The Therapy of Grammar: Language, Meaning, and the Dissolution of Philosophical Problems
§1-7: The Opening Gambit — Augustine’s Picture of Language and Its Limits
The work begins not with a thesis but with a quotation from Augustine’s *Confessions*, describing how he learned language as a child: he watched adults name objects, connected the words with gestures, and gradually built up his vocabulary. This seems like a perfectly natural account of how language works — a picture of words as labels for things, and meaning as the object a word stands for.
Wittgenstein immediately undercuts this apparent naturalness. He characterizes Augustine’s passage as describing “a particular picture of the essence of human language” — namely, that “every word has a meaning. This meaning is correlated with the word. It is the object for which the word stands.” The opening move is not to reject this picture as false, but to suggest that it is too narrow: it describes “a system of communication; only not everything that we call language is this system.”
Here Wittgenstein introduces his first major methodological device: language-games. He asks us to imagine a primitive language where a builder calls out “block!”, “pillar!”, “slab!”, “beam!” and his assistant brings the corresponding object. This is a complete language in itself — a whole activity of giving orders and executing them. The point is devastating in its simplicity: Augustine’s picture works for nouns like “slab,” but it says nothing about how words like “five,” “there,” or “here” function. It also says nothing about the *activity* in which words are embedded. To know what a word means is not to have an image in mind but to know how to use it in the flow of life.
The famous conclusion of §43 — “For a *large* class of cases — though not for all — in which we employ the word ‘meaning’ it can be defined thus: the meaning of a word is its use in the language” — is not yet announced here, but the ground is being prepared. Wittgenstein introduces the builder’s language as a “language-game” precisely to shift our attention from the mental act of meaning to the public practice of using words. The builder’s cry “Slab!” is not a label for a slab; it is a move in a game, an order that gets someone to act. To understand it, we do not need to look inside the speaker’s mind but at the social practice that surrounds the utterance. This is the first fissure in the Augustinian picture — and it will widen into a total reorientation of philosophy.
§8-32: Extending the Game — Primitive Languages as Philosophical Tools
Wittgenstein now builds a series of increasingly more complex language-games, each designed to isolate a different feature of how language actually operates. He introduces a language with numerals (“a”, “b”, “c” for one, two, three), and shows how learning to count is not a matter of correlating words with mental numbers but of being trained to produce the right word at the right moment. He then adds color-samples, and shows how the word “red” is learned through pointing to a sample — but the pointing itself is part of the language-game, not an external act of ostension that magically transfers meaning.
A crucial moment comes at §30-32, where Wittgenstein takes up the concept of ostensive definition — the act of pointing to an object and naming it. He argues that this cannot be the foundation of language because the learner must already be in possession of a “language-game” for the pointing to work. “One has already to know (or be able to do) something in order to be capable of asking a thing’s name.” The gesture of pointing is ambiguous: do I point to the color, the shape, the material, or the object? To understand the pointing, the learner must already have a grasp of the grammatical role the word will play — that “red” is a color-word, not a thing-word, for instance. Ostension presupposes what it is supposed to explain.
This is not merely a psychological observation about learning. It is a logical point about the structure of understanding. If meaning were secured by a mental act of pointing, no amount of pointing would ever be enough to guarantee that the other person has grasped what we mean — because the pointing itself is always open to multiple interpretations. What actually secures meaning is the *practice* in which the word is embedded: the shared form of life, the common training, the regular responses. Wittgenstein here dissolves the ancient philosophical problem of how words connect to the world: the connection is not a metaphysical link but a practical one, established by the way we are trained to go on.
§33-64: The Problem of Understanding and the Illusion of Mental States
Wittgenstein now turns to a question that has haunted philosophy since Plato: what is it to *understand* a word, to grasp its meaning? The temptation is to think that understanding is a mental process — a flash of insight, a picture in the mind, a feeling of familiarity. Wittgenstein devotes a series of intricate investigations to showing that none of these mental accompaniments can be what understanding *is*.
At §33-35, he considers the case of understanding a rule like “Add 2.” When someone says “Now I understand!”, what has happened? There is a temptation to think something special occurred in their mind. But Wittgenstein asks: “But is understanding only a process that accompanies saying something?” He points out that we might say we understand a word when we hear it, yet be unable to give its definition — and we might be able to give the definition but not really understand how to use the word in actual sentences. “Understanding” is not an experience; it is a capacity, an ability to go on in a certain way.
This argument reaches its sharpest formulation in the famous “beetle in the box” passage (though that comes later at §293, the groundwork is laid here). Wittgenstein is attacking the idea that meaning is a private mental object that only I can inspect. He argues that if I could have a private sensation that only I know, then no public language could ever name it — because I could never teach you what “S” means, since you cannot see my sensation. Conversely, if a public language *can* name sensations like pain, then the sensation itself cannot be a purely private object. The grammar of “pain” is not the grammar of a hidden inner thing; it is the grammar of a public expression — groaning, crying, saying “It hurts.” The inner state is not the *meaning* of the pain-word; the public circumstances of its use are what give it sense.
This is not behaviorism. Wittgenstein is not denying that people have inner experiences. He is denying that those inner experiences play the role of *grounding* meaning. The meaning of “pain” is not a private object that I match to the word; it is the role the word plays in a public language-game of expressing and responding to hurt. The philosophical picture of the mind as a container of private meanings is an illusion produced by a misunderstanding of grammar — by taking the surface grammar of “I have a pain” (subject + verb + object) to indicate that there is an inner object, “a pain,” that I possess. Wittgenstein’s therapy consists in showing that this surface grammar misleads us about the depth grammar — the actual use of the word in our lives.
§65-88: Family Resemblances and the Attack on Essentialism
Having dismantled the Augustinian picture and the mental-state theory of meaning, Wittgenstein now confronts what might seem to be the natural consequence of his arguments: if words do not get their meaning from referring to objects or from mental acts, then what *do* they have in common? What is the essence of language? The temptation is to think that all language-games must share some common property that makes them language — just as all games must share some property that makes them games.
