I was incredibly disheartened to find a link to a blog post questioning, yet again, whether social science counts as science, this time by John Horgan at Scientific American. I’ve taken on the myths surrounding my career before, and quite frankly I’m getting sick of it. So this time, I’m going to pick myself up off the floor of frustration and hopefully help move the discussion beyond the same uninformed stereotypes we’ve all heard a million times before. Taken to the extreme, I feel as inaccurately portrayed as the scientist with crazy hair and colored test tubes.
Before I delve into the nitty-gritty, I’d like to tackle the definition of science. There are a number of mostly narrow definitions out there. The one I ascribe to is evidence-based. The research I do is theoretically-grounded, connects research methods to that theory, makes observations using those methods, and then draws conclusions based on that evidence. While this may sound general, science is a broad approach that rapidly sub-divides by discipline and philosophy from there. Now to the less philosophical part…
First, I take offense that all of the social sciences get lumped together into one category. Right off the bat, it shows an un-nuanced analysis of the state of the social sciences writ large. It’s a bit like lumping all of the natural sciences together – physics, chemistry, biology, astronomy, and others. Remember, we social scientists come from different disciplines with different traditions, sets of methods, pet peeves, and history – specifically, political science, economics, sociology, anthropology, and geography. That’s not even breaking down the categories as far as NSF goes, as they add law & society, social studies of science, and coupled human and natural systems, among others.
Speaking of NSF – funding for the social sciences across the board has only shrunk the same amount NSF has shrunk (a sadly large amount). Political science was unfairly targeted because – shocker – they turn a critical analytic lens to the political leaders of this country and the world. This is perhaps the time to discuss the needed unmelding of politics and science, but that’s another discussion altogether and one that has been tactfully taken on by many others. Another very real reason we receive less funding? We don’t need reagents for our research that cost thousands of dollars at a go. I think Horgan is conflating price with societal value. Any economist would tell you it’s more complicated than that.
The most egregious oversight in this article is the confounding of the terms ‘soft’ and ‘hard’ science with qualitative and quantitative. Many of the social sciences are very quantitative – namely, sociology and economics – some qualitative, like anthropology, and others a mix. Horgan quotes from an equally confounded Economist article that states “data from social networks are making social science more scientific.” I think what they both meant to say is “more quantitative”. There’s a ton of Big Data emerging from social networks and computer archives of personal data. That means there will be many more scientists mining this data using statistical techniques. That is not to say that we weren’t scientists previous to this boon in data or that this data is inherently more valuable just because it comes in the form of numbers. Statistics can be even more subjective than qualitative approaches, as taught in a tongue-and-cheek way in early statistics classes with the book “How to Lie With Statistics“.
Horgan also has a funny definition of soft science, accusing these ‘softies’ “of trafficking in theories so lacking in precision and predictive power that they don’t deserve to be called scientific”. I interpret this to mean that any research, then, lacking predictive power is not science? So studying string theory, the inner mechanisms of molecules, or all of descriptive biology doesn’t count as science? The controversy Horgan refers to was a media story over whether anthropologists (not all social scientists) consider themselves scientists. As one of the most diverse of the social sciences, the American Association of Anthropologists put out a formal statement applauding the field’s plurality and supporting anthropologists who identify as scientists. But that debate is specific to anthropology.
The concern over the repercussions of sociobiology that Horgan refers to is true – practitioners of that particular theory were too quick to institute a number of unjust policies – and is a black mark on social science. But the problem stems largely from the fact that E.O. Wilson, a biologist without training in the social sciences, promoted a huge social theory without the nuance required. In the end, it was social scientists who put the story straight.
The more recent blending of neurology and many of the social sciences is not a similar reaching of researchers into the “hard” sciences. Instead, it’s a product of newly available technologies that can now tie chemical and physical changes in the brain to well-documented social trends. It’s a brand new discipline that promises to bring greater understanding to the human condition. But dismissing it as a group of wannabes is a bit like dismissing biochemists because they contaminate the ‘pure’ disciplines of biology and chemistry.
