Language, Truth and Logic

"But it must be understood from the outset that we are not concerned to vindicate any one set of philosophers at the expense of any other, but simply to settle certain questions which have played a part in the history of philosophy out of all proportion to their difficulty or their importance." (134)

Language, Truth and Logic is a brief and charmingly audacious effort to retire metaphysics and its related issues. Ayer is a mid-20th-century exponent of the Anglo-American analytical tradition in philosophy (including the work of Bertrand Russell and others) which seeks to reduce the discipline to applications of logic. His arguments are sympathetic to the earlier empiricists and positivists, but show more sophistication in pointing out and show more sometimes surmounting their shortfalls. I am most in accord with his "emotive theory of values" as a method of dispensing with the philosophical concern over ethics.

Ayers' professed opposition to "schools" in philosophical discourse reminds me of the ultra-Protestant Plymouth Brethren "coming out of sect" in 19th-century England: they paradoxically insist on a narrowing of their field while claiming to transcend distinctions within it.

The 1946 introduction to the second edition consists of Ayers reconsidering and fine-tuning many of the details in the body of the text. Accordingly, I saved it to read until finishing the original eight chapters. In retrospect, however, because of the intricacies of the arguments, a reader would be better advised to read the 1946 remarks in sequence after each individual chapter.

Although mystics (and magicians, to a lesser degree) are unlikely to find this book easy or pleasant, it would be an invaluable supplement to their intellectual diets. After passing through this crucible, they might proceed to the more congenial offerings of a thinker like Gregory Bateson. show less

“… in so far as statements of value are significant, they are ordinary ‘scientific’ statements; and […] in so far as they are not scientific, they are not in the literal sense significant, but are simply expressions of emotion which can be neither true or false” (103)

“The difference between the man who uses language scientifically and the man who uses it emotively is not that the one produces sentences which are incapable of arousing emotion, and the other sentences which have no sense, but that the one is primarily concerned with the expression of true propositions, the other with the creation of a work of art” (44)

Ayer’s principal aim in this book seems to be that of defining the work of philosophy, and it is show more decidedly not in making metaphysical statements about first principles or in proclaiming any other a priori truths that we treat as out of reach from ordinary tests of verification. The role of philosophy is to clarify our thinking and expression, symbolically, by grounding it in and making it traceable to what can be ascertained empirically about the world.

Ayer advocates a focus on verifiability as a condition by which propositions can be understood as significant (i.e., pointing to the world) and capable of producing knowledge because they can be proven, empirically, to be either true positively (verified in the strong sense of the term) or with some probability (verified in the weak sense of the term). Statements that cannot be verified directly or for which we can not envision ways or conditions under which to verify them, cannot be knowledge because they are “senseless” (i.e., not available to the senses or capable of supporting predictions about what that sensory experience would be). It is this focus on verifiability (as opposed to falsifiability) that I understand to distinguish his approach from garden-variety logical positivism.

The benchmark of verifiability means that any statement that is found to be true only inductively contributes to our overall sense of the world and truth. The more often we make consistently verified observations, the more assured we can become that we are making guaranteed true statements. But we never arrive at guaranteed knowledge of the world, only that which is more and more probable. It is based on the belief that in probability that we find enough that is provisionally true to go on. In this way, I don’t think that Ayer is talking about inductive reasoning so much as he is talking about abductive reasoning, which involves making provisional claims to understanding on the basis of incomplete information and then revising based on the experience of application to find what is true and consistent to take another step forward.

The degree to which statements are coherent within a regime of truth (e.g., science) allows us to rationalize what must be true or what is likely to be true because 1) we understand that all such claims can be verified if needed, 2) that verification has already been done before, 3) we see confirmation in our experience of the world that those truths we accepted as provisional hold with the sense field we are observing, and 4) that we can empirically verify that other people are acting in relation to the world that is consistent with our own rational sense of the world.

I’m sympathetic to what Ayer is attempting to do here, but it also seems like he goes a bit too far and not far enough on some points:

Paradox of Verifiability: Ayer’s model of verifiable knowledge holds that we can know something if we can make verifiable truth claims about it. These should either be truth claims that are directly verifiable and can be proven true or false with sense data. Or we can know things about the world that could be verifiable under “certain conditions” (145). That is, what is true and what is probably true are allowed to stand as facts.

The problem is that “certain conditions” is under-specified. In the spirit of Ayer’s argument, “certain conditions” are supposed to account for situations where if I wanted to, for example, verify that Pluto exists in our solar system, I could do so provided that I was able to get there in my lifetime, have sufficient air, have sufficient warmth, and a suitable vessel to be close enough to Pluto confirm, “yup, there it is.” But the same kind of argument could be made for verifying that gremlins or unicorns exist if the conditions for verification can be specified (e.g., you have to truly believe they exist before you can see them, and only at midnight on the third full moon of the year, etc.). It has the potential to get a bit silly. I think that it has to be necessary to apply the test of verification to the “certain conditions” of verification themselves. Going with the example, we can confirm a condition like the third full moon of the year but can we verify “true belief”? When we start applying the rule of verification to the conditions of verification, however, we can end up with a paradox of never having conditions of verification that do not themselves need to be verified before they can stand as conditions of verification. So, perhaps we swing in the other direction and set a hard limit on what can be verified: anybody must be able to be meet the conditions of verification and be present to make independent observations. So, if we cannot figure out how to travel to a distant planet or to the center of the earth, etc. then we have to admit that there are aspects of the world that are beyond our ability to verify or to conceive of how to verify, meaning that they cannot stand as things that we know. Ayer’s model either restricts what can be know too much or is too permissive.