Wittgenstein’s response at §65-71 is one of the most celebrated moves in the *Investigations*. He denies that there is any single feature common to all the things we call “games”: “Look for example at board-games, with their multifarious relationships. Now pass to card-games; here you find many correspondences with the first group, but many common features drop out, and others appear.” He continues: “And we can go through the many, many other groups of games in the same way; can see how similarities crop up and disappear.” Instead of a common essence, we find “a complicated network of similarities overlapping and criss-crossing: sometimes overall similarities, sometimes similarities of detail.”
He introduces the famous concept of family resemblances to capture this: the various things we call “games” are held together not by a single thread running through them all, but by the way resemblances overlap — just as members of a family may share eye color with one, nose shape with another, gait with a third, without any single feature being present in all of them. The concept “game” has blurred edges, and that is not a defect but part of how our concepts actually work. “And this is how it is with *language*,” he adds — the word “language” itself is not a unity but a family of overlapping language-games.
This argument has profound consequences for philosophical method. If concepts like “language,” “meaning,” “understanding,” and “game” do not have sharp boundaries, then the philosopher’s traditional task of giving definitions — of stating the necessary and sufficient conditions for something to be a game or a language — is misguided. The demand for exactness, for a crystalline purity, is itself a symptom of philosophical confusion. At §88, Wittgenstein makes this explicit: “The more narrowly we examine actual language, the sharper becomes the conflict between it and our requirement.” The requirement is for “a pure intermediate realm” — a world of perfect logic — but language itself is messy, contextual, and embedded in forms of life. Philosophy’s job is not to correct this messiness but to describe it, to let language be what it is, and to dissolve the illusions that arise when we demand that language conform to a picture of ideal purity.
§89-133: Philosophy as Therapy — The Nature of Philosophical Problems
The first major section of the *Investigations* culminates in a sustained reflection on what philosophy itself is and what it can achieve. Wittgenstein diagnoses the root of philosophical confusion: we are bewitched by language. “Philosophy is a battle against the bewitchment of our intelligence by means of language” (§109). The problem is not that we lack facts or theories; it is that our grammatical forms mislead us. We take the surface of language — the fact that “being” and “nothing” look like nouns, that “I have a pain” looks like “I have a coin” — and we construct metaphysical systems on the basis of these grammatical analogies.
The remedy is not to construct a better theory but to *describe* — to assemble reminders of how words are actually used in ordinary situations. “We must do away with all *explanation*, and description alone must take its place” (§109). The philosopher’s task is therapeutic: to show the fly the way out of the fly-bottle. When we are stuck in a philosophical puzzle — about the nature of time, or the meaning of a word, or the reality of the external world — what we need is not a new argument but a clear view of the grammar that is trapping us. “A philosophical problem has the form: ‘I don’t know my way about'” (§123).
This is a radical reorientation of the philosophical project. Philosophy is not a quest for hidden truths or ultimate foundations. It is a form of clarification. The philosopher does not discover new facts about the world; he arranges what we already know — the ordinary use of language — in a way that dissolves confusion. “The clarity that we are aiming at is indeed *complete* clarity. But this simply means that the philosophical problems should *completely* disappear” (§133). The goal of the *Investigations* is thus not to produce a system of propositions but to produce understanding — a kind of perspicuous representation that shows us how our language actually functions. When we achieve this clarity, the problems vanish, not because they have been answered but because they have been shown to rest on misunderstandings.
§134-142: The Criterion of Simplicity and the Test of Understanding
We now enter a transitional section where Wittgenstein examines what it means for a proposition to be “simple” or “complete” — and in doing so, he begins to dismantle the assumption that there is a single, ideal logical form underlying all propositions. The discussion follows directly from his earlier therapeutic claim that philosophical problems should disappear rather than be solved. Here, he takes up the notion that a proposition must have a definite sense, and asks what that actually means.
“Let us ask: ‘In what sense is a proposition a picture?’ — But this is perhaps better put: ‘What is it that we call a proposition?'” (§135)
This shift is crucial: Wittgenstein is not asking for a definition but for a description of how we actually use the word “proposition.” When we look at ordinary language, we find that what counts as a simple proposition varies with context. The sentence “The cat is on the mat” seems simple enough, but what about “The cat is meowing”? Is that simpler? Or “The cat”? Wittgenstein’s point is that the search for a single logical unit — the “simple proposition” that corresponds to an atomic fact — is itself a grammatical confusion. The criterion of simplicity depends entirely on what we are doing with language at a given moment. A proposition is complete when it does the work we need it to do, not when it meets some abstract logical standard.
This connects to his earlier critique of essentialism. Just as “game” has no single common property, “proposition” has no single essence. The attempt to find one is not an empirical investigation but a grammatical investigation — one that, when conducted properly, reveals the diversity of our linguistic practices. The philosopher who insists that every proposition must have a definite sense is like someone who insists that every line must have a specific length before we can say what it is. Wittgenstein dissolves this demand by showing that the sense of a proposition is determined by its use in a language-game, not by a hidden logical structure.
§143-155: Learning, Understanding, and the Series of Natural Numbers
Wittgenstein now turns to one of his most famous examples: the pupil learning to continue a series of natural numbers. This passage is a sustained meditation on what it means to “grasp” a rule, and it directly challenges the idea that understanding is a mental state or process that causally produces correct behavior. The example is deceptively simple: a teacher shows a pupil how to write down the series of natural numbers, and the pupil eventually “gets it.”
The moment of insight — when the pupil suddenly understands how to continue — seems to be a discrete mental event. But Wittgenstein asks us to look more carefully. What actually happens when the pupil “understands”? Do we see something in his mind? No — we see him continue the series correctly. The understanding is manifested in the performance, not hidden behind it. Wittgenstein draws a sharp distinction between the experience of “suddenly understanding” and the criterion for having understood: “The criterion for his understanding the series is the way he continues it” (rough paraphrase of §146). The feeling of “Aha!” is not the understanding itself; it is a symptom that may or may not accompany the real phenomenon.
