Just like the execution of a program can be described in Machine Code, 1's and 0's or even as electricity, or any other layer of abstraction.
From a scientific perspective, it's much more important than that. Neuroscience has empirical evidence, theories, and is a science. Psychology has anecdotes, no testable, falsifiable theories, and, for lack of empirical evidence, cannot be a science.
> Basically, I think you're painting a false dichotomy.
So is the director of the NIMH:
http://www.newyorker.com/tech/elements/the-rats-of-n-i-m-h
Quote: "America’s psychiatrist-in-chief seemed to be reiterating what many had been saying all along: that psychiatry was a pseudoscience, unworthy of inclusion in the medical kingdom."
> The dream would be to be able to put all these levels of abstractions together, and have good mappings between them.
Yes, absolutely, and I think this will eventually happen within neuroscience. But focusing on the mind, something inaccessible to empirical investigation, is not the road to that goal.