Data Scientist or Data Humanist?

By Bill McLellan, Senior Consultant, BI Solution Architect

How? you might ask did a humanist get into data? I didn't lie on my resume, but it’s true I didn't start out wanting to do data when I grew up. As a kid, everyone including me thought lawyer. I had a knack for incessantly asking questions and arguing "to the pain" but also frequently changing my mind. And then there was the infamy in 5th grade of reading encyclopedias in the bathroom. Today it's Wikipedia instead of World Book (remember the 90s?), but little else has changed. In making the switch from the humanities to data science I've never felt like a fish out of water.

Two words sum up the less-than-obvious link in my mind: logic and meaning. Or, as they're called in philosophical and computational linguistics, I'm obsessed with syntax and semantics.

Here's my completely (un) reasonable career path:

At Covenant College (of course, a liberal arts school) I double majored in philosophy and history. After graduating I worked jobs in sales and then marketing. Quickly, I gravitated away from sales toward marketing analytics. I found ways to analyze customer and financial data, clarify thinking, standardize technical vocabulary, and create reports that helped decision makers understand what was really going on in their business.

During my early employment in analytics, I also pursued an advanced degree in theology (a humanity requiring no programming skills, believe it or not 😉). But since pastoring after the financial crash didn't make a living, I continued to surprise myself with employable data skills. After graduating seminary, instead of pursuing pastoral ministry I used math, data, and programing skills learned on-the-job by day and in-the-book by night to make my final transition from humanities to science and technology.

Data science, business intelligence, and analytics are all about using formal language and visualizations to model the patterns, anomalies, artifacts and endeavors of human endeavors. Data scientists create and interpret data-texts with the goal of greater a) knowledge, and b) power. If I had to pick any goal that motivates humanists, scientists, and business people alike, whether or not they've read any Foucault, power-knowledge would be it.

Humanity has already developed computers great at logic-syntax-power, but if we're going to advance in teaching machines to understand meaning-semantics-knowledge, we're going to need more artists and humanists who make the shift to science and technology. People with backgrounds in the liberal arts, who also learn to code, design software systems, and lead data projects, just get the following crucial concepts:

1. Datasets -- no matter how rectangular – are also texts usually written for one purpose but useful for many others
2. Sometimes people don't mean what they mean to mean
3. The past has not always resembled its past but has always flowed out of it
4. You usually have to look behind documentation to see what's really going on
5. Accusing people of logical fallacies fails rhetorically every time: we make decent teammates, too

It's been said that scientists study what always happens whereas humanists study (and artists create) what only ever happens once. Each of these two disciplines, the humanities and the sciences, matters in data science and business intelligence, and more than a few lucky souls like me don't have to pick just one.

‍