This chapter is far less demanding than the previous one. For that reason, I think I only need to say a few things before you start. As a reminder, I intend for everyone to read these free previews before you read the chapter. In them, I do not summarize for you what Goodman says, but rather I supply what I hope is useful information for making sense of the chapter on your own. I intend for the material behind the paywall to stimulate discussion after you have finished the chapter. I thus aim at two very different audiences: Those who want to dip into Goodman on their own with a little bit of help and encouragement, and those who want to go deeper and engage with me and other readers, as might happen in a seminar.

Subscribe now

Suppose there are two paintings in front of you. On the left is Rembrandt’s painting, Lucretia, not a copy, but the original painting itself; on the right is a brilliant forgery of it. You can not tell them apart. The question Goodman raises is simple: Is there an aesthetic difference between them if there is no detectable difference? Goodman does not treat this question as one of value but of fact.

Goodman’s method in this chapter is typical of analytic philosophers seeking to solve a problem. He raises the question, offers an answer, demonstrates the first answer’s inadequacy, extracts a lesson from that first failure, then offers a refinement. This process continues until the point of diminishing returns, whereupon he stops and assesses (or announces) what must come next. Sometimes, as happens here, the inadequacy of an answer is due to a lack of clarity of the question. Thus, both the question and the answer get better as he proceeds. By the end of the chapter, he will have a rough answer and a program for further development in the following two chapters.

In the last chapter, I pointed out that Goodman was not offering a definition, because a definition would include the expression, “if and only if.” In this chapter, he does offer a definition, but not of a word in common use. He invents a technical term, “autographic,” and offers a stipulative definition, that is, a definition for the term as he will use it in this book, regardless of what it may mean or may come to mean in other contexts.

Let us speak of a work as autographic if and only if the distinction between original and forgery of it is significant; or better, if and only if the most exact duplication of it does not count as genuine.

He also appropriates a linguistics term, allographic, to mean non-autographic. He then follows up with some examples. This autographic/allographic distinction holds out some hope of letting us classify or think about the different arts more clearly. In the rest of the chapter he probes the strengths and weaknesses of this proposed new classification scheme. It should come as no surprise that the seemingly neat and clean distinction turns out to be fuzzier than it appears to be at first. It remains to be seen whether this way of dividing up the arts is fruitful or not. It is worth noting that the distinction is still being used in philosophy or art.

Read more

Sorry about this week’s reading. This chapter is probably the densest one of the book. It took me three sittings, so you should plan accordingly.

The chapter pretends to be about how one thing can expresses something else, but it really develops a vocabulary for discussing the many ways that one thing can mean something else, expression being just one of those ways. To fully explain “expression,” we need clarity about “reference,” “label,” “denotation” from Chapter I, but also concepts like “metaphor,” “possession,” “exemplification,” and much else. All of this gets hashed out in great detail in this chapter.

Goodman’s conclusion doesn’t appear until the chapter summary on the last page, but it’s helpful to know where you’re headed as you trudge down the path. Here it is:

A few things are going on in this formula that need to be pointed out. Analytic philosophers approach definitions differently than dictionaries usually do. Rather than defining a word by giving a synonymous expression, philosophers prefer giving truth conditions for short sentences that use the word. That is, they look at a word in a context, rather than isolated. So, instead of saying what the verb “express” means, Goodman prefers to look at a sentence form like “a expresses b,” and to say what would have to be the case for any sentence of that form to be true (i.e., a list of truth conditions for the sentence).

In the above formula, the names of any two objects can be substituted for the variables a and b to produce a sentence. So the formula says what you are committing yourself to when you say that something expresses something else. Suppose you believe that Picasso’s painting, The Old Guitarist, expresses desolation. By plugging “The Old Guitarist” and “desolation” into the formula, you get this.

One goal of chapter two is to bring the reader to the point where that formula makes sense. Note that it is not a definition but a formula only. An actual definition would begin, “The Old Guitarist expresses desolation if and only if …” Instead, the three numbered statements in the formula only state the necessary conditions for the painting to express desolation. They are not jointly sufficient. That is, even if you think that all three conditions obtain (i.e. hold true), you are not committed to the view that the painting expresses desolation. But if you do think that the painting expresses desolation, you are committed to the view that the three conditions obtain.

Since you haven’t read the chapter yet, the formula shouldn’t make a lot of sense. You must first learn what Goodman means by “possession,” “denotation,” “metaphor,” and “reference.” But you should keep the formula in mind as you read the chapter and each of the necessary conditions get explained.

Unlike in Chapter I, Goodman is not putting forward any thesis in this chapter, nor does he want to weigh in on a theoretical debate in aesthetics; he just wants to offer some precise definitions that all sides of any debate can use without prejudicing the outcome. By adopting the same vocabulary, opposing sides won’t talk past each other. What makes the chapter so mind-boggling is that Goodman doesn’t just give a few definitions; he considers every possible combination and permutation of the terms as they interact with each other. In all this mixing and matching, however, I don’t think there are a lot of surprises—a few jokes, maybe, but not many outrageous claims. After all, he’s just spinning out the consequences lurking within the proposed definitions. But lord can it get tedious.

His definitions hearken back to some basic concepts drawn from set theory, so you need to have a grasp of those or you will get lost almost instantly. Set theory is the abstract study of collections of objects and the relations between them. Here are some terms Goodman employs.

Sets are groups of particular things, called elements. The tires on my car make a set containing only four elements: namely those four particular tires. But the elements of a set don’t have to be near each other. Anything that exists at any time can be an element of a set; all you have to do is declare that there is a set containing it. Thus, the set of Julius Caesar and my left foot is a set with exactly two elements.

Particular things have properties. Blue is a property of many things. Thus, I can talk about the set of all blue things. Sometimes, we want to compare sets with each other, so we have a few terms for doing that. A set consisting only of some of the elements of another set is called a subset of that larger set. For example, the set of all turquoise things is a subset of the set of all blue things.

The elements of a set are called the set’s extension. Two sets that contain exactly the same elements as each other are said to be coextensive. So, the set of tires on my car and the set of tires on my wife’s care are two sets, because their elements are selected on a different basis. But they are coextensive, since we only have one car between us. My tires are her tires.

Well, I hope this gives you a leg up on the reading. Don’t get too discouraged by the first two sections of this chapter. It gets funnier, especially when he starts talking about metaphors in the fifth section, “Facts and Figures,” which is not about statistics, by the way, but about literal and figurative statements. While talking about metaphors, he sometimes gets carried away and says things like, “a metaphor is an affair between a predicate with a past and an object that yields while protesting.”

Subscribe now

Read more

Someone looked up at the sky one night and thought, “You know, those seven stars look just like a bear.” Does that thought ever bother you?

For Aristotle, the basis of all art was mimesis, or imitation. For the most obvious cases—paintings, drawings, mosaics, sculptures—artists shaped their materials to look like a recognizable something, even if that something was purely fictional. With his strong preference for all-encompassing systems of thought, Aristotle went further and claimed that plays, epic poetry, songs, and even music were mimetic. The represent things by resembling them in ways appropriate for their media.

Ever since we’ve been theorizing about representation, resemblance has played a significant role. But the two notions don’t always coincide.

Portraiture, to take one example, has always served the useful function of letting people now what such and such a nobleman or noblewoman looked like. King Henry VIII wanted to know what Anne of Cleves looked like before agreeing to a political marriage, so he commissioned Hans Holbein the Younger to paint her portrait. Apparently, Holbein did what most portrait artists do: he made the representation flattering. So Henry was sore disappointed when he finally met the woman he was to marry. The portrait of Anne of Cleves undeniably represented her, but it did not resemble her to Henry’s satisfaction.

Caricature artists can, with a few strokes of the pen, capture a likeness that can be instantly recognized. They con somehow represent a person in some hard to state way. We think of it as resemblance, but really, how can a few ink lines on paper resemble a flesh-and-blood person? And identical twins resemble each other almost perfectly, but they don’t represent each other.

Obviously, something else must be going on in visual representations besides resemblance. Maybe it involves resemblance plus something? Or maybe we’ve just been wrong all these centuries and resemblance has nothing to do with how pictures represent the things they do. Goodman wants to champion the latter answer.

In true analytic style, Goodman kicks off his theory of visual representation by showing in chapter one that whatever you used to think about representation falls apart under careful questioning. He questions the common view that a picture represents something by looking like it. But he’s not satisfied just to show that some pictures don’t look like what they represent. He keeps pushing to discover where the resemblance theory pushes back. He counters one objection after another, so that, by the end of the chapter, he has swept aside every holdout, including trompe l’oeils, sculptures, and even the rules of perspective.

So how do pictures represent? He won’t be giving a complete answer in this chapter, but the book’s subtitle hints at what’s to come: An Approach to a Theory of Symbols.

Before you read the chapter, here are a few things to know.

