The software/hardware metaphor is usually considered as misleading as every other model of mind we've come up with.
I don't agree. My guess is it's an unusually good model -- one rooted in the physics of computation. Anything sufficiently complex can compute, which is, souls aside, the same as running a mind...
... in an alternative abstract universe closely related to the one described by the Navier-Stokes equations, it is possible for a body of fluid to form a sort of computer, which can build a self-replicating fluid robot ...
... A central insight of computer science is that, whenever a physical phenomenon is complex enough, it should be possible to use it to build a universal computer ...
Our minds have emerged to run on our desperately hacked and half-broken brains - in hundreds if not tens of thousands of years. In evolutionary terms that's insanely fast (and did it really never happen before?). Minds route around damage and adapt, as much as they can, to both adolescent transformation and adult senescence; they run and run until they slowly fade like a degraded hologram. It's no wonder minds are so diverse.
When that diversity intersects with the peculiar demands of our technocentric world we get "Traits that Reduce Relative Economic Productivity" -- and we get poverty and suffering. We get disease, and so we need names.
We need names because our minds can't reason with pure patterns -- we're not that smart. With names we can do studies, make predictions, select and test treatments.
Names are treacherous though. Once our minds create a category, it frames our thinking. We choose a path, and it becomes the only path. It might be a good path for a time, but eventually we have to start over. Over the past ten years researchers and psychiatrists have realized that our old "DSM" categories are obsolete.
So how could we start over? One approach, informed by the history of early 20th century medicine, is to classify disorders by underlying physiology. That's where terms like 'connectopathy' come from, and why we try to define mind disorders by gene patterns.
We need to do that, but lately I've wondered if it's the wrong direction. If minds really are somewhat independent of the substrate brain, then we may find that disorders of the substrate only loosely predict the outcomes of the mind. Very similar physiological disorders, for example, might produce disabling delusions in one mind and mere idiosyncrasies in another.
So maybe we need another way to attach labels to patterns of mind. One way to do this would be to create a catalogue of testable traits for things like belief-persistence, anxiety-response, digit-span, trauma-persistence, novelty-seeking, obsessiveness, pattern-formation and the like. My guess is that we could identify 25-50 that would span traits that are currently loosely associated with both normal variation and TRREPs like low IQ, schizophrenia, and autism. Run those tests a range of humanity, then do cluster analysis and name the clusters.
Then start from there.
See also:
- Be the Best You can Be: Rebooting pyschiatry: time for a new set of disorders (2013)
- Gordon's Notes: The evolutionary history of minds (2008)
- Be the Best You can Be: Redefining mental retardation and autism - the revolution ahead (2006)
- Be the Best You can Be: Very annoying: the diagnostic criteria for Asperger's are not all that useful (2005) - this was part of the body of work that led to retiring the diagnosis of "Asperger's Syndrome"