The new forums will be named Coin Return (based on the most recent
vote)! You can check on the status and timeline of the transition to the new forums
here.
We now return to our regularly scheduled PA Forums. Please let me (Hahnsoo1) know if something isn't working. The Holiday Forum will remain up until January 10, 2025.
Social Science Is Real! So Why Do Us Nerds Look Down On It So?
Posts
I am deeply wounded by the judgement of people on the Internet!
But really, while there're certainly sector of political science / international relations / regional studies that tend toward highly empirical analysis and very rigorous predictive "formulae" (ie. some strains of realism), I both don't find myself interested in them and don't think that's the approach the field ought to take if it comes as a result of demands to "harden" interpretive disciplines. Max Weber has this excellent analogy of a guy chopping wood. All the empiricist sees is a dude chopping wood. He sees the kind of axe he's using and whether or not it's the best axe for this particular type of wood; he sees the guy's stance, and whether or not it's ergonomically efficient; etc. The one thing he can't tell you is why the fuck the guy is chopping wood. Is the guy a lumberjack? Is he just letting off some steam? What series of events brought this dude to this place to do this thing?
That's the benefit of interpretive analysis, imho: the why*.
*aka Why the Fuck.
That could just be symptomatic of the self-selecting sample that chose to participate in this particular thread, said the political scientist.
So the people defending social sciences are a self-selecting sample while the ones "against" are a meaningful representation of what all nerds think? There is something fishy here ¬¬. Either this thread and these forums are representative or they aren't. But not both at the same time, or we will go into the complex field of quantum argunomics.
Its my impression that if supersymettric partners and no other odd particles show up then we're pretty much back to square one, as all the other candidate theories also expect similar particles existing in the same energy range, and the standard model doesn't make a lot of sense without additional physics going on to explain why certain things in the observable universe are the way they are.
I dunno, I took this more as a "Hey, what's the deal with..." thread than a "Thorough and Exhaustive Survey of Phenomenon X" thread.
With the more traditional sciences you have to get into pretty small details before you have to worry about huge amounts of uncertainty. Their descriptive and predictive power for every-day life is basically 100%.
Social sciences are 'more likely', 'less likely', or even less certain like 'may contribute to'.
"An object in motion WILL stay in motion." vs. "Abuse in childhood increases the likelihood of being an abuser as an adult." It's just a far weaker conclusion, and subject to anecdotal rejection.
Add in the fact that especially with Psychology, the base laws of it have been rewritten every few decades for the last hundred years or so, and often with some what Appeal to Innovation reasoning. The luminifireous ether isn't in disuse because it by current standards considered 'sexist/racist/cis-gender biased etc', and the fundamental phenomena it was used to explain (the wave-like propagation of light) is still a phenomena.*
While psychology and sociology are littered with theories explaining why Women/Coloreds/Gays suffer from conditions X/Y/Z. Which not only is the theory thrown away, but the condition itself is not considered a real phenomena now. Look at all the DSM fighting over what is and isn't a medical condition. When the knowledgeable parties can't even agree on the phenomena how can you accept their conclusions on the causes?
*e: to clear this up a bit more, It seems like a lot of formally accepted as truth theories have been tossed aside because they don't conform to a sufficiently gender/race/etc neutral framework, or at least their rejection is now presented that way. While "Jew's are good with money" is obviously a racial stereotype, has anyone actually done a study to confirm that Jewish people are in fact statistically average at managing their finances?
When you have this constant churn of previous conclusions, being rejected for reasons presented as more "that's ABCist" than counter-study X shows Y. Suddenly the rigor of all the conclusions are suspect. Plus you know, having presented that conclusion as fact a few decades ago.
Except that even hard sciences break down at certain levels (subatomic particles completely disobey Newtonian laws, for the most obvious example), due to our incomplete understanding if the universe.
Social sciences are no different in that respect: we have an incomplete picture of the human condition, so perfect answers aren't available. Saying, "Perfect answers aren't available, so fuck it," is the creationist copout.
