Ayer's Language, Truth and Logic: Logical Positivism
A. J. Ayer's account of empirical verification, its limits for metaphysical claims, and the different work done by logic and ethical language.
PhilosophyAll chapters Reading tools Notes & comments
Meaning and verification › What Would Count as Finding Out?Full chapter · Sections in this chapter1 / 8 ↓
What Would Count as Finding Out?
When A. J. Ayer wrote Language, Truth and Logic, he asked readers to consider whether there are mountains on the far side of the Moon. He could say what a person would have to observe to find out. The technology available to him did not let him get a look for himself. The claim still made sense as one about the world. Whether someone can perform a particular check today is a different question from whether there is anything the check would settle. Language, Truth and Logic, ch. I
Ayer sets this example against a statement purporting to describe an Absolute that is somehow involved in development without developing itself. A reader might be able to repeat the formula, or learn where its author uses it. That would leave a question about what would make the formula true or false as an account of the world. The difficulty for Ayer goes beyond a shortage of evidence. If the statement is equally compatible with every conceivable observation, he asks what gives it the standing of a factual claim. Language, Truth and Logic, ch. I
The book first appeared in 1936. The introduction added to the second edition in 1946 revisits the verification principle and changes how it is formulated. Reading the two formulations together shows both what Ayer wanted the test to accomplish and how hard it was to set up. He wanted to leave room for mathematics and for the indirect evidence scientists use, while drawing a boundary around what can count as a literally meaningful assertion about matters of fact. First-edition catalogue record · Language, Truth and Logic, second-edition introduction
Factual Significance and the Possibility of Verification
Suppose someone says a packet arrived this morning. Looking where it was meant to be delivered might bear on the claim. A video could show somebody leaving it there, and a dated receipt could help establish when it arrived. Observations that would lead you to accept the claim are different from some observations that would lead you to reject it. The negative case needs care. An empty doorstep now may be compatible with successful delivery and removal later. We can understand the claim and ask what evidence would affect our judgment while still being uncertain about what happened. Language, Truth and Logic, ch. I
This is why Ayer distinguishes verification in principle from an available procedure. Consider a claim about where that packet spent the previous night. There may be no photograph, and no record may have survived. In principle, a person inspecting the right storeroom at the right time could have seen it. Specifying the conditions of such an observation says what the claim commits us to, even when it is too late to perform the check. Ayer asks for a connection to possible experience. A practical barrier changes the opportunities for inquiry without taking away that connection. Language, Truth and Logic, ch. I, Moon example and discussion following it
There is a further distinction between evidence that would settle a statement and evidence that could make it more probable. Imagine a claim about the behavior of every specimen of a certain alloy under stated conditions. In whatever limited trials we perform, the untested cases remain open. Those trials can still provide evidence. In Chapter I, Ayer initially distinguishes strong verification, in which experience could conclusively establish the truth, from weak verification, in which experience could support the claim short of certainty. He requires weak verification for factual significance in this initial account. In the second-edition introduction he recognizes a tension. If all empirical claims are hypotheses subject to further experience, his earlier description leaves no class of strongly verifiable claims. He goes on to distinguish claims about what is being experienced that he thinks can be conclusively verified. Language, Truth and Logic, ch. I and second-edition introduction
A factual claim can have an experiential test even if the experience could support it without settling it. In Ayer's first formulation this is enough for the claim to count as verifiable in the relevant sense. Language, Truth and Logic, ch. I
Logic, Mathematics, and What a Test Can Show
If experiential evidence mattered to every meaningful sentence in the same way, it would be hard to explain a proof. In Chapter IV, Ayer asks what we would say if we thought we had five pairs of objects but counted only nine. The result might lead us to check how many things we grouped, whether anything moved, or whether we counted accurately. Neither the grouping nor the result of a recount readily bears on the proposition that two times five is ten. It bears on which arithmetical description fits the objects. Language, Truth and Logic, ch. IV, example of counting five pairs