Since analytic philosophers dwelt on how words work, they had to come up with some precise, technical terms. Natural language won’t do, because it is filled with ambiguous terms that blur some distinctions. So...

Quotation marks. Analytic philosophers talk about words, objects, and how words and objects relate to each other. Sadly, we use the exact same letters to talk about Spike as we use to talk about Spike’s name. So quotation marks around a term is your clue that you are to focus on the name itself, rather than the thing named. (Linguists, in contrast, usually use italics for that purpose). The word “Spike,” when used as a proper name, points to a particular dog. The word “dog” points to the set of all dogs, of which Spike is a member. “Spike is short,” is about a dog’s height. “ ‘Spike’ is short,” is about the length of Spike’s name. Got it?

Reference. When a word or expression points to something, we call that something the word’s “referent” (or sometimes its “reference”). Common nouns, proper nouns, noun phrases, all have the potential to refer to physical or abstract objects. They could also fail to refer to anything at all. “The whole number between four and five” has no reference, despite being the sort of expression that seems like it could.

Denotation. Goodman’s use of the term “denotation” can be somewhat confusing. It’s also a type of pointing. He borrows the notion from set theory where the elements of a set are called the set’s extension. “Extension” sounds like it means something like “size,” but it’s just the elements themselves. The extension of the set of all current popes is not one, but is Pope Leo XIV, the man. So when a term denotes the extension of a set, it points simultaneously to each member of the set. “Spike” denotes a particular dog. The predicate, “is a dog,” denotes each and every dog in the world, all at once, a multiple denotation.

Hyphenated nouns. Goodman needs a way to classify terms without presuming that they have references. His solution is to have hyphenated phrases function as though they are single words. Thus, a centaur-picture is a type of picture, not a picture of a centaur. There are plenty of centaur-pictures around, and we can usually see right away whether a given image is a centaur-picture or not. We never have to ask whether there are any centaurs to know a centaur-picture when we see one. Centaur-pictures are not unicorn-pictures. But here’s the twist: The difference between a centaur-picture and a unicorn-picture doesn’t depend on some sort of difference between centaurs and unicorns, since neither of those things exist. Things that don’t exist don’t have any real differences. Centaur-pictures and unicorn-pictures denote the same thing, namely, the empty set. But neither of them refer to anything. Nevertheless, there is something about all centaur-pictures that lets us group them together, even if we can’t pin down exactly what they have in common. Wittgenstein would say that centaur-pictures bear a family resemblance to each other. The most important thing to remember is that such hyphenated terms cannot be analyzed. They have no internal logical structure. Think of it this way, there is an “ant” inside “lantern,” but that’s not because ants are inside lanterns. The a-n-t inside “lantern” has no meaning. Similarly, there is no relationship between “unicorn-picture” and “unicorn” other than some of the same letters being used in both terms.

Oh, one other thing. Goodman’s footnotes are worth reading. Social scientists tend to use endnotes that are nothing but citations. Philosophers, though, like to stick things into footnotes that are interesting and important but would interrupt the flow of the main argument. They use footnotes rather than endnotes because they don’t want you to ignore the note.

Well, have a go at Chapter I and see what you think.

Read more

Welcome to the latest Slo-Mo Philosophy read-along adventure: Languages of Art, by Nelson Goodman.

Goodman was in the first wave of American analytic philosophers whose works started appearing shortly after the Second World War. At one point in his life he ran an art gallery and his interest in art carried through his life and career. He made contributions to several different philosophical areas. Writing in a time before publish-or-perish, Goodman only published when he had something strikingly interesting and original to say. As stated in the article about him in the Stanford Encyclopedia of Philosophy (SEP), “Goodman was more interested in solving philosophical problems than in his celebrity as a philosopher. He authorized only two interviews, did not write an autobiography, and rejected the invitation to be honored with a volume in the prestigious Schilpp Library of Living Philosophers.” For this reason, biographical details are somewhat scarce.

Languages of Art is groundbreaking and represents a major turning point in analytic aesthetics. As stated in the SEP article on Goodman’s aesthetics, “Goodman’s general view is that we use symbols in our perceiving, understanding, and constructing the worlds of our experience: the different sciences and the different arts equally contribute to the enterprise of understanding the world.”

Share

I think I’ve finally settled down to a format for these read-alongs, so here is what you can expect.

This read-along is broken up into six sessions, one for each chapter. Goodman has kindly subdivided the chapters into short sections, which you should find helpful if you want to tackle the readings in more than one sitting.

I think that struggling to understand an original work of philosophy is far more valuable than memorizing a bulleted summary of its content. So I do not summarize the readings each week.1 You have to read Goodman yourself. I’m sure there are various places on the internet where you can find philosophical works like this one restated. I’m pretty sure that others have done a better job of simplifying than I ever could, so I won’t waste my time trying. Nevertheless, all philosophical writings rely on a foundation of language and prior philosophical knowledge. That’s where I think I can help you.

The weekly posts will consist of two parts. The free content will contain helpful information that you might need before you read the chapter. Behind the paywall will be further topics and questions for discussion. It’s best to look at that stuff after reading the chapter. Comments are open for paying subscribers, whom I encourage to share their questions, perplexities, objections, counterexamples, and other comments. I know that readers benefit greatly from the threads in the comment section.

One quirky feature of analytic philosophy is that one sentence may be perfectly lucid, the next one perfectly opaque, the one after that painfully obvious, and so on. This book is a classic of analytic philosophy, so you will almost certainly find yourself at some point stepping from solid ground into quicksand without realizing it until you are lost. This means that you must stay alert for the moment when that shift has taken place for you. When it happens, either stop and reread as often as it takes, or skip ahead and come back later. Do not keep on reading with glazed eyes until you pass out.

How do you know when you’ve stepped into the quicksand? When some familiar word suddenly makes no sense the way it is being used. Goodman makes perfect sense. He doesn’t waste words, and he doesn’t use obscure words. He is so precise he may seem to be talking nonsense. Trust me, he’s not. You have to adjust the way you read. To paraphrase John Houseman in The Paper Chase (1973), “You come in here with a skull full of mush; you leave thinking like a philosopher.”

Here’s the schedule of readings

• Friday, April 17, 2026: Chapter 1, “Reality Remade”

• Friday, April 24, 2026: Chapter 2, “The Sound of Pictures”

• Friday, May 1, 2026: Chapter 3, “Art and Authenticity”

• Friday, May 8, 2026: Chapter 4, “The Theory of Notation”

• Friday, May 15, 2026: Chapter 5, “Score, Sketch, and Script”

• Friday, May 22, 2026: Chapter 6, “Art and the Understanding”

I’ll post my comments introducing Chapter 1 this Friday.

This is the final installment of the Wittgenstein read-along. I have found Philosophical Investigations to be exceptionally hard to read, mostly because of its unpolished, fragmentary style. I expect you have felt the same way. After this, I will take a one-week break, and then start the next read-along: Nelson Goodman’s classic, Languages of Art: An Approach to a Theory of Symbols. This book was originally published in 1968; revised in 1976.

Goodman challenges us to think of works of art as things to be read, like a text, rather than primarily as things to stimulate emotions or aesthetic experiences. He proposes that artworks function as rule-governed systems of symbols. This approach to understanding art is so far removed from our normal approach that you should find it conceptually challenging and probably even infuriating. He proposes and defends some unlikely theses like representation is not resemblance and expression does not require feeling. Love it or hate it, this book will challenge everything you thought you knew about art.

I hope you will join me in this next project, and spread the word to others who might be interested. If you can’t get a copy of the book before we begin, each post will be available online, waiting for you to catch up.

But back to Wittgenstein. There don’t seem to be many unfamiliar concepts in this final reading. He continues to compare the way people talk about thoughts, feelings, memories, and so forth with what we can observe in ourselves when we think, feel, or remember. What our introspection reveals is that the way we talk about our own psychology doesn’t match up very well with our honest self-observation. We borrow a vocabulary from other contexts and use it in a psychological context, resulting in a huge conceptual muddle. For example, we discovered last week that we use the word “see” to mean what we do when we recognize a color, but we also use it for what we do when we notice an aspect. But these are very different things. This week, he points to some more evidence: I can choose to see the duck-rabbit drawing as a duck or as a rabbit; but “seeing” a color is less voluntary.

§256: Seeing an aspect and imagining are subject to the will. There is such an order as, “Imagine this!”, and also, “Now see the figure like this”; but not “Now see this leaf green!”.

He then asks us if we could even imagine a person who could not see aspects of things. Even if such an inability existed, it would not be the same sort of inability as not being able to distinguish red from green.

Wittgenstein has not systematically developed a thesis in this book. He has given voice to myriad doubts, suspicions, and puzzlements which all lend support to the idea that language is more than a vessel for transferring information from one person to another. It is a form of life which makes sense only within a life-context that includes much more than language. What he calls a “language game” is a little chunk of that language-plus-context, imagined apart from the rest of life, that allows us to examine how one fragment of language functions. Throughout the book, one examination after another has revealed that longstanding philosophical puzzles are not deep at all, but simply confusions that come from using language imported from one context into another.

