An Investigation of the Laws of Thought
Authors: George Boole
Year: 1854 | Walton and Maberly, London
Source: boole-1854-an-investigation-of-the-laws
What it is
Boole’s 1854 book created Boolean algebra — the mathematical system that underlies all of digital computing. By treating logical reasoning as algebraic manipulation with variables that can only take the values 0 and 1, Boole unified logic and mathematics and laid the foundation for everything from circuit design to programming languages to neural networks.
The fundamental idea
Before Boole, logic was the province of philosophy. Aristotle’s syllogisms (All men are mortal; Socrates is a man; therefore…) were analyzed verbally, not mathematically.
Boole’s claim: logical operations are algebraic operations on classes, using a special algebra where every variable satisfies:
x² = x
This single equation — which ordinary algebra does not obey — forces x to be either 0 (the empty class, False) or 1 (the universal class, True). It is the axiom that makes Boolean algebra different from regular algebra.
The algebra
Boole assigns symbols to classes (sets of things with a common property):
| Symbol | Meaning |
|---|---|
| x | the class of men |
| y | the class of mortal things |
| 1 | the universe (everything) |
| 0 | the empty class (nothing) |
| xy | intersection: things that are both x and y |
| 1 − x | complement: things that are not x |
Operations:
- AND →
xy(both must be in the class) - NOT →
1 − x(everything outside the class) - OR →
x + y − xy(either one, Boole’s original; later simplified to inclusive OR)
The fundamental law x² = x holds because “men who are men” = “men” — membership doesn’t stack.
Logic as equation solving
Boole showed that any logical argument can be translated into equations, solved algebraically, and the conclusion read off. No intuition required — just algebra.
Example: “All men are mortal” becomes:
y = x + v(1-x) [x is a subset of y, for some undetermined class v]
which simplifies to conditions that force any x to also be in y.
This was Leibniz’s dream realized: a calculus ratiocinator — a mechanical method for deciding logical truth.
The connection to probability
In the second half of the book, Boole extends the same algebra to probability:
- P(x) = probability of event x
- P(xy) = P(x ∩ y)
- Logical relationships between events become equations constraining their probabilities
This was an early attempt at a unified formal framework for uncertain reasoning — the ancestor of probabilistic logic and Bayesian networks.
How it reached computing
Boole himself never imagined a connection to physical machines. That connection came 84 years later when Claude Shannon, a 21-year-old MIT master’s student, realized that:
- A relay contact has two states: open (1) and closed (0)
- The algebra of relay connections obeys exactly Boole’s postulates
- Therefore, circuit design is Boolean algebra
With Shannon’s insight (1938), Boole’s abstract mathematics became the design language for every digital circuit ever built — and eventually for every processor, memory chip, and GPU.
The path from Boole to AI
Boole 1854 → Boolean algebra: logic as mathematics (0/1)
Shannon 1938 → Circuit algebra: logic as hardware
McCulloch 1943 → Neural algebra: logic as neurons
Turing 1936 → Computation: what logic machines can do
Deep learning → Large-scale parallel Boolean-style computation
Every activation function in a neural network is a smooth approximation of a Boolean threshold. Every attention mask is a selection over a Boolean index. Modern AI systems are, at their mathematical core, extraordinarily complex Boolean circuits — with learned weights instead of fixed connections.
Boole’s original philosophical goal
Boole was not trying to build machines. He was trying to answer a philosophical question: what are the laws of human thought? He believed that the structure of logical reasoning — which seems like a faculty of the human mind — could be captured in formal symbols. His title reflects this ambition: An Investigation of the Laws of Thought.
The irony is that in formalizing the laws of thought, he created the tools by which machines would come to simulate them.
See also
- shannon-1938-symbolic-analysis-of-relay-and — Boolean algebra applied to electrical circuits
- mcculloch-1943-a-logical-calculus-of-the — Boolean neurons: logic gates as brain models
- turing-1936-on-computable-numbers-with-an — What Boolean machines can compute