Natural, formal and social sciences are all “real” sciences.
Some social sciences are “real” sciences. Others not so much.
Social sciences are not “real” sciences.
0voters
Basically, do you draw a sharp distinction between natural and formal sciences (mathematics, physics, chemistry, biology) on the one hand and social sciences (linguistics, psychology, sociology, economics) on the other ? In other words, do you consider them as fundamentally different endeavours ?
Or do you see both as based on the same intellectual principles and relying on the same underlying processes, with methodologies differing only due to the nature of their subjects of study ?
I am not the STEM type, I did my stint in women’s studies, sociology, and social work.
I think there’s a legitimate desire and attempt to study us (i.e., ourselves), and it combined with a desire to embrace the rigor and precision of the physical sciences. (That latter has not by any means been confined to the social sci folks; if you want to see a really intense example of it, check out the field of pyschiatry).
There’s a huge observer problem that they tried to ignore. The hard sciences mostly don’t have to worry about that, and where they do, they’re better equipped to be aware of it and deal with it conceptually and theoretically. The rock in yonder strata or the chemical in that beaker over there tend to “be what they are” and to not exhibit a lot of “depends on the observer and the exact position of the observer and how the act of observation changed the thing observed” behaviors. Microparticles and their position and momentum, not so much, but the physicists acknowledged and dealt with that. Likewise at the other end of the scale, macro “entire universe” structures and processes and the necessity of considering certain things that come into the picture when the object of study is the entirety of all that is.
When the object of study is us, the observer problem is that we’re inside the object of study as participants, as participants whose own intentions affect the formulation of models and hypotheses, whose own intentions will be affected by any understandings gained (or, in some cases, would be affected by them if those understandings were self-permitted to be gained, if you catch my drift). To borrow a term from astronomy, the parallax problem here is huge. How the various topics look depends vastly on the observer’s position (social as well as temporal-physical). Envying the hard sciences, the social science folks tried to back off from the subject of study far enough to achieve some modicum of objectivity, forgetting or not properly realizing that they are always inevitably a part of the subject being studied. That (as portrayed so excellently in MC Escher’s picture of the portrait gallery) the social structure and social interactions that need to be considered contain a person who is spending time studying the social structure and the social interactions and drawing conclusions. Or, actually, a whole bunch of them, the immediate self being merely one of them.
But IMHO the social sciences are an area of examination where it’s more difficult to tease out valid data via the scientific method than other areas. As a result there is a lot more speculation and hypothesizing based on poor or no data than in other areas. Not that such stuff doesn’t go on in the STEM sciences, of course.
Trying to study people and society scientifically is a worthy goal. But it’s a really, really difficult goal, for many reasons (including but not limited to the observer effect and biases @AHunter3 mentions; even aside from those, they’re still just plain really complicated subjects). In a few cases, some social scientists have been really clever, really skilled, and/or really lucky enough to be able to get some scientific results anyway. In most cases, though, they just plain haven’t been able to, and it’s hard to fault them for that inability. And the inability to get good results has, unfortunately, led a lot of people in those fields to instead pursue poor results.
I adhere to Karl Popper’s view of science. The strength of science comes from the ability to test hypotheses and attempt to falsify them. If a theory is unfalsifiable, it’s not science.
Science progresses because it builds on a foundation of objective theories that are constantly tested for conformance with the real world. It doesn’t rely on consensus or on the existence of smart people or experts who pass judgment on ideas, but on actual tests against reality.
Every Nobel winner in a science can agree that something is true, but if a test shows that it’s not, they will have to eat crow. Absent such tests, a ‘science’ just drifts with the political winds and the existence of strong personalities on one side or another.
The problem with many social ‘sciences’ is that they are little more than collections of observations, best practices, consensus opinions, and groups of ‘experts’, but without actual falsifiable and testable claims. This causes them to twist with political winds, devolve into factions that disagree with each other, follow blind alleys for long periods of time, and in general make little real, quantifiable progress.