This is a profound move against the inner/outer picture that dominates so much of philosophy. The inner state — the feeling of understanding, the mental image, the private sensation — is treated as the real thing, while the external behavior is merely its outward sign. Wittgenstein inverts this: the behavior, the public performance, is the criterion for understanding; the inner experience is at best an accompaniment. “How do I know that he understands?” — not by looking into his mind, but by watching what he does. The rule is not a mysterious entity that causes behavior; it is a normative pattern that we apply in practice.
§156-171: Reading, Decoding, and the Illusion of Mental Processes
The discussion of rule-following now extends to reading. Wittgenstein asks: what does it actually mean to “read”? The phenomenon seems straightforward: we look at marks on a page and produce sounds. But when we examine the situation more closely, the boundaries become blurred. A beginning reader who sounds out letters is not reading; an experienced reader who recognizes whole words is reading. But where is the line? Is there a specific mental process that constitutes “reading” as opposed to merely reciting?
“The words ‘reading’ and ‘meaning’ are used in different ways. It is not as if there were one activity called reading that we engage in.” (close paraphrase of §156-158)
Wittgenstein constructs a series of thought experiments to undermine the idea that reading involves a specific inner experience. Consider a person who is given a page of nonsense script and, by luck or training, produces sounds that match the words printed there. Are they reading? We would say no — because they are not actually deriving the sounds from the marks. But what does “deriving” consist in? Not a feeling of effort or attention; a skilled reader feels no effort at all. The criterion for “reading” is not an inner state but a complex of publicly observable criteria — the correct correlation of marks and sounds, the ability to handle new cases, the absence of guessing.
This leads to a broader point about meaning and understanding. When we hear a word, we often feel we “know” its meaning immediately — but this feeling is not itself the meaning. Wittgenstein compares the experience to seeing a familiar face: we recognize it without going through a process of comparison. But the recognition is not a separate mental act; it is simply the way we respond. The mistake is to think that every linguistic performance must be accompanied by a corresponding mental process. “If God had looked into our minds,” Wittgenstein writes in a famous image, “he would not have been able to see there whom we were speaking of” (§ — close paraphrase). The meaning is not hidden in the mind; it is laid out in the use.
§172-184: The Continuation of the Series and the Normativity of Rules
Returning now to the series example with greater depth, Wittgenstein examines the claim that a rule determines its own application. The idea is seductive: once we grasp the rule “add 2,” the entire infinite series is somehow already determined — every step is fixed in advance. But what does “fixed in advance” mean? When we wrote down the first few numbers, were all the subsequent numbers already present in some way? This is not a question about psychology but about grammar.
The temptation is to say that the rule contains its applications as the rails of a track contain the path of the train. Wittgenstein calls this “a mythological description of the use of a rule.” The rule does not compel us; we compel ourselves by choosing to follow it in a particular way. There is no metaphysical necessity that ties “add 2” to “1000, 1002, 1004” — it is a practice, a custom, a form of life. The rule is a signpost, and the signpost does not tell me which way to go; I simply do what I have been trained to do.
This is the famous rule-following paradox: “How can a rule tell me what to do at this point? Whatever I do can be made to accord with the rule if I interpret it in the right way.” But Wittgenstein’s solution is not to find a better interpretation; it is to see that interpretation must eventually give way to action. “To follow a rule is a practice,” he writes. “And to think one is following a rule is not to follow a rule” (§202). The certainty of rule-following does not come from mental acts or logical compulsion; it comes from the shared, public activity of human beings who have been trained in a common technique. The normativity of rules — the fact that we can be correct or incorrect — is grounded not in the mind but in the community.
§185-201: The Paradox of Rule-Following Deepened
We are now deep inside the Investigations’ most famous and difficult cluster of arguments, extending from roughly §185 into the 240s. The problem of rule-following, introduced through the arithmetic example of the pupil who continues the series “+2” oddly after 1000, has been sharpened into an explicit paradox. Wittgenstein is pressing the question: what determines that one way of continuing a series is correct and another incorrect, if any finite set of examples can be interpreted in multiple ways?
The pupil in §185 has learned to count by twos up to 1000, but when asked to continue beyond that point, he writes 1000, 1004, 1008, 1012. He thinks he is following the same rule; he produces an answer with what seems to him perfect consistency. Wittgenstein does not treat this as a mere pedagogical failure. He insists that what the pupil does reveals something about the nature of rule-following itself: no amount of explanation, no pointing to further rules, can definitively close the gap between a rule and its application. At §198 we read:
“But how can a rule show me what I have to do at this point? Whatever I do can be made to accord with the rule if I interpret it in the right way.”
This is the moment of maximum crisis. If every action can be interpreted as following or violating a given rule, then rule-following seems to collapse into anarchy — or worse, into the illusion that we are all following rules when in fact we are simply doing whatever we please. Wittgenstein’s interlocutor voices the skeptical anxiety that threatens to undo the very possibility of normative correctness. But Wittgenstein’s response is not to produce a better interpretation that will finally close the gap; it is to show that the search for interpretation is itself the source of the problem. Each interpretation, to be understood, would require a further interpretation, leading to an infinite regress. The solution, he insists, must lie elsewhere.
What Wittgenstein offers instead is a radical shift from the mental to the practical. At §201 he writes that there is a “misunderstanding” in the skeptical line of thought. The paradox arises only if we think that understanding a rule is a mental act that precedes and determines every application. But if we see rule-following as a practice — as something we do, not something we interpret — then the regress stops. The rule is not a ghostly mental entity that casts its shadow forward across all possible cases; it is a technique, embodied in actions that have been trained and that are shared with others. The community’s agreement in judgments, not in opinions, is what makes rule-following possible.
