A common criticism of science is that it must make foundational assumptions that have to be taken on faith. It is, the critic asserts, just one world view among other, equally “valid”, world views that are based on different starting assumptions. Thus, the critic declares, science adopts naturalism as an axiom of faith, whereas a religious view is more complete in that it also allows for supernaturalism.
This argument assumes a linear view of knowledge, in which one starts with basic assumptions and builds on them using reason and evidence. The fundamentals of logic, for example, are part of the basic assumptions, and these cannot be further justified, but are simply the starting points of the system.
Under scientism this view is wrong. Instead, all knowledge should be regarded as a web of inter-related ideas, that are adopted in order that the overall web best models the world that we experience through sense data.
Any part of this web of ideas can be examined and replaced, if replacing it improves the overall match to reality. Even basic axioms of maths and logic can be evaluated, and thus they are ultimately accepted for empirical reasons, namely that they model the real world.
This view of knowledge was promoted by the Vienna Circle philosophers such as Otto Neurath, who gave the metaphor of knowledge being a raft floating at sea, where any part of it may be replaced. As worded by Quine:
We are like sailors who on the open sea must reconstruct their ship but are never able to start afresh from the bottom. Where a beam is taken away a new one must at once be put there, and for this the rest of the ship is used as support. In this way, by using the old beams and driftwood the ship can be shaped entirely anew, but only by gradual reconstruction.
In this scheme, doctrines such as naturalism are the result of adjusting the web to match reality, rather than being foundational axioms that cannot be questioned. Any idea within the web can be questioned; one does that simply by substituting its negation into the web, and seeing whether that improves or reduces the web’s explanatory and predictive power.
You cannot throw everything up in the air and reconsider the whole ensemble at once — you would have no footing to do so — but you can, piece by piece, re-consider any part of the web and so gradually adjust the web towards a better fit to empirical reality. And that is exactly what science does.
I like the web metaphor too, but that’s all it is: a metaphor.
That got me thinking, and I concluded that, no, the web is not just a metaphor, it is real. I argue here that this “web of ideas” can be identified with the web of neural connections that each of us have in our brains.
First point: where are our ideas encoded? Let’s take basic ideas such as 1 + 1 = 2, or the axiom of logic modus tollens. They are encoded in the neural network in our brains. They may be written down on paper, but symbols on paper only have meaning when interpreted by our brains.
Second, consider how brains first evolved. They would have started, deep in our evolutionary past, as simple logic-gate circuits, doing things like: “if light-sensitive cell A is active, then fire up cilium B to move in that direction”. As animals evolved the processing of information and the decision-making got more complicated, to the extent that humans now have a network of 1011 neurons with about 1014 neural connections.
All the ideas we use to reason and make decisions must be encoded in that web of neural connections — there is no other place. Thus, to the extent that we use maths, logic, critical thinking, and to the extent that we interpret information and then use reasoning to come to conclusions, all of that is in the web.
The patterns in the web will result from a mixture of: (1) evolutionary programming, now encoded by natural selection into the genes; (2) the development of our brain as a foetus and child, though the interaction of the genetic recipe with its environment, and (3) sensory information received over our lifetimes, and interpreted by the learning mechanisms resulting from (1) and (2).
Our world view, the set of ideas we use to interpret information and to form opinions, is thus encoded in that neural web. Quite literally, our “web of ideas” is a real, physical web, located within our skulls. If you were to employ certain principles of logic and reasoning to arrive at a conclusion, what is actually happening is that electrical and chemical signals are whizzing around in particular patterns within that web.
Evolution will have programmed the web to be a pretty good match to reality, since the whole point of a brain is to make decisions that help propagate the genes, and a brain will be best at that if, overall, its processes are a pretty good model of the external world.
For that reason, natural selection will have continually adjusted the web’s recipe to make it better match reality (at least, the range of reality that we would have experienced in everyday life over evolutionary history). What science tries to do today, adjusting the web of ideas to best match empirical reality, is exactly what natural selection will have done by trial and error over evolutionary time.
Indeed, when explaining evolution in The Selfish Gene, Richard Dawkins used a metaphor akin to Neurath’s raft. Dawkins compared our set of genes to a rowing crew, each member collaborating and contributing to the team effort. A coach could swap out any crew member in order to improve the overall result. No crew member is an unquestionable “axiom”.
Hence, what we regard as basic truths of logic and other “a priori” truths that seem obvious to us, will have been produced by evolution precisely because they model the world we live in. We can thus ask, are any beliefs foundational, in that we cannot challenge them? From evolution’s point of view, no there aren’t. The recipe for the web is under genetic control, and evolution just makes random mutations to the recipe and tries them out, seeing whether they produce better or worse results. Evolution can swap out any member of the rowing crew.
Any basic belief might be replaced by a mutation turning into the negation of that belief, and if that worked better then it would tend to result in more descendants and so would prosper. Otto Neurath’s idea that science can challenge and test any and all aspects of the web of ideas, swapping alternatives in and out, is exactly what evolution will have been doing over eons.
By the time we get to adulthood our brains will have become more fixed, and it may be that by adulthood many of our beliefs are — quite literally — hardwired into us. Thus we might find it near impossible to even imagine alternatives to, say, modus tollens, or to ideas such as 1 + 1 = 2. But those beliefs will still have been arrived at as a result of the reality-driven genetic recipe and the empirical environment in which we developed and grew up. Indeed, the whole evolutionary point of the brain is to produce a model of the local environment for use as a real-time decision maker.
Whenever we conceive of a “foundational axiom”, we can question it. All we have to do is ask what the consequences would be if it didn’t hold. To do that we substitute the negation of that axiom into the web, and see how the modified web does for explanatory and predictive power. Most likely, replacing basic axioms with their negation would produce nonsense, or all possible outputs, or something else equally useless. The explanatory and predictive power of the web would likely fall apart. If that happens then we have verified the original axiom as matching empirical reality. Absolutely nothing about the web of ideas is unquestionable — save only, perhaps, if we lack even the imagination to question it.
There is one way in which identifying science’s “web of ideas” with someone’s brain breaks down. Science is a cooperation of hundreds of thousands of people, and any one person can know only a tiny fraction of today’s science. Thus we need to extend the picture to a web of webs, an interacting network of people’s brains, each exchanging information with others in the web. (Similarly, one should regard evolution as operating on the gene pool of the species, not just the genes in any individual.) This extension, though, only reinforces the evolutionary explanation of the origin of our basic ideas.
The above view roots a “metaphysical” view about epistemology, about how we know things, in the scientific understanding of where we humans came from. It thus aligns the epistemology firmly with scientific truth. The web-of-ideas is not “just a metaphor”, it is firm reality existing physically within each of us, having originated in our evolutionary past. And that consilience gives confidence that it is the right idea. In essence, our scientific understanding now settles our metaphysics. But then to a scientismist those were never separate domains anyhow.