything to which we have access is an abstraction. I like the idea that language (linguistic expression) is itself a beginning of infinity wherein the best statement is the best label for an abstraction. Thus, statements are a form of knowledge which are also subject to evolution. Perhaps, this is what Wittgenstein meant by "meaning is use" and "philosophy is language games" and "there are no philosophical problems, just philosophical puzzles".
Seems to me Popper has correspondence but this is better thought of as approximation. We endeavor to identify error and in so doing get a better guess. I agree with Kant we never know a thing in itself. As for realism and truth, I think we best consider utility. A better guess has better utility in the real world.
A statement like "this is a dog" may represent a definite proposition because any particular context in which such a statement may be uttered is bundled into the respective proposition it represents. This is exactly why when someone points at a dog and says "this is a dog", you don't have any difficulty interpreting what proposition that statement is representing. The statement is just a tool to point to the proposition like your finger is a tool for pointing at the dog. Both are imperfect tools, but it doesn't matter, because the proposition they represent has no such ambiguity. The challenge of communication is in both parties honing in on the same infinite proposition through the tool of uttering finite statements. Because propositions are only imperfectly represented in our brains, using neurons and electrochemistry, even our minds only map a finite segment of an infinite object. Our minds can explore that object, but they cannot fathom it. Since our knowledge consists of the infinite abstractions rather than our just our finite representations, our knowledge must always go beyond our own comprehension of it.
At 32:24 David talks of the set of VR environments being the set of Turing computable functions and he goes on to say that the "abstractions we can speak of are a much larger set than {the computable}". So does not Charles's request at 32: 12 amount to asking for comments on whether one day, since mathematicians and physicists are physical systems employing abstractions via physical instantiations, we will be able to produce a VR rendering of such people speaking of yet further (non computable) abstractions? As I understand it, the link if there is one amounts to asking whether we will one day program an AGI.
I think there is a succinct way convey this idea. Language, of any type, including mathematics, is asymptotic. There is the objective reality, or truth, such as possible quantities; i.e. 5. But all the ways we try to communicate, even with the greatest precision, can never be the thing itself. We can, theoretically, with infinitely increasing precision, disambiguate the proposition, the being of 5, but it will always fall short of the absolute truth of 5. Asymptotic approach.
'Correspondence' is a confusing name for this theory of truth. It is clearer to say that 'truth' is 'accurate description or representation' (as most of our theories are tacit and never put into words or other symbols). It is then more clear that a description/representation can attain 'degrees of accuracy' (a similarly less confusing expression than 'verisimilitude'). Many (most?) of our everyday theories are perfectly true (I am sitting on a chair; I am typing; the sky overhead is cloudy; etc.). But they remain conjectures, which my senses and reason test from moment to moment.
We should perhaps bear in mind that 'the correspondence theory of Truth' is how this theory has been known for many years in philosophy. Renaming it to conform to your preferences dosn't help. If you don't like what you suspect are implicit assumptions better to review the literature to make sure you are right then present your arguments in that context to persuade those of us who are familiar with the term and its history.
It is not advocated that it be renamed. A clearer account of what is meant by it is offered. You offer no relevant criticism but, in effect, ask for more ‘support’ for that account to persuade the ‘cognoscenti’. Oh dear.
David's theory is similar to Hilary Lawson's theory of closures from analytical philosophy. They both presume that statements expressed in natural language can never precisely refer to abstractions/reality. Hilary and David both have Oxford roots.
What did Popper think about the "Many Worlds Interpretation" of quantum mechanics? (I realize this is off topic here but, given that those are David Deutsch and the Carl Popper Society, it's a very spontaneous question!)
I'm not sure if I understand what has been said ... Can someone correct me if I am wrong. David believes that statements are mere representations of our abstractions (which includes a vast array of classifications of knowledge), and only our abstractions have a correspondence to what is pristinely true. Since statements are representations, we can only ever utter them in the form of sound waves generated from speech or physically instantiate these statements as words on paper. And so, there is always a gap between the true propositions that correspond to reality and statements which amount to approximations of these truths (as they will contain errors).
