Are all models wrong?

[Skepdick]

WOT

Surely responding with "WOT" to posts that are a "Waste Of Time" is a waste of time? So the question of intent/telos/purpose still perplexes me!

Why are you doing it then?

I wouldn't want it thought that I don't read posts that address what I've said.

So you are actually conceding that my posts address what you've said?!? That's even more perplexing!!!

Would you prefer it if my posts didn't address what you are saying?

[Peter Holmes]

Surely responding with "WOT" to posts that are a "Waste Of Time" is a waste of time? So the question of intent/telos/purpose still perplexes me!

Why are you doing it then?

I wouldn't want it thought that I don't read posts that address what I've said.

So you are actually conceding that my posts address what you've said?!? That's even more perplexing!!!

Would you prefer it if my posts didn't address what you are saying?

Perhaps 'quote' is better than 'address' - though I think punching someone is also a form of address.

In fact, you don't actually address or even acknowledge refutations. You just slide off to another tendentious or flat ridiculous point that demonstrates your lack of understanding. And it's tedious to have to keep showing you your mistakes, because you refuse to learn and correct them. Of course, you can just ner-ner and say I do exactly the same. So we're at an impasse, as usual. My WOR response means 'there's nothing here worth responding to, as far as I'm concerned'. ('But I have read it, and so I'm not dismissing it a priori.')

[uwot]

I am sufficiently certain (to say it) that the statement is perfectly coherent to Peter. That's why he's saying it.

Also, I am sufficiently doubtful (to say it) that the statement can't possibly be coherent to you, given that you are in a vacuum as to any particulars of Peter's knowledge or his wife.

That's where we differ. PH doesn't have to prove to me that he knows enough about his wife for me to find his saying 'I know my wife' coherent. Having a wife myself, I am quite happy to believe that some of what he knows about his wife would correspond with some of what I know about my wife. Since "the statement is perfectly coherent to Peter" and to me, the fact that we are talking about two different wives doesn't, in my view, make his saying 'I know my wife' incoherent.

You are in the exact same vacuum that I am in. If I don't have any particulars - neither do you, so whatever it is that you call 'understanding' of Peter's position seems rather superficial.

Terribly sloppy logic again, Skepdick: it doesn't follow from the fact that you "don't have any particulars" that neither do I.

Those facts, I suspect, would include knowing her birthday, some of her background and history, some of her likes and dislikes, interests, hobbies, work situation, some physical features; the sort of information that even a moderately successful relationship is dependent on.

This is common sense - you aren't saying anything that needs saying.

This is something I would describe as incoherent. By labelling it "common sense", presumably you accept that there are things PH knows about his wife. Unlike you, I do not think that any such facts need to be attached to his saying 'I know my wife' in order to make that claim coherent.

The statement mentions knowledge, but does not communicate any of it.

Well, until he said it, I didn't know he had a wife.

It's not your place to decide my level (or lack) of understanding, far less everybody else's.

I know that. Which is why I am not deciding/judging your (or anybody's) level of understanding.

Clearly I have misinterpreted this then: "I am sufficiently doubtful (to say it) that the statement can't possibly be coherent to you."

Everything that I said is a description of my own, personal state of mind. I am stating the intent behind my words. I am not prescribing how you SHOULD interpret my words. I am merely stating that if you are interpreting my meaning differently to how I intend it to be understood, then you are probably making an interpretative error.

Well, given that "natural languages are near-infinitely flexible", it is a challenge.

My attitude towards that sentence is "Whatever!"

Seven pages for "Whatever!"?

...I didn't ask you whether you've met Karl Popper or not. I asked you whether you know him or not.

I was hoping that the question would receive a yes/no answer.

No.

[Skepdick]

In fact, you don't actually address or even acknowledge refutations.

Because I am not playing the game of "arguing" and "refuting arguments". You are playing that game and you expect me to play it too.

I am outright telling you that I am not playing that game and I have no intention to play it.

