Probability and Evidence

by Alfred Jules Ayer

In Probability and Evidence , one of Britain's foremost twentieth-century philosophers addresses central questions in the theory of knowledge and the philosophy of science. This book contains A.J. Ayer's John Dewey Lectures delivered at Columbia University, together with two additional essays, "Has Harrod Answered Hume?" and "The Problem of Conditionals."

Genres: Philosophy, Logic

144 pages, Hardcover

First published January 1, 1972

About the author

In 1910, Sir Alfred Jules Ayer was born in London into a wealthy family. His father was a Swiss Calvinist and his mother was of Dutch-Jewish ancestry. Ayer attended Eton College and studied philosophy and Greek at Oxford University. From 1946 to 1959, he taught philosophy at University College London. He then became Wykeham Professor of Logic at the University of Oxford. Ayer was knighted in 1970. Included among his many works are The Foundations of Empirical Knowledge (1940), The Problem of Knowledge (1956), The Origins of Pragmatism (1968), Metaphysics and Common Sense (1969), Bertrand Russell (1972) and Hume (1980), about philosopher David Hume. Later in life, Ayer frequently identified himself as an atheist and became active in humanist causes. He was the first vice president of the British Humanist Association and served as its president from 1965 to 1970. He was an Honorary Associate of the Rationalist Press Association from 1947 until his death. He was also an honorary member of the Bertrand Russell. In 1988, Ayer had a near-death experience in the United States after choking on salmon and subsequently losing consciousness. He wrote of his experience in “That Undiscovered Country” (New Humanist, May 1989): “My recent experiences have slightly weakened my conviction that my genuine death, which is due fairly soon, will be the end of me, though I continue to hope that it will be. They have not weakened my conviction that there is no god. I trust that my remaining an atheist will allay the anxieties of my fellow supporters of the British Humanist Association, the Rationalist Press Association and the South Place Ethical Society.” He died shortly after at age 78 in London. D. 1989.

More: http://en.wikipedia.org/wiki/A._J._Ayer

http://plato.stanford.edu/entries/ayer/

http://www.britannica.com/EBchecked/t...

http://www.ucl.ac.uk/~uctytho/AyerbyT...

http://badassphilosophers.tumblr.com/...

http://www.informationphilosopher.com...

Reviews

[Andrew Noselli]

Although I did read a book called Logical Positivism that was edited by A.J. Ayer earlier this month, he only wrote a short introduction to that volume, and so I essentially used Probability and Evidence as an introduction to his corpus, so I don't believe it's the best way to begin studying his work.  After reading this book, not only do I question why I continue to study logic, which I admittedly have little training in -- especially in that it is a subject which borders on mathematics, a subject-area and indeed an entire form of thinking in which I am hesitant to tread, having failed the AP Examination for Calculus in high school; I have no other higher mathematical education except for an introductory course in statistics at Rockland Community College, where I learned to compute the standard deviation from the mean and so forth -- but I also question why we continue to write books of logic at all.  To explain my position, let me make the following argument: In my opinion, the principles of logic were laid down by Aristotle and it seems that they are as solid and unchangeable as the U.S. Constitution; however, Ayer, playing the academic game that logicians get paid for, introduces some extra-logical concepts which mainly deal with the projection of truth-contents onto categories and container-groups which do not belong among the main body of classical logical propositions at all, strictly speaking.  [Note: whenever I hear induction mentioned, I always think of old friends and the simple-minded arguments we had regarding the theory of evolution, and this book contains a lot of grist for that mill, which I wish I could discuss with him at some point - perhaps it will have to wait until the after-life?]  Maybe I'm wrong and somewhat ill-equipped as a logician to make this statement but I think that, generally speaking, the study of logic has become more or less a propaganda exercise when it appears in the public sphere; specifically, my problem with how logicians such as A.J. Ayer write their books is that they appear to want to get the public to accept the ineluctability of history as a substitute for the politics of truth, especially in the minds of people conditioned by only a televised simulacrum of physical contact.  Two stars.

[Otto Hahaa]

Not so sure if this is relevant anymore.

I am not a philosopher, but I am interested in probability and evidence in my work. So I thought this might be useful to read before trying to read something more modern, like "Bayes or Bust". There are a lot of interesting stuff here, but I was wondering if do I really need to read this, or should I try to find something more up to date? The other book with the same name, but by Paul Horwich, might have more low-hanging fruit, so to speak. In both books, a lot is made of black ravens. The black ravens problem is a typical philosophical problem, because at first you think what the heck are these philosophers complaining about, and then you realize that it is an actual problem and complain why don't these philosophers solve it so you don't have to lose your own sleep over it. Or something like that.

Might be more relevant if you are a student of philosopher and really need to understand what people thought of these problems in early seventies.

[prarobinson]

So far so good. Man, Hume was a trouble maker! A.J. Ayers is such a clear and thoughtful writer, though, that I can feel sunlight peeking through the clouds. I highly recommend it.

[Benjamin Webb]

Goodreads reviews for technical philosophy are heavily skewed by non-specialists expecting a modern introduction to statistics or decision theory. Read through that lens, Ayer’s 1972 book looks outdated because it doesn't cover modern Bayesian networks or computer science applications.
However, from an epistemological and historical perspective, it remains exceptionally valuable.
Evolution of Bayesianism: Since 1972, probability theory in computer science, medical diagnostics, and AI has moved almost exclusively toward complex formal Bayesian modeling. Ayer doesn't cover modern objective Bayesianism or algorithmic decision theory.
Modern discussions often focus on quantum probability, Bayesian epistemology in AI, or formal causal inference. Ayer's framework reflects the logical analytic debates of mid-20th-century Britain.
But it is exceptionally valuable as a masterclass in conceptual clarity: Ayer doesn't let mathematics mask bad logic. His breakdown of the frequency theory vs logical probability exposes fundamental flaws in how people define probability that remain true today.
Chapter 2 remains one of the sharpest, most accessible treatments of the problem of induction in print. It tackles why we think past data justifies future expectations—a foundational problem that modern statistical algorithms simply assume away.
Chapter 1 clarifies the difference between a true 'law of nature' and a mere accidental correlation. This distinction is critical for any scientific realist evaluating evidence.
Skip it if you want a modern guide on how to calculate probabilities, build decision models, or apply Bayesian methods to data science. Read it if you want a concise, razor-sharp, analytical inquiry into what it strictly means for evidence to 'support' a claim, written by one of the 20th century's clearest prose stylists