Is physics the easiest science to develop?

[Tosh5457]

Hello.

Social sciences, like economics, are much harder to develop than physics. Mathematical models in economics and finance fail miserably, nobody can make mathematical models that work.
Did physicists get the easy part in science? Nature has deterministic and very simple laws. Physicists can even use mathematics to develop theory, because the laws are so simple.

I viewed physics (as most people do) as a very difficult science to learn. But we just study and apply the very simple and strict nature laws, how can that even compare to the difficulty of social sciences, that have to explain very complex systems and human behaviour?

Edit: Oops I wanted to post in the physics section. Sorry about that...

 

[Office_Shredder]

Physicists can even use mathematics to develop theory, because the laws are so simple.

Are you referring to Newtonian mechanics, or research level physics here? I would argue physicists can use mathematics to develop theory, even though the laws are quite complex

 

[MissSilvy]

Social sciences being more difficult and complex than physics? Now this is an argument I have thankfully never heard before.

Edit: Perhaps there is merit to considering this question. Humans are uninteresting and messy beasts who individually defy modeling and prediction. The reason math applies so poorly to social 'sciences' is because they are poor sciences, not because they are difficult. There are not as clearly defined causal, mechanistic, and mathematic relations among people as there are among particles or stars and such. Also, have you ever tried to derive solid state physics equations from just Newton's laws and Maxwell's equations? I assure you, the laws are beautifully simple but enormously complex in their applications and interpretations.

 

[Pengwuino]

Well I suppose it's not too crazy of an argument if you define what you mean by "difficulty". Now, I'm HIGHLY skeptical of the idea that most economic models are overwhelming failures (please justify this). However, physics, to me, is very transparent. For the most part, you understand the laws and when you can figure out a good theory, it's always going to work in it's domain of validity.

Economics and finance, on the contrary, seem to be so chaotic (and I don't mean that in the technical sense) that it does seem "more difficult" than physics. Since humans aren't exactly known for being predictable to the extent an electron is predictable, I can see how economics can be ridiculously hard if you try to make a model complex enough.

 

[Tosh5457]

The reason math applies so poorly to social 'sciences' is because they are poor sciences, not because they are difficult. There are not as clearly defined causal, mechanistic, and mathematic relations among people as there are among particles or stars and such.

Why does that make social sciences poor sciences? It's not possible to define strict relations between causes and effects in social sciences, because it deals with human behaviour. It doesn't make these sciences poor, that's what makes them difficult.

Now, I'm HIGHLY skeptical of the idea that most economic models are overwhelming failures (please justify this).

The mainstream models are neoclassical models, which failed and have been falsified.
No model (except marxian models) predicts that capitalism fails in the long term. US' total debt is overwhelming, the productive side of the economy is very small and is shrinking to the financial (and non-productive) sector. Japan's economy already stagnated 20 years ago, and US and Europe economies are going for the same situation. Which model predicted this? Marxian model that is, but it's also a bad model in other aspects.
Then of course, which model predicted the meltdown of US financial system? This was the most important economic event in 80 years and no model predicted it? Almost everybody thought US' economy was going just fine right before this major problem.

 

[maverick_starstrider]

Why does that make social sciences poor sciences? It's not possible to define strict relations between causes and effects in social sciences, because it deals with human behaviour. It doesn't make these sciences poor, that's what makes them difficult.



The mainstream models are neoclassical models, which failed and have been falsified.
No model (except marxian models) predicts that capitalism fails in the long term. US' total debt is overwhelming, the productive side of the economy is very small and is shrinking to the financial (and non-productive) sector. Japan's economy already stagnated 20 years ago, and US and Europe economies are going for the same situation. Which model predicted this? Marxian model that is, but it's also a bad model in other aspects.
Then of course, which model predicted the meltdown of US financial system? This was the most important economic event in 80 years and no model predicted it? Almost everybody thought US' economy was going just fine right before this major problem.

