Soft Sciences Vs. Hard Sciences

[teo123]

And of course, you can't really apply mathematics to linguistics, but there is probably a method or two you can apply to make it harder. What those methods are, I don't know.

Well, like I've said many times, this really depends on what we mean by mathematics. Mathematics doesn't need to deal too much with quantities. In physics, the word math usually signifies arithmetic, calculus (fundamental laws of physics almost always include integrals and derivatives) and algebra. But that's not all that math studies. In fact, mathematics is often defined as "the science of formal systems". What linguistics deals with often fits that definition perfectly.
Now, how do I think we can harden the part of linguistics I was studying?
Well, like I've said, I think the study of the toponyms should, first of all, rely more on the known laws of sound change (and those could be said to be mathematical laws), and much less on our very flawed intuition about sound changes. In other words, we should not accept etymologies that contradict those laws, as many, if not most of the, etymologies that ascribe Croatian placenames to Latin do.
Second, I think it should abstain from using the unknown to explain the unknown, which is what's usually going on when some toponym is being ascribed to a Pre-Indo-European/Pelasgian language.
Third, I think that the obvious patterns in the toponyms that cannot be explained using the well-attested languages should not be simply dismissed as a coincidence, especially since the P-values of many of them are amazingly low. That is, I think there is a bias against the idea that many if not most of the Croatian toponyms come from a single unattested language that can be partly reconstructed based on those toponyms.

[Red]

Well, like I've said many times, this really depends on what we mean by mathematics. Mathematics doesn't need to deal too much with quantities. In physics, the word math usually signifies arithmetic, calculus (fundamental laws of physics almost always include integrals and derivatives) and algebra. But that's not all that math studies. In fact, mathematics is often defined as "the science of formal systems".

Right, it depends on the field. High-level math doesn't even have numbers.
But you're wrong in stating math is a science. Math is derived from Logic, and Science is derived from Math. Science is just applied mathematics.

There is no universally accepted definition of math though.

What linguistics deals with often fits that definition perfectly.

There you go again.

[teo123]

But you're wrong in stating math is a science. Math is derived from Logic, and Science is derived from Math. Science is just applied mathematics.

Philosophers of math and logic don't actually agree on whether they are to be called sciences. In our philosophy classes, we were taught that there is a spectrum between logical sciences and empirical sciences. There are also some serious arguments for mathematics being partly empirical, for instance, the world in which the Euclid's Fifth Postulate (and therefore the things such as the Pythagorean theorem) is conceivable, yet we take the Euclid's Fifth Postulate and what follows from it as true, perhaps because of what we know from experience. Some would argue that even logic is empirical, but I don't think that idea should be taken seriously.

There you go again.

Linguistics does often deal with formal languages, what's wrong with saying that? This is obviously true of morphosyntax (human languages are usually taken to be context-free grammars), but it's also true of phonology and historical linguistics, since sound changes can be modeled as a type of a formal language called regular expression. Comparative linguistics also often deals with graph theory. Yes, as far as I know, there is no calculus in linguistics, and I explained what I think why, and why I'd be skeptical of somebody using advanced calculus outside of physics.

[teo123]

OK, sorry, @Red, now I see the first part of my last post would be very unclear to somebody who is not aware of the Non-Euclidean geometry, and I don't know if I can expect you to be. It was late at night when I was writing that.