§202-207: Practice, Custom, and the Rejection of Private Obedience
Building directly on the resolution of the paradox, Wittgenstein now makes one of his most decisive claims. At §202 he draws the line sharply between genuine rule-following and its mere simulacrum:
“To follow a rule is a practice. And to think one is following a rule is not to follow a rule. Hence it is not possible to obey a rule ‘privately’: otherwise thinking one was obeying a rule would be the same thing as obeying it.”
This passage is the hinge on which the entire private language argument will turn. The phrase “to think one is following a rule is not to follow a rule” captures the logical gap between subjective certainty and objective correctness. If I believe I am following a rule, but there is no way to distinguish this belief from actually being correct, then the distinction between correctness and error collapses. A rule that can only be checked “in my own mind” is no rule at all, because it provides no standard against which my performance can be measured.
The next sections elaborate what this means concretely. Wittgenstein imagines someone learning a game by watching others play; the learner may think she understands the rules, but her attempts to play reveal that she has not grasped them. The community’s reactions — corrections, puzzled looks, confirmations — are what teach the rule, not any private mental event. At §206 he writes that “the common behavior of mankind is the system of reference by means of which we interpret an unknown language.” This system of reference is not a theory; it is the shared, embodied way of life that makes language possible. The community of practice is the only tribunal before which the question of correct rule-following can be brought.
The implications are devastating for any philosophy that tries to ground meaning in private mental experiences. If meaning were a matter of inner states accessible only to the individual, then there would be no difference between having the right inner state and merely thinking one had it. Language would become a solitary fantasy. Wittgenstein is not denying that we have inner experiences; he is denying that they can serve as the foundation for the normative structure of language.
§208-219: The Criterion of Agreement and the Form of Life
Wittgenstein now shifts from the negative task of dismantling the private rule model to a positive characterization of what rule-following actually rests upon. The key term introduced in these sections is “form of life” (Lebensform), which appears at §219 in one of the most quoted passages of the entire Investigations:
“When I follow a rule, I do not choose. I follow the rule blindly.”
The word “blindly” is not an admission of irrationality; it is a recognition that at the deepest level, rule-following is not based on reasons that can be further justified. If someone asks me “Why do you continue the series this way?” and I have exhausted all explanations, I eventually reach bedrock. My spade is turned. I can no longer give a reason; I can only say “This is simply what I do.” The certainty of rule-following is not epistemic — it is not that I know something with special certainty — but practical. It is the certainty of action, not of belief.
This connects directly to the concept of agreement that runs through these sections. At §240, Wittgenstein writes that “disputes do not break out (between mathematicians, say) over the question whether a rule has been obeyed or not. People do not come to blows over it.” This is not because mathematicians have special access to eternal truths; it is because their training, their practices, and their forms of life are sufficiently aligned that agreement in judgments is the norm. The agreement is not in opinions about mathematics but in the techniques of calculation, the ways of going on that are taught and shared. Disagreement at this level would not be a mathematical dispute; it would be a sign that the parties were not playing the same language-game at all.
The form of life is the background of shared human activity — eating, walking, pointing, counting, reacting to pain and pleasure, building, trading, teaching — that makes specific language-games possible. It is not itself a language-game or a set of rules; it is the pre-theoretical, pre-linguistic foundation on which all rule-governed activity is built. Wittgenstein is not giving a metaphysical theory of this foundation; he is pointing to it as something we must simply acknowledge. At §217, he writes: “I have reached bedrock, and my spade is turned. I am inclined to say: ‘This is simply what I do.'”
§243-255: The Private Language Argument Begins
We now enter what is arguably the most celebrated and controversial stretch of the Philosophical Investigations: the private language argument. The argument does not appear as a single, crisp syllogism; it unfolds through a series of thought experiments and reminders about what it would mean for a language to be essentially private — that is, a language whose words “refer to what can be known only to the speaker, to his immediate private sensations” (§243).
Wittgenstein begins by distinguishing between a language that is merely *contingently* private (one that I have not yet taught to anyone else) and one that is *in principle* private (one that *cannot* be taught to anyone else because its terms refer to inner experiences that only I can have). The latter is his target. He asks us to imagine a sensation for which no public criterion exists — a sensation that I cannot describe in the usual public vocabulary of “pain,” “tickle,” “ache,” but that I try to name with a wholly private sign.
The famous “beetle in the box” thought experiment appears at §293:
“Now someone tells me that he knows what pain is only from his own case!—Suppose everyone had a box with something in it: we call it a ‘beetle’. No one can look into anyone else’s box, and everyone says he knows what a beetle is only by looking at his beetle.—Here it would be quite possible for everyone to have something different in his box. One might even imagine such a thing constantly changing.—But suppose the word ‘beetle’ had a use in these people’s language? If so it would not be used as the name of a thing. The thing in the box has no place in the language-game at all; not even as a something: for the box might even be empty.”
The point is devastating. If the meaning of the word “beetle” were tied to the private object in each person’s box, then communication would be impossible: no one could check whether their beetle matched anyone else’s. But in fact, people do communicate using the word; therefore the word cannot be functioning as a name for a private object. The object in the box — whatever it is — drops out of consideration as irrelevant. The language-game works perfectly well without it.
§256-271: The Impossibility of Naming a Private Sensation
Wittgenstein now makes the argument more concrete by considering the act of naming a private sensation. He imagines wanting to keep a diary about the recurrence of a certain sensation. To do this, I concentrate on the sensation and write the sign “S” in my diary whenever it occurs. But how do I know that I have identified the sensation correctly the next time? There is no independent criterion of correctness. I cannot check my memory against anything public; I can only have another memory. At §258, the argument reaches its critical point:
“I will impress it on myself!—But how do I impress it on myself? I am to produce the sensation again. And that is something. (It is not, presumably, a mere illusion.) But can I really produce it? I do not know. And what is this ceremony for? For what purpose? How does it help?—And indeed it can help me only if I trust myself. And it is no use to say ‘I surely know what I mean by “S”!’—That only makes sense in a game where there is a check on what I mean.”
