Soft Sciences Vs. Hard Sciences

[teo123]

BrianSoddingBoru4 on AtheistForums thinks that neuralbeans'es objection "Perhaps the nouns in the Croatian language have a significantly lower collision entropy than all the words in the Aspell spell-checking dictionary. Have you checked for that?" is a serious problem with my interpretation of the Croatian names of places. So, I've started a thread about it on two Internet forums to see if anybody is aware of some research on the topic, or at least how I can make an experiment to test that without spending days compiling a list of nouns in the Croatian language:
https://linguistics.stackexchange.com/q/47201/20821
https://www.forum.hr/showthread.php?p=9 ... st99311279
To me, what neuralbeans is saying seems like an obvious ad-hoc hypothesis to defend the mainstream interpretation of the names of places in Croatia: it is inventing the reasons why an experiment wouldn't work. Why would the collision entropy of different word classes (nouns, verbs, adjectives...) be different in the Croatian language? I can see why it would be in the Swahili language (because of the noun classes, verbs can start in some consonant pairs that nouns cannot start with), but I fail to see why it would be in the Croatian (or English) language. And why would the collision entropy of nouns be lower, rather than higher? Seems like a baseless ad-hoc hypothesis, right? And it's not a burden of proof on me to do some complicated experiment because of someone's ad-hoc hypothesis.

[teo123]

What do you think @brimstoneSalad, can at least the following thing be stated with reasonable certainty about the names of places in Croatia?

At a rate far higher than chance, the first two consonants of the river names in Croatia tend to be 'k' and 'r', respectively. The p-value of that pattern is somewhere between 1/300 and 1/17, depending on the assumptions you make about the collision entropy of parts of the grammar of the Croatian language. However, mainstream linguistics provides no explanation for that pattern.

Maybe that's the only thing about the names of places in Croatia that can be stated with reasonable certainty. I mean, sure, there are some things that I think all the scientists who have studied Croatian toponyms agree on (Illyrian was an Indo-European language, "Colapis" meant "winding water" in Illyrian, "Serapia" meant "flowing water"...), but those things are without p-values.

[teo123]

A Quora user called Michal Pietrusinski told me that I am suggesting some irregular sound changes when I am proposing that "Karašica" comes from Illyrian *Kurrurrissia, and that that's especially problematic since I am proposing that those irregular sound changes occurred twice (for both rivers named Karašica). He didn't specify which "irregular" sound changes he has in mind, and Quora isn't letting me ask him via a comment. As far as I understand the history of Croatian language, if *Kurrurrissia was borrowed into Croatian around the 7th century, it would turn to *Karaš(-ica) in modern Croatian via regular sound changes.
So, I asked about it on forum.hr.
I seriously doubt that I actually made such a mistake, because why didn't Dubravka Ivšić warn me about that when I sent her an early version of my paper Etimologija Karašica? If my hyothesis is impossible (or at least very unlikely) from the perspective of historical phonology, why not use that as an argument, rather than using much weaker arguments?

[teo123]

A forum.hr user called DarkDivider claims that there was a phonotactic law operating in Proto-Slavic that made it impossible for more than three syllables with yers to be consecutive. That would make my supposed Proto-Slavic name for Karašica, *Kъrъrьsьja (a borrowing from Illyrian *Kurrurrissia), impossible. Of course, DarkDivider cited no source for that claim.
In my opinion, he is doing the same thing I was doing back in 2016, when I was inventing laws of optics which supposedly prove the Moon Landing photographs were fake and inventing the laws of fluid mechanics which supposedly prove rockets cannot exist.