What Happens When Wittgenstein Asks "What Happens When ...?"

E.T. Gendlin
University of Chicago

Wittgenstein insisted that rules cannot govern speech, because they are formulated only from the practice of speaking, and only by it. He also rejected observational reports as a basis for understanding language. It has long been an open question, just where Wittgenstein is speaking from.

Before we can consider this question, we must be concerned with a prior problem: It is generally agreed that Wittgenstein rejected the usual assumptions of objectivism with its basis in external observation, as well as the intellectualism which finds its basis in rules, convention, or construction. But most discussions of Wittgenstein involve some of these assumptions nevertheless, not because his rejections are missed, but because no other discourse is available.

As David Pears says, "Most recent accounts of Wittgenstein ... start from the assumption that there really ought to be a definite rule dictating the use of every word ... [1] Others assume that statements must be based on observation. The account of Wittgenstein then consists in saving as much of these assumptions as one can. Or, if they cannot be saved at all, Wittgenstein appears as if he were concerned with what cannot be said, something "ineffable." It seems that one cannot avoid beginning with assumptions he rejected, then trying to account for their rejection, always still in a discourse that assumes them. But this is not due to any oversight of ours. It is Wittgenstein who has made discussions about his work difficult if not impossible. He said he could not say, only show. He was convinced that what he showed could not be talked about. We must examine his reasons, but this requires the very discourse which they make impossible.

The problem is not only ours; it was also a problem for Wittgenstein. He knew that he was usually misunderstood, but he did not think it possible to speak about the showing he was doing. To do so, one would have to present language as if it were an object that could be presented. To substitute such an artificial representation would be misleading, and would miss how words actually work — not as representations, nor based on representations.

Wittgenstein does his showing by means of language, of course. What mode of language is employed in only showing? This question must be asked, but Wittgenstein did not think it could be answered. He thought that language could not turn to talk about this showing use of language. He thought that the attempt to say would lose the showing mode.

In the first part of this paper I will propose a mode of discourse that does not assume what Wittgenstein denies. With the aid of this I will return to the question of Wittgenstein's standpoint. Perhaps we can then at least begin to say where he is speaking from.

So far it has become clear that we face a dilemma: If we talk about Wittgenstein's showing, we exceed the bounds he set for himself, but if we do not, then we cannot make sense of his position. How to navigate between these two pitfalls is the problem. Rather than pretending to solve it, I will traverse the problem in very small increments, pausing at each juncture to examine exactly what in Wittgenstein we may have violated.

At no point do I intend to exceed Wittgenstein in just the way that he prohibited, by substituting representations for the actual working of language. I think we can develop another way. It would be something that he does not do, but it might enable us to say what he does,

1. Affirmative uses:

First I will stay within what Wittgenstein does. The violation will consist only in selecting certain of his phrases; we will still use them only as he did. Later I will mark where we exceed what he was willing to do with them.

It is often said that Wittgenstein dispelled mistakes but did not assert anything. This is not quite so. He said that he could only show, but let us notice: He did assert that he could show. We also find him constantly asking questions and answering them with examples that involve quite affirmative statements Let me call your attention to some characteristic phrases with which Wittgenstein asks and answers himself.

For example, while he denies that a concept or common feature governs the different uses of a word, he also affirms that they share only a family resemblance. This is obviously a metaphor. Later we will ask how it differs from the metaphors against which he constantly warns.

In a great many instances in the Investigations Wittgenstein asks "What happens when someone says ...?" (Was geschieht?) and "Ask yourself ..." and "Let us look and see what [this] really consists in" (for instance 578). [2] There are many variants of this question. Then he cites many possible happenings.

He says, also affirmatively, that what happens is "more intricate" (verwickelter 182) than a single referent would be. The greater intricacy of what happens is shown by the long series of examples. Later we can try to say what they show, but Wittgenstein did not think one could do that.

For the moment our only violence has been to select and set out certain of his phrases: "only show," "family resemblance," the question "what happens?" with the examples, and "more intricate," all of which are affirmative or at least not negative.

2. Using his words to speak about his use of them:

Now let us move a step further. We will need some words with which to speak about his use of these words. Rather than importing a discourse other than his own, I will use only Wittgenstein's affirmative words for the next phase, and only in his sense. But let us be conscious that we are moving further than he would, as soon as we employ the words to speak about how they are used.

Let us ask what happens when Wittgenstein asks the question: 'What happens when ... ?'

We exceed Wittgenstein by turning the phrase to ask about its own use, but let us not exceed him in how we answer. Let us answer it as he would, and by using only these phrases. What happens? Well, ask yourself, and look and see what [this happening] really consists in.

To do as he would ask, let me cite two of his passages, In the first, Wittgenstein denies (as he does so often) that a meaning is a single object that exists and can be referred to. Here he takes up a case when the existence of a single meaning-referent seems almost unavoidable. When we struggle to find the right words to express something, isn't that a meaning-referent? We assume that we have the meaning there, since we are trying to find the "right" expression. But Wittgenstein asks:

"What happens when we make an effort, say in writing a letter, to find the right expression for our thoughts? This phrase compares the process to one of translating or describing: The thoughts are already there ... and we merely look for their expression. This picture is more or less appropriate in different cases. — But cannot all sorts of things happen here? I surrender to a mood and the expression comes. [his italics] Or a picture occurs to me and I try to describe it. Or an English expression occurs to me and I try to hit on the corresponding German one. Or I make a gesture, and I ask myself: What words correspond to this gesture? And so on." (335)

Consider a second passage: Wittgenstein is arguing that there may be, but need not be, an inner process when we say we "expect" someone or something.

