Historical Linguistics is one of the more brutal subfields in Linguistics, probably because you can hardly prove much of anything.
It involves looking at languages and arranging them into families and then arranging them in the families in a proper fashion. So an essential aspect of Historical Linguistics is the discovery of new language families and the elaboration of existing ones. The former is pretty much over in this field because this silly discipline has decided that there will be no more large or old language families discovered. Nonsensically, this has resulted in an utterly idiotic proliferation of insipid “isolates” which are languages that cannot be proven to be related to others. But actually, long-rangers have already stacked most all of the world’s languages into decent families and in their view there are no isolates left.
In addition, there are all sorts of idiotic small families with a couple to separate members, and said family is not related to anything else. I guess nothing’s related to anything then! The bizarre fact is that this preposterous fake science takes great pride in this silly nihilism. Obviously every language is related to every other language ultimately because surely language arose only once in mankind’s history.
Nevertheless, Linguistics insists that this obvious fact is not proven, so I guess it’s not even a fact. Instead the dead solid truth is that somehow there scores of isolates and silly small language families that have no relations. Surely that is a false conclusion. The only way it could be true is if language arose scores of times all the way down to a few thousand years ago.
There were scores of bands of humans who had no language whatsoever except grunts and sign language, and they all independently developed language scores of times in the last ~50,000 years. It was an incredible case of parallel development, the most amazing the world has ever seen. Because this is the only way that Linguistics’ crazy conclusion could be true. So Linguistics is now stating essentially is that this is what happened – language being independently developed all over the world down to the last several thousand years. Dumb, huh?
Historical Linguistics also involves the reconstruction of dead languages or earlier aspects of existing languages. The dead languages have left no record and are often 7-10,000 years old. The earlier phases of existing tongues also have often left no record.
So it is unprovable guesswork guessing at what ancient languages looked like, with no real way to prove if anyone is right or wrong because the languages no longer exist.
On top of that, the field has become mired in stick in the mud conservatism such that I doubt if any new ancient language families are going to proven in my lifetime. The conservatives keep moving the goalposts, and no evidence is ever good enough. Linguistics is ecstatic about this because endlessly moving the goalposts so you can never prove anything anymore means that Linguistics is now really groovy and scientific and this cures their physics envy.
Really it’s just another fake science in the social sciences, although a lot of the more basic work is indeed factually and empirically based. So the field encompasses a lot of excellent empirical based work. In addition, there are a number of preposterous leftwing shibboleths that everyone in the field has agreed are settled truth. Linguistics has adopted these silly ideas because they are leftwing and PC, and the field is at the heart of SJW Central Command. Mixed in with these silly politically based agreed upon facts (for which there is typically no evidence whatsoever) there is this prideful stubbornness and ultra-conservative attitude in Historical Linguistics because the way to be all sciency is to deny forever more any new language families. Because that cures our physics envy and makes us feel all sciency.
Actually many of the long-rangers have gathered excellent evidence for their work, all of which is rejected. For instance, Altaic now has a 1,000 page etymological dictionary of all things and there are many reconstructed forms and a great deal of commonality in basic morphology, core vocabulary, pronouns and language structure. We also have quite a few actual paradigms which are impossible to derive in unrelated languages. The long-rangers churn out many papers and here is where the real science is. They are doing dramatic work and proving a lot of new things.
On the other hand, the fake science folks on the other end chant over and over in Gregorian fashion, “You didn’t prove it. You didn’t prove it. You didn’t prove it.” No matter what evidence is assembled and presented, the response is always this autistic nihilism of “You didn’t prove it.” The arguments of many of the deniers have been destroyed already. The deniers now take the preposterous position that there has been mass borrowing of personal pronouns in Asia and the Americas in particular. Such mass borrowing of personal pronouns would have had to have taken place on a scale almost never seen on Earth. In fact, personal pronouns are borrowed only very rarely. In Altaic we have pronoun paradigms cascading down through person and number, all lined up like the Marines in perfect formation.
This is waved away with “You didn’t prove it.” In fact, the standard line in Linguistics as voiced with complete seriousness by one of the top linguists in the field is that the stunning pronoun paradigms in Altaic were all borrowings. That statement is insipid on its face. It doesn’t even qualify as theory because it’s not even possible. They might as well say, “Bats flew out my butt” as there was mass borrowing of entire pronoun paradigms.
In addition, Altaic has a huge amount of core vocabulary in common including forms that match in say Turkish and say Evenki. Apparently the Evenki and the Ottomans borrowed from each other. How? Bats flew out my butt.
Typically and for many decades now, all of these cognates in core vocabulary are said to be borrowings. There are specialists who spent most of their careers ferreting out these “borrowings” most of which are actual cognates. These men frittered away a lot of their careers on a theory that is obviously false. For the only way Altaic could not be true is if this vast amount of borrowing actually took place. The level of borrowing of core vocabulary postulated for Altaic is on a scale that is far beyond the language borrowing we have seen anywhere else on Earth. In other words, it didn’t happen. Bats flew out my butt. Once again it fails even the hypothesis stage because hypotheses are supposed to be plausible and anti-Altaic fails that those grounds alone.
Being a Historical Linguistics conservative is the hip and cool thing to be in Linguistics, and the peer pressure in the field is worse than an eighth grade playground. If you take a liberal position that says that some ancient language family like Altaic exists, the peer pressure on you as a fraud, idiot, kook, crank and loser is unbelievable. I am amazed that there are any liberals left promoting daring new ideas on ancient language families.