The ceremony of concentration and inner pointing gives the illusion of establishing a connection between the sign and the sensation, but this connection is actually empty. The problem is not that I might misremember; it is that the distinction between seeming to remember correctly and actually remembering correctly has no application here. There is no difference between thinking I have the same sensation again and actually having it — and where there is no difference between seeming and being, the concept of correctness loses its grip.
The diary-keeper might protest: “But I do have a criterion! The sensation itself is my criterion. I recognize it.” Wittgenstein’s reply is that recognition presupposes a standard of comparison. In public language, the standard is provided by shared practices: we can point to an object, we can show someone the color red, we can calibrate thermometers. But in the private case, there is nothing to calibrate against. The sensation appears, I write “S”, and then later another sensation appears and I write “S” again — but whether these are the “same” sensation is simply a matter of what I decide. The word “same” has lost its normal meaning.
§272-282: The Grammar of Sensation Words
Having dismantled the idea that sensation words get their meaning from privately observed inner objects, Wittgenstein now re-examines how sensation words actually function in our language. At §272, he writes:
“The essential thing about private experience is really not that each person possesses his own exemplar, but that nobody knows whether other people also have this or something else.”
This “nobody knows” is not a factual limitation that might be overcome by better technology; it is a grammatical remark. The grammar of sensation words does not allow for the kind of discovery that would settle the matter. If I say “I am in pain,” the question “Are you sure it’s pain and not something else?” does not have the same sense as asking “Are you sure that’s a table and not a chair?” The first question is nonsensical in ordinary language, not because we have already confirmed the answer but because the logical space for such doubt does not exist.
Wittgenstein develops this in connection with the criterion/ symptom distinction. Pain-behavior — moaning, wincing, crying out — is not merely a symptom from which we infer the presence of pain; it is a criterion. Criteria are not evidence that could be overridden by further evidence; they are the standards by which the application of a concept is governed. When a child cries and we say “He is in pain,” we are not making a hypothesis about an inner state that is only contingently connected to the behavior. We are applying a concept whose grammar is woven into the entire fabric of human life — into our responses of sympathy, our practices of care, our linguistic training.
At §281, Wittgenstein makes the point with characteristic sharpness: “Only of a living human being and what resembles (behaves like) a living human being can one say: it has sensations; it sees; is blind; hears; is deaf; is conscious or unconscious.” This is not a biological fact but a grammatical one. The concept of sensation is tied to the concept of a living human body in a way that makes sensation-language part of our natural history. The inner is not hidden behind the outer; it is expressed through it.
§283-295: The Asymmetry of First-Person and Third-Person Sensation Language
A crucial feature of Wittgenstein’s account emerges in the asymmetry between “I am in pain” and “He is in pain.” These are not two sides of the same descriptive coin. At §246, he writes that the first-person utterance “I am in pain” is not a report about a state of affairs but an expression — something closer to a moan or a cry than to a description. The third-person utterance, by contrast, is genuinely descriptive and is based on observable criteria.
This asymmetry has been a source of endless philosophical confusion. Philosophers have thought that because “I am in pain” seems to have the same grammatical form as “He is in pain,” it must be a report of an inner state that the speaker has privileged access to. Wittgenstein’s reply is that the first-person utterance does not rest on observation at all. I do not find out that I am in pain by introspecting; I simply am in pain. The utterance is not based on evidence, and it is not subject to the kind of error that reports are. If I say “I am in pain,” it makes no sense to ask “Are you sure?” — not because I have infallible knowledge, but because the utterance is not a knowledge-claim in the first place.
At §290, Wittgenstein writes:
“What I do is not, of course, to identify my sensation by criteria; but to use the same expression. For this is the game: an expression of sensation.”
The word “expression” is carefully chosen. A cry of pain is not a description of an inner state; it is a natural, pre-linguistic response. When the child learns to replace the cry with the words “I am in pain,” she is learning a more sophisticated form of expression, not learning to describe something she previously could not name. The language-game of sensations is grafted onto natural expressive behavior. This is why the private language argument is so powerful: it shows that the entire enterprise of naming inner sensations as if they were private objects rests on a misunderstanding of how sensation language actually works. The inner stands in no need of an inner ostensive definition; it already has a place in the public language-game through its connection with natural expressions.
§296-307: The Illusion of the Inner Object
Wittgenstein now turns to diagnose the philosophical picture that makes the private language seem necessary: the picture of the mind as a container filled with objects that only the owner can see. At §304, he presents one of the most delicate moments in the argument:
“And yet you again and again reach the conclusion that the sensation itself is a nothing.—Not at all. It is not a something, but not a nothing either! The conclusion was only that a nothing would serve just as well as a something about which nothing could be said. We have only rejected the grammar that tries to force itself on us here.”
The paragraph is a masterpiece of philosophical balance. Wittgenstein is not denying the reality of sensations; he is not a behaviorist who thinks that pain is nothing but behavior. The sensation itself is real — it hurts, it matters, it organizes our lives. But its reality does not consist in its being a “something” in the sense that the philosophical tradition imagines: an inner object that stands in a relation of reference to a word. The philosophical picture that treats sensations as objects of inner perception is the source of confusion, not the sensations themselves.
The “nothing” that would serve just as well as the “something” is the emptiness of the private object hypothesis. If the word “pain” referred to a private object, then the object’s role in the language-game would be exactly that of the beetle in the box — something that could be absent or constantly changing without affecting communication. But in fact, the language-game of pain is not threatened by the unknowability of other people’s inner objects; it does not depend on them. The grammar of pain is public through and through, and it functions perfectly well without the metaphysical inner object.
At §307, Wittgenstein makes the point explicit: “If I say of myself that it is only from my own case that I know what the word ‘pain’ means—must I not say the same of other people too? And how can I generalize the one case so irresponsibly?” The solipsist’s claim that only he knows what pain is turns out to be not a discovery but a grammatical confusion: he is trying to use a first-person expression as if it were a third-person description, and then drawing metaphysical conclusions from the asymmetry that results.
