Chapter 12 is written as a series of clauses and is separated into seven subcategories, which have been listed and summarized below.

1. Introduction

The central thesis of this chapter asserts that there cannot be a physicalist causal theory of the human language. Here, physicalism refers to the belief that everything knowable by humans has to have a physical form. In the context of language, physicalism gives credence to behaviorism, the idea that any action performed by an organism, including the use of language in communication, is but a response to stimuli. In other words, Popper rejects the idea that human language can only be understood as behaviorism.

2. Four Major Functions of Language

In this section, Popper references the work of German psychologist and linguist Karl Bühler, who assigned three major functions to the human language. These are presented in a strict hierarchy, where the higher functions cannot be present without the lower functions, but the lower functions do not require the higher ones to be true.

The first and lowest is the expressive or symptomatic function, where language is the result of an attempt to express voice physical sensations or other experiences. The second is the stimulative or signal function, which is the use of language to make requests. The third is the descriptive function, whose purpose is to communicate ideas.

To Bühler’s thesis, Popper adds a fourth function, which is the argumentative function. It is different from the descriptive function because it requires the use of logic and reason. It is also the highest function because it encompasses all the other functions: It is an expression for being an “outward symptom of an internal state;” it is a signal because it seeks to provoke a reply; and it is a description as the view it supports is a state of affairs (295).

3. A Group of Theses

Science and philosophy both make use of the descriptive and argumentative functions of language, but behaviorism and physicalism can only be limited to the two lowest functions. This is because two people may exhibit the same outward linguistic behavior, but they may have done so for completely different reasons. Since human intent has no physical form, it cannot be seen in a person’s behavior. For example, two people may both praise a political candidate, but one may only be repeating what he has heard (expressive function) and the other might be preparing to persuade their listener (argumentative function). Neither the physicalist nor behaviorist theories could tell the difference; they are hence bound to either ignore the existence of the higher functions or dismiss them as outliers of the lower functions.

4. The Machine Argument

Popper demonstrates that the higher functions cannot be simply dismissed as special cases of the lower functions. Intent, despite having no physical shape, matters, particularly in differentiating humans from machines. A thermometer, for example, can be said to express, signal, and even describe its internal state (the temperature it reads). However, it cannot rationally be attributed any intent—it is not the thermometer that volunteers this information out of its personal will, but rather reflects the intent of its maker.

This is why people do not argue with machines; the machine’s behavior is not characterized by intent. By extension, then, the use of language in any serious social situation must assume the existence of intentionality in the listener. Some have argued it is impossible to know whether other minds truly exist; Popper believes that the existence of language itself solves the problem.

5. The Causal Theory of Naming

Popper asserts that it is impossible to establish a causal relationship between an object and its name, which then implies that there cannot be a causal physical theory of the two highest functions of language—description and argumentation.

He provides the following example to explain his theory: A machine that is coded to name every ginger cat “Mike,” might respond by saying “Mike” if it recognizes a ginger cat. However, this does not imply that it realizes there is a direct causal relationship between ginger cats and the name “Mike.” This is because, prior to the cat’s appearance, the machine is already ready to respond with “Mike” if given the right stimuli. There are no objective physical situations in which it is “correct” to respond to ginger cats with naming them “Mike.” The rule has to be established by the machine’s maker, and the choice of name is entirely arbitrary.

Since naming is a descriptive function and it is impossible to establish a causal and physicalist realization of the name-relation, it follows that the two highest functions of language—description and argumentation—can neither be causal nor physical. In other words, naming ginger cats “Mike” is not the result of an objective physical truth in which ginger cats must be named thus, but the result of human intent.

6. Interaction

Popper then takes this a step further: He points out that logical relationships are abstractions that do not have physical forms. For example, a person living in house A and who is told that Mike has been adopted into house B, might say “Mike is not here.” This is not merely a response to his five senses having failed to detect Mike, but a deduction based on information he was given, which do not take a physical shape. If people’s actions can be influenced by their logical reasoning and by grasping abstract relationships, then there is no reason to believe that language or any other human action is merely a response to outside stimuli.

