Sciences and Humanities: Moving on from the ‘Two Cultures’

by Jeroen Bouterse

It is a commonplace to say that a divide has occurred in modern academia between the sciences and the humanities. In the anglophone world, this diagnosis is often traced back to a lecture by the British scientist-novelist Charles Snow, who pointed out in 1959 what he saw as a lamentable gap between ‘two cultures’: the literary and the scientific culture. Snow’s Rede lecture has become the main point of reference for later commentators, who often sigh in frustration that in spite of Snow’s warnings, the divide has deepened or widened.

That we have grown so used to the ‘Two Cultures’ framework is unfortunate, however, for multiple reasons. For one, Snow’s lecture wasn’t about the sciences and the humanities. (He never even uses the term ‘humanities’ in the Rede lecture.) His worries were about literature, about certain writers who got their views on ethics and literature all wrong; not so much about liberal arts or humanistic scholarship. That’s not to say that literature and the humanities are unrelated, of course; but they are not always the same thing either, which is why Snow has little to offer us by way of explanation of the sciences-humanities divide. That, in fact, is a second reason why Snow is a less-than-ideal key witness: there is a lot of lamentation and exhortation in his lecture, and very little definition and analysis.

A third reason, and I would say the most important one, is that whatever the virtues and shortcomings of Snow’s model, the omnipresence of the ‘two cultures’ framework comes at the cost of a richer historical perspective. (That is a typical humanistic concern, of course.) People did think seriously about the relationship between the sciences and the humanities before and after Snow, and collapsing all the results of that thinking into the category of the ‘two cultures’ means giving yourself over to an unselfconscious cliché about the modern intellectual landscape.

When I say that people thought about this before Snow, I am primarily thinking of the decades around 1900, and I am thinking in particular of several German philosophers, who took some time to argue that there was a real and deep division between two parts of modern ‘science’ (the German term suggests the natural sciences not as much to the exclusion of other sciences as does the English). Kantian thinkers such as Wilhelm Windelband or Heinrich Rickert saw this distinction in a mostly positive light. Some of their arguments withstand scrutiny better than others; all of them are worth considering simply for the fact that they explicate intuitions that inform our current thinking about the sciences and humanities even now.

Their arguments were a response to the ‘unity of science’-movement of the century: positivism. The idea was that the accomplishments of the natural sciences in the 19th century were due to their successful search for laws, and that the study of the human world had something to gain from imitating this way of working. John Stuart Mill devoted a large portion of his System of Logic to showing that there was nothing about human psychology, society or history that precluded systematic knowledge in the form of law-like regularities. The ‘moral sciences’ (which in the German translation would become Geisteswissenschaften, or humanities) could be sciences like any other.

Towards the end of the nineteenth century, newly emerging disciplines such as psychology and sociology actually went to work, searching for the patterns and laws that governed human life. Some enthusiastically proclaimed that a scientific psychology could serve as the foundation for all the humanities, in the same way that physics seemed to be fundamental to all the natural sciences.

All this made some humanists and philosophers uncomfortable. In 1894, the Kantian philosopher Wilhelm Windelband decided to address the issue in the philosophical manner: by adding a distinction. Empirical science, he said, can do two things: it can try to generalize and search after regularities and laws, or it can refrain from doing that and satisfy itself with concrete description of particular phenomena. Or, to use Windelband’s own terminology: it can be nomothetic or idiographic.

Now, with regard to any object, either mode is legitimate in principle: we can strive after general laws of human psychology. This would be the natural-scientific mode of working. Windelband’s point is now that this mode cannot, does not, should not enter in competition with the idiographic sciences, which try to paint a picture of human life in its uniqueness. There is value in the concrete, in the particular, in the once-only-ness of human history and human works of art. However awesome and shiny the regularities are that nomothetic sciences may uncover, our knowledge of those regularities alone is never going to cut it. On the contrary: Windelband seems to suspect a trade-off between the level of abstraction from concrete events, acts and creations, and our appreciation of their value. This is not a problem when we talk about uniform physical entities such as the “colorless and mute world of atoms”, but it is a problem when we want to describe the richness of human lives – our descriptions simply cannot be abstract and concrete at the same time.

Now, your head may be exploding with objections to this reasoning – and in retrospect, Windelband’s claim that the essence of the humanities is not to search for patterns may have done more harm than good to their image. In his own time, one problem was that nineteenth-century philosophers liked their science objective as well as empirical. That is to say: scientific claims ought to digest their empirical input in a non-arbitrary way. The positivistic ideal met this criterion: it said that the more general a law, the better. But by what measure, if any, can we assess the success of an idiographic science?

Addressing this question, Heinrich Rickert developed the distinction between nomothetic and idiographic sciences into one between natural and cultural sciences. Natural sciences deal with natural, unintentional processes. By their very definition, these have no meaning – they don’t have authors that intended them to signify, represent or embody anything. Our scientific interest in these things is general: we want to understand as much of them as possible in as broad a sweep as possible.

Cultural sciences, on the other hand, deal with more-or-less conscious creations of the human mind, with intentional acts or artefacts. These objects embody values, and they are understood best as such: as value-laden creations, not as cases of a more general pattern. Rickert isn’t prone to giving concrete examples, but we can think of him as saying, for instance, that someone who compares word frequencies in Shakespeare to those in other authors isn’t so much wrong as missing the point. The essence, after all, is understanding a work of literature as something of cultural value.

Is that science, though? It is an empirical endeavor, obviously; but is it objective in any sense? According to Rickert, it can be if we get our philosophy of value in order. In that case, we will have a clear view of what is worth our while in human history, art, literature, and of what isn’t.

Again, this move may curl your toes – and again, other scholars came up with different solutions. Still, Windelband and Rickert have something to teach us about the divide between the sciences and the humanities. Most importantly, that this divide need not be one between science and non-science. Our Kantian philosophers are uncompromisingly committed to the idea that the humanities (or cultural sciences) are and ought to be empirical disciplines, and that they ought to be objective. You can expand the list at will, and they would often have agreed: science needs to be rational, factually accurate, methodical, et cetera. Whatever ideals flow from the ideal of empirical science, they apply to the humanities as well as to the natural sciences.

However, they add: on reflection, some features that we intuitively associate with science make sense only relative to background assumptions about the aims of particular disciplines. It turns out that the positivistic ideal of ever-more-general laws does not cover all our possible knowledge aims. Sometimes, we want to know something specific or concrete. Well, then, in those cases we shouldn’t assess the success of an academic discipline by a measure borrowed from another one.

Windelband and Rickert overstate their case, suggesting and attempting to sanction philosophically a kind of in-principle incommensurability between the natural sciences and the humanities that simply isn’t there. But interesting things happen along the way, which deserve a place in current thinking about the sciences-humanities divide.

Taking Snow and the ‘Two Cultures’-narrative as our only guide to this divide will leave us exasperated about the inertia of the so-called Second Culture, wondering what could possibly be the matter with these humanists: why aren’t they all swiftly rearranging their disciplines to look more like real science? Reading almost any other thinker than Snow will give us at least a sense of what could be at stake: of why people could conceivably think that there is something to protect from a uniform view of scientific goals and methods. And, hopefully, it will remind us that those goals and methods have changed over time, in the sciences as well as in the humanities.


Rickert, H.J., Kulturwissenschaft und Naturwissenschaft: ein Vortrag (Freiburg 1899).

Windelband, W., Geschichte und Naturwissenschaft (Strassburg 1894).