Both are valid in their own way and often benefit from each other.
I'm doing alot on political science methodology right now.
The statement that there aren't any properly defined terms that most people can agree on is just wrong.
Modern social sciences exist since the 1950s basically. There have been some very important breakthroughs during thse 60 years.
Comparing it with the "hard" sciences that can just do test-retests all day long is not very useful.
A lot of modern "hard" sciences also have the same sort of incidences of probability outcomes. I ended up on a pub trivia team last night with a guy with a doctorate in genetics who is studying autism. He's hopeful that he's going to stumble upon some specific gene that will have a very high correlation between incidence of autism between his control or experimental group, but he's not optimistic because it's not even clear that autism is one specific condition. There could be two separate and distinct things that we are grouping together as autism. It could be a thousand different things.
The fact that his next publication is going to almost certainly have a conclusion expressed the same way a political science journal will express its conclusions--as a series of multiple but inconclusive probabilities--doesn't mean genetics isn't a science, right?
This reminded me of something my uncle told me:
"People think they're a Picasso, oh so complicated and refined. Everybody wants to be unique. Look at me, how difficult I am to understand, how fine my lines and nuanced my colours. While in reality, they're a fucking stickman, and people are mostly the same."
Why is it that prediction should be the measure of virtue in science? Is there some essential characteristic of science that involves prediction? If not, why should we prioritize it in this discussion?
"We believe in the people and their 'wisdom' as if there was some special secret entrance to knowledge that barred to anyone who had ever learned anything." - Friedrich Nietzsche
"We believe in the people and their 'wisdom' as if there was some special secret entrance to knowledge that barred to anyone who had ever learned anything." - Friedrich Nietzsche
Demographers advise people making public policy.
While that's funny, I don't really agree with the sentiment. You don't have to be Picasso or Da Vinci to be worthwhile or to do great stuff, and those fucking guys - while talented - were still human beings with all kinds of faults (I don't know my classics as well as I should, but wasn't Picasso a totally misogynistic asshole who's other favorite hobby was beating his wife?). People are unique; you don't need a Goddamn 'Indigo Child' or whatever the current special snowflake popular trend is, or to relate with the latest Oprah anecdote, in order for your life to be a rich, complex thing.
But for whatever reason there's this idea that if we can understand it, well then somehow it's diminished, or it's not rich anymore, or it loses it's charm, even though that's almost never the fucking case (you don't suddenly lose all appreciation for music when you study and understand it - well, unless it's bad music that you used to love. :P )
This doesn't appear to answer any of the questions that I asked.
"We believe in the people and their 'wisdom' as if there was some special secret entrance to knowledge that barred to anyone who had ever learned anything." - Friedrich Nietzsche
Ehhhh. Define 'square one'. No supersymmetry means that pretty much all of the currently accepted standard model extensions are off the table, so in a sense we're back to square one. But your comment about the standard model not making sense is hogwash. The standard model is a perfectly good tool for describing and predicting behaviors in the vast majority of physical modes. It just has a number of holes that we have no explanations to fill (if none of our proposed extensions that fill said holes are, in fact, correct, that is).
Why is there such a disparity between the strengths of the various forces? Why isn't the Higgs vastly, vastly more massive than it is? Why are there so many kinds of quarks, and why are they so disproportionate to one another in terms of mass?
The standard model fails to answer those questions, along with failing to contain a description of quantum gravity, but for QCD (the unifying theory of strong-weak-electromagnetic force interactions), the standard model does a good job. There is obviously physics beyond the standard model, since there are these big, open questions. Supersymmetry is a very nice, potentially compact answer to a number of problems. If we don't find it then we're, effectively, in exactly the same place we've been since quantum chromodynamics, as far as getting a full and complete description of the universe. There are quantum gravity theories that are independent of supersymmetry (and, in fact, don't aim to unify gravity with the other forces at all; just describe its actions at the quantum scale), so it's possible we'll find that SUSY is wrong but that one of the QG's is in.
