I seriously warn against digging "natural vs. artificial" trenches any deeper than they already are.  These distinctions vary with culture, and they are as artificial as anything that they purport to contain. It's all natural or it's all artificial, but wherever you say otherwise is where you've dropped in the tiny miracle of "will." That is, as I've called it elsewhere, the second Big Bang when suddenly some things but not others get the chance to envision future possibilities to drive their artifice.

Computers are tools that we have built, just as dams are tools that badgers have built. I don't know who might call the badger dam unnatural.

Metaphors themselves are artifice by definition, and I don't know what a "more natural" metaphor might accomplish better than a "less natural" or "non-natural" metaphor.

No metaphor can lead us all the way through - it can maybe serve as a leverage for more intuitive understanding a phenomenon, but not as it's explanation.

Neurosciences are getting advanced enough (see e.g. Dehaene, Stanislas (2014). Consciousness and the Brain) to let us dig for real answers, not just metaphors or approximations.

Artificial systems, like computers, are often used as metaphors, but we tend to forget that most products of civilization were created to *complement*, not to *resemble* our natural qualities. For example, movie makers discovered over a century ago that they can flash images in quick sequence, to have them perceived as a moving picture by the amazed viewers. It took close to a century to find out that this exactly complements the way visual networks in the brain process their input - as a sequence of quick bursts of input.

The same with computers - they were not designed to resemble the brain, but to complement it, to help overcome e.g. the memory capacity limits, and slow speed of conscious processing. Therefore, it might be a mistake to try to use computers as a parallel for a brain. Before any external stimulus becomes conscious, it undergoes a heavy, parallel processing by the brain - and as of today, I do not see any  reasonable metaphor for this parallel processing system in the computer world. If anything, we might try to use an analogy of statistical analysis & random sampling to get some intuitions on how ambiguous, messy and imperfect input gets converted into a consistent (we can only see a single version of an ambiguous image at a given moment), sustainable (we can think of an object after it ceases to be visible) conscious representation of the external world.

As for natural metaphors: as long as we are unable to identify any other natural phenomenon capable of using language to communicate abstract ideas, we are at loss here - there might be nothing else in the world that we could use as a natural metaphor for how the brain works. We can use simple neural networks to better understand how small bits of unconscious processing work, but I seriously doubt we can find a good metaphor, or analogy, to how human brain works as a whole.

Thanks for your responses!

Hello Grzegorz. I agree that a metaphor cannot entirely explain a phenomenon. Nonetheless, even the terms you used to exemplify visual perception (“processing”, “input”, etc.) carry the influence of a computer metaphor. Furthermore, we’ve modelled the nervous system as a telephone operator, a magnetic field, a hologram, a network of tubes and pipes, and many more in the history of mental theorizing. Each metaphor instrumentally shaped the practices and explanations rendered by its contemporary scientists. So, I do believe our analogies are an important consideration.   

I also agree that technology is largely meant to complement cognition and behavior. However, I do not think that this point is central to the present discussion...as a system, complementary or fundamental to the brain, can act as a metaphor for the brain.

Hi Damian. I agree that the distinction between artificial and natural is trivial at a basic level. Your point is also an important one to add to this discussion, since we want to avoid falling into an artificial vs. natural system discussion for its own right. However, I do believe the distinction is meaningful in the context of this question. We borrow the basic principles, concepts, terminology, and methodologies of a theory or metaphor when we apply it to a phenomenon.

The difference between a computer explanation and beaver dam explanation is considerable. Building a theory around each requires a whole different kind of science (computer science vs ecology in the above example). In light of your comments, I will recast my question.

What system within the purview of natural science does the brain most closely resemble?

Brandon,

Thanks for your detailed comments.

> Furthermore, we’ve modelled the nervous system

I think this helps in narrowing down the question: do you mean *any brain* or *human brain*? Because, if you mean any brain, e.g. fish, amphibian, or especially the simplest invertebrate brains, then we might have, with a little success, managed to model these using the tools you mentioned.

If, however, we focus on human brain - or, to be more precise, on a brain capable of conscious processing (as we do not know with certainty that this is a pure human quality), then let me disagree. Science *attempted* to model such thing, but so far without any success - maybe because the prevalent approach has been to treat the brain as a pretty homogeneous computing machine.

We've learned to build computers using specialized subsystems for graphics, audio, etc. We have not yet, however, managed to make them "speak a common language" and understand each other easily - while this might just be one of milestones for modeling a brain in a machine.

Hence, I still insist that no today's computer / hologram / etc can be used as a reasonably accurate model of a complex brain. 

But, none of the above answers your question. I think no close metaphor of natural kind is available. A very remote metaphor might be, in my opinion, a large colony of bacteria in a changing environment. Each bacteria acts independently, resulting in a massive capacity for parallel processing (which is also a characteristic of the brain). At the same time, evolution provided several mechanisms that integrate the most prevalent reaction expressed by some part of the colony into a consistent (re)action of the colony as a whole - which very roughly resembles integrating ambiguous input into a single perception, which generally is consistent with the interpretation determined by majority of the pre-processing neural networks in the brain. If some part of the colony is subject to a toxic substance, and some other subject to a destructive light wavelength, they might both produce a common type of a "danger" signal - a bit like in a brain, where different stimulus modalities are translated into a message that can be integrated at higher processing levels.

Nah. I don't like it any better, Brandon. What's "natural" science?

Computers organize flows just like beaver dams do. Computers are part of human ecology, and dams are part of beaver ecology. I understand the inclination to make computers out to be non-natural, but again, seeking "naturalness" is just plain problematic.

Thanks again for your contributions, Damian.

The natural sciences are a group of disciplines that contains the life sciences (biology, medical science, etc.) and the physical sciences (physics, chemistry, etc.). There is a clear separation between the purview of the natural sciences and that of the computer sciences.

I did not intend on making any strong claims about the fundamental difference between natural and artificial kinds...though it is an interesting debate in its own right. This is, as you've pointed out, a philosophical quagmire. 

If the difference between natural and artificial is considered as a disciplinary distinction, as the difference between natural and computer sciences is, then it can be clearly determined. A beaver dam is certainly not a phenomenon that stems from the computer science tradition.

Grzegorz, thanks again for your comments.

Your example of a colony of bacteria is very interesting. I believe It highlights the important relationship between structural and functional descriptions of the brain.

Certainly, there must be similar properties of the dynamic interactions within networks of neurons and bacteria. Perhaps the spread of activation is similar to the dynamics of bacterial swarming (please note that I am no expert on this subject!).

But, as you've eluded to in your "danger signal," it is unclear whether a structural analogy of the brain might map onto functional description of cognition.