Saying social "sciences" are more "difficult" makes a number of incorrect assumptions: firstly, that these things are being approached with the same force of mathematical rigour as other science (there exist mathematicians who try and do such but the average "quantitative" economist (a specialist in econometrics) is operating at a mathematical level roughly equivalent with grade 12 of high school). Secondly, saying they've got the hard job implies that success is a possibility. It's the old joke: "economists have correctly predicted 9 of the last 5 recessions". The vast majority of "economic theory" is really just gut intuition by people who have about as successful a track record as your average card player and for every idea you can always find a tenured professor who agrees with you. In addition, if you read any economic theory you will immediately notice two things: The level of operation is extremely low-level (Conclusions drawn solely from assumptions like "people will always try to maximize their profit") and based on underlying axioms that are quite obviously wrong (such as "with every action human beings will try to, quantitatively, maximize their gain"). Something like economics tries to hide such shortfalls by preaching an illusion of quantitativeness, most obviously manifest through excessive use of graphs that could never, EVER actually have any meaning. Could we ever, EVER know the quantitative form of a SUPPLY curve? Hell no! We'd need an infinitum of alternate realities where each one has a different price level on soap or whatever and all other variables are magically held constant (ceterus parabus). The LRAS curve is vertical? How many economies have you seen taken to infinity? It's really a classic case of "the emperors new clothes", you give a fancy name to shading in some fraction of a graph or calculating it slope (despite that fact that it could never be known) and playing that it provides some actual insight.

In addition, all mathematical attempts are doomed to failure because even the most naive, first flush attempts result in chaotic equations (in the literal, non-layman sense, i.e. they're non-linear and extraordinarily sensitive to initial boundary conditions). Plus, those who implement economic policy ALL base their decision making on balls-out, swing for the fences, incorrect assumptions. Neoclassical (INFINITE forecast error? please), Monetarists (Oh yeah, the national bank TOTALLY has a handle on the economy), Keynesians (Stagflation anyone?), etc.

The real question isn't whether social sciences are the harder sciences, but rather whether they're sciences at all. If they could come up with a single, just one, correct model then maybe there'd be something to talk about. But as of yet? A big no.

... That and every econ major I've ever met was simply clueless (that's my euphemism), but that's probably just my own experience.

 

[maverick_starstrider]

Actually, now that I've taken a moment to think about it I think I can peg down a much bigger schism between science and social "science". No social science can explain PAST data, yet that never stops social "scientists" from weighing in on FUTURE results. There is not a single theory, something like "If X,Y and Z occur in manner A then B will result", which can be consistently reconciled with past data ("except in this case, that case and that case (and all these other cases)). In science, such a stance would be inforgivable. If your theory of *blank* doesn't correctly predict the results of every previous experiment that has ever tested it than you'd never think about making a PREDICTION about the results of a future experiment. It'd be nonsense. Yet this is what psychologists, sociologists, and economists do every day.

However, perhaps more important than this is the fact that there exists no outlet for experiment in social "science". One can never make a society, or raise a human being, or a culture in some identical controlled manner so that one can conclusively record the effect of some external force on them. The old adage "we learn history so that we never repeat it" is, in some ways, a crock. Is communism inherently invalid as a political philosophy? Well every society who has tried it has ended up the worse for wear... and yet every society that's tried it was coming off some violent revolution or civil war or an oppressive autocracy, would democracy have fared better as a model for "picking up the pieces"? How could we ever know. We can't repeat events and, to borrow a phrase form science, the "initial conditions" are never the same. In science, making assumption off a handful of data, none of which were obtained with the same initial conditions, would be garbage for establishing a trend. And yet that's a tuesday for social "sciences".

 

[ZapperZ]

Hello.

Social sciences, like economics, are much harder to develop than physics. Mathematical models in economics and finance fail miserably, nobody can make mathematical models that work.
Did physicists get the easy part in science? Nature has deterministic and very simple laws. Physicists can even use mathematics to develop theory, because the laws are so simple.

I viewed physics (as most people do) as a very difficult science to learn. But we just study and apply the very simple and strict nature laws, how can that even compare to the difficulty of social sciences, that have to explain very complex systems and human behaviour?

Edit: Oops I wanted to post in the physics section. Sorry about that...

What you described here tends to indicated that physics is easier to falsify! Because it must be well-defined, both in terms of its theoretical formalism and experimental verification, it makes it clearer in the sense of when something is valid, and when something can be tossed out.

Whether that makes it easier to 'develop' is debatable.

BTW, this IS posted in the proper forum. It doesn't belong in the physics forums because topics posted in those forums must have actual physics content, i.e. they must be posts on physics, not ABOUT physics.

Zz.