For example, Freud dominated psychology for many decades, but now his stuff is considered mostly junk. Economics devolves into ‘camps’ like the ‘fresh water’ and ‘salt water’ economists, and they can’t agree on even basic economic principles like what causes inflation. If your political leanings affect your conclusions and there is no objective test to determine if an idea is actually objectively correct, you aren’t doing science no matter how much math you use.
It might be a bit broad to call out an entire field as being either science or not science. In many of these fields there are areas where real science is done, and areas where it isn’t.
Here’s a good example of the problem. The field of behavioural science makes plenty of predictions about how people will respond to various events and stimuli. Their history of prediction is terrible, but they can blame it on the complex environment they are working in, where there are so many variables and conditions that you can always find a reason why your prediction failed. This is true of economists as well.
So then someone comes along and looks at all their ‘scientific’ papers and their results, and this is what happens:
Many of these fields also have a severe ‘replication crisis’, meaning people can’t even replicate the results of others’ experiments.
In my mind, mathematics is at least as much of a fundamentally different endeavor from the natural sciences as the natural sceinces are from the social sciences.
You may be right. I’d say that social sciences are sciences, but they’re sometimes pursued and written about in non-scientific ways (perhaps like the natural sciences were in the pre-scientific days between Aristotle and Galileo).
There has been a huge push (including research funding) to apply the same sort of of mathematical modelling, simulation, and data analysis used, for example, to tease out conclusions from particle accelerator observations in the social sciences, but nobody said it was easy.
The blunt fact is History is much more difficult than any of the hard sciences. Take, for example, “rocket science,” often used as the ur-example of hard science. I’m sorry, but the moon is a pretty freakin’ big target and you’ve got billions of dollars and hundreds of thousands of people working to get there. In contrast, we have (adopts movie trailer voice) “One lone historian, armed only with their brain, note cards and a pencil, must recreate an entire world…” See also, “Who knows what evil lurks in the hearts of men? The Historian knows…and it is a terrible burden.”
My daughter has a PhD in psychology, and one Christmas her favorite present was a MatLab manual I gave her. She’s not a touchy-feely type.
I hear from her all the ways people can screw up experimental design and data analysis, so it does not surprise me that many predictions don’t work out so well.
But the physical sciences are not immune to this problem. The mass of the electron was wrong for years, and slowly converged to the true value after repeated experiments. I think Feynman noted this. And a favorite creationist trick is to send samples for radiometric dating where they didn’t follow the known protocol, and then claim the technique doesn’t work.
Yup, and what’s worse is that it’s not always appropriate to the subject of study.
As the saying goes, to somebody with a hammer, everything looks like a nail. The so-called “scientific method” of experiments testing precisely specified hypotheses under rigorously controlled conditions is an excellent and extremely effective hammer, and science has discovered metric assloads of valuable results and invaluable knowledge of the world by using it.
But sometimes people yield to the temptation to believe that a hypothesis is more precisely specified than it is, or that experimental conditions are more rigorously controlled than they are, and demand for the findings the same sort of acceptance that would be accorded to experimental results in, say, the physical sciences. (Boy, did we ever used to see a lot of this behavior on these boards when self-described “race realists” came in with some new study purporting to link social characteristics and race, for example. Boy, did we ever.)
Social-science issues still definitely need to be studied and thought about. And for certain narrowly focused questions amenable to rigorously controlled experiment, we should of course take advantage of the methods of experimental science to do so.
But pretending that social-science study is more “scientific” than it is, or claiming that its subject matter is only worthy of study to the extent that it can be treated with scientific rigor, is counterproductive.
Quantitative methods are also used in astrology. Whether a field uses math or not is utterly irrelevant to determining if it’s a science. In fact, the use of math often obscures the fact that there is no rigor or reality underneath it.
He did. He brought it up in ‘Cargo Cult Science’, as a demonstration of the hazards of relying on ‘experts’. The experiment in question was the Millikan oil drop experiment. Millikan got the wrong answer because he used the wrong value for the viscosity of air. So his answer was too small. But Millikan was a ‘great man’, so a curious thing occurred. I’ll let Feynman describe it:
But here is the key takeaway: They DID eventually converge on the correct number, because they had objective tests for correctness which were falsifiable. The fact that it took a while shows how powerful group signals can be as a force preventing the search for truth. The only way around that is to insist on objective, falsifiable tests of a theory, regardless of how important the person who presents it is, or how big a ‘consensus’ of scientists you’ve managed to gather around it.