I don't get why propositions have to be infinite. A simple proposition can refer to a bunch of entities and assert that they are related in some way. Its relation to the entities and to the type of relation makes it a finite abstraction. What I think is more appropriate to say, and even then I'm unsure, is that technically one would need an infinite STATEMENT in order to avoid the possibility of error in someone interpreting that statement as a proposition. But once one has gone through the interpretation, the proposition they find through it is finite and simple.
I'm not entirely sure what you mean: we don't have access to propositions because they are entirely unambiguous, and to remove all ambiguity from a statement, the statement would have to be infinitely long.
@@OxfordKarlPopperSociety In writing this I had in mind something a previous comment said: "The challenge of communication is in both parties honing in on the same INFINITE PROPOSITION through the tool of uttering finite statements." I didn't see why the property of being infinite was applied to propositions.
@@OxfordKarlPopperSociety Another thing. Given a statement, we may be wrong about what proposition it's meant to stand for. But in making a guess about what it stands for, don't we access *some* proposition? Granted, maybe the wrong one, but still an instance of the class of propositions?
Interesting conversation but I think some opportunities were missed because David was concerned with maintaining talk of truth in terms of correspondence. Rather than asking, "if truth is the correspondence of propositions to facts, how do we deal with the ineffability of propositions in language?", one might ask, "if language is the medium of propositional expression, what must truth consist in?" The answer to the second question of course could not be correspondence, given the problems that David points out with language, but that doesn't mean conclusively that the question must be based on a false premise. Instead, the issues that David raises can point the way to Quinean holism, which doesn't need an ineffable metalanguage. In fact, I suspect that David's approach would break down when trying to fill in his notion of how abstract terms would refer to chunks of reality, as I doubt that they could genuinely be constructed without reference to the linguistic expressions we use to approximate them, leading to the problem of analyticity that Quine dismantled in Two Dogmas.
David has no pretension to claim we can know with any certainty that statements do represent abstract propositions, he just has a theory that they do. Abstract propositions can refer to reality, and scientific theories show that they can, via the use of mathematical structures which the theories conjecture correspond to the world.
Post Modern French philosophy - something like "there is no absolute truth in anything we say" and "its power plays" But if I say to someone "Waitrose is open right now", that's true in a practical sense and very useful. So lets not go overboard on bad French philosophies.
@@OxfordKarlPopperSociety Yes, thank you. First of all, judging by the physionomy ... The characteristic facial features are obviously similar. And the statures could also provide clues. What is amazing is that even the way they speak is very similar in both people. Aren't both Professor David Deutsch and Professor Martin Rees cosmologists? So if it is not the relationship, we should look for other, unknown causes of the striking resemblance, which I find in the dark of astonishment ...
David repeatedly says verisimilitude isn't a useful concept because there's no way to quantify the truth in a proposition. You can use it whilr casually speaking as a synonym for the phrase that "our theories get closer to the truth", but it only introduces the possibility of mistake that someone will interpret you in a more formal way, which is a problem since it isn't possible to formalize the concept.
@@PicturesJester Justification, which is included in virtually all definitions of the word knowledge, isn't coherent in a "formal" sense either. Should we abandon that word? Or how about the word truth itself? No doubt people will think you're talking about the raw, unobserved, absolute state of reality kind of truth. People's interpretations of my words are their problem, not mine. All I can do is imbue my words and symbols with as accurate an intent of my meaning as possible, and if I judge veristimilitude has value? It has. It's especially relevant in this context because when you abandon certainty, some approximation of the truth is what you're left with, which is what the word describes. Listen to what he says at 54:00, his vision of progress in knowledge completely maps onto it.