And, I mean, hell, we found the Higgs and explained the neutrino problem. It's not like physics has been twiddling its thumbs for 30 years.
Generally speaking, prediction is useful. Finding a rule that happens to fit all existing data but cannot predict what new data might arise is less useful.
Say I give you a set of numbers: 5, 3, 4, 3, 7, 1, 2, -4, 187. It's trivial to come up with math function that fits all those data points. Say I do so... now what? What was the point? What if I tell you that there's a tenth number? If your function fits the data but cannot predict the tenth number, is it a very useful function? I'd say not.
Honestly, I'm trying to think of a counter example, and I can't think of a single way in which a non-predictive model can be useful in pretty much any field. Any rule or hypothesis that purports to explain existing phenomena, or existing data, generally does so with an eye towards predicting future phenomena. Even things like history - to the extent it's useful as more than trivia, it's so as to understand past events in a way to guide future actions. Theories about evolution help us predict what sort of fossils or findings we might discover in the future, even if we're only finding examples of things that happened eons ago. Maybe there's a counter example of a kind of science that has no predictive power whatsoever, but the ones I can think of are like some of the varieties of string- or M-theory discussed earlier, in which theories are designed only to fit existing data and can't really predict anything verifiable.
So yeah, if you want to give me an example of non-predictive useful science, I'd love to hear it.
...I can't think of anything non-predictive that would be considered science. Even most junk science is 'predictive', or at least pretends to be.
Creation Science (TM) ? :P
I was gonna comment on how a lot of this discussion ignores much of philosophy of science, and also on how there is a nominalism streak going through parts of it (it's more science if it resembles this science!) but I decided to have a beer instead.
Good beer though.
Uh...um...archaeology, maybe? Geology?
I don't think it's absolutely necessary for a science to predict what's going to happen in like 2 years or whatever, although generally I think there needs to be the ability to predict...something? Right? Your hypothesis needs to predict the outcome of an experiment. The validity of the data your experiment produces is only valuable to the extent that they are reproducible, right? Even in archaeology, if your working theory is that ancient Egyptians during 400 BCE traditionally burned the remains of dead slaves on funeral pyres near their job sites, you're going to expect that all other remains of slaves found in that area from that time period will also have been burned. So you're at least making that prediction.
Geography / Cartography?
I mean, that's the closest one I can think of, because you're just directly mapping whatever you find (I think? I'm not a professional cartographer, and have never met one).
A lot of evolutionary research technically doesn't predict anything. The fact that a species died out THIS MANY thousands of years ago does not directly predict shit. It's perhaps useful data for other research that might predict stuff, but then what you're really asking is if all this science being done is a net-benefit to the human race.
I think Loser's point is that even if one is not doing science for the purpose of prediction that it's silly to say they're not doing science. While it can be said that all science adds to some predictive power of science it is nonsensical to argue that that is what makes it science. Studying a supernova is itself science, the fact that such research helps us predict better is the reason for doing science and not a requirement of science.
...What 'stuff', specifically, is non-predictive? I mean, the example you provided is perfectly predictive (in that case, the prediction would be, "At [X] million years ago, we should be able to find [Y] species, because that's when [Z] species was adapting to a new environment,")
Studying/observing a supernova also involves prediction - everything from the more mathematical predictions of contemporary astronomy, to the old predictions of, "Hey, I think that bright object might've been something that happened to a star,"
I mean, Computer Science doesn't provoke the same ire. Nor does Domestic Science.
I have a feeling that there is too much of a tie between useful and predictive. I guess I could talk about how scientific advancements like lasers also allowed for aesthetic experiences (which are valuable), and thus they are useful for something that has nothing to do with prediction. But I feel that you will simply say that they are useful for prediction as well. I don't think that I can find anything scientific that just cannot be used for prediction now, nor ever will be used for prediction in the future. I mean, some genetic manipulations aren't strictly speaking predictive, but they could be used in such a way in the future.