7. Conclusion

The author concludes that people who fear being called obscurantists have embarked on an unnecessary project of attempting to separate “objective” language from its “metaphysical” counterpart. However, this duality is in fact a fake problem and attempts at rectifying it are just another form of obscurantism.

Chapter 13 is a reply to three criticisms raised by Professor Wilfrid Sellars regarding Popper’s paper, Language and the Body-Mind Problem, summarized in the previous chapter.

Sellars first raises a question about whether Popper is confusing true facts about physical states (such as “it is a fact that today is rainy”) from quasi-facts not based on physical states (such as “it is a fact we must complete our homework”) when he states that “the statements of physics and of psychology are […] two ways of talking about the same facts” (300). To this, Popper replies that the distinction is unimportant in the context of his use: As long as both physics and psychology use them in the same way, his argument stands.

Next, Sellars challenges the idea that the act of referencing something cannot be defined in terms of behaviors. To this, Popper replies that he has never put forth such an idea in his paper, and that if he did, he recants it completely.

Finally, Sellars claims in his review that Professor Ryles has defended the idea of dualism. Popper believes it is his writing that has given rise to this misunderstanding, and in fact, Professor Ryles does not agree with establishing a duality between language used in physics and in psychology.

Popper reiterates that it is not necessary to separate language into an “objective” system for stating scientific facts and another for “quasi-facts” for philosophy. The mind-body dualism is a fake problem. There are no words objectively correct for describing things with physical appearances, and no words objectively incorrect for describing states or other intangible things. Human intent imparts upon language a descriptive and argumentative function, which can be used, along with logic or reason, to speak of both physical and non-physical things.

This creative chapter tells the story of a dialogue between Theaetetus and Socrates. It is a story Popper crafted to illustrate how indirect self-reference in language can be meaningful, despite claims to the contrary.

After a few exchanges, Theaetetus asks Socrates the following question: “What did I say between your last two interruptions, Socrates?” To this, Socrates answers correctly that Theaetetus said, “The very next question which I am going to ask you is an extraordinary one, although expressed in perfectly ordinary language” (304).

Then, Theaetetus asks if Socrates understood the meaning of his question, to which Socrates replies that yes, Theaetetus was referring to something he will ask Socrates shortly. Theaetetus continues: “And what was this question of mine to which my warning referred? Can you repeat it?” Socrates then accurately repeats what Theaetetus said earlier: “What did I say between your last two interruptions, Socrates?”

Finally, Theaetetus asks if Socrates understood his meaning in the last question he asked. Socrates affirms that he obviously understood, since he answered correctly.

This little story demonstrates that indirect self-reference (the warning refers to a future question Theaetetus was about to ask, and the question Theaetetus finally asked referred to the earlier warning he made) can still be properly understood; it is therefore not entirely meaningless.

The story continues to show that even direct self-references need not be paradoxical or meaningless. In fact, the truth or falsity of some statements that directly self-reference can be established by empirical testing or logical reasoning. Socrates asks how this is possible. Theaetetus then offers the following statement: “What I am saying now is meaningful.”

Socrates immediately understands this and explains its logic. The statement Theaetetus just said directly refers to itself. Assuming that Theaetetus’s statement is false, then it can be clearly established that what he just said is actually not meaningful. By clearly establishing the fact that it is false, the statement is proven to be falsifiable, which means it is actually meaningful. This proves that even direct self-reference in statements can be meaningful if they can be used as logical or empirical falsifications.

Theaetetus in the story proposes that any utterance whose meaning can be understood is valuable. Similarly, any question for which an answer can be given must also be meaningful. If meaning is only assigned to utterances that have physical shape or have already been empirically confirmed, then this excludes all hypotheses and conjectures as meaningless absurdities, as they have not yet been proven (or, in this case, falsified). Since it is clearly irrational to stipulate that hypotheses are all meaningless until they are proven, this chapter concludes that self-referring assertions are not merely worthless absurdities.