@brimstoneSalad, will you respond to my post or not? My questions about what you think the economics says about the '08 depression, what makes you think economists agree on that, and what you think the Keynesian economics actually says are honest questions. I am happy to hear the other side of the story (for instance, yours) if there is one.
If not, I think that's understandable, but I'm not sure that's excusable. You think like a meat-eater: you believe potentially-harmful stuff (like that we need a government) and are convinced the evidence for that comes from a wide range of sciences, but you have an, at best, superficial understanding of those sciences. You want to believe that the government can be based on science. In fact, you believe it can be based on a science even more reliable than, for example, linguistics. In fact, you seem to believe the world governments are even now partly based on sciences, that a significant part of those thousands of pages of regulation is indeed based on extensive empirical and theoretical research. And, you are also convinced you know enough about linguistics to say how linguists should do their jobs, and when I point out what you are saying doesn't make a lot of sense, and that it's been, thanks to people like you, tried without success, you resort to mockery and insults and then stop responding. Time well spent, right?
Oh, and since you've brought up that study that supposedly shows the scientific consensus disappears once a science gets "softer" in the other thread, do you bother pointing me where that lack of scientific consensus resides in linguistics? Like I've said there, I am pretty sure nobody of those who study Croatian toponyms doubts the basics of historical linguistics, such as the existence and the basic properties of the Proto-Indo-European language, and I am also quite sure nobody of them doubts even the finer details, such as the Laryngeal Theory and the Havlik's Law.

[Red]

OK, sorry, @Red, now I see the first part of my last post would be very unclear to somebody who is not aware of the Non-Euclidean geometry, and I don't know if I can expect you to be. It was late at night when I was writing that.

Please don't patronize me, I know what Non-Euclidean geometry is.
Look, if you really think you figured out a way to harden linguistics, why not call the Nobel Prize Committee? They'd be thrilled by a discovery such as this.

[teo123]

Please don't patronize me, I know what Non-Euclidean geometry is.

I assumed you know, but I can never be certain with that stuff. I also took for granted you were aware of the arguments used against animal testing of medications, it turned out you weren't.

Look, if you really think you figured out a way to harden linguistics, why not call the Nobel Prize Committee?

I suspect that to be a trolling question, but I am going to respond anyway: that's not even remotely how the Nobel Prize works. Computer scientists don't get a Nobel Prize no matter what they achieve, because there is no Nobel Prize in Computer Science. But one of the most prestigious award in Computer Science is called Turing Award.
Second, it's not at all commonly accepted, even by those who believe in this soft-science-hard-science-spectrum/distinction, that linguistics is a soft science. In fact, it's usually accepted to be, just like psychology, a science where social sciences and natural sciences meet. Phonetics certainly has much more in common with natural sciences than with social sciences.
What prizes do linguists get for their work, I don't know right now and I honestly don't care about that. The motivation for doing science needs to be intrinsic, and not extrinsic (which is also why government funding of science doesn't work).

[brimstoneSalad]

@brimstoneSalad, will you respond to my post or not?

I already responded to you. You need to quote me if you want me to see your responses.

What do you mean by "epistemological"? It's more of a poorly defined claim about the empirical world.

The facts about science in practice may be empirical, but it's still based on fundamentally non-empirical principles.

Is it possible for them to reach by reason alone that it exists and that it's a reliable science? I don't think so.

Based on biology it is. You know things eat and derive nutrition from those things. And that, based on chemistry, knowing how chemicals react, based on physics, and so on.
You would be able to derive a conceptual framework.

Now, whether the field was politicized and full of corruption or dogma you would not be able to deduce by reason alone.
A science can depart from its ideal in practice, but that doesn't say anything about the underlying potential.

We should be able to reason that chemistry is probably a softer science than physics merely by their relation to each other.
It could be in some world that physics is infested by quacks and pseudoscientists and that chemistry is not, but that's a different issue and does not speak to the underlying principles involved.

How could you determine how reliable that is by reason alone (as the word "epistemological" implies)?

Again, based on relationship to other sciences and dependencies on various variables those deal with.

All we can do is to say from the experience that it appears to be reliable, people who do that strive to make their theories coherent (both internally and with other well-accepted theories in linguistics) and sometimes they are confirmed later by other sciences (as in the case of the Antun Mayer's interpretation of the name "Asseria").

The latter case is concordance and that's what you want. That actually confirms something and hardens a science.
A mere framework that doesn't contradict itself isn't science, it's ad hoc hypothesizing. Again, the same deal with Flat-Earth models.