§308-315: The Purpose of the Investigations
These sections serve as a methodological pause, in which Wittgenstein reflects on what his investigations have accomplished and what they have not. At §309, he writes:
“What is your aim in philosophy?—To show the fly the way out of the fly-bottle.”
The fly-bottle is the philosophical problem itself — the conceptual trap created by misleading pictures of language that we are tempted to adopt. The Investigations do not offer solutions in the traditional sense; they offer clarifications. The goal is not to answer the question “What is the meaning of a word?” with a theory, but to dissolve the question by showing that it rests on a misunderstanding. The fly does not need to fly higher or discover new facts; it simply needs to see that the bottle has an opening it had not noticed.
Wittgenstein contrasts his method with the traditional ambition to discover essences or foundations. At §308, he warns: “The first step is the one that altogether escapes notice. We talk of processes and states and leave their nature undecided. Sometime perhaps we shall know more about them—we think. But that is just what commits us to a particular way of looking at the matter.” The commitment to seeing understanding, meaning, and sensation as “processes” or “states” is not forced on us by the phenomena; it is a grammatical decision that we make without realizing it. The task of philosophy is to bring these decisions to light and to show that alternatives are possible.
The private language argument has not, Wittgenstein insists, proved that there is no inner world, or that we cannot talk about our sensations. It has shown that a certain picture of how that talk works — the picture of inner ostensive definition, private objects, and privileged access — is incoherent. The fly has been shown the way out not by being given a new theory of the mind but by being reminded of how the language of sensation actually works in the ordinary traffic of human life. The inner remains; only the metaphysical picture of the inner has been dissolved.
§316-331: Thinking as a Process
The final sections of this portion turn to another great source of philosophical confusion: the concept of thinking. Philosophers have treated thinking as a peculiar inner process — silent, invisible, occurring in the mind or the brain — that accompanies speech and gives it meaning. Wittgenstein’s aim is not to deny that thinking occurs but to examine what role the concept of thinking actually plays in our language.
At §316, he considers the idea that thinking is a “stream” or “activity” that can be observed introspectively. He asks: when you think, what do you actually observe? “Thinking is not an incorporeal process which lends life and sense to speaking,” he writes at §339, “and which it would be possible to detach from speaking.” The picture of thinking as a ghostly process that runs alongside speech is a temptation precisely because it seems to explain how speech can be meaningful — as if the meaning were a mental act that the words merely express. But this picture leads to an infinite regress: if thinking is a process that gives meaning to speech, then what gives meaning to thinking? Another process?
At §330, Wittgenstein approaches the matter from a different angle:
“Is thinking a process?—’Well, if I say that it’s a process, it’s a process; that’s just the word I use.’—But the question is: what do we contrast it with? If we call it a process, we are already committed to a certain grammar. But the point is to examine what we mean by ‘thinking’—to see what language-game we are playing.”
The confusion arises because the grammar of “thinking” resembles the grammar of “running” or “digesting” — it takes the same grammatical form of a verb describing an activity. But the criteria for thinking are quite different. If I am asked “Are you thinking?” I do not answer by introspecting to see if a certain process is occurring inside me; I answer by considering what I am doing, what I am saying, what I am writing. Thinking is not hidden behind these activities; it is manifest in them.
The crucial insight is that thinking is not a private accompaniment to public behavior but is itself woven into the public fabric of language. At §329, Wittgenstein writes: “When I think in language, there aren’t ‘meanings’ going through my mind in addition to the verbal expressions: the language is itself the vehicle of thought.” This is not a reduction of thought to speech; it is a liberation from the picture of thought as a mysterious inner realm that needs to be connected to language. Language is not the clothing of thought; it is the body of thought.
§332-339: The Grammar of Thinking and the Rejection of the Mental Copy
We are now deep into Wittgenstein’s sustained dismantling of the picture of thinking as a hidden inner process. Having established that language is the vehicle of thought, he turns to clarify what this means for our understanding of the relation between thinking and speaking.
At §332, Wittgenstein considers the temptation to say that thinking is an “activity” that accompanies speaking. He asks us to imagine someone who, when asked what thinking is, replies: “It is a peculiar activity of the mind, which takes place independently of speaking.” The immediate response is to point out that this tells us nothing; it merely substitutes one mysterious picture for another. The real work is to examine how we actually use the word “thinking” in concrete situations. Wittgenstein observes that we say “I think” in many different contexts — when hesitating, when expressing doubt, when formulating a hypothesis — and that these uses do not share a single common essence but are linked by family resemblances.
“Thinking is not an incorporeal process which lends life and sense to speaking, and which it would be possible to detach from speaking, rather as the Devil took the shadow of Schlemiel from the ground.”
This remarkable simile captures the entire thrust of the argument. The image of the Devil taking Schlemiel’s shadow is from Adelbert von Chamisso’s *Peter Schlemiel’s Miraculous Story*, where the protagonist sells his shadow to the Devil and then finds himself unable to participate in human life. The shadow, which is not the substance of the person but is inseparably connected to them, becomes a haunting absence when detached. Wittgenstein uses this to show that thinking is not an independent process that can be detached from speaking; it is as inseparable from speech as the shadow is from the body. To imagine thinking as an “incorporeal process” that lends life to speaking is to create a mythical inner realm that does no explanatory work. The passage illuminates that the supposed “inner” process of thinking is not a cause of speech but is manifest in speech itself. When we say someone is thinking, we are not inferring an invisible activity; we are describing what they are doing — hesitating, calculating, deliberating, formulating.
§340-342: The First-Person Authority and the Impossibility of Error
The discussion shifts to a crucial asymmetry in our use of mental language: the difference between saying “I am in pain” and “He is in pain.” Wittgenstein is concerned with why it makes no sense to say “I think I am in pain” or “I believe I am in pain” — why the first-person present tense seems to have a special status.
