Quine's web of beliefs

Two dogmas of empiricism

Quine wanted to show and refute two dogmas of modern empiricism, which are:

1. There is a clear distinction between analytic (based on meaning instead of external facts about the world) and synthetic (based on external facts about the world) statements.

2. Reductionism of statements to logical constructs like the term "bachelor" can be reduced to "unmarried man" and therefore create logically true statements like "No bachelor/unmarried man is married" which are true regardless of which are used.

Reductionism

Quine differentiates between two classes of analytic statements:

1. Logically true statements like:

"No unmarried man is married."

The logical particles which are not interchangeable are: "un-", "no", "and", etc.

"A logical truth is a statement which is true and remains true under all reinterpretations of its components other than the logical particles." Therefore, no matter what we use to replace the variable "married", the truth value will be preserved.

2. Statements which are true by synonymy:

"No bachelor is married."

These statements can be turned into logical truths (1. Class of analytical statements) by replacing synonyms for synonyms. In this case "bachelor" can be replaced by "unmarried man".

Quine argues that this second class of analytic statements is based on synonymity, but we often misinterpret synonymity with extensions. "Bachelor" and "unmarried man" refer to the same set of individuals and therefore they have the same extension. Similarly, "creature with a heart" and "creature with a kidney" may have the same extension in the world.

However, having the same extension does not guarantee synonymy because the may not be interchangeable in all contexts without changing the truth value. For example in the case of beliefs. Person x believes that all bachelors are messy but they might have a false understanding of the word "bachelor". Similarly, in the case of necessity when adding the condition of necessity the truth values can change like "Necessarily, 8=8" and "Necessarily, 8= Number of planets", especially considering the change from the former belief of the number of planets being 9.

Similarly, our belief that "bachelor" and "unmarried man" are synonymous could be based on accidental facts or based on other beliefs which can be revised, not on a necessary relation of meaning.

We lack proper characterization of the second class of analytic statements, specifically the notion of synonymity and meaning and therefore of analyticity in general.

The web of beliefs

Quine argued that just as synthetic statements depend on external facts about the world, so do analytic statements depend on meaning and synonymity. Meaning and synonymity are not clearly defined and rely on definitions that are circular.

Therefore, even the most fundamental beliefs are just based on other beliefs and can be revised. Everything is connected like a web and can be revised. The beliefs closest to the edge of the web are closely connected to external experience. External experiences don't necessarily affect a specific point of the web but the entirety of the web, meaning any external information can cause any point(s) within the web to be revised. Our natural tendency is to not disturb the system or change it as little as possible especially the central beliefs.

However, any belief even the ones on the edges can be held onto regardless of external experience. On the other hand, even the most central beliefs can be revised even by minimal input from external experience.

Even scientific theories can therefore be seen as falsifiable tools within the system, rather than truths outside the system. They can be revised and the system can be changed.

Personal perspective

I think this pragmatic approach explains paradigm shifts in science, other fields of study, personal changes and changes in relationships between people. It also reflects the philosophical principle of scepticism while retaining a rational and practical approach to our understanding of the world and our belief system.

Source