You *could* make a model that explains everything about the flat Earth, it would just take a lot of work and involve some very radical divergences from established theory. If you worked hard enough, you could even force compatibility with other things. That doesn't mean anything.

Softer sciences don't *have to* be filled with ad hoc hypotheses, it's a correlation, but humans being how they are they want to pretend to know things they don't and can't plausibly confirm so there we go.

OK, the study you linked in the other thread talks about the supposed lack of a scientific consensus in social sciences.

That's a correlation that comes from the low hardness.

You could absolutely get nearly 100% consensus in a soft science, though. Just need enough group thinking. Look at the level of consensus in theology that a god exists, for example. There's also the bias that an atheist is unlikely to go to school to study theology.

This has actually been a problem in psychology for a while, with a lot of poor studies that have just been accepted for a long time *because* it's a soft science so it's hard to disprove them, but also because there's so much to "discover" that there isn't as much incentive to challenge existing ideas.

A high level of consensus, though, suggests we should look at something more deeply because there *may* be something there.

Where is that in linguistics? I don't quite see it.

It's not for you to subjectively analyze based on your feeling on the matter and claim, out of your ass, that there's high level of consensus.
If you *really* want to claim there's a higher or equal level of consensus in linguistics, then use the same methods that study used and apply them to linguistics (not just a niche within linguistics studying Croatian).

Remember when I told you to do your homework in the Flat-Earth thread and show the math for your model?
Well, do the same here. You can read that study and its methods, and apply it to the published papers in linguistics.
Show some proof that linguistics has a higher level of consensus than clinical psychology, for example. Your general feeling on the issue is not evidence.

I have no doubt that there are a few matters there's consensus on, like the historical existence of root languages, which could be predicted by biological genetics.
Obviously people dispersed from common ancestors, and we can trace lineage through genetics pretty reliably. We also have well established that languages diverge by drifting and mixing (based on good written evidence) and there's no reason to believe they're being completely replaced by constructed languages; although sometimes they might be replaced by conquest, a pidgin is more likely. There would also be genetic evidence of a foreign conquest.
Matters like root languages can be harder science because we can use a harder science to predict them (like biology).

you believe potentially-harmful stuff (like that we need a government)

STOP straw manning me. If you continue to do so you will get banned. We have been over this multiple times, and now you're basically spam-mentioning me with straw-mans of my own arguments.

You're the one who claimed murder SHOULD be legal. You're claiming we DON'T need government, and that the world would be better without governments. This is an unknowable claim because anarchy isn't a thing we can assess.

Your belief is loaded with with the claim that "anarchy" even could exist in any stable and sustained way without degrading into what amounts to decriminalized organized crime/mafia and tribal governments which amass power until they become feudal states, which is the process we have historically seen from human beings.

You're the one making the claim, and if you persist in bringing me into threads in order to straw man my arguments to me we're going to have a problem.

Stick with the Soft science vs. Hard science and stop bringing up your anarchism bullshit, OR start a thread addressing your central claims about anarchism by presenting an actual model of how you think anarchism is supposed to work (Again, like I called you out on presenting your Flat-Earth model). It's all well and good to complain about government, but as much as government sucks you need to present an alternative, because history has shown us that the alternative to one government is a power vacuum which just forms another government, and often a shittier one.

START A NEW THREAD. Then make sure you actually put up and outline your claims and how you think it's supposed to work.

[Red]

I assumed you know, but I can never be certain with that stuff. I also took for granted you were aware of the arguments used against animal testing of medications, it turned out you weren't.

And arrogant Teo is back again!

Are you referring to Bite Sized Vegan's video? I am well aware that what she was saying was nothing but myth, and you might realize the same if you put aside your libertarian biases.

I suspect that to be a trolling question, but I am going to respond anyway: that's not even remotely how the Nobel Prize works.

That was the joke, since what you put forth in order to apply mathematics to linguistics is completely asinine. Surely you don't think you were the first one to try such a thing?

