# Problem Driven Epistemology Problem Driven Epistemology is in one way exactly what it sounds like: it is an [Epistemology](Epistemology.md) that is driven forward by [Problems](Problem.md). But in a sense it is far deeper than it first appears. [David Deutsch notes that the concept of a problem](https://nav.al/deutsch-files-ii) may be the most important concept that Popper introduced. It provides another way of understanding [Epistemology](Epistemology.md). And [Problems Create a Logic of Reasoning](Problems%20Create%20a%20Logic%20of%20Reasoning.md). All previous epistemologies assumed that knowledge is [Justified](Justification.md) true belief. This misconception arises due the fact that because we want to rely on knowledge, wherever it comes from must also be reliable too. But once you think of science and rational thought generally as being about problems, you lose the need to talk about where it comes from. The problem is simply there to be solved and the solution is what you want - not the justification of the solution by going back to first principles. This allows us to arrive at the momentous conclusion: [Explanations Are Justified By Their Superior Ability to Solve Problems They Address](Explanations%20Are%20Justified%20By%20Their%20Superior%20Ability%20to%20Solve%20Problems%20They%20Address.md). ## Is Problem Driven Epistemology Circular? Now you may very well ask: is this [Circular Reasoning](Circular%20Reasoning.md)? At first glance it appears that we are making two claims that have a sneaky interdependence between them: 1. Problem-solving is justified because it yields knowledge 2. The knowledge gained is justified due to it's origins of solving some problem While these claims are interconnected, a close examination reveals that the reasoning here is not actually circular. The thread that keeps us tied to the ground is that our problems are not *arbitrary*. They are chosen for a variety of reasons, but in general they are chosen because we wish to make genuine [Progress](Progress.md) in some way, shape or form. This is usually done via improving our understanding. Notice that the entire reason we wanted to [Justify](Justification.md) our [Knowledge](Knowledge.md) in the first place is because we want to be able to *rely* on it. Rely on it how? To solve *specific problems that matter to us*. These problems are connected to reality and our place in it. We can see what happens if we don't attempt to make progress in a connected way via a thought experiment. Let us imagine a society on a remote island where "knowledge" is valued, but its justification rests solely on a [Self-Referential](Self-Referential.md) loop between [Problem Solving](Problem%20Solving%20Process.md) and [Knowledge](Knowledge.md) claims. We can call this *The Island of Circular Reasoning*. In this society problems are posed without any grounding in a broader framework of understanding. They arise not from a genuine desire to comprehend the world or solve real issues, but rather as *arbitrary* puzzles or challenges designed to test the islanders' problem-solving abilities. Notice that this shares much in common with [Parochialism](Parochialism.md). Solutions are valued primarily for their ingenuity and elegance, regardless of their connection to reality or their potential for generating further insights. A solution might be celebrated for its mathematical sophistication or its logical consistency, even if it has no practical application or explanatory power. Knowledge is seen as a collection of solutions to these arbitrary problems. There's no emphasis on connecting these solutions to a broader understanding of the world, and the justification for accepting a knowledge claim rests solely on its ability to solve a particular problem. The justification for problem-solving itself is derived from the knowledge generated. Islanders are encouraged to solve problems because doing so is believed to produce valuable knowledge. However, since the knowledge is *only valued for its [Parochial](Parochialism.md) problem-solving capacity*, the justification becomes circular. And there is the weak link: the *knowledge is only valued for its problem-solving capacity*. It is not valued for any other other reason. Consider how knowledge works in the real world. It is valued for many things: to improve our lives, to improve our understanding of the world; most generally, to create [Progress](Progress.md). In a real sense this is linked to the fact that our problems are rooted in reality - they are not *arbitrary*! The issue with this approach can be summarized in two words: arbitrary and disconnected. The knowledge created has an arbitrary nature to it, because the problems were arbitrary. And the knowledge is disconnected from understanding the world because the problems were not connected to understanding the world. So while there is some "progress", it is only in the weak, [Parochial](Parochialism.md) sense of progress towards solving arbitrary problems and building the knowledge to do so. If we take a step back and think about how the actual problem solving process works, we can yet again use two words to describe what makes it so different from the island of circular reasoning: progress and understanding. We don't pick problems arbitrarily. We chose problems for all sorts of reasons, but they almost always revolve around the desire to make progress in some way that is connected through the world, and that requires we increase our understanding. Lets update our original claims that seemed to have an interdependence. We can make the implicit explicit: 1. Solving non arbitrary