Ayer calls a proposition analytic when its validity depends solely on the definitions of its constituent symbols. The validity of a synthetic proposition depends on facts of experience. The mathematical statement in the example can be understood with the meanings and rules of arithmetic. The claim that a particular pile contains ten pieces entails a factual commitment. Seeing one person hide a piece before the count could change our answer to the second question. It would help establish that the terms of the arithmetical example had been misapplied in this case. Language, Truth and Logic, ch. IV
Another of his examples concerns whether any ants are parasitic. A person could learn a new fact about ants by examining their behavior. The disjunction that some are parasitic or that none are holds regardless of how the inquiry turns out. What one learns in deriving a logical consequence can still be new to the person who derives it. Ayer's argument is that it does not thereby describe another feature of how ants behave. Logical and mathematical work can bring out consequences that were previously hard to see without depending for their validity on further observations. Language, Truth and Logic, ch. IV, the ant example and discussion of tautologies
When Grammar Suggests the Wrong Question
Ayer also asks how a person might come to mistake a sequence of words for a factual claim. He compares saying that martyrs suffer with saying that martyrs exist. The two English sentences can look as though they perform the same work. One of them ascribes suffering to people of a certain kind. The other says there are people of that kind. If we mistake existence for a further property of a martyr, an intelligible question about whether there are any martyrs turns into a search for a special feature possessed by the martyrs who exist. Ayer takes this to illustrate how a philosophical problem can grow out of a superficial similarity in grammar. Language, Truth and Logic, ch. I, examples of existence and predication
Another of his examples is the difference between saying that dogs are faithful and saying that unicorns are fictitious. In the usual reading of the first, a speaker ascribes a quality to animals. Ayer thinks the familiarity of that reading can tempt a person to say that unicorns must exist in some way in order to have a special quality called being fictitious. To compare the two readings, imagine someone admitting that no unicorn has ever been found and acknowledging that they occur in stories. Those claims can be assessed by asking different questions from the ones we would use when considering the behavior of a dog. Clarifying what the sentence does may dissolve the demand for an unfamiliar sort of animal to bear its predicate. Language, Truth and Logic, ch. I, example of fictional unicorns
This part of the book makes room for two distinct outcomes of analysis. An apparently puzzling statement can turn out to have a different logical form from the sentence in which it was expressed. In such a case the task is to make a comprehensible claim easier to recognize. In another case the proposed reading may leave us unable to say what a speaker is undertaking to report, even after the wording has been made clear. The second outcome is the one Ayer needs if the verification principle is to rule out a class of apparent factual claims. The first is a reason to examine each use before passing judgment. It leaves open whether someone could give a clearer formulation of a position Ayer would reject. Language, Truth and Logic, ch. I
Reworking the Criterion
Weak verification leaves open a problem about what size of contribution a statement must make to the evidence. You can take a sentence that gives you no procedure for finding out whether it is true and join it to a claim that plainly does. Suppose someone says both that a stone left in the sun will warm and that the Absolute is content. The conjunction leads to the expectation about the stone. The second conjunct has done nothing to produce that expectation. The example shows why merely being able to obtain an observational consequence from a collection of statements is an insufficient test of each member of the collection. Language, Truth and Logic, second-edition introduction
Ayer's later formulation distinguishes direct and indirect verifiability. Direct verification covers observation statements themselves, as well as statements that yield an observation statement when combined with other observation statements that would not yield it on their own. An indirectly verifiable statement, together with admissible premises, must yield a directly verifiable statement that those premises alone do not yield. The admissible premises have to be analytic, directly verifiable, or independently established as indirectly verifiable. Ayer is trying to ensure that a statement under examination genuinely adds something while allowing other claims to help connect it to observation. Language, Truth and Logic, second-edition introduction, direct and indirect verification
Consider a claim about a pressure difference inside a sealed piece of equipment. The description can concern something you cannot make out by looking directly at the equipment. If a gauge works as specified, a given pressure difference could give its needle a reading that would otherwise be unexplained. The connection depends on other premises about the gauge and the setup. Those premises need a basis of their own for the arrangement to provide a meaningful check. Ayer makes room for theories stated in terms that do not directly name observable things by allowing the connections to run through separately warranted premises. The gauge is an illustration of why that allowance matters. Language, Truth and Logic, second-edition introduction, discussion of unobservable theoretical terms