Again, it's our desire to think we've discovered something new, so it makes us feel special. It's one of the same reasons why you thought the Earth was flat. You're imagining and finding PATTERNS (something conspiracy theories always look for, and not actual evidence) for things that simply aren't there.

[teo123]

I already responded to you.

That your post was written hours after I said "Why aren't you responding?". After you haven't had responded to my post for five days, while still being active on the forum, what it is that I should take from that? Maybe that I finally started asking the right questions, so you, just like JROA on the TFES forum, run away?

You *could* make a model that explains everything about the flat Earth,

Not in any meaningful way. At best, you can make something like the Euler's concave Earth model, which he proposed as a thought experiment to show how the same observations can be described by different formulas with different natural interpretations. Then those formulas would just be more complicated, but they would make the same predictions as a Round-Earth model.

Just need enough group thinking.

Yeah, because nobody would publish a scientific paper proving the Havlik's Law wrong if it were wrong... Wait, it actually did happen. And then another law was discovered to explain those apparent exceptions from the Havlik's law (that, at some stage in the Chakavian dialect of Croatian, the yers in the first syllable were always vocalized). And, about a century before that, almost the same thing happened for Germanic languages (see the Grimm's Law and the Verner's Law).
I doubt that you are being serious now.

Look at the level of consensus in theology that a god exists, for example.

A good proxy for that would be how many theologians study eastern religions, about the same number of them won't accept that a god exists. So, it's not a near-anonymity here. Plus, it's really just an illusion of consensus created by a poorly-defined question. If you just replace "god" with its definition, "an omnipotent, omniscient and a benevolent being", the level of consensus is going to drop drastically.

Show some proof that linguistics has a higher level of consensus than clinical psychology, for example.

I don't know about the level of consensus in clinical psychology, and that's not really what we were discussing. You claimed that there is higher degree of consensus in economics than in linguistics. So, how many economists agree, for example, that minimum wage is harmful or that inflation is caused primarily by an increase in money supply?
https://en.wikipedia.org/wiki/Economics#Agreements
Yeah, you don't need a double-blind study to understand that the level of consensus in linguistics is greater than that.

like the historical existence of root languages, which could be predicted by biological genetics.

I think you know what my answer is going to be here. Genetic arguments play no role in historical linguistics.
A textbook example, the Finnish people have genetically much more in common with the Scandinavian Germanic people than with the Samoyedic peoples in Siberia, yet their language is not demonstrably related to the Scandinavian Germanic languages, but it is demonstrably related to the languages of the Samoyedic peoples in Siberia.

You're the one who claimed murder SHOULD be legal.

Because the claim that murder shouldn't be legal is an incredibly loaded assertion. Let's see how many unproven (and, if you ask me, ridiculous) assumptions there are behind it:
a) Police does more good than harm.
b) Courts do more good than harm.
c) Prisons do more good than harm. That is, putting a murderer to a place from which he will return with even more psychological problems, which made him murder in the first place, somehow protects the society.
...

If you continue to do so you will get banned.

That's right, seek to create an echo-chamber for the Auguste Comte's philosophy in today's world. And repel people interested in social sciences from animal rights advocacy. How can that be a good thing?

Your belief is loaded with with the claim that "anarchy" even could exist in any stable and sustained way without degrading into what amounts to decriminalized organized crime/mafia and tribal governments which amass power until they become feudal states, which is the process we have historically seen from human beings.

Yes, unfortunately, a political system can't be sustained if there is a significant number of people who want to revolt against it. But that's not a problem exclusive to libertarianism at all. If a great number of people doesn't believe in democracy, as was the case in Venezuela, democracy can't be sustained.

Stick with the Soft science vs. Hard science and stop bringing up your anarchism bullshit

My friend, this is somewhat connected. If there can indeed be a government based on science, then we should strive to establish it. If our current government is indeed partly based on science, then perhaps we should not seek to overthrow it. If there can't be a government based on science, then all any government can do is forcing someone's subjective opinion onto people, and that's authoritarian and wrong.

That was the joke, since what you put forth in order to apply mathematics to linguistics is completely asinine.