But that wasn't what I was asking specifically. I wanted to know if the ability to predict was some way in which we SHOULD construct a schema of value in science. I know that we can, but is there some justification?
"We believe in the people and their 'wisdom' as if there was some special secret entrance to knowledge that barred to anyone who had ever learned anything." - Friedrich Nietzsche
One the breakdown on not knowing, is so much shallower. Newtonian physics covers basically all your daily observeable interactions.
Two,Its not a just breakdown of not knowing the answer , there's a lot of disagreement on whether the phenomena be asked afteare real
Literally, exactly, 100% what you are saying is "these fields aren't science because they're harder than other science".
You would also be wrong to do this, since you have no reason to suspect that the function you're working with is invertible you do not have any way to determine which is the best explanation based on data. We go back to people not understanding why people have to stress "correlation is not causation" because they seem to have forgotten when correlation is causation and why we use these metrics when we do.
Here is an important question. "Is it possible to carry out an experiment which will imply your hypothesis is correct?" and it will have an unsurprising answer (Pro tip: Its "no").
Correlation is, like any experiment by any physicist, simply a non-refuting result. What do we call non-refuting results when we have "reason to believe" that the theory is correct?
"reason to believe" is in quotes here for a reason.
The same data from the same star is the same data it is not more data. What you achieve when you do that is information about the quality of your measurement, not information about the star.
The short answer is likely self selection. People who are good at mechanical operations in simple worlds with simple rules tend to gravitate towards doing mechanical operations in simple worlds with simple rules.
Engineering, and the other "geek disciplines" are more or less, applied rules. Sure there are creative problem solving but the rules never change. Engineering, "computer science", etc etc etc.
Sciences, as in actual sciences, are disciplines in which the rules are, essentially, always changing. Or, alternately "if the rules are stable then what the fuck are the rules?". This draws a fundamentally different type of people in.
This differentiation probably starts due to how introductory applications are taught. In the "geek disciplines" the tolerance for what we know the rules to be is very small and we can teach simple things that appear to be always true. Does F=MA? No. But for an engineer its always close enough, the rule never changes how we can apply the rule to build this bridge.
In the "social sciences" we don't teach in this manner and "geeks" self select out of those disciplines. The problem with the "want to feel better than others" answer is that well, everyone wants to feel better than others so why do nerds to it over "hard" sciences.
(Though strictly speaking, I'm not entirely convinced that "less predictive power" is always a component of the science rather than of the underlying system seeks to model. It can be, no doubt, but if there's a natural law that says everything is deterministically predictable we sure as hell havn't found it yet)
It's simpler than all this. Science strives to know reality, how things work. In knowing how things work you need to have a model that adjusts to reality, and the main test of whether a given model adjusts to reality or not is its predictive capability.
As an example, we know Newtonian equations describe classical physics accurately because you can use those equations to drop a howitzer shell on the head of that guy over there. If you got the same set of equations but you either had no way to test them or everyone who tests them got different results, we could safely say that we can't trust that particular model to fit reality well.
Well the idea that we can compare predictive power between the sciences is kind of dumb. I mean how many inches away did we get from predicting GDP last year? If we are measuring something immensely small does that mean that a .000001 tolerance we have make our prediction error hundreds of thousands of percents?
I.E. even if we want to use a unitless measurement the measurements are meaningless.
Even in physical geography there is predictive science when researchers look at interactions between humans and the environment. Human geography includes tons of predictive science. And most mapmaking these days involves mapping demographics, social characteristics, etc. to find spatial patterns so we can predict future outcomes.
"A Picasso" as in "a painting", not the person.
But it's ridiculous to me that people can prefer not understanding something.
Engineers were more likely to talk to us at parties.
A lot of the rest is just justification.