I love this line in the final section (xiv):

§371 … [I]n psychology, there are experimental methods and conceptual confusion.

By which I take him to mean that, at least within psychology, the sophisticated, sciency methods of data collection and analysis have raced ahead of any workable conceptual framework that would give the data meaning. The methods are great; the foundational concepts, not so much. The same could probably be said of all the social sciences.

Read more

Section x deals with what Wittgenstein calls “Moore’s Paradox.” This is inspired by a remark by G. E. Moore, an influential British philosopher. He noticed that it seems paradoxical, even if not technically contradictory, for someone to say something like this: “It is raining, but I don’t believe it is raining.” Or for someone to say something like this: “It is raining, but I believe it is not raining.” (Substitute any statement of fact for “it is raining,” and the result is still paradoxical.) Such sentences have a paradoxical feel to them when said in first person, since people usually only assert things they believe. On the other hand, it seems perfectly natural to say, “It is raining, but Fred does not believe it is raining.” I can assert that someone else has beliefs that don’t fit the facts. Likewise, I can critique beliefs that I once had: “It was raining, but I did not believe it was raining.” None of those variations are troublesome. Moore mentioned the issue in passing. Wittgenstein took it much more seriously, and later philosophers, starting with Jaako Hintikka, have taken up where he left off and analyzed these “Moorean” problems to death. There was, as a matter of fact, an entire volume of essays published in 2007, entitled, Moore’s Paradox: New Essays on Belief, Rationality and the First-Person, by Mitchell S. Green and John N. Williams.

Section xi deals with the visual phenomenon of gestalt shifts. This is the section where the duck-rabbit makes its (their?) appearance. When you look at the drawing (in §118), you can see a drawing of a duck facing to the left or you can see a drawing of a rabbit facing to the right. The phenomenon becomes interesting when you switch from one to the other. Gestalt psychologist Gerhard Köhler claims that when you make the switch, the image becomes organized differently, and therefore you are seeing two different objects. Or at least that is how Wittgenstein interprets him. Wittgenstein does not actually mention Köhler (except in passing in §180c) but it’s pretty clear that Köhler is Wittgenstein’s target.

Wittgenstein proposes instead that the person who switches between duck and rabbit is attending to different aspects of the same image. Köhler says the object of vision has changed, that is, one sees a rabbit at one time and a duck at another. Wittgenstein says the object stays the same. One is simply looking at something and “seeing” different aspects of it. Wittgenstein ultimately blames the source of Köhler’s confusion on language (big surprise). We use the word “see” when we look at the duck-rabbit, and also use it when we look at colors. The word, “see,” is a homonym: two meanings but one spelling and one pronunciation. Only context can separate the meanings. When we say we are seeing two different colors (say, when we switch from looking at the sky to looking at a meadow), we use “see” in one sense. But when we say we are seeing a duck and then seeing a rabbit, we are using the word to mean something different. There should be two words for when we see an object and when we see an aspect of an object.

In this reading, Wittgenstein explores this notion of seeing an aspect, and he will continue doing so in the next. He discusses foreground-background shifts, shifts between seeing a drawing as two-dimensional and three dimensional, and other such shifts. I’m afraid I don’t know what illustration he is talking about when he discusses the interlocking hexagons. Sorry.

Read more

Wittgenstein is still interested in how we talk about mental states, but he is focusing more on the states themselves. The edition I’m using numbers the small sections, as in Part I, but older editions don’t. Other than that, text is pretty much the same. We’re reading the first nine units, i–ix.

One of the favorite methods of analytic philosophers was to take a sentence that sounds normal, then replace one word in it with a synonym and ask if the sentence still sounds normal. If it sounds wrong somehow, there must be something interesting going on with that word that got replaced—something is not “preserved” upon substitution. Wittgenstein has used this method a few times, and he does it again in this reading. In unit i (§3), for instance, he says:

“For a second he felt violent pain.”—Why does it sound odd to say: “For a second he felt deep grief”? Only because it so seldom happens?

Here, I think Wittgenstein expects you to say that it’s not because it seldom happens. While violent pain and deep grief seem like they would be similar in important ways, you can’t talk about them in the same way without offending your intuitions about language usage. When we talk about deep grief, we do so in contexts that involve a long duration, whereas pain-talk can involve much shorter times. The rules of all the different language games in which we discuss emotions can have subtle variations.

These variations bring home the fact that the connection between emotion-language and emotion is a lot more complex than we normally think it is. If we say we are angry, are we reporting a preexisting feeling, or are we working ourselves up to a feeling? Or is saying we are angry one way of showing our anger?

The memory of a sensory experience, like a pain, is not itself a pain. It’s a memory. The pain existed in the past, but the memory exists now. The same goes for emotions. A memory of a past emotion is a memory, not an emotion. So it seems fair to ask if the memory of a pain or an emotion is accurate. A lot of memories don’t accurately reflect what actually happened. So, Wittgenstein, being Wittgenstein, wonders if memories of dreams work the same way. Suppose you don’t dream, but you see that other people will start narrating stories in first-person voice right when they wake up, and they call these stories “dreams.” Do you have to believe that these stories accurately describe experiences they had while asleep? Is “believe” even the right word? Maybe you just accept the stories. There is no way to do a fact-check on someone’s dreams like you can with certain memories. Maybe the language games we all play when we talk about dreams have different rules from those we play when we talk about last night’s pains or emotions.

If you look up almost any word in the dictionary, you will see that it has several definitions. When we use words, we typically use them one of their many ways, and the listener also takes them to have one meaning. Yet the word could actually mean any of those dictionary definitions, and variations of contexts open up even more possibilities for shifting the meanings of the words. Wittgenstein asks us if the words we use have a sort of aura about them—a cloud of possible meanings. Is that what speaking feels like? Probably not. But if it is, wouldn’t there have to be a way we quickly settle on one meaning or another for every single word?

Well, maybe not. Wittgenstein says in vi (§ 37):

The meaning of a word is not the experience one has in hearing or uttering it, and the sense of a sentence is not a complex of these experiences. —(How is the sense of the sentence “I haven’t seen him yet” composed of the meanings of its words?) The sentence is composed of the words, and that is enough.

Wittgenstein suggests that it’s not words but entire sentences that leave us with whatever feeling of confidence we have about their meanings. The last sentence in the quoted passage is especially interesting. He’s saying that sentences are functions of their words, but the meanings of sentences are not functions of the meanings of their words. Think about that as you read these sections.

Read more

It seems to me that this week’s reading should require very little help from me. By now, if you’ve been following all along, you should be comfortable with the fragmentary exposition, or at least you have become reconciled to it.

In the first section of this week’s reading (§610), Wittgenstein says,

Describe the aroma of coffee!—Why can’t it be done? Do we lack the words? And for what are words lacking? —But where do we get the idea that such a description must, after all, be possible? Have you ever felt the lack of such a description? Have you tried to describe the aroma and failed?

Wittgenstein doesn’t get very fancy here—he doesn’t have to. “Describe the aroma of coffee!” The aroma of coffee is distinct, easily recognized, and not something we can easily describe without mentioning...coffee. Clearly, coffee smells differently for different people. Some people find it pleasant, and some are nauseated by it. And you would think that such a wide diversity of reactions would call for words, but no, all we have are feelings in search of words.

This observation about coffee was one that I had about wine. I spent a long time trying to master the vocabulary of wine tasting, and it struck me how totally inadequate our words are when it comes to smells. We have hardly any terms that specifically describe odors. Instead, we use the names of the things we are smelling to describe their smells: lemon, grass, skunk, strawberry, and so forth. Or we borrow words from other senses and apply them metaphorically: dark, bright, sharp, loud, green. So, unless you are comfortable with obscure words like nidor, mephitic, or graveolent, (none of which you would use to describe a wine!) you’re pretty much stuck with objects and metaphors.

I think Wittgenstein is right about this. Even when we can instantly distinguish between two smells, we are at a loss for the words we know ought to exist.

What he does next, though, may seem like a bit of a leap. He wants to take feelings that we often do describe in words and ask if they are not equally as nameless as smells. In the rest of this week’s reading, Wittgenstein explores how we talk about the feeling of having an intention. In the first few sections he explores the differences between predicting an event and stating an intention. Predicting that your arm will rise feels very different from stating that you intend to raise your arm. But what is that difference? Don’t you just raise your arm, or do you really have some sort of intention like, “I now intend to raise my arm”?

Wittgenstein asks us to look inside ourselves at the supposed moment we describe as intending to do something and asks us to honestly recollect what was going on inside us. He suggests that all we will find is an inarticulate feeling which we subsequently have dressed up in words. Because we don’t have a vocabulary for these feelings, they only acquire their verbal content as an afterthought, in memory.