At §340, Wittgenstein asks us to consider the grammar of “I think” and “I believe.” He points out that if someone says “I believe it is raining,” we can ask “What is the evidence?” But if someone says “I am in pain,” we cannot ask for evidence of their own pain. This is not because they have private access to some inner object; it is because the utterance “I am in pain” is not based on observation at all. The word “pain” in the first person is not a report of a private mental state but a criterion-governed expression. It is an avowal, not a description. When a child learns to say “I have a pain,” they are not learning to name a private sensation; they are learning a new way of expressing pain, replacing the natural cry with a linguistic expression.
“It is not a something, but not a nothing either! The conclusion was only that a nothing would serve just as well as a something about which nothing could be said. We have only rejected the grammar which tries to force itself on us here.”
This passage at §304 is the culmination of the private language argument’s implications for sensation language. The temptation is to say that pain is either a “something” (a private object) or a “nothing” (a fiction). Wittgenstein rejects both alternatives. Pain is a real phenomenon, but its reality is not that of a private inner object. The grammar of “pain” in the first person is that of an expression, not a report. When I say “I am in pain,” I am not describing a mental object to which I have privileged access; I am giving voice to my suffering. The philosophical confusion arises when we try to treat this expression as if it were a description of a hidden entity. The rejection of the “something” does not lead to the conclusion that pain is unreal; it leads to the recognition that our language about pain does not function the way we think it does.
§343-352: The Grammar of “I” and the Refutation of the Cartesian Subject
Wittgenstein now turns to the grammatical peculiarities of the first-person pronoun itself. The questions at stake are among the most fundamental in philosophy: What does “I” refer to? Is there a subject of experience that is distinct from the body? The Investigations offers a radical alternative to both Cartesian dualism and behaviorist reduction.
At §343, Wittgenstein examines the sentence “I have a body.” He points out that this sentence is not analogous to “I have a car” or “I have a house.” The grammar of possession breaks down when applied to the body because the body is not an object that I possess in the same way. I do not observe my body from the outside and then claim ownership; I am my body in a sense that cannot be captured by the subject-object distinction. The philosophical picture of the “self” as an inner entity that possesses a body is a grammatical illusion.
“The human body is the best picture of the human soul.”
This famous remark at §178 (though the numbering is slightly different in some editions — the remark appears in the context of the discussion of aspect-seeing and the relation between inner and outer) is not a piece of metaphysical speculation but a grammatical observation. It shows that our understanding of another person’s inner life is not based on inference from behavior to a hidden mental realm but on a direct recognition that is mediated by the body. The face is not a sign of the soul; the face is the soul made visible. When we see someone’s expression of pain, we do not first see a contorted face and then infer pain; we see the pain in the face. The body is not a veil hiding the inner; it is the very manifestation of the inner.
§353-363: The Inner and the Outer — The Dissolution of the Boundary
This section deepens the attack on the inner/outer distinction as a metaphysical boundary. Wittgenstein is concerned to show that the distinction between inner and outer, as philosophers have traditionally drawn it, is a grammatical confusion rather than a discovery about the nature of mind.
At §353, Wittgenstein considers the example of someone who says “I have a pain in my foot.” The philosophical temptation is to think that the pain is an inner object located “in” the mind, and that the foot is merely the place where the pain is felt. Wittgenstein shows that this picture is incoherent. The pain is not in the foot in the same way that a nail is in the foot; it is felt in the foot. The grammar of “in” shifts when we move from physical objects to sensations. This does not mean that sensations are “in” the mind; it means that the question “Where is the pain?” has a perfectly straightforward answer in ordinary language — “in my foot” — and the philosophical attempt to locate it somewhere else is a confusion.
“It is a grammatical remark that a ‘private’ or ‘inner’ experience is something that only I can know; or that only I can have it.”
The scare quotes around “private” and “inner” signal that Wittgenstein is not endorsing these terms but examining their grammatical function. The statement that only I can know my own pain is not a discovery about the nature of privacy; it is a grammatical rule about how we use the words “know” and “pain.” We do not say “I know I am in pain” because the concept of knowledge implies the possibility of doubt, and there is no doubt in the first-person case. But this is not because of privileged access to an inner realm; it is because the first-person expression of pain is not a knowledge claim at all. The asymmetry between first-person and third-person sensation language is not a metaphysical fact about the mind but a grammatical fact about how our language works.
§364-375: The Problem of Other Minds and the Criterion of Behavior
Wittgenstein now confronts directly the problem of other minds — the skeptical worry that we can never know what another person is feeling because we only have access to their outward behavior, not their inner experiences. His response is not to provide a proof that other minds exist but to dissolve the problem by showing that the skeptical question is based on a misunderstanding of how “know” functions in this context.
At §364, Wittgenstein asks: “If one says ‘I know that he is in pain,’ what kind of justification does one have?” The answer is not that we infer inner states from outward signs but that we see the pain in the behavior. The criterion for someone being in pain is not the behavior considered as evidence for something hidden; the behavior is the criterion for pain in the sense that it constitutes our grounds for the ascription. This is not behaviorism, which would reduce pain to behavior; it is the recognition that the connection between pain and pain-behavior is not contingent but grammatical.
“Just try — in a real case — to doubt someone else’s fear or pain.”
This challenge at §303 cuts to the heart of the skeptical position. Wittgenstein is not offering a philosophical argument against skepticism; he is pointing out that skepticism about other minds is not a doubt that can actually be entertained in real life. When we see someone writhing in agony, we do not wonder whether they are really in pain. The philosophical doubt arises only when we abstract from the concrete situation and picture behavior as “mere” behavior from which we must infer inner states. But this picture is already a philosophical construction. In actual practice, we do not need to infer pain — we see it. The skeptical problem is not a genuine problem to be solved but a confusion to be dissolved by attending to how our words actually work.