Finally, Horgan makes a valid point that studying humans is different than other disciplines because we can talk to, and possibly alter, our research subjects. That’s why there’s a careful ethical practice in place for human subjects research, both in the clinical and social sciences. But moving on, many other disciplines also perturb their observed. Marine biologists literally drill holes in the dorsal fins of sharks to attach satellite tags, field ecologists often rip out pieces of the ecosystem to see what will happen, and my favorite example from physics – the Law of Observation also known as the Observer Effect, which states that you can’t observe a particle without changing its quantum state and therefore the fundamental behavior of the particle. Hence a major divide between theoretical and experimental physicists. So perhaps it is all scientists who should think about making their claims with humility.
Related to this point, the conclusion that social scientists are therefore obligated to do applied research because we can talk to our research participants is analogous to the long-standing argument over whether funding should support the basic sciences that has recently re-emerged in the wake of the US sequester. I’ll not delve into that piece of controversy, but point you to a reality that is the state of this particular argument.
And, furthermore, to extend the argument that not only should social scientists be limited to applied research but they should also be called engineers – well, that’s just a personal affront. For someone who works on environmental sociology, sure, I’d like some of my study sites to move towards more sustainable cultural practices. But sometimes I’m not there to make the world a better place, just understand it better, figure out what makes society tick. That may be useful to tomorrow’s activists, or it may just contribute to satisfying that part of human nature that makes kids ask “why?”
In the end, I hope that people use Horgan’s article as a jumping-off point for a fruitful discussion about the social sciences. I think the fundamental misunderstandings that I frequently encounter, the ones that put many practicing researchers on the defensive, stem from a lack of basic education in the social sciences during elementary education. Remember when you had chemistry, biology, physics, and “social studies” in high school? A more well-rounded palate would have offered geography, sociology, and anthropology alongside physics. So while children learn the laws of Newtonian physics, they also learn social theories. Alongside dissecting frogs, they learn how to write a good survey. In that world, I won’t have to justify the validity of my field to those who grew up without a full suite of scientific experiences.
“In classical physics science started from the belief — or should one say from the illusion? — that we could describe the world or at least parts of the world without any reference to ourselves.” — Werner Heisenberg, _Physics and Philosophy_.
As one who holds a master’s in psychology (from Horgan’s school, where I had taught human factors concepts to engineering students), I appreciate Heisenberg’s recognition of human subjectivity when it comes to research — from experimental design up through the interpretation of results.
Everyone, including scientists of all stripes, is a product of the times. In his article “False Premise, Good Science” (in _The Flamingo’s Smile_), Stephen Jay Gould described how Lord Kelvin had undermined the entire discipline of geology with a theory that subsequent studies of radioactive decay ultimately disproved. First, Kelvin had calculated the Earth’s age in a way that corresponded to Biblical scripture. His religion drove his theorizing. Second, many geologists followed Kelvin’s lead because physics was considered the “queen of the sciences.” Wrote Gould, “Geologists should have trusted their own intuitions from the start and not bowed before the false lure of physics.”
Perhaps a respect for and greater support of social science as science might help mitigate errors made by those in the “natural sciences.” In the end, human nature and human subjectivity drive all human endeavors.
Perhaps the all too common flagrant disregard for proper scientific method is to blame here. Far too many times, the social sciences publish “groundbreaking” articles based off statistically insignificant data samples. While a group of subjects, say roughly a dozen, is just fine for preliminary investigations it is all too common to see such research used as a basis for general application to the populace as a whole.
This is a serious issue in that the mass media has a tendency to run with such announcements. One must keep in mind that the bulk of the population has only a cursory understanding of the sciences as a whole and are prone to believing anything they read or hear. This trend is further exasperated by the failure of the researchers to state “more research is needed”. The tendency to let a statistical model prove a point and then apply said model to a populace needs to stop. Real science would use the statistical models to reinforce findings from a large, control based sample size instead of a a dozen random subjects from one locale in one country.
