Want a truly stimulating, no-holds-barred career filled with fascinating projects? It’s the data scientist. What’s the data scientist? Well, the computer science guys would have us believe that it’s a completely new career, built on big digital data, requiring cutting-edge technology skills. That’s an amusing half-truth that leaves out other historically significant approaches to the career. Actually, the data scientist is not a new career and a lot more than technology understanding is required. But more about that later.
So according to the techies what are the requisite skills? Part hacker, part analyst and part communicator. And, according to an article by Harvard Business Review’s Davenport and Patil, no university has yet created a program for such folk. The most basic skills, they write, are the ability to write code, communicate in language that non-techies can understand and demonstrate their insights. That’s an intriguing set of competencies that are nearly nonexistent in a single person.
Career and job definition secrets
Here’s a very important secret about career definitions—and for that matter, definitions of any sort. Definitions are always slippery and career definitions tend to be underlain with power issues. The individual and group that gets there first gets to define the career. So you have lawyers, physicians, CPA’s, PhD’s and etc. with required education and credentials, based upon tests. After a while these professionals accept their role in the existing order of things because they see or imagine no alternative to it. Some see it as so natural and unchangeable they value it as divinely ordained and beneficial.
But all of these professional definitions raise a number of interesting questions: Who is benefiting from this definition? Whose interests are being served? Whose interests are being silenced? How did it come to pass that this is commonly accepted as true? What would happen if we rejected this truth claim and substituted another? Is there another way of looking at a profession? Does the profession already exist, but in another form?
Yesterday’s big data scientists
Davenport and Patil say that data scientists today are “akin to the Wall Street “quants” of the 1980s and 1990s. We can go back a lot further than that to define the big data scientist. Historically, big data was the province of the CIA, built up from surveillance, foreign newspapers and broadcasts and human contacts. The information varied in its reliability, not unlike business data. Currently, there are a dozen functional areas of information gathering by “analysts” as well as analytic methodologists who keep up to date and apply new methodologies to data. As the information scope has been digitized, the CIA now uses the terminology of “data scientist,” but that scientist accesses the expertise of its “analysts.” If you were to look at the background of that glut of data careers, you’d find physicians, economists, political scientists, the military, sociologists, psychologists, etc. The data scientist overlap into other disciplines makes a narrowly defined career significantly improbable. The dominant traits of these people are curiosity and associative thinking—and with that Davenport and Patil agree.
The role of communication
Davenport and Patil draw one interesting conclusion with which, having interacted with these people often, I agree fully. Although data scientists are imaged as technology people today, within five years their more enduring competency will be the ability to communicate “in language that all their stakeholders understand—and to demonstrate the special skills involved in storytelling with data, whether verbally, visually, or—ideally—both.” To a significant degree, communication expectations in such settings are very demanding, requiring an understanding of how to interact with differing audiences and individuals, how to create story and analogy formats, as well as extensive, even colorful vocabulary. These soft skills are far more difficult to learn than the “hard” IT skills. Indeed, many go into the IT field expecting not to have to communicate regularly. But, there you go.
I note that the HBR article identifies their career example as a physics PhD. That implies exceptionally well-rounded, liberal arts, scientific, pattern thinker with a graduate degree. So don’t take the “data scientist” expectations too seriously—other than to be well-rounded educationally, with a basic understanding of coding, math for algorithms, statistics and a lot of communications, rhetoric and literature to support the needs for translation.