You just slide off to another tendentious or flat ridiculous point that demonstrates your lack of understanding.

Peter, I have been telling you that I don't understand what you mean from the get-go. I am not even sure why you think that it required a "demonstration"? That seems like a really stupid game.

It is precisely because I don't understand is why I keep asking you to use the sentence in a broader context. You are still avoiding a simple request!

And it's tedious to have to keep showing you your mistakes, because you refuse to learn and correct them.

Peter, it's tedious to interact with black-and-white thinkers such as yourself. You udon't know what a "mistake" is.

What you seem to think is "showing and correcting my mistakes" amounts to me not doing what you want me to do. That's entirely my prerogative - you have zero influence or authority over me or anybody.

It's pretty arrogant to even assume the role of an "error corrector".

Of course, you can just ner-ner and say I do exactly the same.

Your only "error" is your bone-headedness. I keep asking you to use a simple fucking sentence in broader context, and here you are - going on some tangents on how I it's a mistake to do so...

So we're at an impasse, as usual. My WOR response means 'there's nothing here worth responding

There IS nothing here "worth responding to". Because I am still waiting for you to clarify when and where you might find yourself using a sentence such as "I know my wife".

Once you actually clarify I will actually respond. Until then it's like pulling teeth with you.

[Skepdick]

That's where we differ. PH doesn't have to prove to me that he knows enough about his wife for me to find his saying 'I know my wife' coherent. Having a wife myself, I am quite happy to believe that some of what he knows about his wife would correspond with some of what I know about my wife. Since "the statement is perfectly coherent to Peter" and to me, the fact that we are talking about two different wives doesn't, in my view, make his saying 'I know my wife' incoherent.

I am not asking Peter to "prove" anything. You seem to be so deeply stuck in your philosophical ways.

I am not coming from a place of doubt - I am coming from a place of uncertainty. I neither believe nor disbelieve Peter BECAUSE the sentence "I know my wife" conveys no information - it contains no particulars.

Had he said "I know my wife likes chocolate" I would be content.
Had he said "I know my wife goes to book club on Tuesdays" I would be content.
Had he said "I know my wife has three sisters" I would be content.

But the sentence "I know my wife" conveys nothing of interest whatsoever. You would never use it in a sentence on its own. Which is probably why Peter refuses to demonstrate such use.

Terribly sloppy logic again, Skepdick: it doesn't follow from the fact that you "don't have any particulars" that neither do I.

If it were logic it would be sloppy. What's sloppy is just your misinterpretation. It's a statement of observation, not a statement of logical inference.

Your source of knowledge regarding Peter's wife is exactly the same of mine - this very thread. It follows then you have exactly all the particulars about Peter's wife that I have. I happen to know that I have zero particulars about Peter's wife beyond the trivialities such as "I know she is married to Peter Holmes."

If you have more than zero, I'd be curious to know where you acquired that knowledge from.

This is something I would describe as incoherent. By labelling it "common sense", presumably you accept that there are things PH knows about his wife.

I accept that there are some things PH knows about his wife.
I also accept that there are some things PH doesn't know about his wife.

And so until any given particular is classified into one of the above categories, the sentence "I know my wife" is vague enough so as to be incoherent.

Unlike you, I do not think that any such facts need to be attached to his saying 'I know my wife' in order to make that claim coherent.

That's entirely your prerogative, but it boils down to whether you think being informative/clear is a necessary condition for the coherency of a statement.

I say yes, it is necessary.
You say no, it's not necessary.

We can agree to disagree.

Well, until he said it, I didn't know he had a wife.

And now you know Peter Holmes has a wife. But you still know nothing about her.

Correction. You know that Peter Homes' wife is married to Peter Holmes.

If that's knowledge, I'll trade you for toilet paper.

Clearly I have misinterpreted this then: "I am sufficiently doubtful (to say it) that the statement can't possibly be coherent to you."

Was I in correct in my prediction? You are yet to furnish the particulars of your comprehension.