In the later sections, he explores the way we talk about past intentions, particularly those that we didn’t act upon. A good case to explore is where you were interrupted when you were about to say something. After the interruption, you may claim you were going to say blah-de-blah-de-blah. But were you really about to say blah-de-blah-de-blah? Were all those words fully formed in your mind and ready to come out as imagined? Or did you just have a feeling of being ready to speak and then explain the feeling by calling it “an intention to say blah-de-blah-de-blah”?

Now that we’ve finished Part One, it seems a good time to shift our attention from language to the mind. So the next three readings will turn to that. We will cover the remainder of the book in three more sessions. Some editions call the material that follows this week, “Part Two,” but the most recent edition calls it “Philosophy of Psychology: A Fragment.” In it, we will meet Wittgenstein’s famous duck-rabbit. Almost everyone is familiar with the picture, but very few know the context in which he used it.

If you haven’t upgraded to a paid subscription yet, you should know that the next work we will start reading slowly in April will be Nelson Goodman’s, Languages of Art. This work makes many highly original contributions to aesthetics and the philosophy of art within the analytic tradition. He covers concepts such as symbol systems, forgery, notation, exemplification, and representation. There will be a lot to discuss and to challenge, so order your copy now. (I’ll be using the edition that came out in 1976, rather than the 1968 original version).

Read more

This week is the penultimate reading for part one. It should give you a few aha! experiences as Wittgenstein draws your attention to how contexts affect the meanings of many things. As before, he hops around from one topic to another, but the insights bear a family resemblance to each other.

We’re all familiar with how malicious critics will take the remarks of an opponent out of context and create a different context which completely changes the meaning. Wittgenstein doesn’t worry about how contextual cues may be misused; he simply notices that every word we speak is uttered at a place and a time that has an effect on the meaning. What words came before it? What words came after? They always lend a tone, mood, feeling, atmosphere, to the word in question.

It’s not just words; facial expressions also acquire meaning from context. Section §539, in which Wittgenstein discusses this, reminds me of the Kuleshov effect, which is something I used to talk about in my philosophy of film classes. The Russian filmmaker, Lev Kuleshov, spliced together some footage of an actor’s expressionless face looking at the camera and some other images. He showed the actor first, then the other image, then the actor again. Depending on what the other images were, viewers attributed different emotions to the actor. Another effect he studied, called “creative geography,” arose from splicing together, for example, an exterior shot of someone entering a building and an interior shot of that person coming into a room to create the impression that the room was inside the building and the action was continuous. This impression would arise even if the shots were filmed at different times and places. Fans of Dr. Who know exactly what I’m talking about.

In several sections of this week’s reading, Wittgenstein writes about mental states or states of mind. Some examples are (1) expecting something and hoping for something, early in the reading, and (2) later on, tacitly believing or knowing. He’s trying to get us to see that we often talk of these states as involving a mental “object,” like an expectation, a hope, a belief, or a bit of knowledge. But if you start to ask pointed questions, you will have a hard time finding that so-called mental object. Take belief. If you sit down in a familiar chair, you believe that it will support you. You don’t think, “I believe this chair will support me,” though. Instead, you have something like a feeling of familiarity or maybe confidence or maybe just exhaustion that accompanies your plopping into the chair. But if someone were to ask you later, if you believed at the time that the chair would support you, you would agree. Obviously, what you’re calling a belief and putting into words never did occupy your conscious mind as a statement of belief.

In other sections, Wittgenstein asks you to reflect on how certain very familiar words get used. They may feel like they mean the same thing every time you use them, but look closer. Take “not,” for example. Does negation work the same way every time you employ it? Or don’t double negatives sometimes cancel each other out, and at other times express emphatic negation? And what about numbers? Does “one” mean the same thing when you say this stick is one meter long and when you say there is one person in the room? And what about the word, “is”? We sometime use it as a copula and sometimes as an identity. In “This rose is red,” we use it to attribute a color to an object. But when we say “Two plus two is four,” we are saying that “two plus two” and “four” are just two different names for the same number. Isn’t it strange how we don’t have to struggle with such differences in meaning for these multipurpose words when we use them? In the context of use, it just feels right to have them perform different functions. What seems like one word, upon closer inspection seems like completely different words.

Finally, he asks you to guess the time and then describe what just happened in your mind. You probably did not consult an inner timepiece and read off the time. Maybe your mind whirled around and around and came to rest on a time that felt right. Whatever is going on seems like the mind being in, or more accurately seeking and finding, a certain state, right before you state your guess.

So I might characterize this week’s material as continuing to explore how context shapes meaning and how our efforts to explain a state of mind inevitably introduce words where there may have been none.

Read more

With the completion of the beetle-in-the-box passages, we have covered all the main issues that have attracted philosophical interpretations and debates. The remaining sections are investigations into a variety of related issues of thought, intention, understanding, and more.

This week’s reading circles around—and once or twice directly touches—a huge problem in the philosophy of language. I can compress this problem into a single word: “not.” There is no “not” in reality, but there is in language. If my friend Mike is not in the room, there is only a room filled with furniture and other things. I may search the room up and down, looking for a missing Mike, but all I can find will be things that are present there. The absent Mike has no reality within that room. What makes this a problem for language is that in the sentences, “Mike is here,” and “Mike is not here,” the name “Mike” has the same positive presence within both sentences. We can’t restate the sentence, “Mike is not here,” as “Non-Mike is here.” They don’t mean the same things. Non-Mike, which might be a rug, can be here at the same time as Mike. But Mike can’t be here and also fail to be here simultaneously.

A photograph, movie, painting, or audio recording of Mike standing and talking in the middle of the room represents someone who is in the room. A photograph, movie, painting, or sound recording of Mike not being in a room just represents those things that are in the room. But they don’t represent the absence of Mike. People can create clever representations that, when explained or interpreted, may lead you to conclude that Mike’s absence is somehow implied. But it will only be language that gets you there, not sensory data.

So, I think the most important thing you need to recognize in most of the present reading is that when Wittgenstein hops around from one topic to another, it all relates somehow to the way we talk about missing things. For instance, he starts by talking about desires and then moves on to expectations. The language we use to describe objects of desire suggests that there is something missing that would satisfy that desire. Expectations are described in positive terms. Thus, saying that we expect a train to arrive is to mention the arrival of a train that has not arrived. As he says in §445, “It is in language that an expectation and its fulfillment make contact.”

Wittgenstein moves to another way we talk about what is not present when he asks, “What does man reason for?” (§466). Reasoning applies most often to that which is not directly experienced. We reason about the future by offering evidence from the present or the past. How can that work except through language or calculation? The future doesn’t exist yet, but sentences about the future do. And the objects or events they refer to may or may not ever exist. But the words do.

So, if you keep in mind the nonsense (i.e., the non-sensed things) that language can be about, then you should be able to see connections between the various topics from this reading. As Wittgenstein says in §464, “What I want to teach is: to pass from unobvious nonsense to obvious nonsense.”

Read more

This week’s reading resumes our investigation into the language we use to describe mental events like sensations; he’s particularly interested in the way we talk about having a sensation. He asks crazy questions about how you might imagine a rock or a chair having a pain or thinking to itself. Of course that’s ridiculous, yet the language we use to talk about our own thoughts or pains transfers very easily to talking about the thoughts or pains of other persons; and as long as we’re talking about thoughts that aren’t our own or pains that aren’t our own, what’s to keep us from locating those pains or thoughts in the body of a chair?

That language involves the grammar of possession. If we talk of my thoughts or Abigail’s thoughts, why not the chair’s thoughts? Where are “my” thoughts located? We may have been told that the brain is what’s “doing the thinking” but introspection doesn’t reveal that to us. Even if every neural activity associated with a thought takes place in my brain, those are just neural activities. They are associated with thought, but not identical with it. We know the two are not identical because they have vastly different properties. Neural activity is neither true nor false, for example, whereas thoughts are.

Wittgenstein challenges the assumption that subjective experiences simply are electrochemical events. The sorts of questions he raises are part of what is sometimes called “the hard problem of consciousness.” Questions about neurons and brain mapping are part of the easy problem—easy because they should sooner or later be answered with the tools neuroscientists already have. Questions about subjective experiences, however, may prove much more intractable.

When we say “I have a pain,” or simply “this pain of mine,” we are using the grammar of ownership. However, pains or any other sensation are not alienable possessions. If I own something, I should be able to give it to someone else. But I can’t give you my pain or my perception of the color green. The fact that I can’t give you my pain, even though I own it, is a puzzle that Wittgenstein suggests arises from grammar. My book, my car, my freedom are things I can hand over to you. Not so with my pain or my thinking. The grammar of “mine” and “yours,” may not reflect the reality. My pain and your pain could hardly be more different. I observe or infer that you are in pain, for example, but I am in pain without observing or inferring it. I have no evidence of pain; there is just the pain itself.