@@Google_Censored_Commenter When David talks about knowledge, he never means a justified belief, or a justified claim. Yes, we should completely abandon the notion of knowledge as justification. His notion of progress has nothing to do with our having more justified beliefs, or our already justified beliefs having a firmer grounding justification. He just meant we are able to give more explanations that solve problems we couldn't before. Those explanations aren't justified, they are just good in that they solve problems without raising more problems than they solve.
16:31 “We’re striving to eliminate error … we’re not aspiring to be able to utter true statements. We’re aspiring to eliminate false statements from the things that we say, which always leaves more false statements and we can’t be sure we haven’t eliminated a true statement”
17:56 “We guess that our statement is some kind of indicator of the abstraction”
21:00 “My problem is: if truth is correspondence with the facts, how can a statement - which is incapable of being true or false - correspond to a fact, which is either there or is not?”
We can listen David hours.
A statement like "This statement is false" does not represent a proposition, because when you try to find the proposition that it represents there is nothing. Statements are imperfect tools for pointing at propositions, just like your finger is an imperfect tool for pointing at physical objects. They're usually imperfect for their finite ambiguity, but also they can fail to represent any proposition at all, not unlike how your finger can be pointing toward nothing in particular.
Statements represent propositions and propositions correspond to reality. A statement is true when the proposition it represents corresponds to reality. Statements are always insufficient to correspond to propositions, so they can only be representations of a proposition. We can improve our representations, but they're always just stand-ins. Statements and models must be constructed from a finite physical substrate, while the propositions they represent are infinite in their logical content. Propositions are abstract objects that correspond to other objects, including physical and abstract objects.
Did you hear the last question though @ 1:26:53 ? Does it make sense to you?
Once again, a fascinating musing from David. The only problem with the theory that comes to mind is how do statements exist, ontologically speaking, if they are not themselves abstractions (propositions)? David mentions that statements are physical entities (like audible vibrations or pen marks) that get interpreted into abstractions, but the concept of "physical" is itself an abstraction that we think describes/corresponds to reality. I just wanted clarity on this issue, as the idea currently seems internally inconsistent. Perhaps, statements are abstractions that exists at a different level of emergence? the realm of "meaning"?
Maybe the solution lies somewhere in the fact that statements are imprecise guesses at abstractions which themselves imprecisely correspond to reality. After all, everything to which we have access is an abstraction. I like the idea that language (linguistic expression) is itself a beginning of infinity wherein the best statement is the best label for an abstraction. Thus, statements are a form of knowledge which are also subject to evolution. Perhaps, this is what Wittgenstein meant by "meaning is use" and "philosophy is language games" and "there are no philosophical problems, just philosophical puzzles".
Seems to me Popper has correspondence but this is better thought of as approximation. We endeavor to identify error and in so doing get a better guess. I agree with Kant we never know a thing in itself. As for realism and truth, I think we best consider utility. A better guess has better utility in the real world.
A statement like "this is a dog" may represent a definite proposition because any particular context in which such a statement may be uttered is bundled into the respective proposition it represents. This is exactly why when someone points at a dog and says "this is a dog", you don't have any difficulty interpreting what proposition that statement is representing. The statement is just a tool to point to the proposition like your finger is a tool for pointing at the dog. Both are imperfect tools, but it doesn't matter, because the proposition they represent has no such ambiguity. The challenge of communication is in both parties honing in on the same infinite proposition through the tool of uttering finite statements.
Because propositions are only imperfectly represented in our brains, using neurons and electrochemistry, even our minds only map a finite segment of an infinite object. Our minds can explore that object, but they cannot fathom it. Since our knowledge consists of the infinite abstractions rather than our just our finite representations, our knowledge must always go beyond our own comprehension of it.
Very interesting, thank you.
At 32:24 David talks of the set of VR environments being the set of Turing computable functions and he goes on to say that the "abstractions we can speak of are a much larger set than {the computable}". So does not Charles's request at 32: 12 amount to asking for comments on whether one day, since mathematicians and physicists are physical systems employing abstractions via physical instantiations, we will be able to produce a VR rendering of such people speaking of yet further (non computable) abstractions? As I understand it, the link if there is one amounts to asking whether we will one day program an AGI.