I get it, it is all about publishing or perishing, but at what cost? No amount of support will change the public view of the social sciences. Respect is earned, not dictated. Small sample sizes and over generalizations are not “good” science, well structured, carefully crafted, peer-reviewed or not.
In short, if the social sciences need to re-invent their image, they need to start with the scientists themselves and use ALL the proper, valid methodologies.
The primary domain of “failure” of the soft sciences is in testability and falsifiability. Since dealing with living things entrains a moral component, proper experiments are often impossible to infeasible. Does that compromise the practice? How high is up?
I can’t help but take your comment personally. Sure, there are bad researchers in the social sciences. But this is true of all disciplines (the publish or perish is true). Marine mammal researchers are happy to get an n of 4. Why? Because their study subjects are often endangered and hard to access. Does that make the work they do any less valuable?
And do the practices of some make us all dismiss-able? No. I had over 100 water-quality-related people (in a town of 3,000) participate in my dissertation work – both statistically valid and having reached the point of saturation. Do I try to generalize my work in one town to the country? No. A less personal and more poignant example – the census bureau is one of the largest employers of social scientists. Does the entire country count as not statistically valid?
Point of clarification, also – not all science is experimental. That is, the scientific method does not equal science. Plenty of observational studies undergird the work in almost every discipline, from Newton’s falling apple to Darwin’s finches.
I can’t help but get a bit prickly that you’re writing off several disciplines, with plenty of smart people who follow the rules of science, because of a narrow view of what they do.
What exactly is a “soft” science?
Science has no value outside that placed on it by people. Does anything published in the social sciences add value to society or to individuals? It certainly has in fisheries. Economists and other social scientists have been arguing for decades to stop using command-and-control regulations and switch to TURFs, IFQs, and other less centralized regimes. See http://www.sciencemag.org/content/321/5896/1678.full for an example of the benefits.
I’m a fisheries economist and serve on numerous technical committees with mostly marine biologists. Can I age a fish using an otolith or run a stock assessment model? No. Can the marine biologists fully dissect the potential outcomes of a new system of regulations? No. The system runs far better now than it did before we were brought into it.
“the scientific method does not equal science” — well said!
Timely and related – What’s wrong with the scientific method – courtesy of Wired.
Just because an object is inanimate does not make it any more or less influenced by the scientist doing the science. In response to Horgan, JOF, and Bruce, perhaps you could post one of your favorite actor-network theory articles for them to read. They all seem to view science in black and white (or as purely objective), which is hardly the case. I think that’s one of the greatest attributes of the social sciences…that they understand and examine the gray areas that others either don’t even see or don’t care to waste their time seeing.
“Far too many times, the social sciences publish “groundbreaking” articles based off statistically insignificant data samples. While a group of subjects, say roughly a dozen, is just fine for preliminary investigations it is all too common to see such research used as a basis for general application to the populace as a whole.”
Are you basing this assertion on peer-reviewed journal articles, or on the mass-media summaries of those articles? I couldn’t possibly count the number of times I’ve seen some grandiose pop media headline like “Study finds that men cheat more than women!” only to find that the peer-reviewed article was much more limited in scope, like “Gender differences in perceptions of infidelity among college students.” A researcher might be very careful to accurately describe the scope of validity of their research and to avoid hasty generalizations, but journalists seem to be less conscientious about that.
This is a particular problem not just for psychology, but for any scientific field with mass market appeal – cancer research (“We applied a botanical extract from the Rubus genus to tumor cells and it slowed their growth by 10%” becomes “Blackberries newest weapon in fight against cancer!”), or engineering (a lab manages to build a graphene square a few millimeters wide, and some magazine starts talking about body armor for soliders).
Who cares what others think? Ask interesting questions (interesting to you), do good work, and the cream will rise to the top.
Let the haters hate. Your work will stand or fall over time on its own merits.
The ‘scientific community’ can be a voting machine in the short term, but will be a weighing machine if you take a longer view.
I’m rather surprised that *anyone* still pays attention to what John Horgan writes about science, after his 1996 stinker “The End of Science”. Seriously, has the man ever said anything that was (a) original, and (b) true?
