This article has a few flaws.<p>First, I disagree with the definition of parallelism. So the whole argument becomes moot to begin with.<p>Second, the author conveniently ignores that languages with side-effects can also be automatically parallelized by their compiler, which achieves the same effect as the parallelism described for functional languages. Which makes the point of the article moot again.<p>In conclusion (to me at least), this article adds to the confusion rather than offer the clarity it was aiming for.