I think part of the defensiveness on the part of social scientists (medical anthropologist here) is that people are often more willing to accept absolute garbage research if it seems "more rigorous". Evo psych is more or less the #1 test case for this. Now, over time sciency-sounding garbage does tend to get exposed for what it is, but it can take quite a while. And then when you mix it in with the politics of race, colonialism, etc. etc. ... the "Psychology Today" style of 'hard social science' can do a lot of damage.
So, while part of it is just sensitivity on the part of social scientists, they have some pretty concrete concerns w/r/t to the question posed above by Australopitenico, among others: "Why do people assume that when you say something is a soft science, you're dismissing it as invalid?" In fact, that often happens. If you look around, a lot of the response to the Chagnon/Sahlins controversy takes exactly that tone: that clearly Chagnon is a worthy addition to the NAS because he's more sciency -- even if that means distributing machetes to rival groups and deliberately getting them furious with eachother by breaching social etiquette, then calling the resulting violence a demonstration of essential, primordial human nature.
Most people don't get any formal experience with social sciences until they are in college, usually just a required core Psych or Sociology 101-type course. It's basic stuff, and a lot of it is so abstracted, compressed, and dumbed down that there just isn't that much useful information. Easy A's and things like that.
Meanwhile, the 'hard science track' freshman is probably taking Calculus II, some sort of Chemistry / Physics class, etc. These are classes that are building on twelve+ years (with math) and usually a couple of high school courses (Chem / Physics). You are actually doing experiments, getting lab time, and doing science that's not THAT fundamentally different than what 'real' scientists are doing. To most people, those are 'difficult' classes (the math is hard meme) but for a science nerd freshman Calc / Chem / Physics are pretty basic stuff...especially looking back after another four years.
In comparison, Psych and Sociology seem basic and simple, and seem like anyone can do them. Quite simply, because they can...they are only scratching the surface. Those classes aren't much more than a simple introduction to the foundations of those sciences.
Imagine if we spent twelve years teaching the basics of sociology / psychology, and first exposed people to Chemistry / Physics / Math their freshman year of college. You would know advanced modeling methods, the affects of media and propganda on populations, survey methods, etc. But you would be taking one class in math where you learned multiplication and division...or a Chemistry class where you spent two months learning about states of matter and what the periodic table is.
Of COURSE you would have a low view of the 'hard sciences' - they would seem like backwards and basic courses.
I'm not sure if the answer is requiring students to be more well rounded and take more courses from Psych / Sociology, or finding ways to have more cross-disciplinary projects where people from different majors and sciences could be exposed to more of each other's tools and techniques.
And never discount narcissism and the special snowflake effect. Nerds seem to be especially vulnerable to those 'afflictions'.
It took people a hell of a long time to figure out that first one, and for good reason. Objects consistantly have outside forces acting on them, to the point where we could actually say with perfect accuracy that "an object in motion WILL NEVER stay in motion". Physics isn't more rigourous, it's just much easier to oversimplify.
I'm sure you're saying something here but I have no idea what it is. Correlation can lend support for a proposed causative operation; it can't simultaneously indicate that relationship and be proof of it in the same data set. It's question begging. "Oh, I have 6 cubes and 3 of them are blue, I bet collections of cubes always contain half blue ones and... look! These ones I have are half blue!"
As for the other bit, yeah, it's a shitty, basically untestable hypothesis. That was kind of the point. The history of sociology is rife with these kinds of shitty little stories based on what observations "suggest". Girls like the color pink because, during our species' formative epoch, women performed the 'gathering' part of 'hunter-gatherer' and red berries were good to eat. That's a real thing that real sociologists actually published and weren't immediately laughed out of the science club for. Except that pink as a feminine color is a concept that's existed for less than two centuries, prior to which the opposite was true. And that's just in European (and European-colonized) countries; other parts of the world don't draw the distinction at all.
Or maybe you're agreeing with me? I honestly can't tell.