A reminder from E Rutherford “All science is either physics or stamp collecting”. As quoted in Rutherford at Manchester (1962) by J. B. Birks;
The problem as I posit it here, is the importance and or relevance ascribed (or not ascribed) to imaginative organization, data collection, analysis, correlation and theorizing. True this is stamp collecting, but some day generations hence, with luck, it will be physics. I am sure many many folks can tell me bunches about the relevance and importance ascribed to the tag “science” – as an American I want to know if you can make money from it, and if so – – study away and call it what you will.
An apology for a disjointed comment but I wanted to add something from the original Horgan article in Sci Am –
“Here’s a more specific suggestion: Social scientists should consider identifying not with the harder sciences or the humanities but with engineering.”
Actually as a practising engineer I would loudly applaud this sentiment. Science moves forward largely due to engineering. Engineering is not science but the tools built for scientists by engineers ….. well the feedback loop should be obvious.
I think Horgans point is well made especially in context of engineering vs science. Practically we all suffer too often from the request for scientific answers to what are actually engineering problems?
Respectfully Floyd Roberts
Of all of Horgan’s arguments, the suggestion that social scientists should be engineers was by far the worst and stems from a fundamental ignorance of what social science is and what practicing social scientists do.
I also continue to have no idea what exactly a “hard” or “soft” science is, and suspect that most people using those terms actual mean “science I value” and “science I don’t value”.
I concur with your definition/opinion of hard vs soft science within the present context. Since I was enamoured of the social scientist as engineer comparison, I postulate that I too suffer from the same fundamental ignorance of what social science is. The banishment of ignorance being desirable could you point me towards a reference or elucidate further? I have been working from the comparative reference that physics is the the study of physical phenomena, going forward to social science is the study of social phenomena. I maintain NOT that social phenomena cannot be quantified, simply, that the mutability of the subject (I maintain that evolution is not finished, but the laws of physics are) and the current data (social science data is of necessity currently a very small subset of human experience) suggests that like medicine there are a large number of engineering challenges to be overcome before data may be applied/ Laws? formulated, to build foundationary structures. Extrapolating, I suggest that the reputation (such as it is) of science comes from the relative permanence of such foundational constructs.
If we’re really making these distinctions, I’d rather separate physics and leave the rest of science as science. Physics is really more like math anyway.
The problem is when this attitude becomes prevalent among people deciding on grants and organizational research decisions, important social science research risks not getting done.
Social science. I’m sure it is treated like alchemy in the restrictive so-called formal science indoctrinations. But then, Physics, chemistry, and many other disciplines came from alchemy. True alchemists were the first curious people who challenged their perception of reality counter to those who called themselves the repositories of real knowledge in the past. I’ve always been interested in the studies of people interactions on a large scale though it seems that today’s (not so) open minded scientists may consider it a waste of time at best and black magic at worst. But even my extremely minor forays into how people interact in increasingly large groups have allowed me over the years to with scary accuracy to predict the actions of entire countries merely because of my unscientific observations. What kinds of benefits could we using social science methods could be had in the future. I suspect that a thorough study of humans as invariably social creatures might even be able to cure the war mentality we all seem so predisposed to follow through an insanely overdeveloped survival instinct called greed.
Amy is right to emphasize the diversity of social sciences, which range from highly quantitative to fully qualitative, from experimental to prescriptive, and so forth. In contrast, many people writing critiques of social science generalize from small and oddly selected samples, just anecdotes really, and show no sense of this diversity.
As a sociologist who often collaborates with natural scientists, I find that the statistical toolkit and complex-data perspectives brought from social science add something new to many projects, opening possibilities for integrated analysis of physical and social-domain data — with results that (oddly) seem more acceptable to natural than social-science journals.