I want to share something with you. This morning, while scrolling through the Guardian, I came across an article about a new book by Michael Pollan, author of The Omnivore’s Dilemma. The book is called A World Appears, and is scheduled for release on February 24 of this year. The article’s title and subtitle caught my attention: “Inside voice: what can our thoughts reveal about the nature of consciousness? — Scientists and philosophers studying the mind have discovered how little we know about our inner experiences.” The article is a long excerpt from Pollan’s book. Apparently, he participated in some empirical research to discover what our minds are doing when we’re not carefully composing our cleaned up version for reporting to others. What will our minds reveal to us if we take them by surprise, as it were? Pollan was given a beeper that would go off at random times of the day. He was to jot down what he was thinking at the exact moments the beeper went off. Pollan shares his personal insight from participating in the experiment:

I’d always assumed that my stream of consciousness consisted mainly of an interior monologue, maybe sometimes a dialogue, but was surely composed of words; I’m a writer, after all. But it turns out that a lot of my so-called thoughts—a flattering term for these gossamer traces of mental activity—are preverbal, often showing up as images, sensations, or concepts, with words trailing behind as a kind of afterthought, belated attempts to translate these elusive wisps of meaning into something more substantial and shareable.

To me, that sounds a lot like what Wittgenstein is trying to help us realize Here’s a link to the article. It is well worth the effort to read it.

I knew that mysterious cosmic forces were at work when I read the following lines, describing an example of Pollan’s thoughts when the beeper interrupted him:

My wife, Judith, and I are seated on the couch, reading. She stops to read aloud a passage from The Man Without Qualities: “Knowledge is a mode of conduct, a passion.” Beep: Huh? What does that even mean, and why does she think I need to hear it?

Really? What a line for me to read at this moment in The Decade Project! The Man without Qualities, by Robert Musil, is the book I’m working through right now, and it’s good to hear of at least one other person reading it. If Judith had been reading Anna Karenina or Wuthering Heights, I would have thought nothing of it, but The Man Without Qualities? What were the chances? When I finish it in a couple of weeks I should have some more reflections to share, but I do remember that line she read to him. It’s from the first volume and I can confirm that Judith was right, Michael did need to hear that passage.

Read more

Here we are, about halfway through the Philosophical Investigations, and if you are still with me you may be totally perplexed; or you may be having one numinous insight after another; or you could be just about convinced that this is all nonsense. Whichever description fits you, I commend your perseverance.

When I was a freshman in college, I fell in with a group of upperclassmen who were studying Heidegger. I sat and listened to them go back and forth, arguing about dasein and existence—which they pronounced “existenz”—and marveled at how they seemed to understand each other. One person would say something, another would counter it, the first would excitedly retort. Hand gestures, facial expressions, tones of voice all played a role, too, not just words. None of this made any sense to me. The same thing happened when I sat with students of the works of St. Thomas Aquinas. They would argue about esse and essentia and the merits of Gilson versus Maritain. At first, I merely observed all this conversation-like behavior, but I eventually tried out a few phrases that I thought might work. Building on a few successes, I eventually joined these language communities.

Wittgenstein would call such conversations language games. They are like games in that participants can make moves that the others would recognize as legitimate or illegitimate. After a while, I learned how to make mostly legitimate moves, although I doubt I could have articulated a strict set of rules that I picked up on.

Such academic conversations strike outsiders as odd performances, but Wittgenstein wants us to see how our everyday talk is no less a language game than these abstruse discussions. For instance, suppose a friend asks us why we accepted a lower offer for a car than our asking price. We may then explain ourselves by stating our intentions, needs, beliefs (about the buyer), knowledge (of the car), and many other internal “things” within us or the buyer. This sort of talk is no less a language game than talk about existentialism or Thomism. It’s just that it’s a language game we have all mastered many years ago. Philosophers call this particular language game—i.e., explaining behaviors by inner states—“folk psychology.” Are there really such things as beliefs or desires? Maybe or maybe not, but we sure do know how to use belief-words and desire-words in conversation, don’t we?

What we’ve been reading so far has nudged us into looking with a skeptical eye at the language games we play when we talk about word meanings, rule-following, reading, and learning. This week, we are going to look at how we talk about sensations like pain or color. When we talk about our sensations, we so easily fall into “playing by the rules” that Wittgenstein has to startle us into wonderment. He does so by taking the grammar of pain-talk out of its usual context. Instead of talking about pain—my pain, a pain I and only I experience, a pain that I cannot compare to anyone else’s pain—try talking about your beetle in a box, the one that you and only you ever see. Everyone has a beetle in a box that they keep with them and never show to anyone else. Beetle-language works by the same grammar as pain-language. Does that mean that pains are just as thing-like as beetles? That seems highly implausible, given that synapses firing all over the brain are not thing-like. Nevertheless, the way we have learned to talk about pain has trapped us into thinking about it in certain ways. We describe headaches, scrapes, and kidney stones as pains. But is there anything they share other than the word? Does every pain necessarily fall somewhere between a smily face and a frowny face?

There are many puzzles about mental states like sensations. It’s the job of philosophy, as Wittgenstein says, in §309, “To show the fly the way out of the fly-bottle” through an analysis of language games.

Read more

If you have been feeling yourself slipping quietly into a coma during this read-along, it’s time to perk up. This week’s passages introduce the famous Private Language Argument (PLA), a line of reasoning that has inspired several books and hundreds of articles. The PLA continues into the next week’s readings as well, with the equally famous beetle-in-a-box example.

We’ve been exploring meaning as use for a while now, and language games have always involved interacting with others. It’s beginning to look like words are always learned and used collaboratively, in ways that involved actions, gestures, facial expressions, changes in tone of voice, and countless other clues that, together with the words themselves, constitute communication. Language seems to be, at least in practice, a very public activity. So Wittgenstein wonders if language is always public? If “always,” is that just by accident, or is a private language impossible for some reason?

The problem to be solved is stated in §243 and explored more or less directly until §315, which is in the middle of next week’s reading. To start with, Wittgenstein again takes the simplest case he can come up with: a symbol that one uses to indicate a non-public sensation. Imagine someone, I’ll call him Fred, who writes a letter “S” on his calendar every day that he experiences this sensation. The sensation has no external criterion by which anyone could determine whether Fred has correctly recognized the sensation or has correctly used the symbol. Fred could be lying, confused, delusional, or wrong in many different ways, but without an objective test for the sensation, only Fred could know if he is mistaken. But sensations pass, so the only way for Fred to check his own use of “S” yesterday is to confer with his current memory of yesterday or look at his calendar. But a sensation and a memory are not the same, and a mark on the calendar is just a mark. What could he compare the memory with to test its validity?

In §246, Wittgenstein says this:

In what sense are my sensations private? — Well, only I can know whether I am really in pain; another person can only surmise it. — In one way this is false; in another, nonsense.

Typical of his style, the question is Wittgenstein’s own question; the answer, set off by dashes, is how he expects his readers to respond; and the next sentence is his conclusion. Of course he goes on to explain his conclusion in greater detail and with many analogies. But it should help you to know that when Wittgenstein uses the words “false” and “nonsense” he is saying two very different things. When most people say that a sentence is “nonsense” they just mean it “is laughably false,” or “is false and the person who utters it deserves rebuke.” But Wittgenstein and most other philosophers say a sentence is nonsense if it has no meaning. That is, a nonsense sentence looks like a statement, but really isn’t one. To be false, a sentence must make a statement that is contradicted by the facts. To be nonsense, a sentence can’t make a statement at all.

This distinction between something’s being false and its being nonsense is very important for understanding much of what follows. His analytical focus will be on the word “know.” He’ll make us wonder if there is a difference between being in pain and knowing that one is in pain. It is false that other people can only know Fred’s sensation by surmising it. It is nonsense to say that Fred can know he is in pain.

I hope that these comments will give you everything you need to follow Wittgenstein’s PLA, which proved to be a turning point in the philosophy of language. My goal here, as always, is not to summarize the text, but to help you read the original work for yourself. I strongly urge you to read Wittgenstein first and only then see what others have said.

If you want to go deeper or check your own understanding, there is no lack of material on the Internet that summarizes or interprets the PLA. Here is a link to what Wikipedia has to say, and here is a much more thorough analysis and critique from the Stanford Encyclopedia of Philosophy.

Read more

These sections (165–205) start out with Wittgenstein trying to figure out, from introspection alone, what is going on when he reads. It’s worth mentioning that he means “reading aloud,” not “reading silently.” At time T1, a reader looks at some marks on paper, and then at time T2 utters some sentences, and we call that “reading.” The marks on the paper do not make a sound, but the reader does. What is going on between T1 and T2? Once again, Wittgenstein engages in an imaginary dialogue in which one voice offers answers and the other voice asks questions and raises objections.