 

[Vagn]

What you described here tends to indicated that physics is easier to falsify! Because it must be well-defined, both in terms of its theoretical formalism and experimental verification, it makes it clearer in the sense of when something is valid, and when something can be tossed out.

Whether that makes it easier to 'develop' is debatable.

BTW, this IS posted in the proper forum. It doesn't belong in the physics forums because topics posted in those forums must have actual physics content, i.e. they must be posts on physics, not ABOUT physics.

Zz.

Physics also doesn't tend to have the ethical issues that the social sciences have to deal with, experimenting on the psychological effect of pain would presumably be a lot easier if you could inflict pain upon the experimentee! Obviously doing so is unethical though, which I'd guess limits the ability to falsify hypotheses.

 

[arildno]

The main problem is that the vast majority of social "scientists" are NOT content with investigating those areas of human activity in which they COULD have defined clear criteria of falsification, for example.

Rather, such areas are deemed as "unimportant", and that proper "research" should be about large-scale social phenomena, about which the social scientists love to fantasize and theorize.

They are a plague, and are actively preventing proper, solid social research from ever developing.

Besides, their insane, inane fantasies have deleterious political effects when sought to be implemented as policy.

 

[ZapperZ]

Physics also doesn't tend to have the ethical issues that the social sciences have to deal with, experimenting on the psychological effect of pain would presumably be a lot easier if you could inflict pain upon the experimentee! Obviously doing so is unethical though, which I'd guess limits the ability to falsify hypotheses.

I disagree. I don't think it is a matter of involving "ethical issues" or not. There are ethical issues in biology, yet, it is still more definite and more well-defined than social science, economics, etc.

Again, the biggest difference between the two is the falsifiability. That character, in physics, allows for a very high degree of certainty of accepted physics ideas and theories.

Zz.

 

[Pythagorean]

We must pluck out their forebrains to prevent them from feeling the scalpel.

 

[rootX]

Question is who here is qualified to judge
1) social scientists?
2) physicist?I don't see it is easier or even meaningful to compare different disciplines.

 

[maverick_starstrider]

Question is who here is qualified to judge
1) social scientists?
2) physicist?


I don't see it is easier or even meaningful to compare different disciplines.

Well there's always the small matter of success. Can field X accurately predict phenomena within their purview? For social "science" this would be a resounding no.

 

[rootX]

Well there's always the small matter of success. Can field X accurately predict phenomena within their purview? For social "science" this would be a resounding no.

I will only be willing to hear about what is social science from someone who is an expert in this field, and that's what I was referring to in my above post.

I believe you are trying to describe social science in physics language which to me does not make any sense.

"If X,Y and Z occur in manner A then B will result", which can be consistently reconciled with past data ("except in this case, that case and that case (and all these other cases)). In science, such a stance would be inforgivable. If your theory of *blank* doesn't correctly predict the results of every previous experiment that has ever tested it than you'd never think about making a PREDICTION about the results of a future experiment. It'd be nonsense. Yet this is what psychologists, sociologists, and economists do every day.

This is not correct IMO. If social scientist knew all the input parameters (assume here X,Y, and Z are only input parameters) and those parameters occurred only in the manner (A) then B will always result.

Problems arise because you never know if X,Y, and Z are only the input parameters. Second that you can never do controlled experiments.

Please note that I am not an expert in social science so not qualified to talk how it works.

 

[Pythagorean]

from my physics-first interdisciplinary view, social problems are just very complex physics problems.

 

[atyy]

from my physics-first interdisciplinary view, social problems are just very complex physics problems.

How many sciences are there?

3 - physics, statistics & mathematics or 2 - physics and mathematics?

 

[Office_Shredder]

from my physics-first interdisciplinary view, social problems are just very complex physics problems.

Explain how physics is going to actually inform us on social problems. Interdisciplinary actually requires those disciplines to contribute usefully.

 

[maverick_starstrider]

I will only be willing to hear about what is social science from someone who is an expert in this field, and that's what I was referring to in my above post.

I believe you are trying to describe social science in physics language which to me does not make any sense.


This is not correct IMO. If social scientist knew all the input parameters (assume here X,Y, and Z are only input parameters) and those parameters occurred only in the manner (A) then B will always result.

Problems arise because you never know if X,Y, and Z are only the input parameters. Second that you can never do controlled experiments.