The notion of social science vs actual hard science is a difficult bridge to cross, with multiple gray areas so I would instead suggest that the dichotomy is indeed a false one, that there is no real conflict in terms of what constitutes ‘science’ or indeed a scientist. Moreover, I would instead embark on the thought that in the so-called hard sciences, many researchers suffer from the prejudice that they are being encroached on by other disciplines such as Anthropology and I suppose Psychology also. And while the study of science in the narrowest terms’ work may be superficially more ‘sexy’, that gives them no plinth on which to stand above others in such fields as economics. However, not only are the old guard of physics, biology and chemistry to blame, but the social scientists themselves. This is simply because they are perhaps jealous of the Einstein’s and theoretical physicists of the world and jump up and down saying “we are a science too”. And while I respect the endeavour of scientific discovery, the more attractive sciences are always going to win. Yet, it is perhaps a weakness of mine that if indeed sociology is a science, I see no clear drawing of the line in terms of what cannot be a science. Thus the word ‘science’ is only characterised as what exactly? utilisation of the scientific method? Linear, thoughtful and rational reasoning? Oxford Dictionary “the intellectual and practical activity encompassing the systematic study of the structure and behaviour of the physical and natural world through observation and experiment”. So the real problem is stretching Science to include social phenomena? If so, then this debate is rather synonymous to the definition of marriage. And whilst indeed social science is less sexy as some would find our gay brothers and sisters, it is no less valid. As a disclaimer I would like to say i am but a humble 16 year old student who studies mainly social science.
Wow, seriously? Physics is the most fundamental of all the sciences. Yes, it is also the most dependent upon mathematics, but that is because physics/chemistry/engineering/etc. are described by mathematics. But literally everything that exists can be described (perhaps not with present knowledge) by physics. Every chemical reaction, every biological occurrence, every neuron connection. Everything that goes into social science or any science can be boiled down to a physical occurrence. To swat it away as “more like math” is short-sighted at best.
It’s certainly possible to do science in complex environments, or we wouldn’t think of biology as a science. I suppose from the pov of a chemist or physicist they wonder, with all those variables in play, how you can ever be sure you’ve found a causal relationship.
I’ve never done research with the questionnaire, but I watched up close as other people tried to do so. It seems to me that collecting testimony as data is always going to be fraught. Any scientist is capable of faking data, but when the data can fake itself there is a problem. The element of subjectivity is so much higher.
Another problem the social sciences have is the case of the media reporting research conclusions like “time heals” (this was genuinely reported some years ago). Apparently if you just wait emotional upsets calm down. Science reporting is pernicious, social science reporting doubly so.
I think you overplayed your hand invoking the observer effect. It only applies to observing single subatomic particles. For lumps of matter you can see (even with a microscope) it doesn’t apply – it is statistically smeared because grams of matter contain septillions of atoms, let alone electrons. I think social scientists need to beware this sort of thing – if you are criticising one group for not understanding your work, you don’t help your case by showing that you don’t understand the group you are criticising. Macro-mechanics and quantum mechanics are entirely different beasts.
“The research I do is theoretically-grounded, connects research methods to that theory, makes observations using those methods, and then draws conclusions based on that evidence.”
I have no idea what social science theories are considered to be, but if predictions and test are not part of the methods and not meant to be part, the absence of ways to tell what is wrong means it will be the same as pattern search (stamp collection) and the theories are not scientific theories. We have plenty of testable evidence for how that works (doesn’t work), phrenology et cetera as for science, astrology et cetera as for social phenomena.
The outsider’s test is to look for different schools of ideas where one should expect one successful theory. AFAIK that is what social sciences fails on.
One could “opinionate” on other sciences, but one can also criticize science for what it is and what it isn’t. I assume those who are concerned with understanding science and protecting its quality (I’m both) are interested in criticism, and if they can engage the practitioners involved it becomes more constructive for all parties.
I’m interested in astrobiology, it is criticized not for testability (much) but for being superfluous (re planetary science, microbiology and what not). It is a valid concern.
At the same time science is the market of ideas, so external criticism means little. If social science is valuable, it will survive. Astrobiology is flourishing despite the valid criticism – it is valid but evidently erroneous.
Many social science disciplines are the most scientific ones of all. Gels? Blots? faked up irrelevant cell-based “systems”? Please.