What? Look, linguistics often deals with mathematics of formal languages. And you would probably know that if you looked into any research paper in morphosyntax. Here is but one example of that:
http://www.eecs.harvard.edu/~shieber/Bi ... eber85.pdf
Or should I assume you know about that, but are being dishonest, and not feeding the troll by not replying any more?

[brimstoneSalad]

That your post was written hours after I said "Why aren't you responding?".

I didn't read this post until after responding to the other.
I tend to respond to posts before reading newer ones.

If you continue down this conspiratorial and accusatory path you're going to find that I have less and less patience for you.
You need to apologize if you want me to continue to respond to you. I'm sure @Red is perfectly capable of handling your arguments, but I'm not really in the mood for your bullshit.

At best, you can make something like the Euler's concave Earth model, which he proposed as a thought experiment to show how the same observations can be described by different formulas with different natural interpretations. Then those formulas would just be more complicated, but they would make the same predictions as a Round-Earth model.

You don't need a concave Earth if you make the claims precise enough or formulas complicated enough. Fancy ridiculous prism-kinda sky dome could refract light from sun and stars in pretty much whatever way you want, and then you have the overt hologram claims.

Yeah, because nobody would publish a scientific paper proving the Havlik's Law wrong if it were wrong... Wait, it actually did happen.

Then that shows disagreement in the field (lack of consensus) apparently.
Are you, or are you not, going to do a statistical analysis of published papers to prove your claims?

Yes or no.

If not, then you're just talking out of your ass.

And then another law was discovered to explain those apparent exceptions from the Havlik's law

OR... another ad hoc exception was fabricated.

I doubt that you are being serious now.

I'm perfectly serious. Obviously linguistics is a softer science.
You just don't seem to understand what that means in terms of what we should believe or how we should regard it.

Does that mean I default to disbelieving experts in linguistics? NO. Stop being an idiot.
I assume that every claim that has gained consensus in linguistics is correct. I do that on the basis of expert opinion, and because I have no compelling reason to disbelieve it. I just do not hold the kind of certainty as I would hold for chemistry or physics.

You seem to have so much trouble with this concept. Accepting as the null hypothesis that experts in a field are correct about what they say if there's general agreement UNTIL there's compelling evidence to reject it in favor of an alternative.

You have this terrible habit of throwing out mainstream consensus when you decide it's not adequately supported and then falling into whatever random alternative you prefer.
That's not how sane epistemology works.

You should tentatively stick with the consensus NOT just until you decide there's not a lot of good evidence for it, but until you have an actual alternative with BETTER evidence.

A good example of this done wrong is epistemologically poor atheists. They decide there's not enough evidence they can understand for a god and so reject it in favor of nothing without really understanding why theology is wrong (e.g. the supernatural claims for which there IS compelling evidence against, and the logically impossible claims baked into many god definitions).

Plus, it's really just an illusion of consensus created by a poorly-defined question. If you just replace "god" with its definition, "an omnipotent, omniscient and a benevolent being", the level of consensus is going to drop drastically.

That's true, but consensus doesn't have to be that precise. The same is true in any science when you push for a higher level of precision.

Look at how questions on climate change are worded, and how that changes consensus.
Say it's happening, say it's happening at X rate or higher, say it's primarily caused by humans (vs. partially), say it will change weather patterns in a way to negatively affect humanity in X ways.
Consensus is still pretty good, but more claims weaken it.

Like Occam's razor, the more you add the weaker consensus becomes.

How do we measure the exchange value between "God is Omnipotent" and "Climate change is mainly caused by humans" and how those affect the reduction in consensus?
Maybe the former is a much *larger* and more *precise* claim that should be expected to affect consensus more than the latter?

The point is, you have to get a lot deeper into the subject matter, and you need reasons to reject consensus. And importantly, the harder a science the more reason you need. You don't need much to reject theology; any incompatibility or disagreement with physics should be enough.