That’s the hallmark of a real science: Do errors propagate over time, or do they correct? Is there general agreement on the key theories in the field, or do they propagate along political or other lines? Any experimenter can be wrong, and every theory can be flawed. The only way to find the truth is to have falsifiable tests of the theory.
Absent falsifiabity, theories get presented and accepted as consensus, then that gets used as a base for further theories. You build an entire field on a body of sand, regulated and controlled by the ‘great men’ of the field or by committee or consensus.
What are the fundamental advancements in economic theory? Are there any theories that are accepted as fact by people from all political persuations? In my lifetime, Keynsianism was supreme, then it was discredited and monetarism reigned, then Keynsianism came back. Now we have a whole bunch of unintended consequences that call Keynsianism into question again. And people can throw fringe theories out there like MMT that get real traction among economists of certain political stripes. And what you believe seems to have a lot less to do with education in the field and more to do with your political beliefs.
Freud, like Millikan was a ‘great man’. So his ideas held sway in Psychology for what, 70-80 years? And now many psychologists think his theories were nonsense. But why did they hang on so long? Well, because the field is full of non-falsifiable claims and deals with a complex system (the brain) that by its nature can not be learned through reductive analysis and which adapts and changes so what was true yesterday may not be true today. So bad theories can hang around indefinitely if people in the right places like them enough.
See my cite above about the absolutely non-existant power of behavioral sciences to predict behaviour. And yet, no one in the field will take those results as a falsifiable test. They’ll just continue on.
In economics it’s been shown again and again that any forecast beyond about six months has the predictive power of throwing darts at a dart board. But that doesn’t stop us from demanding constant prediction and treating it like it’s ‘science’. We wait with bated breath for the next CBO score (they use math and everything!), ignoring the fact that CBO scores have proven to be useless in the past. We have a great need to find people who will promise to tell us the future. It keeps astrologers, fortune tellers and pundits in business whether or not anything they say is useful or true.
These are what Feynman called ‘Cargo Cult Sciences’.
Economics is the closest social science to real science there is, that I know about, anyway. It seems my vote is in the current majority, and I wonder what other sciences similar voters are considering.
I don’t know how one would measure that, though. As Sam validly points out, the question isn’t whether the discipline has quantitative models described with mathematical rigor, but whether those models accurately represent the underlying reality.
The problem is that social sciences by definition study the behavior of human beings, and human behavior is hard to represent accurately with mathematically tractable models.
I am more academic than most. I have a reasonable understanding of simple science, but also a long standing personal interest in politics, economics, history, sociology, law and psychology.
This question is too simple to be useful. Any discipline can be practical or esoteric, and most knowledge has its place. It is true economics, for example, often relies too heavily on mathematical models which make assumptions much simpler than the real world, and sometimes worse than limited use - because it is used with false confidence or understanding of limits (though this is not always true). Then sometimes social sciences remove appropriate math to make the subject more broadly accessible.
Engineering also uses mathematical models, but much more successfully since physical values are easier to measure and there is more basis. This does not diminish the importance of economics, but there are limits to applying scientific approaches in some subjects not easily amenable to experiment or quantification.
What one deems important may be affected by what one knows, what one dislikes, and what one is good at. The problem with expertise, as opposed to generalized knowledge, is that there is the temptation to view every nail using only the hammers at one’s disposal. Not all expertise is academic.
I’m a STEM scientist* (paper #25 just got published. Huzzah!). Although I should be careful saying that because some people don’t consider computer science a real science either. I voted “some”; however, I don’t think this it is the field but the work. Some work is what one would call “science” and some is less so. However, saying that some of it is less so should not be taken to mean that it is of lesser value.
for now. As some of you know, I am working on starting a new career.
Yeah, I voted “they’re all real sciences” because the question’s too simplistically phrased. Individual scientific fields with their own established methodological norms get to define what does and doesn’t count as “real” science within that field.
But they don’t get to define how society as a whole uses “science” to refer to a wide variety of different fields.