Apparatus of Verification: Following on the previous point, I would have also liked additional information about the apparatuses of verification that Ayer had in mind. He seems to be assuming that empirical verification is limited to the senses. One refutation of my example about verifying that Pluto exists in our solar system is that the New Horizons spacecraft did this for us. And the sense data that we gathered about Pluto came to us through the instrumentation that received the data from the instruments on New Horizons that gathered data from Pluto. There are at least two levels of mediation here. Our sense data as knowing beings who are capable of making propositions about the world is, strictly speaking, a sensory experience of watching a data stream on an earth-based computer interface. Is that a suitable apparatus of verification such that we would be comfortable making knowledge claims about Pluto? Probably most of us would agree that it is, but that is because we are taking things as true a priori based on faith in science and engineering such that we do not feel a need for further verification, beyond QA testing prior to launch, that the New Horizons spacecraft is a reliable extension of our own senses. Why do we not then afford the same trust to other mediating devices that proclaim to allow access to other knowable truths: video footage of Sasquatch, conspiracy theories, truths about the afterlife in religious texts, etc.?

Ethical Knowledge: Ayer claims that normative ethical statements cannot result in verifiable propositions (106). How does one, for example, verify that murder is wrong? And in this we get to one of what I understand to be a well-established critique of Ayer, that his formulation strips all normative weight of moral claims to say that they are not significant because of their appeal to inaccessible and unverifiable matters. As others have put it, he is equating the moral force of “murder is wrong!” with “Boo, murder!”. Where his claim does seem to hold up is that a moral law, like the Categorial Imperative (CI) is accepted as true without appeal to the world of experience. That may be true, but if we take an implication of the CI to be, in the case of murder, that I cannot wish for murder to be ethically allowable because in allowing murder we have to allow that it is ethically fine for someone to murder us, we seem to arrive at point where the proposition “murder is okay” is counter to our experience which would lead us to conclude, when our own lives are at risk, that murder IS wrong. But not really. Even though in a situation like this, we can say that murder is not desirable, or murder is counterproductive to establishing community, or murder is detrimental to civil order. But we still can’t say that murder is wrong because what does “wrong” mean? I get this objection on principle, but it does still seem like the normative force of “murder is wrong” is what animates more verifiable statements like “murder (is wrong because it) is counterproductive to establishing society. So is it true that there is no significant ethical knowledge? I don’t quite believe it. It may just be that the language of ethical claims are too abstracted from the empirical realities that they entail. show less

This is a product of its age, to be sure... But, then again, dunking on metaphysics should be a product of every age

I know this is a seminal text but I'm finding it a tough read. The writing style is dense, reflecting its origin in the 1930's. It's a challenging book, posing the question of how we judge what truth is. Do we work purely from empirical observation, only accepting propositions that can be logically proven? Or is truth something else, wider, deeper? Perhaps truth value or meaning can be found outside or apart from language, and does not only reside within it? This book proposed that metaphysical and religious language was meaningless, as it could not be empirically verified. Ayer himself, later in life, acknowledged that his verification theory, proposed within this book, is false. However, it had great influence when it first came out show more which ensures its place on reading lists and bibliographies to this day. show less

I read the "squashed" version of the book and I can agree that while it is not "wrong", the content of the book is incomplete. But mostly, it is very intriguing and thought-provoking and it's a good start for enquiring about language and it's relationship with philosophy.

A response to all the "blah blah blah" of many philosophers, in which he claims that unless statement is implementable, at least in theory, in the real world, it is actually devoid of meaning.

"A delightful book … I should like to have written it myself." — Bertrand RussellFirst published in 1936, this first full-length presentation in English of the Logical Positivism of Carnap, Neurath, and others has gone through many printings to become a classic of thought and communication. It not only surveys one of the most important areas of modern thought; it also shows the confusion that arises from imperfect understanding of the uses of language. A first-rate antidote for fuzzy thought and muddled writing, this remarkable book has helped philosophers, writers, speakers, teachers, students, and general readers alike.Mr. Ayers sets up specific tests by which you can easily evaluate statements of ideas. You will also learn how to show more distinguish ideas that cannot be verified by experience — those expressing religious, moral, or aesthetic experience, those expounding theological or metaphysical doctrine, and those dealing with a priori truth. The basic thesis of this work is that philosophy should not squander its energies upon the unknowable, but should perform its proper function in criticism and analysis. show less