No but really, what is Data Science?
This piece by Xiao-Li Meng is probably one of the better responses I’ve read so far to the all-too-familiar question of “what is data science?”
The TL;DR:
- data science is not just the field of machine learning or just the field of statistics
- data science is not only about predicting things
- data science is not only about analysis
- data science does not exclusively live in STEM fields
- data science is not a single discipline
- data science is an ecosystem of disciplines, each of them self contained but influencing and interacting with each other.
Meng compares the ecosystem of Data Science to that of “science”, “social science”, or “humanities”. No one would expect someone who works in science or the humanities to know about and excel at all the sub-disciplines contained under that banner, but the fields share commonalities and influence each other.
As a side note, I wonder if the chronic imposter syndrome in data science partially stems from its ecosystem-ness; virtually everyone in data science has at most a handful of disciplines they specialize in, but we all know there are entire fields of study we lack deep experience in. When someone says “I’m a data scientist”, people don’t automatically respond “oh cool, what kind?” but perhaps we should.
It’s a wide definition, and he wrestles with the name Data Science as being not quite right for this big tent under which we’re all working, which I agree with. But I think it’s a more productive model for thinking about who we should be listening to and learning from, and how we situate ourselves in this greater movement. Gatekeeping and useless distinction make the conversation smaller, and while corporate data science teams are driving towards standard, narrower definitions for their titles (in many cases for good reason), failing to engage the whole variety of disciplines in the field may cause us to miss out on the best solutions.