§376-385: The Will and the Grammar of Intention
The discussion now turns to the concept of the will and the grammar of intention. Wittgenstein is concerned to show that the philosophical problem of free will — the question of whether our actions are determined or free — arises from a misunderstanding of how we use words like “voluntary,” “chosen,” and “forced.”
At §376, Wittgenstein considers the sentence “I can will this action.” He points out that the word “can” here does not function like “can” in “I can lift this weight.” It is not a prediction about what I am capable of; it is an expression of my sense of agency. The grammar of “I will” is not that of a report of an inner event that causes action but that of an expression of commitment.
“The grammar of the expression ‘I intend to do it’ is different from the grammar of ‘I intend to do it’ in the third person.”
This observation at §575 (the remark appears in the discussion of intention, though the exact numbering varies across editions) captures the asymmetry that runs through all mental language. When I say “I intend to go for a walk,” I am not reporting an observation of my own mental state; I am expressing my plan for the future. When someone else says “He intends to go for a walk,” they are making a claim based on what he has said or done. The two uses are not related as inner to outer; they are different language-games altogether. The problem of free will arises when we conflate these grammars and treat intention as a hidden inner cause that either does or does not determine action.
§386-393: The Role of Examples and the Grounding of Philosophy
Wittgenstein now reflects on his own method and the role that concrete examples play in philosophical investigation. This is not a digression but a self-conscious articulation of the therapeutic approach that has been practiced throughout the *Investigations*.
At §386, Wittgenstein says that the philosopher’s task is not to advance theories but to assemble reminders for a particular purpose. The examples we have been considering — the builders, the series of numbers, the sensation language, the rule-following — are not meant to prove a thesis but to loosen the grip of philosophical pictures. Each example is designed to show that our language works perfectly well in concrete situations and that the philosophical problems arise only when we abstract from these situations.
“Don’t think, but look!”
This imperative at §66 is the methodological motto of the *Investigations*. It captures the entire therapeutic approach: instead of constructing theories about the essence of language, mind, or meaning, we should look at how our words are actually used in the stream of life. The “looking” is not a passive observation but an active attention to the details of our linguistic practices. When we look, we see that words like “game,” “meaning,” “understanding,” and “thinking” do not have a single essence but are linked by family resemblances. The philosophical problems dissolve when we stop looking for hidden essences and attend to the surface grammar of our language.
§394-404: The Photographic Picture and the Illusion of the Inner
Wittgenstein now returns to the theme of mental imagery and the temptation to think that understanding requires having a picture or image in the mind. This is connected to the earlier discussion of reading and following rules.
At §394, Wittgenstein considers the example of someone who says “I have a picture of this arrangement in my mind.” He asks: “What does it mean to have a picture in one’s mind?” The answer is not that there is a mental object that resembles the arrangement but that the person is able to do something — to describe the arrangement, to draw it, to recognize it. The “mental picture” is not a cause of these abilities but a way of describing them.
“An inner process stands in need of outward criteria.”
This remark at §580 is the culmination of the entire argument against the picture of the inner. It does not deny that there are inner experiences; it denies that these experiences can be identified independently of their outward expression. The criterion for someone’s having a mental image is not their private access to something inner but their ability to describe, draw, or recognize what they imagine. This is not behaviorism; it is the grammatical observation that “inner” and “outer” are not two realms that need to be connected but are conceptually linked through criteria.
§405-411: The Concept of the Self and the Use of “I”
The final sections of the *Investigations* return to the concept of the self and the grammar of the first-person pronoun. Wittgenstein is concerned to show that the philosophical problem of the self — the question of what the self is and whether it is distinct from the body — arises from taking the grammar of “I” too seriously.
At §405, Wittgenstein considers the sentence “When I say ‘I am in pain,’ I do not point to a person who is in pain.” This is a grammatical observation: the word “I” does not refer to a person in the way that “he” or “Ludwig” refers to a person. When I say “I am in pain,” I am not identifying a person and then predicating pain of them; I am expressing pain. The word “I” in this context does not function as a referring expression at all.
“The word ‘I’ does not denote a person.”
This remark is not a metaphysical claim about the nature of the self but a grammatical observation about how the word “I” is used. In the first-person present tense, “I” does not pick out an object; it is part of the expression of the mental state. This is why the problem of the self — the question of what “I” refers to — is a pseudo-problem. There is no something that “I” refers to, but also no nothing: the word has a perfectly clear use in our language, and that use does not require an object of reference.
§412-421: The End of Philosophy — The Return to the Rough Ground
The *Investigations* concludes with a series of remarks that reflect on the nature of philosophy itself and the purpose of the therapeutic method. Wittgenstein is not offering a system or a theory; he is offering a way of seeing through philosophical confusion.
At §412, Wittgenstein says that philosophy is a battle against the bewitchment of our intelligence by means of language. This line captures the entire project: the *Investigations* is not a positive doctrine but a series of therapeutic exercises designed to free us from the pictures that hold us captive. The goal is not to arrive at a correct theory but to achieve a clarity that makes the philosophical problems disappear.
“The real discovery is the one that makes me capable of stopping doing philosophy when I want to.”
This remark at §133 is the goal of the entire work. The ideal is not a comprehensive theory of language, mind, and meaning but a method of investigation that allows the philosopher to see when a question has been answered — or rather, when it has been dissolved. The *Investigations* does not end with a conclusion; it ends with the recognition that the method has been demonstrated and that the reader, if they have followed the exercises, now has the tools to see through other philosophical confusions.
The final remarks return to the image of the fly in the fly-bottle from §309: “What is your aim in philosophy? — To shew the fly the way out of the fly-bottle.” The *Investigations* is not a book that argues for a position; it is a book that opens a door. The reader who has walked through the work with attention has not acquired a set of doctrines but has learned a practice — the practice of looking at language, of noticing the grammar of our words, of resisting the temptation to project philosophical pictures onto the ordinary use of language. The work ends not with a final statement but with an invitation to see clearly, to think for oneself, and to find the way out of the bottle.
