When I was discussing my paper about applying information theory to the Croatian river names on Internet forums, many people told me stuff like "Fuck off with those p-values, go to Alternativa." (Alternativa being a subforum of forum.hr dedicated to things like astrology). That is obviously an absurd thing to say, as p-values are a foundation of the modern scientific method. But why do people on Internet forums say things like that?
I think that is because they have read a lot of tertiary sources (Wikipedia, etymological dictionaries...) and little or no primary and secondary sources. Wikipedia and other tertiary sources of information almost never discuss p-values. So no wonder discussions about p-values sound alien to them. It is very unfortunate.
I was wondering what you thought about that.

This isn't me challenging you, but just asking. Are you sure the p-value approach is appropriate for linguistic matters?

Anyone who takes Wiki as a solid source is being mentally lazy. You have to take each article on its merits regarding accuracy and detail.

I'm not in favor of sweeping generalizations. I tell people that wiki often useful for the breadcrumbs, i.e., the sources cited. Call it step zero in a research effort.

(August 27, 2023 at 9:21 am)FlatAssembler Wrote: When I was discussing my paper about applying information theory to the Croatian river names on Internet forums, many people told me stuff like "Fuck off with those p-values, go to Alternativa." (Alternativa being a subforum of forum.hr dedicated to things like astrology). That is obviously an absurd thing to say, as p-values are a foundation of the modern scientific method. But why do people on Internet forums say things like that?
I think that is because they have read a lot of tertiary sources (Wikipedia, etymological dictionaries...) and little or no primary and secondary sources. Wikipedia and other tertiary sources of information almost never discuss p-values. So no wonder discussions about p-values sound alien to them. It is very unfortunate.
I was wondering what you thought about that.

I don’t think it’s Wikipedia, and I don’t think it has anything to do with secondary or tertiary sources. It’s your ego - you can’t bear to be questioned and any suggestion that you might be wrong or that you’re going about things arse upwards is anathema to you.

It’s not your critics that have a skewed view of science. The problem is a lot closer to home.

Boru

(August 27, 2023 at 10:55 am)GrandizerII Wrote: This isn't me challenging you, but just asking. Are you sure the p-value approach is appropriate for linguistic matters?

Well, if p-values turned out to be so useful in natural sciences, as well as some other parts of linguistics (If I am not mistaken, the Mate Kapović'es PhD thesis was about applying p-values to historical phonology, showing that some supposed law of historical phonology doesn't predict the accents any better than chance.), doesn't that suggest we should at least try to apply them to the names of places?

(August 27, 2023 at 12:21 pm)BrianSoddingBoru4 Wrote:

(August 27, 2023 at 9:21 am)FlatAssembler Wrote: When I was discussing my paper about applying information theory to the Croatian river names on Internet forums, many people told me stuff like "Fuck off with those p-values, go to Alternativa." (Alternativa being a subforum of forum.hr dedicated to things like astrology). That is obviously an absurd thing to say, as p-values are a foundation of the modern scientific method. But why do people on Internet forums say things like that?
I think that is because they have read a lot of tertiary sources (Wikipedia, etymological dictionaries...) and little or no primary and secondary sources. Wikipedia and other tertiary sources of information almost never discuss p-values. So no wonder discussions about p-values sound alien to them. It is very unfortunate.
I was wondering what you thought about that.

I don’t think it’s Wikipedia, and I don’t think it has anything to do with secondary or tertiary sources. It’s your ego - you can’t bear to be questioned and any suggestion that you might be wrong or that you’re going about things arse upwards is anathema to you.

It’s not your critics that have a skewed view of science. The problem is a lot closer to home.

Boru

Whether or not my theory is right, "Fuck the p-values." is an absurd argument. If my theory is wrong, that's because somebody has found (or will find) a model of language more appropriate for this than simple collision entropy measurements and birthday calculations, and that model suggests that this k-r pattern in Croatian river names is not actually statistically significant.

(August 27, 2023 at 1:31 pm)FlatAssembler Wrote:

(August 27, 2023 at 12:21 pm)BrianSoddingBoru4 Wrote: I don’t think it’s Wikipedia, and I don’t think it has anything to do with secondary or tertiary sources. It’s your ego - you can’t bear to be questioned and any suggestion that you might be wrong or that you’re going about things arse upwards is anathema to you.

It’s not your critics that have a skewed view of science. The problem is a lot closer to home.

Boru

Whether or not my theory is right, "Fuck the p-values." is an absurd argument. If my theory is wrong, that's because somebody has found (or will find) a model of language more appropriate for this than simple collision entropy measurements and birthday calculations, and that model suggests that this k-r pattern in Croatian river names is not actually statistically significant.

See?

Boru

Good thing about the sciences, it doesn't give a single flying fuck whether or not some, or all the dingleberries ideas about it are wrong. It's going to work regardless.

(August 27, 2023 at 1:55 pm)BrianSoddingBoru4 Wrote:

(August 27, 2023 at 1:31 pm)FlatAssembler Wrote: Whether or not my theory is right, "Fuck the p-values." is an absurd argument. If my theory is wrong, that's because somebody has found (or will find) a model of language more appropriate for this than simple collision entropy measurements and birthday calculations, and that model suggests that this k-r pattern in Croatian river names is not actually statistically significant.

See?

Boru

See what?