Logic: Science or Anomaly?

1. Defining Anti-Exceptionalism

What does it mean to say logic is "exceptional"? Traditionally, logic has been seen as possessing a unique combination of properties that separates it from the sciences:

• It is a priori: its truths are known independently of experience.
• It is necessary: its laws hold true in all possible worlds.
• It is analytic: its truths are grounded in meaning or convention.
• It is foundational: its principles are the bedrock upon which other knowledge, particularly in mathematics and science, is built.

Anti-Exceptionalism about Logic (AEL) is the view that this picture is wrong. Logic is not special in these ways; it is, in some fundamental sense, continuous with the sciences.

To get a precise handle on this position and its different flavours, please read the following introduction to a chapter on the topic.

As you read, the key distinction to grasp is between:

• Metaphysical AEL: The view that logic's subject matter is continuous with science. Its laws describe the most general, structural features of the world.
• Epistemological AEL: The view that the methods for justifying and choosing logical theories are continuous with those in science.

Our focus today is on Epistemological AEL, as it directly addresses the question of whether logic's relationship to evidence and theory-choice makes it continuous with scientific practice.

2. The Quinean Vision: Logic on the Web of Belief

The most influential argument for Epistemological AEL comes from Quine, whose work we studied last lesson. For Quine, there is no separate domain of a priori justification. All our beliefs—from fleeting observations to the laws of physics and logic—form a single, interconnected web. This web is judged as a whole by how well it predicts and explains our sensory experience.

This view, known as naturalized epistemology, dissolves the boundary between science and philosophy. To explore the full impact of this idea, let's watch a video that delves into Quine's vision.

Quine's holism provides the classic picture of logic's continuity with science:

• There is no sharp distinction between logical and factual truths, only a difference of centrality in the web of belief.
• Logical laws appear necessary and a priori only because of our extreme reluctance to revise them, given the massive disruption this would cause to our overall theory of the world.
• In principle, however, a logical law could be revised on pragmatic, empirical grounds—if, for example, doing so produced a simpler, more powerful overall physical theory (as was once proposed for quantum mechanics).

3. Modern AEL: Abduction and Theory Choice

While Quine's view is foundational, many contemporary anti-exceptionalists adopt a more nuanced position. They often part ways with Quine's strict evidential anti-exceptionalism (the idea that only empirical, sensory evidence ultimately matters). Instead, they champion methodological anti-exceptionalism.

This is the view that even if the data for logic isn't purely empirical (it might include our judgements about the validity of arguments, mathematical proofs, or solutions to paradoxes), the method for theory-choice is the same as in science. Logicians, on this view, don't discover logical truths through a special faculty of intuition. Instead, they propose and defend logical theories using abduction, or inference to the best explanation.

A logical theory is chosen over its rivals if it better accounts for the data while displaying theoretical virtues like:

• Strength: It proves more, without proving things that are known to be false.
• Simplicity: It is elegant and avoids ad-hoc rules.
• Unifying Power: It explains a wide range of phenomena (e.g., both mathematical reasoning and natural language inference).
• Explanatory Power: It provides a satisfying explanation for why certain inferences are valid.

Let's return to the Martin & Hjortland paper to see how this works.

This abductive model paints a picture of logic as a dynamic field of theoretical inquiry. Just as physicists debate competing theories of gravity based on their explanatory power and fit with observational data, logicians debate competing logics (e.g., classical vs. intuitionistic vs. paraconsistent) based on their ability to resolve paradoxes, model mathematical practice, and account for our inferential judgments.

4. An Alternative: Logic as a Non-Natural Science

So far, we have a spectrum: Quine's radical continuity with empirical science, and the more moderate methodological continuity. But is the choice only between logic being "exceptional" and logic being "like physics"?

The philosopher Timothy Williamson offers a compelling third way. He argues against philosophical exceptionalism, but he does so by broadening the concept of "science."

Williamson's distinction gives us a powerful new lens for our evaluation:

• Science (narrow sense): Natural sciences like physics and biology, characterized by experiments, measurement, and observation.
• Science (broad sense): Any rigorous, systematic, evidence-based discipline.

In Williamson's view, mathematics is a paradigm of a science in the broad sense. It is exceptionally rigorous and evidence-based (proof is a form of evidence), but it is obviously not a natural science. He argues that philosophy, and by extension logic, belongs in the same category.

This perspective allows one to be an anti-exceptionalist—to deny that logic is some unique, mystical discipline—without committing to the Quinean idea that it is directly answerable to sensory experience. The choice of a logical theory might be an a priori, armchair activity, but it is still a rational, evidence-based, scientific (in the broad sense) activity of theory construction and evaluation, much like proving a theorem in pure mathematics.

Williamson concludes that the desire for philosophy to be utterly unique is a "narcissistic delusion."

Conclusion

So, is logic continuous with the empirical sciences? We have evaluated several nuanced answers.

• The traditional exceptionalist says "no." Logic is a unique, a priori discipline that provides the fixed foundation for science.
• The Quinean anti-exceptionalist says "yes." Logic is simply the most central, abstract part of our total scientific theory of the world, and is ultimately answerable to empirical evidence.
• The methodological anti-exceptionalist says "its methods are." We choose between logical theories using the same abductive principles of theory choice (simplicity, strength, explanatory power) as scientists, even if the data isn't strictly empirical.
• Williamson's view offers a final synthesis: "No, logic is not a natural science, but it is a science." It is continuous with the general project of rational, evidence-based inquiry, placing it alongside mathematics rather than physics.

Key Takeaways:

• Anti-exceptionalism about logic (AEL) is the rejection of the traditional view that logic is an a priori, foundational, and un-revisable discipline.
• Quinean holism provides the classic argument for AEL, viewing logic as a revisable part of our total "web of belief" that is continuous with empirical science.
• Methodological AEL is a more modern variant, arguing that theory-choice in logic follows scientific methods (like abduction) even if its data (e.g., intuitions, paradoxes) are not purely empirical.
• An alternative perspective, championed by Timothy Williamson, holds that logic is a non-natural science, similar to mathematics. It is rigorous and evidence-based but does not use the experimental methods of the natural sciences.

Preview of the Next Lesson:

The idea that logic is a theoretical enterprise we choose, rather than a fixed system we discover, opens up a profound question: could there be more than one "correct" logic? This brings us to the debate between logical monism (the view that there is one true logic) and logical pluralism (the view that multiple different logics can be correct). In our next lesson, we will define these positions and begin to explore the arguments for each.