You claimed that there is higher degree of consensus in economics than in linguistics. So, how many economists agree, for example, that minimum wage is harmful or that inflation is caused primarily by an increase in money supply?

See above. What's the exchange rate between those claims and how they should be expected to affect consensus.
Economics is also more politically charged than linguistics, which would be expected to motivate higher levels of disagreement.

There are a lot of variables at play, so degree of consensus alone is not the best metric (it's just the easiest empirical one).
What you need to do is a statistical analysis of publications if you want to argue it, though.

I think you know what my answer is going to be here. Genetic arguments play no role in historical linguistics.

Genetics can tell you where people came from and where they migrated, basically trace their location over time.
Not just of *living* people, but also of remains. It leaves a trail which is very predictive, and can give you concordance with statistical analyses of words.

A textbook example, the Finnish people have genetically much more in common with the Scandinavian Germanic people than with the Samoyedic peoples in Siberia, yet their language is not demonstrably related to the Scandinavian Germanic languages, but it is demonstrably related to the languages of the Samoyedic peoples in Siberia.

And why is that?
Why do you think DNA analysis couldn't have predicted that?

You're the one who claimed murder SHOULD be legal.

Because the claim that murder shouldn't be legal is an incredibly loaded assertion.

Why are you so insistent on being this stupid?

Just because you're skeptical of a POSITIVE claim, doesn't mean you assert the OPPOSITE claim.
Why can't you just be agnostic to varying degrees?

Let's see how many unproven (and, if you ask me, ridiculous) assumptions there are behind it:
a) Police does more good than harm.
b) Courts do more good than harm.
c) Prisons do more good than harm. That is, putting a murderer to a place from which he will return with even more psychological problems, which made him murder in the first place, somehow protects the society.

That assumes a world without those things in some form is magically possible. There's no reason to believe it is.
What we're really comparing is a shitty police force vs. even more shitty mafia style enforcers or feudal knights.

I told you to make a thread on this. Don't reply to this stuff again in this thread. It's off topic, this thread is about hard vs soft science.
@Red if you see that teo has brought up this anarchism stuff again here, can you split the post off into a new thread?
I'll probably be scarce for a few days. I also might not be replying to him anymore, because I'm getting annoyed by his taunting (unsure if he'll apologize for that or not).

Yes, unfortunately, a political system can't be sustained if there is a significant number of people who want to revolt against it.

The number of people is proportional to the strength of the government they're revolting against.
How powerful is an oxymoronic anarchy government? Answer: not at all. You need one person so declare his or herself in charge and one other person to follow that person, and with that you have the most powerful force on an otherwise anarchist world. Those two people can overpower pretty much any other individual (a few exceptions might require two followers).

This is not much of an empirical question. The inevitable failure of a true anarchy is more like Earnshaw's theorem: https://en.wikipedia.org/wiki/Earnshaw%27s_theorem

But that's not a problem exclusive to libertarianism at all.

Not exclusive, just massively magnified.
An active government with powers acts like a stabilizing force. Look into active magnetic stabilization; it doesn't take much force to stabilize most systems. Can it be overcome by a large push? Yes, but you get more time in a stable configuration which is all government needs to do. Work *most* of the time to hopefully prevent feudalism which is usually worse.

If you want to argue that, start a new thread.

If there can indeed be a government based on science, then we should strive to establish it. If our current government is indeed partly based on science, then perhaps we should not seek to overthrow it. If there can't be a government based on science, then all any government can do is forcing someone's subjective opinion onto people, and that's authoritarian and wrong.

It's not an either or. We need to push government to use more science and less dogma.
Again, we're not comparing government to some fantasy of utopia, we're looking at what actually happens without any government... which is that a new one pops up that's usually (but not always) worse.
In cases where we can reasonably expect a new government to be better then overthrowing makes sense... but it's usually not, as we've seen with U.S. interventions. It almost always makes things worse, because politicians are living in a dream land of idealized government when reality isn't so simple.

Again, new thread if you want to discuss that. It's only barely related.