Please note that I am not an expert in social science so not qualified to talk how it works.

Well social science is hardly a "deep" field. A psychology or sociology paper in an academic journal is entirely readable and understandable by someone outside the field. The most you'll ever need is wikipedia for buzz words. Something like psychology or sociology isn't like physics where you have a first-year of university and then a second, which can only be taken given that one adequately understands the first year and then you move to a third year which heavily draws on the second (and first) and so on. In something like psych you take a first year course (psych 101) and then that's often the only prereq (other than "being enrolled in year X") for all the other courses, they don't draw on each other. One student might take courses on personality theory and testing, another is more interested in developmental psychology. They don't overlap and their only prereq is likely just psych 101 (unless they're structured something like: developments psychology I, developmental psychology II, etc.)

My mom got her degree in psychology, I can pull any of her old textbooks off the shelf and completely comprehend it instantly, no problem. I mean they'd need to be, let's face it, that dumb drunken frat guy or party girl we knew in uni? They were ALWAYS in sociology or psychology (or history).

As for economics it has a much deeper/richer formalism and arguably requires a great deal more knowledge to understand. I say arguably because I actually KNOW a great deal of economics and most of it is simply junk. It's a lot of vague notions (like "for some goods the demand goes UP when the price goes UP (a Veblen good)); that are made to SEEM concrete and fact by tying them to behaviors of fictional, unknowable curves on fictional graphs and the likes. Something like Econometrics, which I think most econ majors would call the "pinnacle" of difficulty in the field, really requires around grade 12 math (maybe a smidge more stats). This stuff is EASILY comprehensible to joe-blow mr. physicist once the mountains of jargon (there really is quite a lot of jargon) is stripped. I mean rudimentary linear regression is about as far as I've seen an economist go. (mathematicians WORKING in economics obviously can come up with much more elaborate theories, but average joe-blow mr. economist can't understand these models anyways).

 

[maverick_starstrider]

Explain how physics is going to actually inform us on social problems. Interdisciplinary actually requires those disciplines to contribute usefully.

Every once in a while some physicist gets bored and takes a stab at social science. For example, I remember seeing a talk by a guy who was considering democratic voting and how party preference tends to be geographically localized and peoples politics tend to be influenced by those close to them, etc. He basically mapped it to a system of interacting particles on a lattice. Retardedly simplistic model and yet as accurate as anything that comes out of sociology. Surprisingly, he was actually able to recreate many phenomena of voting with his model (such as how a third party effects the political landscape and such) and was able to produce graphs of voting that were qualitatively very similar to actual historical data from Britain (now OF COURSE, he likely CHOSE britain since its voting history most matched his model, but still).

Although I really question how much traction these approaches can ever really have (in the sense that I question what progress could ever be made in the social sciences period.) As a physicists we know that when we have any model that is known a priori to be stripped of all complexity and complicated interactions then we should take all its predictions with a beach worth of salt. Social scientists tend to ignore this, stick their fingers in their ears and just assume that because a model predicts one thing, kinda right, it must predict everything right. It's like the disclaimer in historical movies: Any resemblance of a theoretical result in social science to events historical... is purely coincidental.

 

[Tosh5457]

I think many people here suffer from http://en.wikipedia.org/wiki/Scientism" .

The rigor and the falsifiability of physics exists because it studies the fundamental laws, where no ambiguity is possible. Physics is a "good" science not only because of the physics theorists who developed it, but mostly because it deals with fundamental laws. Other sciences don't have that "luck".
When you study more complex systems, like biology, it's not so easy to make falsifiable theories and claims. And when studying even more complex systems, like psychology and economics do, it gets worse. I don't think it's possible to use the same methods physics or chemistry uses to these areas, and it's not a question of whether social sciences researchers are incompetent or not.

The real question isn't whether social sciences are the harder sciences, but rather whether they're sciences at all. If they could come up with a single, just one, correct model then maybe there'd be something to talk about. But as of yet? A big no.

I agree, social sciences don't use the same methods of sciences such as physics, so they're not sciences if we use the same definition of science we use for physics. That's why they're called social sciences, but the 'science' part in 'social science' doesn't have anything to do with the definition of 'science'.
Now, we can't expect to understand psychology or economics on the same level as we understand physics. So it isn't harder to understand social sciences than it is to understand physics, they're very different areas to compare.