The second voice—the one that tries to answer by explaining what’s happening when we read—comes back several times to the notion that the written text is, in some meaningful sense, “guiding” our vocalizations. If you first read a sentence aloud then compare that experience with the experience of looking at a line of squiggles and uttering a sentence, there is obviously a difference. But what? What makes the experience with the words different from the experience with the squiggles? The circumstances are different, to be sure. The feeling of confidence is also different. But what else? Do we identify each letter, one at a time, recall the sound that it is associated with, then choose to make that sound instead of some other sound? Is that what “guiding” consists of? That seems hardly likely. We don’t even think about what we’re doing, we just do it. But then what is it about just those letters that “lead to” just those sounds?

He gets to the heart of the problem when he notices that nothing we interpret is unambiguous. A picture of someone walking up the stairs is also a picture of someone stepping backwards down the stairs, or a picture of someone on the stairs stomping a cockroach. There’s nothing about the picture itself that compels us to read it a certain way. There may be a popular convention for how to interpret things, but we are not compelled to abide by any convention. So if what we are doing when we read a sentence is interpreting marks on a page, are we thinking, “I’m going to interpret these marks on the page as English words”? That doesn’t match my experience of what I’m doing when I read; but it may be how someone else tries to explain what I’m doing.

When I ask my wife what I think of as a yes-or-no question, she sometimes answers by giving me all the information that should compel a normal person, following normal interpretive conventions, to draw the “obvious” conclusion. Of course I instantly see four or five conclusions one could draw from the same data. So I have to ask again. She accuses me of “just being difficult.” But, really, drawing oddball conclusions from evidence is an occupational hazard for philosophers. We also don’t get picked very often for jury duty.

Someone had to explain to me that this symbol on my dashboard doesn’t mean, “Something exciting is in the cauldron.”

Wittgenstein focuses on the simplest mechanical process in these language games, so as to make the question about reading much more manageable than one about, say, imagining Middle Earth or caring about Anna Karenina’s fate. You don’t even have to understand what you’re reading. Just read a sentence aloud. Yet even that activity is shrouded in mystery, when you try to say exactly what’s going on in your mind that ends with you making just those sounds.

As T.S. Eliot said, in The Waste Land,

Between the idea

And the reality

Between the motion

And the act

Falls the Shadow

As we read these sections, we scratch our heads trying to figure out how the letters on a page influence us to say certain words. Eventually, we suspect that we can’t answer these questions because they presuppose a certain type of answer; and maybe the questions are wrong, not our answers. If language is a practice, as Wittgenstein suggests, someone observing some native speakers can come up with a lot of rules to “explain” what is happening “within” the speakers. But that doesn’t mean that they themselves are deliberately “following” any of those sets of rules.

Read more

In this week’s cluster of sections Wittgenstein transitions from puzzles about language to some introductory puzzles about mental events.

Wittgenstein seems puzzled by things that most of us have never stopped to think about. This can be very frustrating. What puzzles many non-philosophers the most is why Wittgenstein finds language so puzzling at all. And of course “the value question” often kicks in whenever he shares his puzzlement: “What does it matter? I can speak, read, and understand language just fine. What’s the use of understanding how it works.”

This is similar to “Why should I care how electricity works if flipping the wall switch turns on the light?” I would answer both questions the same way: “If you only want to get by, then you don’t need to understand how things work in depth.” But the world and everything in it becomes deeply puzzling when you seriously try to understand it. This deep puzzlement is the source of joy and wonder that draws people to science and philosophy. If you take a drop of water from a pond and look at it closely—very, very closely—mysterious creatures appear. Those creatures have very complex structures and fascinating behaviors. Their behaviors are quite puzzling. Their bodily structures are quite puzzling. Digging further into any of those puzzles only gives rise to more puzzles.

A mere drop of water holds a bottomless well of mysteries. Why would you expect language to be duck soup? Perhaps the most valuable lesson Wittgenstein can teach us is how we too can become puzzled by the utterly familiar.

I described last week how philosophers tried to explain linguistic meaning as some kind of link between words and things. This week, Wittgenstein tries to take seriously the idea that understanding the meaning of a word involves calling up a mental image. So, when someone understands the meaning of the word, “cube,” Wittgenstein wonders how images could possibly help explain how words work. Does our theory of meaning get any better if we say they “have an image of a cube in their head”? We started with a simple theory that a word like “cube” points to actual cubes, so we now say the word causes some sort of mental image of a cube. Does an image point any better than a word does? How sharp or fuzzy is the image? What color is it? Wittgenstein doesn’t exactly ask these questions, but when he shifts his questioning from cubes to images of cubes to projections of images of cubes, he’s expressing how hopeless the picture approach looks when you try to iron out the details. That’s what he gets for trying to take the picture theory seriously.

In section 123 he says “A philosophical problem has the form, ‘I don’t know my way about.’” Then, in the last line of section 153, he says, “I am in a muddle.” That’s how you know we’re really getting somewhere.

It seems to me that Wittgenstein’s flow of puzzlement in these sections looks like the following. What is a word’s meaning if not a thing that the word names? Or a picture of the thing? How does someone “understand” what a word means? Doesn’t understanding a word’s meaning just boil down to knowing how to use the word? Does knowing how to use a word require being able to articulate a usage rule? Suppose someone who says they know how to use a word can’t give a rule for how to use the word; can’t they still know how to use it? Suppose someone knows how to use a word, what was involved in learning how? What’s going on when we learn anything? If we see someone’s test results improving we say they are learning. Is there anything going on inside the learner, or are they just scoring better? If something is going on inside, what does it feel like? At that point, Wittgenstein has opened a huge can of worms that I expect he will be trying to corral for quite some time.

Read more

In the 19^th century and before, scientists and philosophers often operated in the same way. They both tended to make broad, sweeping generalizations and use vague or qualitative concepts. But the hard sciences started to favor quantitative data and increasingly refined mathematics. Scientists built mathematical models of physical systems which treated messy material stuff as though it was a purely geometric or numeric ideal. And they made great progress.

I think that philosophers looked at what the scientists were doing and got envious. The analytical tools that scientists were using were primarily mathematics and various systems of symbolic representation (such as chemical symbols). When philosophers looked at their own tools, they found nothing but natural language which was ambiguous, hopelessly vague, and unquantifiable. They set about trying to “fix” language in hopes of making the same kind of progress with age-old philosophical questions that scientists were making with age-old empirical questions. So they focused on sharpening their analytical tools first and put off using those tools for a while. That went on for well over half a century.

When I was in grad school (in the 1980s), we PhD students were growing impatient with all this tool sharpening. We wanted to show measurable progress on solving “real” philosophical problems. But, even before our time, Wittgenstein was shaking his head in amused disapproval at the analysts. “You’re going about it all wrong!” he seemed to be saying. “The language you already speak is a perfectly good tool. You’re not going to improve it by making it into something else.” (His amusement was due in part to the fact that he had been such a fervent proponent of the language-fixing project.)

In this week’s reading, Wittgenstein continues his critique of the very idea of “fixing” language. I’m sorry, but you can expect some of his paragraphs to be incomprehensible. The audience he wrote for was fully aware of the ideas he was attacking, so he never explained those ideas. But to make things more confusing, but he often expressed himself with playful metaphors. He’s like a beat poet who had once been an accountant. I’m here to help by providing some context.

The discussion of Moses might need some explanation. One failed attempt at fixing a part of language was to say that names are “really” just abbreviated descriptions. “Moses” would then just abbreviate “The man who led the Israelites through the wilderness.” But if that’s the case, then what would it mean to say that Moses didn’t exist? Did the man who led the Israelites through the wilderness not exist? Are you saying that someone else led the Israelites? Do you mean that no one did? There are a number of other problems, but the underlying point is that names cannot mean the same as any description, because you can’t substitute the description for the name in every sentence and have the two sentences still mean the same thing. A true definition (one would think) can always replace the term it defines. “Moses led the Israelites through the wilderness” is informative. It tells us something we might not have known. “The man who led the Israelites through the wilderness is the man who led the Israelites through the wilderness,” tells us nothing at all. Wittgenstein doesn’t come right out and put it that way, but I think that’s the heart of the matter.

In several of the passages here, Wittgenstein makes the point that there are many times when we don’t want to make a vague term precise. Sometimes, a word’s imprecision is exactly right. No cook wants a recipe to state exactly how many salt grains to use when all you need is a pinch. A chemist, however, might need more precision. Each use of a word has its expected level of precision in a given context.

In §93, Wittgenstein criticizes a viewpoint he characterizes by this exclamation: “A proposition—That’s something very remarkable!” If I’m reading him right, I think the exclamation is a cute way to encapsulate a huge body of literature in the philosophy of language. If declarative sentences are connected with facts, that implies a bridge connecting them. A sentence is always in a particular language: German, French, Italian, or what have you. But you can declare the same fact in many languages. Suppose the relevant fact is that snow is white. Then “Snow is white,” and “Schnee ist weiss,” both point to the same fact. How? They both mean the same thing. So the two sentences must express the same proposition, that snow is white. And it’s the proposition that points to the white-snow fact. Facts are states of affairs but propositions are some sort of ephemeral, linguistic entities floating above sentences, connected to the world by pointing to exactly one fact.

