While Glen’s answer captures many important aspects of what the scientific theory is and how the process works, it also misses many aspects.
I choose the example of a scientific theory being the theory that life evolved from inorganic constituents by a process of evolution by natural selection.
A recursive application of evolutionary theory gives us our best understanding of understanding itself, what we are, how and why we think and do as we do.
Evolution as a process starts simple, and rapidly gets extremely complex.
The start is simple enough.
It requires something that can replicate, with some useful degree of fidelity (ie close enough that it can persist, but not perfectly, such that there are variants in the populations that result).
It requires that there be differences in the environment, such that some variants do better in some locations than in others.
Everything that follows is about the differential survival of variants in different contexts.
That process rapidly gets very complex.
As the number of variants increase, then the levels of “strategy” start to happen (strategy in this sense, at this level, is not conscious, but just systems doing what they do). If the risk to individuals in a population comes largely from factors outside the population, then levels of cooperation may emerge and stabilize. Raw cooperation is always vulnerable to invasion by cheating strategies, so there emerges something of an evolutionary arms race between cheating strategies and detection and removal strategies. We see this at every level of biology. In the RNA systems within cells, in the protein systems within cells, in the protein systems on cell surfaces, in the immune systems of complex organisms, in our emotions and our legal and cultural systems (to name just a few).
Cooperation allows for expanding complexity, and the exploration of new systemic spaces of possibility.
Competition tends to drive systems to some set of local minima in the complexity landscape and keep them there.
It seems probable that we are the most complex and the most cooperative systems in existence.
It seems probable that at every level, our understandings and models of the world are driven by many levels of the differential survival of heuristics that worked in our past. All of our ability to make sense of the world, to model the world, to model ourselves and others as actors in the world, seem to be the result of this process of the differential survival of heuristics.
It seems that for every one of us, our experiential reality is a subconsciously created model of reality. That model is partly the result of our biology, partly the result of our “culture”, partly the result of our individual experience, and partly the result of our choices.
Heuristics are not “Truths”.
Heuristics are things that work in practice in particular contexts (at least as well as they do, which in an evolutionary context means well enough to survive better than the alternatives available – often simplicity has a lot going for it, response time is often important in survival contexts – being just a little faster to respond than your neighbor means you are less likely to get eaten, as one example).
For each of us, the heuristics we have worked well enough for our ancestors and those from whom we copied them, that they managed to survive (and survival is not a trivial thing).
That our heuristics worked as well as they did in our deep past, is no guarantee that they will work in our exponentially changing present.
A very simple heuristic is that all things can be resolved down to “True” or “False”, and can be understood in such simple terms.
That is a useful starting place for the exploration of complex systems, it is not any sort of “Truth”.
Reality seems to be far more complex than that. Many levels more complex. Reality seems to demand a more probabilistic understanding.
Reality seems to be a balance between order and chaos, between the knowable and the unknowable – at every level – recursively, potentially infinitely so.
Glen mentioned the post modernists. They seem to have gotten half of a useful idea.
They have understood the idea that all things contain uncertainties, but they have taken that to mean that all things are equally uncertain.
That is not a valid extrapolation; it is not supported by the evidence from biology or systems theory.
Every level of complexity and existence seems to have a requirement for a minimum degree of order required to sustain the boundary conditions that make that level of complexity possible, and sufficient permeability in those boundaries to allow the exploration of the infinities that lie beyond the boundaries, both for the threats and the opportunities contained therein.
There can be no absolute guarantees of security in such a systemic space; and we can learn a lot from our explorations of the past and present (including all of biology, ecosystems, human history, culture, literature etc), and our explorations of the systemic spaces of complexity, algorithms, mathematics and logics.
Hierarchies are a useful way of ordering systems with varying degrees of competency; and they can easily be captured and exploited by cheating strategies on the cooperative entities involved. Hierarchies themselves are not the problem. The problems we have are in the nature of the systems present in specific hierarchies. Those systems must be fundamentally cooperative – all levels, if any of us (even them) are to have a reasonable probability of long term survival.
We are all fundamentally cooperative, and we all have our competitive sides, that can emerge if the context demands it. Our survival seems to be predicated on maintaining contexts that promote universal cooperation. That means delivering reasonable levels of universal abundance and universal freedom, and both of those come with requirements that we each, as individuals, act responsibly in both social and ecological contexts.
When one looks deeply at the systemic structure of the life that we are, and the life that surrounds us, the need for such fundamental cooperation becomes clear for any who take the time to look (and it does take a while – I’ve been doing it for over 50 years).