 

[aegrisomnia]

Social sciences are definitely harder than hard/natural sciences. Come on.

I reject out of hand the arguments that social sciences aren't sciences because "people can't be modeled mathematically". If you can prove that, by all means, it would be the most important psychological, sociological, and probably religious discovery of all time. Seriously. This is tantamount to proving free will exists, solving the mind-body problem, etc.

I also dismiss the argument that social sciences are inferior because their practitioners are less mathematically prepared than their hard/natural science counterparts. This may simply be a chicken-and-egg scenario: practitioners don't waste time learning deep mathematics because, as of yet, the mathematical basis for much of the social sciences is on a much lower level than that of the hard/natural sciences. While it may be a vicious cycle, I stand by my previous point: let a physicist or mathematician get the ball rolling! If you gave sociologists the equivalent of Maxwell's relations, I bet they'd start teaching sociology majors Calculus I-III and Differential Equations.

In a lot of ways, understanding human beings and human society is one of the hardest - and most important - things we could possibly hope to do.

 

[Ryan_m_b]

I've always wondered why "social sciences" were called sciences. I can see something like psychology and at a stretch sociology. I can even see trying to apply those fields to other fields such as economics. What I don't get is why a field that doesn't follow the scientific method is called a science.

If by the definition of "hard" we mean currently most difficult to produce good data from then yes social sciences are "harder", this is thanks to a lack of comprehensive understanding of the fundamentals meaning that A) far from enough data can be gathered and B) no good models can be built.

However this does not mean that either side has any superiority, what it means is that physics is a more mature field with a much better understanding. In biology there are many things that we haven't even begun to scratch the surface of but I wouldn't dream of comparing my work as "harder" than a physicist simply because I have less knowledge of the processes.

 

[tenchotomic]

Hello.

Social sciences, like economics, are much harder to develop than physics. Mathematical models in economics and finance fail miserably, nobody can make mathematical models that work.
Did physicists get the easy part in science? Nature has deterministic and very simple laws. Physicists can even use mathematics to develop theory, because the laws are so simple.

I viewed physics (as most people do) as a very difficult science to learn. But we just study and apply the very simple and strict nature laws, how can that even compare to the difficulty of social sciences, that have to explain very complex systems and human behaviour?

Edit: Oops I wanted to post in the physics section. Sorry about that...

The Laws of physics are neither deterministic nor very simple.It takes years of hard work of hundreds of physicists to arrive at a law which seems "simple" to those who read it after it has been discovered, take for example Newton's law of gravity,in which he simply states the law and left the question "how gravity acts?" to the readers.Only after more than 250 years did people came to know of the mechanism through general relativity.
As far as social sciences are concerned I'm not sure that using mathematics is the right way to go forward.

 

[Pythagorean]

I've always wondered why "social sciences" were called sciences. I can see something like psychology and at a stretch sociology. I can even see trying to apply those fields to other fields such as economics. What I don't get is why a field that doesn't follow the scientific method is called a science.

If by the definition of "hard" we mean currently most difficult to produce good data from then yes social sciences are "harder", this is thanks to a lack of comprehensive understanding of the fundamentals meaning that A) far from enough data can be gathered and B) no good models can be built.

However this does not mean that either side has any superiority, what it means is that physics is a more mature field with a much better understanding. In biology there are many things that we haven't even begun to scratch the surface of but I wouldn't dream of comparing my work as "harder" than a physicist simply because I have less knowledge of the processes.

Well, good psychology definitely uses the scientific method (and is sometimes hard to differentiate from theoretical neuroscience).

The so-called "soft sciences" are in more of a stamp-collecting phase currently, but I don't think they're not sciences. To me, these things are depending on biology to get past "the surface" as you call it. We want to understand and quantify the significance of information transfer laterally across species (intuitively, it has an obvious evolutionary value, but we can't trace the evolution of social behavior as easily as morphogenesis for some of the same reasons that social sciences are difficult, the degree of complexity (and thus, degenerate permutation) is higher than that of biology (which is higher than that of physics).

They (social scientists) still develop some theory and laws (i.e. supply and demand) that till a little bit of the story about lateral information transfer. I myself am excited that it will be in my living days that we reach the next level of understanding in complexity in biology.