A Goal-Oriented Approach to Philosophy

This is a paper I wrote for an Epistemology course at Harvard in 2018. I think it has some important ideas, so I decided to share it here where it is more easily accessible.

Why do we do philosophy? At a very high level, I’m sure many of us have an idea as to why we study philosophy. We care about why, because, if we didn’t, we might be worried that philosophy will not help us to accomplish what we want to accomplish in our lives. This paper argues that the fundamental question of “what do we want from this?” is missing from many philosophical inquiries, and defining answers to this question may lead to faster and more meaningful results. First, I will explain how to use a goal-oriented approach to solving philosophical problems. Then, I will describe how to use this approach in epistemology and propose naïve goals for the field of epistemology. I will conclude by discussing particular issues that a goal-oriented approach will help to diagnose and fix.

Often when doing philosophy, we discuss a particular solution to a problem and the strengths and flaws that such a solution may have. Using a goal-oriented approach to philosophy is simply defining the criteria to determine a “flaw” or “strength” before coming up with a solution. For example, let’s say that a philosophical issue is to define the term “multiplication” and what it means to “multiply” two natural numbers. Without using a goal-oriented approach, we may discuss a solution such as “putting the two numbers next to each other”, so that 2 * 6 = 26 and 20 * 18 = 2018, and note that it has a particular flaw of always overestimating our intuitive definition of what it means to “multiply” two numbers together, but that it does have the strength of always increasing as its inputs increase, a property we hope a better solution to have as well. With a goal-oriented approach, we define exactly what we want from a solution. In this case, when we “multiply” a number A by another number B, we want the result to be the same result as if there were A groups, where each group consists of B objects, and if we were to count all of the objects in all of the groups, we would arrive at A * B. According to this criteria, the solution of “putting the two numbers next to each other” is obviously wrong, as it does not accomplish our goal of being able to count up groups of objects. When we then come across the definition of multiplication which states that A * B can be defined as B added to itself A times, we can easily establish that that definition is a solution because it accomplishes our goal.

Let us now look at how this process may be used in epistemology. Epistemology is generally known as the study of knowledge. Epistemology seeks to answer two fundamental questions: “what is knowledge?” and “what is justification?” It seems that if a flawless answer can be given to both of these questions, we will have found at least one solution to the field of epistemology. But what exactly is it that we want from such a solution? I propose a goal for each of these fundamental questions. With regards to the knowledge question, if presented with a definition of knowledge, I want to be able to use that definition of knowledge in order to determine whether my beliefs or others’ beliefs are certainly true. For justification, I would like a definition which allows me to predict when a person is likely to be correct or incorrect based on the information available to that person. Note that my goal for a definition of justification is less specific, since it includes the word “likely,” which is not very well-defined. However, each of these goals for what I expect from solutions allow us to identify whether a particular response to the fundamental questions of epistemology is a solution.

It is worrying that the goal of epistemology is not agreed upon by all epistemologists, since it may be that different solutions are valid solutions, merely catering to different goals. If we reconsider the example where we are attempting to find a good definition for the word “multiplication,” it is easy to see that different goals as to what a solution should accomplish will yield very different solutions. For example, imagine that someone understood “multiplying” as a term that applied to rabbits, and, given a group of A male rabbits and another group of B female rabbits, wanted the definition of “multiply” to determine how many new rabbits could be expected if the two groups were left together for a year. It may be that a definition of “multiply” that works for adding up groups may also work for predicting rabbit births. However, both people working on the same issue (defining the term “multiply”) have different goals as to what a solution should accomplish. Therefore, when critiquing solutions, they may be evaluating completely different criteria. This will lead to disagreements and a feeling that there is no clear solution, when instead the issue is that the people creating the solutions have different goals. For example, externalist “solutions” to defining knowledge or justification which merely give definitions as to what it means for someone to in fact know something or in fact be justified do not seem to be valid solutions to me, since I cannot use them to certainly determine whether I know or am justified in believing a particular proposition. A different philosopher, with the goal of finding objective notions of “knowledge” and “justification,” regardless of their usefulness, may be completely content with externalist solutions, and not understand or care about my critiques.

The main worry in cases in which multiple people are working on the same problem with different goals is that it slows down the process of identifying solutions. Instead of concluding that a solution is correct and demonstrating that critiques of it require different goals, philosophers may instead try to respond to each other’s critiques, ending up talking past each other and never settling on a solution, even though solutions exist. Using a goal-oriented approach to answering philosophical questions may aid with this issue, since it will clarify what it means for a solution to be correct and limit misunderstandings. However, it is not a perfect solution since misunderstandings may still occur. Two people may state that they have the same goal but internally have different definitions of the words used in the description of the goal. Thus, even though their written goals look exactly the same, they will still lose time debating about the merits of particular solutions until they determine that they need clearer definitions of the terms used to define the goal. So, while using a goal-oriented approach will not always stop useless debate, it may help quickly illuminate crucial differences between philosophers.

There is another deeper concern that may be diagnosed more easily when following a goal-oriented approach: we may not know what our goal is. This issue was brought to light during discussion on Gettier cases. In class, we were presented with multiple solutions to the problem of knowledge, each of which, at first, appeared to solve all of the Gettier cases, thus providing a valid solution for defining knowledge. But we were then able to create new Gettier cases, each of which revealed different issues in the solutions. Situations such as these point towards a pattern that we have an “intuition” as to what our ultimate goal should be, but we can’t figure out how to describe it.

There are two potential methods of addressing this issue. The first is to create a solution for the best goal we can describe, and then find issues with that solution that don’t correspond to our “intuition.” By defining new goals which require solving those issues, we iterate closer to an ultimate goal with each new solution, create a clear path of what is expected of the next solution, and also gain the ability to keep track of the progress made by previous solutions. Note that each of those previous solutions is a valid solution, just with a goal that we later realized did not adequately represent our desire. In class, we seemed to follow a method similar to this, but not quite as explicit. The second method of dealing with a goal we can’t describe is to consider the fact that we can’t describe our goal to be a problem in and of itself. That is, instead of asking “how do we define ‘knowledge’,” a better question may be “what, exactly, do we want from a definition of knowledge?” The goal of a solution to this second question is to define all of the characteristics of a valid solution to the first question, such that any solution to the first question that fits those characteristics will be considered a valid definition of ‘knowledge.’ To the extent that the second question cannot be answered, we may resort to the first method and try to iterate towards an answer.

However, I find it worrying that we do not yet have a valid solution to the second question. (I consider my proposals above as to what we want from definitions of knowledge and justification to merely be suggestions, not full and complete responses.) If we don’t have a concrete idea as to what it means to come up with a “solution” in this field, do we really know what we’re doing? Not only have we yet to produce a completely satisfactory definition of knowledge or justification, but we aren’t even sure we could identify such a definition if we had one.

I believe a goal-oriented approach will help to at least diagnose, if not fix, many of these issues plaguing epistemology and other philosophical fields. It will help cut down on time spent debating over solutions where the issue is not necessarily the solution itself but the merits by which the solution is being judged. It will also help to push philosophers to define end goals for their field and raise the question of whether there is at least one complete and valid solution to the questions they are asking. (Note that just because a valid solution is found does not mean that there aren’t other equally valid solutions for that field to be found as well; a valid solution signals an exciting beginning, not an end.) A goal-oriented approach helps bring structure and provide a clear path for future work, as well as giving a foundation as to why it is that we are working on particular questions. If we never examine why, is it worth doing?