Tightening the rules about the other premises does not by itself show that the test rules out every arbitrary claim. In his review of the second edition, Alonzo Church constructed a logical objection to Ayer's revision. Using observation statements and an arbitrary statement, he showed how combinations could meet the test in a way that lets that statement or its negation through. The point depends on the exact rules for entailment in the revised version. It illustrates the danger in judging factual significance by the logical arrangements into which a statement can be put. How a sentence contributes to a real investigation may be harder to capture than which formulas it can help entail. Church’s review, bibliographic record · Jeff Speaks, reconstruction of the formal objection
What Ethical Language Does
Ayer's separation of factual meaning from other uses of language gives the book's discussion of ethics a particular shape. He compares accusing someone of stealing money with accusing them of acting wrongly in stealing it. The occurrence of the theft can be a factual issue. Calling it wrong expresses moral disapproval on his account. He takes the ethical expression to do work in what we say to one another while adding no separate factual assertion to the accusation. Language, Truth and Logic, ch. VI, example of the theft accusation
Compare that use with saying that a speaker disapproves of what happened. The latter can be a factual claim about the speaker, which could turn out to be false if the person was acting or joking. To express disapproval is to use language in a different way. In the original accusation, a tone of anger can show an attitude without the speaker making a claim about their own state of mind. Ayer also says that ethical wording can be used to arouse feelings and encourage action. The way he discusses truth-telling brings the point out. An ethical expression can carry something of the force of a command or suggestion. Language, Truth and Logic, ch. VI, discussion of disapproval and commands
There is still room for argument about facts that people take to bear on their ethical judgments. Two people may initially disagree about whether a person should return the money. One thinks the cash was stolen. The other thinks a lender received it back. Checking the history could change the dispute. Ayer's harder case occurs when the parties agree on the description and still take different ethical positions. His account separates the dispute about the events from the clash of ethical attitudes. That brings into view a question an opponent can press. If someone offers reasons for reconsidering which kinds of action deserve disapproval, how much of that reasoning can be represented as a dispute about the facts alone? Language, Truth and Logic, ch. VI, discussion of whether ethical disputes are possible
Putting the Test on the Table
Ayer's procedure helps a reader ask what kind of agreement a sentence is seeking. The report that a packet arrived calls for evidence about an event. The claim that there are five pairs in front of you can be checked by looking and counting, while the arithmetic used to describe the count has a different basis. A criticism of someone who hid the packet can express disapproval even when nobody disputes what happened. Separating these functions can expose cases in which people appear to argue about one thing while relying on different uses of the sentence.
The proposal itself gives rise to a different question. What gives a criterion of literal meaning its authority? Applying it to an ordinary report is simpler than explaining why all putative factual discourse should answer to it. Treating the criterion as a proposal about how to use literal significance would place it in a different role from the reports it is meant to sort. But a philosopher who makes that move still owes an account of why this proposal serves as a good boundary, and why it should disqualify the disputed cases. The ability to identify how a particular observation bears on a particular sentence is one result. The success of a single rule for every candidate statement is a further result, and the second-edition revision shows how much more would have to be demonstrated. Language, Truth and Logic, second-edition introduction
Further Reading
- A. J. Ayer, Language, Truth and Logic. Start with Chapter I, then compare the different formulation in the second-edition introduction. Chapters IV and VI treat logic and mathematics, and ethics, respectively.
- Alonzo Church, Review of the second edition of Language, Truth and Logic. The record identifies the original review. Jeff Speaks’s reconstruction sets out the formal argument.
Join the discussionAsk a question or share what you learned.
Keep exploring
Reference
Citation details
Cite or share this article.

Comments
Add to the conversation
A question, a correction, or an example from your own work.
No comments yet. Be the first to share your thoughts!