If you think this sounds weird, you’re in good company. I’m just touching the surface of the complexities that arise from trying to make language excruciatingly precise. There’s a lot more weirdness where that comes from. Wittgenstein is advising us to sweep it all aside, but he’s not just advising, he’s making a strong case. He has to make the case, because he’s challenging what the vast majority of analytic philosophers were doing at the time.

I am not totally confident that this is what he’s getting at in §93, because in the translation I’m reading, the German word “Satz” is translated sometimes as “sentence” and sometimes as “proposition.” However, I trust the translator to have gotten it right.

The last thing I’ll mention is that he challenges the whole concept of following a rule. It’s not that we can never follow a rule, but rather that we can’t fully explain an action as simply following a rule. Following a rule is itself an action that follows a rule. For example, if I capture a pawn with a bishop in chess, it’s not good enough to say I applied the rule, “Bishops move and capture diagonally.” I also have to apply the rule, “Pieces that look like this are bishops.” Then there’s a rule about what counts as a piece, and so forth.

Read more

To understand what’s going on in §§ 37–64, it helps to have some context. Philosophers of language had been trying for years to understand words as somehow referring to objects (not always physical objects). However, there was obviously a problem when trying to make this theory work for words other than nouns. When prepositions, adjectives, verbs, conjunctions, negations, and so forth occur in a sentence, they don’t themselves point to objects, but they must be explained as somehow related to the objects mentioned by the nouns in the sentence. A theory of language that tries to analyze language solely in terms of references to words is a “referential theory of language.” Referential theories actually met with a lot of success at first.

Transitive verbs, for instance, could be thought of in terms of ordered pairs of objects. “John hears Sally,” would thus point to one element of the set of all ordered pairs, (x, y), such that x hears y. [So, “John hears Sally” might be represented in logic as “Hjs,” where “j” means “John,” “s” means “Sally,” and “Hxy” means “x hears y.” Replace the variables “x” and “y” with the names “j” and “s” respectively, and voilà! the English sentence has been rendered into logic.] A project like this—trying to understand all language as making statements about objects—could, if successful, allow anything worth saying to be translated into symbolic logic. In pursuit of this goal, many philosophers worked on the problem from one angle by expanding the power of symbolic logic, and many philosophers worked on the problem from another angle by extracting the inner logic of natural-language sentences.

This and related activities called for an enormous amount of analytical effort, all in the service of removing ambiguity from natural languages (English, German, Chinese, and so forth). The idea was not to provide word-for-word translations from, say, English into “Logicese,” but rather to express the same things in the language of logic that English sentences express. Grammar specific to the natural language might have to give way, and short words might need to be expanded into long logical expressions in order to preserve the meaning. Ambiguities would have to be resolved. Homonyms must go. The task was daunting, but the rewards in terms of clarity promised to be enormous. Imagine how useful it would be to have a language in which it is impossible to be vague or ambiguous: every sentence would have to be either true or false with no in-between. Science would be well served, and the most intractable problems of philosophy would probably simply disappear.1

This is the optimistic background against which you must understand Wittgenstein’s skepticism. Names, as you will see in this week’s reading, proved surprisingly resistant to analysis. A name, properly considered, was supposed to “attach” uniquely to an individual object, somehow. (In reality, of course, several things often bear the same name, but that’s just one of those messy features of language that would disappear once philosophers had cleaned it up.)

When you’re reading the sections for this week, be aware that Wittgenstein’s puzzlement about names is mostly his puzzlement about then current referential theories about names. For instance, if “Nothung” is the name of Siegried’s sword in Wagner’s Ring, what happens to the meaning of the name when the sword itself is shattered? If the meaning of a name is nothing but its reference, and the referent is destroyed, does “Nothung” become meaningless? Well, obviously not, so how can a referential theory of names deal with that issue? That’s the sort of thing he’s grappling with.

On a personal note, I’ll say I have a great fondness for the referential approach to language, quixotic as it is. I took a course in grad school on definite and indefinite articles (“the” and “a” or “an”); I took another course on the logic of non-existent objects; and in my dissertation I proposed a definition of “about” from within the referential framework.

This enormous referentialist project ultimately failed, but not for lack of effort. Analytic philosophy has moved on, in large part due to the reflections you are reading in Philosophical Investigations.

Sections 37 through 64 are, I think, among the most difficult passages that we’ve read so far, mostly because Wittgenstein focuses so narrowly on problems with naming. Just keep in mind that everything in these sections is related to how names function. He cares about “this” and “that” only because they seem to pick out individuals, just like names do, but they clearly use some other mechanism than naming. A simple is the smallest object named in a given language game, and a composite would be an object made up of simples. Everything in these sections relates back to theories of naming.

When we get to section 65, he returns to the notion of a language game. He’s been using the term and providing examples, but for a couple of sections he tries to pin down exactly what a game is. Philosophers have long been obsessed with definitions. We always want to pin down the precise meaning of a term. The Greeks used definition per genus et differentiam, or defining a term by stating the larger class to which it belongs (its genus) and also stating a differentiating characteristic (specific difference) that carves out the precise subclass of the genus. Here’s my favorite example. An arctophile is a collector of teddy bears. The genus is the class of collectors, and the specific difference (teddy bears) is what distinguishes arctophiles from philatelists (stamp collectors), numismatists (coin collectors), or cartophilists (cigarette-card collectors). When Wittgenstein tries to define “game,” he can’t come up with a neat definition that captures everything most people normally think of as a game, yet excludes everything that is not a game. So he throws up his hands and says that only a “family resemblance” connects all games, not any one or two specific characteristics.

The last thing I think I should point out is that some of these passages read like snippets of dialogue. When Wittgenstein wants to argue with another point of view, he puts the opposing voice inside quotation marks. Thus, he can have a back-and-forth discussion. Of course, that can be confusing if you don’t keep Wittgenstein and his devil’s advocate voice distinct.

Read more

Philosophical investigations into what? Into language, thought, and meaning. We use language so naturally that we seldom stop to marvel at what an extraordinary thing it is. The Philosophical Investigations will shake you awake and make you take a good long look at some things we take for granted.

It’s best, in my opinion, to think of these writings as in-progress investigations; they are not reports about past investigations. They are themselves the investigations, and you are an experimental subject. Thus, for as long as you are reading them, the investigations are ongoing. You must not take anything in these pages as a conclusion. Later, certainly, you may reflect on what you read and draw some conclusions, but while you are reading Philosophical Investigations, you should think of yourself as being experimented upon. Don’t read quickly. Stop and process what you’ve read after each numbered section.

Philosophy of language is so hard to understand because we normally use language to talk about things other than language. Ordinary language is transparent, in a sense. We access the world through it. It gets weird when we try to use language to analyze itself. When you start doing that, you may feel like you’re wandering around in a maze. Here’s an example. Consider the following, call it P1:

P1 says nothing about a person. If it says anything at all, it says it about a string of four letters concatenated with each other, namely ‘J’, ‘o’, ‘h’, and ‘n’, in that order. And it says that that string of four letters is a short string. How do you know that P1 is not about a person? The pair of single quotation marks on either side of the string announces that we are not using those letters in a normal, transparent manner.

But what if I were to tell you that P1 doesn’t say anything? What if I’m using P1 to illustrate something rather than report something? By placing P1 on a line by itself between two paragraphs, I am merely displaying some letters, punctuation marks, and spaces, without making an assertion with them. P1 is simply some shapes I want you to consider. For P1 to say something, I would have to use it as a sentence. But actually, I’m just showing you what I mean by a sentence.

When we use words to talk precisely about words we quickly realize that we have to become creative. Creative authors need cautious readers. That’s why you need to pause frequently as you read these sections and let them work on you. See what happens.

Wittgenstein starts off in §1 with a passage from St. Augustine’s Confessions, in which Augustine offers an account of how he came to acquire his first language as a child. This account seems simple enough to be uninteresting. Augustine says grown-ups would indicate or point to an object then say its name, and after a few times, he caught on and started talking. OK. Let’s go with that. Wittgenstein grants, for the sake of argument, that Augustine’s account is correct and then he tries to imagine how it could possibly be correct. Just exactly what must be the case for someone to acquire a first language in the way Augustine describes?

Wittgenstein creates an unrealistically simple language containing only four words (“brick,” “pillar,” “slab,” and “beam”) and imagines a context in which that language could work. By doing so, he makes us focus on only the relevant features of a language made up of nothing but nouns. Using this bare-bones language, when someone says “Brick!” the person he is talking to picks up a brick and carries it from one pile to another. Pretty simple language, right? Problems start right away when we ask how could anyone possibly acquire that language, without knowing any other language already? Dad takes Junior to a pile of bricks, points to one of the bricks, and says, “brick!” how does that lead to Junior doing anything with any of the bricks in the pile? How does pointing function in this language? If he then points to a place where he wants the brick to be placed, does pointing to that place mean the same as pointing to a brick? Isn’t the pointing at the brick part of learning the language (by offering an ostensive definition), while pointing at the place part of directing an action (by issuing a command)? How can Junior know what exactly Dad is pointing at (a brick, a color, a shape)? What makes sticking out an index finger an act of pointing, rather than, say, complaining about a splinter?