Well, given that "natural languages are near-infinitely flexible", it is a challenge.

It's way less challenging when you don't presume interaction is adversarial.

[Peter Holmes]

Skepdick

'But the sentence "I know my wife" conveys nothing of interest whatsoever. You would never use it in a sentence on its own. Which is probably why Peter refuses to demonstrate such use.'

Q: Why do you think she'll like that idea?
A: I know my wife.

Now, please explain what a model that is right would consist of. How much and what kind of information would it contain? I look forward to seeing your answer, and I'll abstain here until you produce it.

[Skepdick]

Skepdick

'But the sentence "I know my wife" conveys nothing of interest whatsoever. You would never use it in a sentence on its own. Which is probably why Peter refuses to demonstrate such use.'

Q: Why do you think she'll like that idea?
A: I know my wife.

Peter, pay attention. This is very much the example I gave you. Under a microscope this is a statement of confidence, not a statement of knowledge.
You are guessing/predicting. There is a non-zero change that your guess is wrong.

Now, obviously - you have more information about your wife than I do so I would trust your guess far better than I would trust mine.
But you know whose "guess" I would trust even more? Your wife!

I can simply ask her if she likes the idea and then I'll know. And where would you and your "knowledge" end up if your wife actually hates the idea you thought she'll like?

Now, please explain what a model that is right would consist of.

It would consist of more information than you currently possess. It would be more complete than your current model. It would make less errors than your current model.

It would consist of all the things that you don't know about your wife. And in particular it would consist of a contextual understanding of why your wife doesn't like an idea you thought she would.

How much and what kind of information would it contain? I look forward to seeing your answer, and I'll abstain here until you produce it.

However much information and whatever kind is necessary in order to minimise erroneous predictions.

The ideal is zero errors, but such model cannot exist in practice - that is why all models are wrong.
But relative to the ideal some models are better than others.

[Nick_A]

Some models have a face and body which introduces the additional variable that it doesn't matter if they are right or wrong.

[surreptitious57]

The ideal is zero errors but such model cannot exist in practice

This could be mapped asymptotically since no matter how close one gets to zero it will never actually be attained
This was what Popper was referring to when he labelled hypotheses as polished conjectures - mere approximations

At the moment the standard model of particle physics is only 4 per cent accurate and the other 96 per cent is simply unknown
Over time however those figures will slowly change but it will never be I00 / 0 because that is just not possible with induction

[AlexW]

George Box claimed that ‘All models are wrong but some are useful’.

We will never be able to tell if a model is right or wrong (nor the percentage of its correctness), simply because we can only judge its correctness by introducing a new/"better" model, which supersedes the previous one. But this new model is still a model and as such affected by the same problems as the one before...
It's like trying to approach infinity - you can not get any closer by proceeding in one or the other direction...

[uwot]

Terribly sloppy logic again, Skepdick: it doesn't follow from the fact that you "don't have any particulars" that neither do I.

If it were logic it would be sloppy. What's sloppy is just your misinterpretation. It's a statement of observation, not a statement of logical inference.

I take it that news of irony has yet to reach your outpost. Why is it that you can blithely draw information from sentences that do not contain that information, and insist that 'I know my wife' contains no information, despite admitting that the implicit information is "common sense"?

Those facts, I suspect, would include knowing her birthday, some of her background and history, some of her likes and dislikes, interests, hobbies, work situation, some physical features; the sort of information that even a moderately successful relationship is dependent on.

This is common sense - you aren't saying anything that needs saying.

[Skepdick]

Terribly sloppy logic again, Skepdick: it doesn't follow from the fact that you "don't have any particulars" that neither do I.

If it were logic it would be sloppy. What's sloppy is just your misinterpretation. It's a statement of observation, not a statement of logical inference.

I take it that news of irony has yet to reach your outpost. Why is it that you can blithely draw information from sentences that do not contain that information, and insist that 'I know my wife' contains no information, despite admitting that the implicit information is "common sense"?

