Lucile Pommier
A recent study supports the theory of the Hierarchy of Sciences. It classifies disciplines according to their complexity, from the hardest sciences to the softest. Although often considered politically incorrect, in the long term, comparing disciplines could change the way we do research and perceive the sciences.
Cet article existe également en français : La hiérarchie des disciplines pour une nouvelle conception des sciences
“Hard” sciences, “soft” sciences… Who has never put quotation marks around these two words? A recently published study available on MyScienceWork might help us stop tiptoeing around the matter. Researchers analyzed the bibliometric parameters of various scientific disciplines in order to create an objective classification of the sciences. In the end, the theory of the Hierarchy of Sciences, which classifies disciplines according to their complexity, made the final cut. Their results could have significant implications for the world of research.
Classifying disciplines along a scale of complexity
According to the theory of the Hierarchy of Sciences, disciplines are harder or softer according to their complexity. The more parameters and interactions are involved, the more complex a given system is, and the softer a particular science becomes. And the softer a science is, the harder it becomes for us to study this system. According to these criteria, human societies, for instance, represent a much more difficult subject to study and from which to draw meaningful conclusions than subatomic particles.
To evaluate this scale of complexity on an empirical basis, the researchers analyzed various bibliometric parameters for more than 29,000 recent publications covering 12 different disciplines: number of authors, age of references cited, diversity of sources, etc. These parameters indicate a given science’s complexity. For example, the harder (and less complex) a discipline is, the more scientific consensus is found in the field, due to the considerable theoretical and methodological foundations that have been established. This consensus corresponds to the number of co-authors and research teams that participated in the study yielding the article.
From softer sciences to harder sciences
Their results showed that there was no dichotomy between natural sciences and social sciences, contrary to the widely held view. There is, instead, a continuous progression from the hardest to the softest. Biology sits in the middle. Some of its sub-disciplines, such as biochemistry, are closer to the hard sciences than others, like zoology.
Bibliometric results of scientific publications according to discipline. (m= mathematics, p=physics, b=biology, s=social sciences, h=humanities).
Sources: D. Fanelli, W. Glänzel
This study gives another response to the recent debate started by a blogger from Scientific American: “Is psychology a science?” According to Daniele Fanelli, an evolutionary biologist who studies scientific bias, and coauthor of the paper, "Any empirical effort can be a science. Psychology, in a sense, is dealing with the highest complexity of all: brains studying brains. For this reason, psychology is the most difficult science."
More work needs to be done, however, to complete these results, as a bibliometric study cannot embrace all of the cultural or philosophical assumptions surrounding research and sciences.
Seeing scientific disciplines in a new light
Without introducing any value judgment between disciplines, these results show that different sciences are not carried out in the same way. A soft science might need more time and more vigilance before a result is validated. "The chance of getting a false positive [in softer sciences] is much higher. This is probably what people don’t like to face," states Daniele Fanelli. Pushed to publish new results in the same timeframe as scientists in other disciplines, American researchers in social sciences have inadvertently submitted more false results. If these differences are not taken into account, phenomena such as this, which Fanelli calls the US effect, might grow out of proportion.
The hierarchy of sciences, highlighted by this Scottish study, encourages a new vision of research and researcher evaluation according to the needs of each discipline. For example, it could call for attributing more funding to replicating the results of the softest disciplines. Media coverage should also be adapted according to the field in question, with more precaution taken with new research in soft sciences. Understanding these differences could be “part of recipe for a better sort of science," Daniele Fanelli concludes.