problems is justified because it yields non arbitrary knowledge growth 2. The non arbitrary knowledge gained is justified due to it's origins of solving some non arbitrary problem It is this "non arbitrariness" that gets us out of [Circular Reasoning](Circular%20Reasoning.md). For it means that there will be external progress. There will be change. We will move forward. We will not simply be stuck in place, static. Now if we just analyze the above sentence structurally, it may appear that it still has a general circular form: 1. A justifies B 2. B justifies A But that is really just a [Positive Feedback Loop](Positive%20Feedback%20Loop.md), a type of [Cyclical Process](Cyclical%20Process.md). We can see that while there is a relationship between the solving of non arbitrary problems and the non arbitrary knowledge created, meaningful progress is occurring along another dimension. We are moving forward in a way that is connected to the issues we face.  Now of course the same thing occurs in the case of arbitrary problems. However, it is not accurate to call that progress in the way that we have been referring to it. It is a far more parochial form of progress.  ## Better Language Removes Any Circularity It is worth noting that if we clean up our language here we can actually remove any sense of circularity before it is ever perceived[^1]. The entire circularity we just investigated centered around the *justification* of knowledge. But why did we want to justify our knowledge in the first place? It was because we wanted to use it to help us *navigate the world*! Notice that this already contains the non-arbitrariness that we just highlighted. The desire to justify knowledge did *not* come solving arbitrary toy problems. In that case the justification is self contained. For example, consider the problem of wanting my kitchen to be clean. The way to solve this is to clean my kitchen. Why was cleaning my kitchen the right choice? Because it was the exact way to solve my problem of having an uncleaned kitchen. There was effectively no ambiguity. And there was no new knowledge created. I already had it in my brain. I knew that based on the Second Law of Thermodynamics my kitchen was never got to reverse the process of getting dirty while dinner was being made, so the way to change its physical state to one of being clean was to put energy into the system and clean it myself. In the case of a toy problem, the way to solve it is to solve it. There is no ambiguity here because I am not hoping that it connects to the broader world in any way. There is no need to *justify* my solution, for the only criterion that I need it to satisfy is that it solves the toy problem! It does not need to hold under different *contexts* in the world, or generalize to different scenarios. I have removed the contextual complexity that *caused us to seek justification in the first place*! We can further reduce the potential circularity confusion if we embrace the definition of [Knowledge](Knowledge.md) put forth by David Deutsch. His definition of knowledge bakes in the *non-arbitrariness*, for arbitrary problems wouldn't reliably exist across the multiverse. ## Relationship to Evolution Problem driven epistemology has much in common with [Evolution](Evolution.md)[^2]. Both are processes of *variation* and *selection* where success is marked by the creation and survival of [Knowledge](Knowledge.md). Biological evolution, starts with a "[Problem](Problem.md)". This is an ecological [Niche](Niche.md), an environment with a specific set of conditions and resources. In other words it is a *context* that provides [Constraints](Constraints%20are%20Foundational.md). The "theories" are [Genes](Genes.md). These genes have varying levels of adaptation to the [Niche](Niche.md). New genes are "proposed" via random mutation. Criticism, or elimination of errors, comes in the form of [Natural Selection](Natural%20Selection%20is%20a%20Constraint.md), where individuals with genes that encoded more advantaged traits are more likely to survive and reproduce, passing these genes on to the next generation. [Knowledge](Knowledge.md) refers to the [Adaptations](Adapted.md) encoded in an organisms genes that enable it to thrive in its niche. Problem driven epistemology posits thats the starting point is a [Problem](Problem.md). Here, a problem arises when we have a set of clashing ideas, inadequate explanations. Variation occurs via *conjecture*. Selection occurs via *rational criticism* (of which experimental testing mode). It is in this way that we see: [There Is No Such Thing As Instruction From Without, There is Only Conjecture From Within](There%20Is%20No%20Such%20Thing%20As%20Instruction%20From%20Without,%20There%20is%20Only%20Conjecture%20From%20Within.md). In evolution, that conjecture comes in the form of random mutation. In problem solving it comes from conjecture in human minds. There is no instruction or guidance from any external source. Conjectures are selected and improved on based on how well they solve a [Problem](Problem.md). Problems are our guide. Problems allow for [Error Correction](Error%20Correction.md). --- Date: 20241109 Links to: [Poppers Theory of Knowledge](Poppers%20Theory%20of%20Knowledge.md) [Problem Solving Process](Problem%20Solving%20Process.md) [Forced To Take a Position](Forced%20To%20Take%20a%20Position.md) Tags: References: * [The Deutsch Files II](https://nav.al/deutsch-files-ii) [^1]: I have chosen not to do this so it can be used as an illustrative example of clear, reasoning thinking. [^2]: See more in [Evolutionary Epistemology](Evolutionary%20Epistemology.md)