I think there is a succinct way convey this idea. Language, of any type, including mathematics, is asymptotic. There is the objective reality, or truth, such as possible quantities; i.e. 5. But all the ways we try to communicate, even with the greatest precision, can never be the thing itself. We can, theoretically, with infinitely increasing precision, disambiguate the proposition, the being of 5, but it will always fall short of the absolute truth of 5.
Asymptotic approach.
50:50. The difference
that you might perceive in imagining whether an ant is on the outside of a donut or inside the donut nor neither.
'Correspondence' is a confusing name for this theory of truth. It is clearer to say that 'truth' is 'accurate description or representation' (as most of our theories are tacit and never put into words or other symbols). It is then more clear that a description/representation can attain 'degrees of accuracy' (a similarly less confusing expression than 'verisimilitude'). Many (most?) of our everyday theories are perfectly true (I am sitting on a chair; I am typing; the sky overhead is cloudy; etc.). But they remain conjectures, which my senses and reason test from moment to moment.
We should perhaps bear in mind that 'the correspondence theory of Truth' is how this theory has been known for many years in philosophy. Renaming it to conform to your preferences dosn't help. If you don't like what you suspect are implicit assumptions better to review the literature to make sure you are right then present your arguments in that context to persuade those of us who are familiar with the term and its history.
It is not advocated that it be renamed. A clearer account of what is meant by it is offered. You offer no relevant criticism but, in effect, ask for more ‘support’ for that account to persuade the ‘cognoscenti’. Oh dear.
David's theory is similar to Hilary Lawson's theory of closures from analytical philosophy. They both presume that statements expressed in natural language can never precisely refer to abstractions/reality. Hilary and David both have Oxford roots.
What did Popper think about the "Many Worlds Interpretation" of quantum mechanics?
(I realize this is off topic here but, given that those are David Deutsch and the Carl Popper Society, it's a very spontaneous question!)
If you search youtube you'll find Deutsch himself covering that topic.
I'm not sure if I understand what has been said ... Can someone correct me if I am wrong.
David believes that statements are mere representations of our abstractions (which includes a vast array of classifications of knowledge), and only our abstractions have a correspondence to what is pristinely true.
Since statements are representations, we can only ever utter them in the form of sound waves generated from speech or physically instantiate these statements as words on paper. And so, there is always a gap between the true propositions that correspond to reality and statements which amount to approximations of these truths (as they will contain errors).
You are all missing the point. It is not about truth. It is about the elimination of error.
Ding Ding !
Presently I favor that the feymen diagrams are so very close to a fitting puzzle that is incredibly near holding together a picture of a cylinder.
Ha! "-It's good explanations all the way down!"
I don't get why propositions have to be infinite. A simple proposition can refer to a bunch of entities and assert that they are related in some way. Its relation to the entities and to the type of relation makes it a finite abstraction. What I think is more appropriate to say, and even then I'm unsure, is that technically one would need an infinite STATEMENT in order to avoid the possibility of error in someone interpreting that statement as a proposition. But once one has gone through the interpretation, the proposition they find through it is finite and simple.
I'm not entirely sure what you mean: we don't have access to propositions because they are entirely unambiguous, and to remove all ambiguity from a statement, the statement would have to be infinitely long.
@@OxfordKarlPopperSociety In writing this I had in mind something a previous comment said: "The challenge of communication is in both parties honing in on the same INFINITE PROPOSITION through the tool of uttering finite statements." I didn't see why the property of being infinite was applied to propositions.
@@OxfordKarlPopperSociety Another thing. Given a statement, we may be wrong about what proposition it's meant to stand for. But in making a guess about what it stands for, don't we access *some* proposition? Granted, maybe the wrong one, but still an instance of the class of propositions?