Those facts, I suspect, would include knowing her birthday, some of her background and history, some of her likes and dislikes, interests, hobbies, work situation, some physical features; the sort of information that even a moderately successful relationship is dependent on.

This is common sense - you aren't saying anything that needs saying.

Because I am not "drawing information" from the sentence. I am hallucinating information given the sentence. That hallucination is my "common sense". All the information you think I have "drawn from the sentence" is coming from my head.

I am projecting my own meaning onto the sentence "I know my wife".

If I believe that knowing Grandparents' birthdays, shoe sizes, music tastes and blood types are all necessary criteria for "wife-knowledge" then that is what I am hallucinating the sentence "I know my wife" to mean. It's the set of all necessities.

But that says absolutely nothing about PH's intended meaning. Because I know that ALL of my "understanding" of PH's sentence is my own hallucination.
I am yet to establish how much of an overlap there is between my set of necessities and PH's set of necessities for "wife knowledge".

Which is precisely why I am asking PH to use the phrase again and in a different context. So that I can confirm or disconfirm the correctness of my hallucination. It's a calibration tactic.

So that I can confirm whether there is an overlap between Peter's "common sense" and my "common sense" understanding of the phrase "I know my wife". Trust but verify!

This is exactly what I was doing here....

This is common sense - you aren't saying anything that needs saying.

I was giving you confirmatory evidence that the items you have listed (your set of necessities) coincides with mine.

But do observe that you are still only mentioning said knowledge. I know that you think "birthdays, work situation, background, tastes, physical features" etc. are necessary for "wife-knowledge". You are telling me about categories of knowledge, not knowledge and I do seem to recall you recently agreeing that categories are conceptual.

You are telling me things about how you've organized your knowledge, but you haven't shared your actual knowledge with me.

[Skepdick]

This could be mapped asymptotically since no matter how close one gets to zero it will never actually be attained
This was what Popper was referring to when he labelled hypotheses as polished conjectures - mere approximations

That is exactly the sentiment captured by "All models are wrong".

And like I said: it's a sentiment expressed from a view-point of idealism, or perhaps even Platonism.

A model that actually gets to zero is an ideal Platonic form. We know it's impossible, but we work towards it anyway.

At the moment the standard model of particle physics is only 4 per cent accurate and the other 96 per cent is simply unknown
Over time however those figures will slowly change but it will never be I00 / 0 because that is just not possible with induction

If I turn a blind eye to the questions "How are you measuring model-accuracy?" and "How are you converting your measurement to percentages?"
I agree with all of the above.

[uwot]

I am projecting my own meaning onto the sentence "I know my wife".

Right, so according to my understanding of 'incoherent', the sentence isn't incoherent.

I know that you think "birthdays, work situation, background, tastes, physical features" etc. are necessary for "wife-knowledge".

Well, no you don't, because no one of those is necessary in my view.

[Skepdick]

Right, so according to my understanding of 'incoherent', the sentence isn't incoherent.

There's nothing in the text "I know my wife" - they are just symbols on a screen. They contain no inherent information.

We use the words that we use because of the effects we expect (and predict) them to produce in our interlocutor's minds given our best understanding of out interlocutor's minds - we use targeted language.

I use the word "cat" because my intention is to get you to hallucinate (read: imagine, picture) a furry, purry domesticated feline, ergo, I predict that when I say 'cat' you will hallucinate a furry, purry domesticated feline.

When it comes to the sentence "I know my wife" I have no expectations of what you might hallucinate. Your criteria for necessity/sufficiency may be vastly different to mine!

I could assume our criteria are identical; I could assume our criteria are different; I could assume that there is an overlap (however small or large) between our criteria; or I can try and determine (in real time) the degree of similarity and differences between that which you hallucinate and that which I hallucinate when we both hear PH utter the sentence "I know my wife".

Well, no you don't, because no one of those is necessary in my view.

OK, if they aren't necessary (to you) are they sufficient (to you)?

Or is the language of "necessity and sufficiency" not even relevant?