The article is generally imprecise, and I completely agree with you. Stating that `... data is inherently dynamic` in this context is plain and deductively wron

The article is generally imprecise, and I completely agree with you. Stating that `... data is inherently dynamic` in this context is plain and deductively wrong. If we can not add type to data, we can't even interpret it as a bit-string because, hey, it's a type! (the type of a string of 1's and 0's) Types are mostly used to reason about a program internally. But sometimes we need to consider the real world. When doing do we need to interpret data. In the Haskell-sense this means assigning types to it. A language like Elm has exemplified how the internal / external data barrier can be implemented for strictly typed languages. Haskell just considers everything a string and lets it be up to the developer to parse it.