In this week’s passages Wittgenstein tries to understand language as though it were a very complex game that we all play pretty well by following certain rules, even if we can’t formulate those rules. To simplify the analysis, he creates tiny versions of the full game, versions with very few “pieces,” “moves,” and “rules.” He calls these mini-languages “language games,” and studies them in hopes of figuring out what sort of thing must be going on when we are using the full version.

These opening passages of Philosophical Investigations touch on a wide range of puzzles about language and thought. What are words? What are names? What are letters? What does it mean to say that a word means something? Do words actually mean anything, or isn’t it only complete sentences that mean things? What’s going on when we define a word by pointing at an object? No matter where you start investigating language, the puzzles multiply at a dizzying rate. There are so many puzzles in fact, that solving any one of them may not be possible without solving all of them together. So you’ll find that Wittgenstein walks around an issue, asking a question here and there, weighing one plausible answer against another to see which approach looks the most promising. He’s getting you to appreciate the contours of the huge puzzle that is language. Don’t expect any conclusive answers right away. But you should find yourself wondering if maybe looking for meanings is a dead end.

In the edition I’m using, the translator puts an asterisk in the left margin to indicate a note about that paragraph at the end of the book. These notes are especially helpful in understanding Wittgenstein’s comments about Frege, as in §22.

Read more

Announcing a new Slo-Mo Philosophy project: Wittgenstein’s Philosophical Investigations.

Starting Friday, January 2, I will begin reading this cornerstone of ordinary language philosophy by Ludwig Wittgenstein. It was a turning point in twentieth-century philosophy that redefined what analytic philosophy should be. I hope you will join me in this venture.

Philosophical Investigations is one of the most important works of philosophy written in the twentieth-century, not because it proves any theorem or defends a thesis, but because it raises thorny problems for how we think and talk about thinking and talking. Since thinking and talking are pretty much all that philosophers do, the problems Wittgenstein raises strike at the heart of the entire philosophical enterprise. Even if you’ve never read anything by Wittgenstein before, you almost certainly have seen or used some of his ideas. The duck-rabbit drawing, language games, family resemblances, and the private language argument are just a few of Wittgenstein’s inventions you may have come across.

I am an analytic philosopher by training, and analytic philosophy in the last part of the twentieth century developed mostly in the shadow of Wittgenstein. However, I haven’t read this book since graduate school, over forty years ago. So I will approach this slow reading more as a fellow learner than as a teacher. Wittgenstein’s musings never fail to spark feelings of wonder in his readers about things they had always taken for granted. To my way of thinking, that means some good times ahead.

Share

Which edition?

Wittgenstein worked for many years on the material now called Philosophical Investigations, but he did not complete his task. His thoughts were too important to be lost in obscurity, so the British philosopher, G. E. Anscombe, assembled passages from Wittgenstein’s manuscripts and translated them into English. The first German-English bilingual edition came out in 1953, two years after Wittgenstein’s death, and there have been a few subsequent editions since then. Ultimately, though, all versions derive from Anscombe’s assemblage. I will be using the revised 4^th edition by P.M.S. Hacker and Joachim Schulte, and I recommend that one. However, any other editions should be fine. The point, after all, is to become inspired by this non-systematic cluster of ideas.

Most editions have the German and the English on facing pages. There are two unequal parts. The longer Part One consists of 693 numbered sections. The numbering is the same in every edition, so there should be no problem locating passages. The short, second part begins after section 693, and the numbering for it is a bit trickier. Even though the numbering is consistent, there may be a few scattered fragments in both parts that appear in one edition but not in others, but I don’t think that will cause much of a problem. I have a couple of early editions that I can refer to if necessary.

The reading plan

I want to start this project at the beginning of the new year so everyone has time to get their own copy. I hope to stick to the following schedule. The numbers refer to sections, which are seldom more than a page long and often just a few sentences. These numbers in Part I are consistent through all editions. So I’ll refer to section numbers instead of page numbers.

• January 2, 2026: Sections 1–36

• January 9, 2026: Sections 37–70

• January 16, 2026: Sections 71–122

• January 23, 2026: Sections 123–164

• January 30, 2026: Sections 165–205

• February 6, 2026: Sections 206–282

• February 13, 2026: Sections 283–355

• February 20, 2026: Sections 356–432

• February 27, 2026: Sections 433–524

• March 6, 2026: Sections 525–609

• March 13, 2026: Sections 610–693

End of Part I

The next 372 sections were called Part II, in earlier editions, but the 4^th edition calls them “Philosophy of Psychology—A Fragment.”

• March 20, 2026 : Sections 1–85 (i – ix in earlier editions)

• March 27, 2026: Sections 86–211 (in earlier editions, start with x and stop at the paragraph in xi, shortly after the triangle illustration, that begins “The aspects of a triangle...”)

• April 3, 2026: Sections 212–372 (in earlier editions, start from where we left off and read to the end of the book)

A note on subscriptions

Anyone may read along for free. Every Friday I will post a few remarks that I hope will help you understand what you will be reading. For paid subscribers, I will open the comments and offer some thoughts for further discussion. I will try to respond to every comment, but I also hope you will engage with each other.

Subscribe now

Well, kids, we’ve reached the end of the road. Congratulations for making it all the way through.

I think every chapter has held at least one surprise for first-time readers, but this one takes the cake. I’m pretty sure you didn’t expect to learn about the eight Sirens standing on the spheres of the planets, each singing one note of a chord, accompanied by the three Fates singing a melody. Nor did you expect to take a trip to that big sorting room in the sky where souls choose their next reincarnation. Am I right?

There is a translation issue here that Waterfield mentions in his notes. Up till now, he has been translating the Greek word “psyche” as “mind,” and he continues doing so; but the arguments at the beginning of this chapter are ones we might normally think of as applying to souls, rather than minds. I use “mind” and “soul” interchangeably.

Socrates thinks that he’s accomplished his main goal of proving that goodness is its own reward and that no external consequences make any real difference. Nevertheless, the way things work out in this life often seems unjust. Socrates agrees that when people cheat, they almost always start out seeming to win. He quickly reassures us that things always do turn around, given enough time. I take him to be saying that the apparent unfairness of life comes from its being too short. That’s why I think he needs to show that soul never dies. Bad guys eventually lose and good guys win. Eventually.

The argument for the soul’s immortality assumes a complete split between body and mind, of course. Each can only be destroyed by its special nemesis—sort of like superman and kryptonite. The body’s kryptonite is ill health, and the soul’s is immorality. But whereas a body can be killed by ill health, a soul can only be harmed—not killed—by immorality. Since nothing but immorality can so much as harm the soul, it can never die.

There is one more problem, though: the soul seems to be a composite, but nothing composite is immortal. I can’t tell whether Socrates accepts this view about composition or whether it is an objection that some people might raise. The idea is that if something is not perfectly simple it can only hold together for so long. It will eventually fall apart. So the soul, if it is a mixture of parts, would also have to decompose. Socrates doesn’t argue against this objection, but instead tries to get around it. He suggests that the soul, in its journey through life, picks up things that are different from it like a ship picks up barnacles. To see the pure soul, you have to strip away all the accumulated grime from a life on Earth. Presumably, that excess baggage comes from the functions the soul needs while encased in a body: the functions we previously called the appetitive part and the spirited part.

Finally, Socrates launches into the bizarre myth of Er, who dies and comes back to life again. During the time between death and reanimation, he sees the workings of the cosmos and what happens at the soul recycling plant.

My interpretation of this myth is that Plato wants to throw a bone to those readers who can’t do the hard work of following the arguments. “If you can’t understand my reasoning,” he seems to say, “Let Socrates here tell you this nice myth you can believe instead.” But then, the myth isn’t all that easy to follow either. As with the numerological stuff in an earlier chapter, there are some very complex descriptions in this chapter that remain mysterious to modern interpreters. Waterfield’s notes are again very helpful.

For those of you who have free subscriptions, I hope you have enjoyed this slow reading of a major philosophical classic. Now that you’ve read it from cover to cover, I suggest you celebrate by watching the movie, The Village (2004), directed by M. Night Shyamalan. Even if you’ve already seen it, you might experience it differently after having read the Republic.

If you decide at some point to upgrade to a paid subscription, all these posts will still be in the archives, and I will still respond to any comments you leave. The same goes for the first slow reading project of Descartes’s Meditations on First Philosophy.

The next classic slow-reading I’ll be hosting is Wittgenstein’s Philosophical Investigations. I’ll be posting something about that project soon.

Read more