Meditations

[Isobel The Sorceress]

I think Occam's razor principle is near to what Redafro said. I see it as a statistical principle: what is more probable, "Tomorrow it will rain" or "Tomorrow it will rain at 15:00"? The first one covers more cases, so it is more likely to happen. In the best of cases (if you know that beyond all doubt, if it rains tomorrow, it necessarily will rain at 15:00), the two assertments are equally probable. So it will be always better to choose the most simple one. But here "simple" just refers to "covering more possible cases", not simple to understand, or expressed with few words, or anything like that.

And of course, as Redafro pointed out, if you oversimplify it you end up with things like "Tomorrow something will happen", that covers virtually all the cases but is completely useless.

This is all wrong. Occam's razor is NOT about probabilities or predicting the future!

It's about using the simplest possible theory to explain things, until a new (possibly more complex) theory is found that has more explanatory power.

[Vortex]

But theories are about predicting things, if a theory can't predict anything, then it is not falsifiable, and would be deemed pseudoscience by most modern standards.

And what exactly do you mean by "simple", then? My definition of simplicity agrees with that of many authors:

Simple statements, if knowledge is our object, are to be prized more highly than less simple ones because their empirical content is greater; and because they are better testable.

... the simplest hypothesis proposed as an explanation of phenomena is more likely to be the true one than is any other available hypothesis, that its predictions are more likely to be true than those of any other available hypothesis, and that it is an ultimate a priori epistemic principle that simplicity is evidence for truth.

[Isobel The Sorceress]

But theories are about predicting things, if a theory can't predict anything, then it is not falsifiable, and would be deemed pseudoscience by most modern standards.

Theories are about explaining why things happen. You can't predict anything if you don't know the cause.

And what exactly do you mean by "simple", then? My definition of simplicity agrees with that of many authors:

Example: If you throw a ball up in the air, it will fall back down. Why?

Theory A: The law of gravity.
Theory B: The law of gravity that was created by God.

Theory A is more simple, since the fact that God created the law of gravity adds no explanatory power to the theory.

[Vortex]

Theories are about explaining why things happen. You can't predict anything if you don't know the cause.

But then many things would enter in the category of theory, like mythological explanations of the sun. I think a theory must be supported by experiment, and it can't be supported by experiment if it doesn't predict anything. It must also explain why things happen, but the predicting part is essential to a scientific theory.

Example: If you throw a ball up in the air, it will fall back down. Why?

Theory A: The law of gravity.
Theory B: The law of gravity that was created by God.

Theory A is more simple, since the fact that God created the law of gravity adds no explanatory power to the theory.

Then your definition of simple is "what has no statements that add no explanatory power"? Under that definition you are right. But the definition of simple I have seen in the formulation of Occam's Razor that I know is "what makes the fewest assumptions". And that, in a logical-statistical background, means that it covers more possible events, and thus is more likely to be true, which supports the epistemological view of science approaching to the truth. That's also an empirical reason to support that definition of Occam's Razor.

EDIT: I found this in Wikipedia. It might be interesting to our discussion. Specifically the last paragraph. It might mean that we both are right but I don't know much about Turing machines, so I can't say.

[Rooster5man]

There's a difference between a Theory and a Prediction. For example, Submachine Theories: Although they are labeled as "Theories," some are Predictions, i.e. "In Sub10, we'll leave the SubNet."

Although you can consider it a "Theory," it's more a Prediction of what's to be, even though both a Prediction and a Theory are an idea.

A Theory has to have some evidence behind it in order to consider it a Theory - As you said, research and background information. Without anything to back up your claim, and if it hasn't happened yet, it's merely a Prediction.

However, if it's happened and there's no evidence to back it up, I don't feel comfortable in labeling it as a "Theory," but I don't know what else to label it as.

[Vortex]

There's a difference between a Theory and a Prediction. For example, Submachine Theories: Although they are labeled as "Theories," some are Predictions, i.e. "In Sub10, we'll leave the SubNet."

Although you can consider it a "Theory," it's more a Prediction of what's to be, even though both a Prediction and a Theory are an idea.

A Theory has to have some evidence behind it in order to consider it a Theory - As you said, research and background information. Without anything to back up your claim, and if it hasn't happened yet, it's merely a Prediction.

However, if it's happened and there's no evidence to back it up, I don't feel comfortable in labeling it as a "Theory," but I don't know what else to label it as.

Well, the schema I have in my head is like this:
-What you call predictions I call hypotheses.
-A theory must contain testable hypotheses, and also must be self-consistent and have explanatory power (if not, it would just be a relation of two quantities or something along these lines, without semantic content).
-Things that have happened... I'd call them facts

[Rooster5man]

Prediction ~ Hypotheses, I would agree with that, and wouldn't you agree Hypotheses aren't Theories? Yes, they may make up Theories, but a single Hypotheses is not a Theory and thus cannot be applied to Occam's Razor.

[zombyrus]

(I don't have anything to say, and I actually am no longer very interested in Occam's Razor, but I direly want to bump this thread back up.)

[Rooster5man]

We can always change the subject.

[Anteroinen]

Do you have any pet peeves, grammatical or otherwise? Do you consider them justified.