Interesting conversation but I think some opportunities were missed because David was concerned with maintaining talk of truth in terms of correspondence.
Rather than asking, "if truth is the correspondence of propositions to facts, how do we deal with the ineffability of propositions in language?", one might ask, "if language is the medium of propositional expression, what must truth consist in?"
The answer to the second question of course could not be correspondence, given the problems that David points out with language, but that doesn't mean conclusively that the question must be based on a false premise. Instead, the issues that David raises can point the way to Quinean holism, which doesn't need an ineffable metalanguage.
In fact, I suspect that David's approach would break down when trying to fill in his notion of how abstract terms would refer to chunks of reality, as I doubt that they could genuinely be constructed without reference to the linguistic expressions we use to approximate them, leading to the problem of analyticity that Quine dismantled in Two Dogmas.
David has no pretension to claim we can know with any certainty that statements do represent abstract propositions, he just has a theory that they do.
Abstract propositions can refer to reality, and scientific theories show that they can, via the use of mathematical structures which the theories conjecture correspond to the world.
David would write SCIENCE FICTION books/novels. It suits him, I guess. 😊
Post Modern French philosophy - something like "there is no absolute truth in anything we say" and "its power plays"
But if I say to someone "Waitrose is open right now", that's true in a practical sense and very useful.
So lets not go overboard on bad French philosophies.
They weren't necessarily bad philosophers, though they certainly wrote some bad philosophical arguments (especially some not-even-wrong arguments).
Not sure that if true 'it is most likely to be found somewhere in Popper' is likely to satisfy most serious philosophers :)
Where did he say that?
How come?
@@l1berty53000 3.13
Because that isn't an argument. And btw - my remark was intended to be taken lightly.
@@PeteGriffiths_petegrif Ah, right. I think that was a self-deprecating joke on his part.
What about the proposition: most men are mortal. Could we just leave that without qualifiers?
A proposition must be true or false. This one is demonstrably false.
Ist David Deutsch der Sohn von Martin Rees?
Is David Deutsch the son of Martin Rees? No.
@@OxfordKarlPopperSociety Yes, thank you.
First of all, judging by the physionomy ...
The characteristic facial features are
obviously similar.
And the statures could also provide clues.
What is amazing is that even the way they
speak is very similar in both people.
Aren't both Professor David Deutsch and
Professor Martin Rees cosmologists?
So if it is not the relationship, we should
look for other, unknown causes of the
striking resemblance, which I find in
the dark of astonishment ...
Poor volume
I'm sensing a distinct lack of my favorite word in these conversations: verisimilitude
David repeatedly says verisimilitude isn't a useful concept because there's no way to quantify the truth in a proposition. You can use it whilr casually speaking as a synonym for the phrase that "our theories get closer to the truth", but it only introduces the possibility of mistake that someone will interpret you in a more formal way, which is a problem since it isn't possible to formalize the concept.
@@PicturesJester Justification, which is included in virtually all definitions of the word knowledge, isn't coherent in a "formal" sense either. Should we abandon that word? Or how about the word truth itself? No doubt people will think you're talking about the raw, unobserved, absolute state of reality kind of truth. People's interpretations of my words are their problem, not mine. All I can do is imbue my words and symbols with as accurate an intent of my meaning as possible, and if I judge veristimilitude has value? It has.
It's especially relevant in this context because when you abandon certainty, some approximation of the truth is what you're left with, which is what the word describes. Listen to what he says at 54:00, his vision of progress in knowledge completely maps onto it.
@@Google_Censored_Commenter When David talks about knowledge, he never means a justified belief, or a justified claim. Yes, we should completely abandon the notion of knowledge as justification. His notion of progress has nothing to do with our having more justified beliefs, or our already justified beliefs having a firmer grounding justification. He just meant we are able to give more explanations that solve problems we couldn't before. Those explanations aren't justified, they are just good in that they solve problems without raising more problems than they solve.