More than fifty years ago, the English novelist and physical chemist C.P. Snow gave his classic lecture The Two Cultures at the University of Cambridge, describing the intellectual and cultural gulf that had already opened between the sciences and the humanities.

Both the sciences and the humanities provide valuable tools to understand the world, and each toolkit is incomplete without the other. Snow lamented the scientific and technical ignorance of humanists and the humanistic ignorance of scientists, as major obstacles to human advancement.

Decades later, the gulf between the sciences and the humanities remained, perhaps becoming even wider. However, synthesis between the two cultures was so valuable, those at the crossroads found remarkable opportunities.

In the 1970s, Steve Jobs created his company Apple in a way that deliberately blended outstanding engineering with extraordinary design. Jobs attributed Apple’s success to this marriage of the liberal arts with computing.

The modern discipline of data science is a trendy buzzword for an emerging field. But data science simply means the art and science of data analysis. And when done well, it stands firmly at the nexus between the two cultures, the sciences as well as the humanities, drawing extensively from both.

How Data Science Blends the Two Cultures

Often, data science is simply described as an engineering field. It is frequently folded into computer science and engineering departments at universities and into engineering teams at companies. While engineering is a major component of data science, calling data science engineering erases essential parts.

For example, a major tool of data science is machine learning, the use of algorithms to make predictions or decisions based on existing data. Of course, this includes engineering problems. Implementing the algorithms falls clearly within the discipline of engineering.

So does the enabling of the algorithms, which are often used to analyze Big Data, to run efficiently on a powerful computer cluster. So does the processing of the data into the correct format, which is often the most time-consuming and resource-consuming step. And so does the testing of the process to ensure it runs smoothly.

But ultimately, the data must come from somewhere, and its analysis must be based on some type of premise. In fact, the quality of the data set often matters more than the quality of the algorithm. There is even a term for this — “Garbage in, garbage out.” In other words, a faulty data set or a specious premise behind the analysis will yield nonsense results, and no algorithm can avert this.

Obtaining the data set and deciding how it should be analyzed is a qualitative, not quantitative, endeavor. Rather than being an engineering problem, it belongs to the domain of the humanities.

It involves an understanding, in descriptive rather than quantitative terms, of the data set’s characteristics. If the data is of humans, it involves understanding them qualitatively. And intuitive logic is one’s guide to how the data should be analyzed.

The Two Cultures on a Data Science Team

The larger the organization, the more siloed teams can become. However, even in the largest organizations, the same people may be working on the scientific, engineering, and humanistic aspects of a Data Science analysis.

Or, people in each of the two cultures may be working very closely together. This is because of the need for frequent and clear communication, across the entire team, about the details of the data and the process of analysis.

So, not only is Data Science for people who are adept in engineering and the sciences, but it is also for people who are well versed in the social sciences and humanities. If Data Science is acknowledged to be as much a part of the social sciences and humanities as it is a part of science and engineering, the development of the field, its education, and its training may go a long way toward bridging the gap between the two cultures.