Anti-Exceptionalism About Logic and the Regress Problem

The Rules of the Game: Why Logic Might Be Trapped in a Loop

Imagine you are trying to build the ultimate rulebook for a game. You want the rules to be perfect, fair, and undeniable. In the world of science, we usually figure out the best rules by looking at the game being played, testing different rulebooks, and seeing which one explains the action best. This is how we choose between theories in physics or biology. But there is a special corner of science called logic. For a long time, people thought logic was different from everything else. They believed logic was the "referee" that never made mistakes, the set of rules that existed before the game even started, and the only way to prove that any other rule was right.

However, a group of thinkers has started to say, "Wait a minute." They argue that logic is just like any other science. They call this Anti-Exceptionalism. Their idea is that logical rules aren't perfect or unchangeable; they are just theories we use to make sense of reasoning, and we should pick the best one using the same tools we use for everything else: looking at the evidence, keeping things simple, and seeing what works best. This is called Abduction (or "inference to the best explanation"). It's like being a detective: you look at the clues and guess which story fits them all together.

But here is the catch: to decide which logical theory is the best detective story, you have to use logic itself to check the clues. If two detectives (two different logical theories) disagree on a basic rule, how do you decide who is right without already assuming one of them is right? This paper explores whether this "Anti-Exceptionalist" idea can actually solve the problem of choosing between different logics, or if it gets stuck in a loop where you need a rule to pick the rule, and another rule to pick that rule, forever.

---

The Detective's Dilemma: When the Rules Fight Each Other

This paper, written by Yi Ding, tackles a tricky problem facing the "Anti-Exceptionalist" view of logic. The author argues that while the idea of treating logic like a normal science is appealing, it hits a massive wall when two logical theories fundamentally disagree. The paper suggests that the method used to choose between these theories—the "detective work" of abduction—gets stuck in an endless loop, much like a famous puzzle in philosophy called the "Problem of the Criterion."

The Setup: Logic as a Science
The paper starts by explaining the Anti-Exceptionalist view. Imagine logic isn't a holy, unchangeable set of laws written in the stars. Instead, imagine it's a toolbox. We have different tools (different logical theories) like Classical Logic (the standard kind) and Intuitionistic Logic (a stricter kind that doesn't accept certain shortcuts). The Anti-Exceptionalists say we should pick the best tool based on Abductive Criteria:

1. Simplicity: Which tool is easier to use?
2. Strength: Which tool explains more things?
3. Fit: Which tool matches the data (like how people actually reason or how math works)?

The goal is to be neutral, like a judge who doesn't pick a side until they hear the evidence.

The Trap: The Theory-Laden Clues
The paper argues that this "neutral judge" doesn't actually exist. Here is why: The "evidence" you use to judge the tools is itself shaped by the tool you are holding.

Imagine a Classical Logician and an Intuitionistic Logician are arguing.

• The Classical Logician says, "My tool is simpler because it has fewer rules."
• The Intuitionist says, "No, my tool is simpler because it doesn't assume things exist that we can't prove."

Who is right? The paper says it depends on what you value. But here is the kicker: What counts as "simple" or "strong" depends on the logic you already believe in. The Classical Logician sees the Intuitionist's refusal to accept a certain rule as a weakness; the Intuitionist sees it as a strength. The "clues" (the data) are "theory-laden," meaning they are colored by the theory itself. You can't use the clues to pick the theory because the clues were already picked by the theory.

The Infinite Loop: The Meta-Logic Problem
This leads to the paper's main finding: a Regress Problem.
To decide between Logic A and Logic B, you need a "Meta-Logic" (a higher-level logic) to evaluate the arguments.

• If you use Logic A as your Meta-Logic, you are simply assuming Logic A is right to prove Logic A is right. That is circular reasoning.
• If you use Logic B, you are assuming the opposite direction.
• So, you need a Logic C to choose between A and B. But then, how do you choose Logic C? You need a Logic D. And then a Logic E.

The paper shows that this goes on forever. You are stuck in a loop where you need a rule to pick the rule, and another rule to pick that rule, with no starting point.

Why the Usual Fixes Don't Work
The author tests four ways people usually try to escape this loop and shows why they fail for logic:

1. Starting with specific facts (Particularism): In other fields, you might say, "I know I have two hands, so let's build a rule from that." But in logic, the "facts" (like "this argument is valid") are exactly what the two sides disagree on. You can't start with a fact if the two logicians can't agree on what a fact is.
2. Starting with a method (Methodism): You might say, "Let's just use the 'Abduction' method." But the paper argues that using the Abduction method requires logic. If you don't have a logic to run the method, the method can't start.
3. Giving up (Scepticism): You could say, "Okay, we can't know which logic is right." But if you can't know, then the Anti-Exceptionalist's own advice ("use abduction to choose") is also unproven.
4. The "Common Ground" Escape: Maybe both sides agree on some basic rules (like "if A and B, then A") and can use that as a neutral starting point? The paper says no. Even those basic rules are contested by other types of logicians (like Relevant Logicians), and the "common ground" isn't strong enough to do the heavy lifting of proving complex theories.

The Conclusion: A Modest Limit
The paper does not say that Anti-Exceptionalism is false. It doesn't say logic isn't revisable. Instead, it suggests that the Anti-Exceptionalist view has a specific, structural blind spot.

The author concludes that while we can treat logic like science in many ways, we cannot use a "neutral" abductive method to solve fundamental disagreements between logics. The method of "inference to the best explanation" cannot step outside the box to pick the box. The paper suggests that if we want to choose between these deep logical differences, we might have to rely on broader philosophical goals or values, rather than a purely neutral scientific checklist.

In short, the paper finds that the Anti-Exceptionalist dream of a perfectly neutral, scientific way to choose between all logics hits a wall. The "detective" needs a magnifying glass to find the clues, but the magnifying glass itself is one of the things being investigated. The loop is real, and the paper argues that the standard ways out of the loop don't work for logic.