... 'I am expecting him' would mean 'I should be surprised if he didn't come' and that will not be called the description of a state of mind. ... But we also say 'I am expecting him' when I am eagerly awaiting him. We could imagine a language in which different verbs were consistently used in these cases. (577)

When I sat down on this chair, of course I believed that it would bear me. I had no thought of its possible collapsing.

But, 'In spite of everything he did I held fast to the belief ... [that he will come]' Here is thought, and perhaps a constant struggle to renew an attitude.

I watch a slow match burning, in high excitement follow the progress of the burning and its approach to the explosive. ... This is certainly a case of expecting. (575-576)

If we employ only Wittgenstein's phrases, and do so only in his fashion, we can minimally talk about how these phrases are used. What "happens" when he asks this? We can say he tells what happens (or the sort of thing that may happen) and this shows what "happens" means. So it also shows what "shows" means, and what "more intricate" means. Later we may become able to use other words to say what these words mean, but now we are only showing it, (and only showing what "showing" means).

When Wittgenstein asks what happens when we use these expressions, he is far from denying what most people call "subjective" events. He does not consider them subjective, but he constantly answers his question by speaking of them. He shows that what may happen is more intricate than one categorizable inner process. [3]

Wittgenstein is even more interested in cases when no inner events happen, although we seem to refer to one, for example when we say that we "remember," "read," "expect." He brings this home by contrasting cases in which no mental events happen with cases in which they do. Wittgenstein clearly says that mental events do happen in some cases, not in others. [4]

Can we call Wittgenstein's array of what may happen "examples?" What do they exemplify? What do they have in common? So far, let us say only that each group of happenings may happen when we use the same expression (for example "expecting"), but these uses share only a family resemblance.

So far we have answered always again only with his own words. Before we go further, let me say why it seems wise to remain for a while just with Wittgenstein's own words. Suppose we try saying more about what happens, for example where it happens. We might say it happens in a situation. Wittgenstein does sometimes talk of the situation (for example 337), but this is easily misunderstood as something observed from the outside. We would need to argue that "situation" must be understood in a certain way, Wittgenstein's way. If we say that the happening happens "in situations," how could we bring it about that the word "situation" would be used in Wittgenstein's way? We would have to cite his examples, (some of which are externally observable and some of which are not). So it would not help us at this stage, to say that this all happens "in a situation," since this use of "situation" would still require the examples. We would have to define "situation" by what happens when Wittgenstein answers the question "What happens when we ..."

Since any other words we might import here would require this return to the examples of what happens, we might as well use the original phrase about itself. Since other words would depend on the examples of what happens, we would gain nothing by importing them. Let us remain for a while with "what happens" and use the phrase itself to say (as Wittgenstein does not) how this phrase gets its meaning. We answer: "What happens?" means the sort of things that the examples show.

3. What happens in the examples and what happens here with them?

Let us go a half step further simply by recognizing that we have answered two questions at once, when we just said that the examples show what may happen. We pointed to the happenings which are shown in the examples, but in doing so we also used the phrase to say what happens here, namely that the phrase gets its meaning from the examples. The latter is what happens when Wittgenstein asks "What happens when we say ... ?"

Our phrase works in two ways: It says what happens when we say "I expect him," and this is the only use of it that Wittgenstein undertakes. But we have also turned the phrase to ask how the phrase itself is used and acquires its meaning in Wittgenstein's use of it. Indeed, we cannot help but do both at once. Both happen if we merely say that the meaning of "happens" depends on Wittgenstein's examples. This already tells both what happens when we expect, and what happens when Wittgenstein tells what happens when we expect. But can not all sorts of things happen here? To do as Wittgenstein does, let us really look and see what may happen here. We are being asked to look for something. We are given examples. We grasp that "happens" is to be understood in terms of the examples. We may appreciate the examples, perhaps find ourselves thinking of other examples like these. We may be puzzled by the examples, not grasping what is being exemplified. Or, we may think that we have grasped what they show, only to be horribly disappointed because the next example cannot fit what we had surmised. We may be frustrated and give up trying to find a thread through this diffuse series. At times perhaps much else may happen as well.

One might object that what happens when we read the text is not the happenings cited in the text. They are different, but not different in kind, and not on different levels. Words about words depend on the ongoing practice just as much as all words do, including the words of a text about words. Our phrases such as "understood in terms of the examples" and "cannot fit what we had surmised" are not different in kind from "expecting," or "trying to find the right expression for our thoughts." To be sure, it is in a doubled way that we ask "What happens when he asks 'what happens,'" but the phrase depends on what may happen when we use it, and we must look for this in the same way here too.

Although we are using only his words, our use of them in discussing them adds a dimension to Wittgenstein which he did not employ. But there is no single line between what he did and what we are adding. Rather, we are traversing a zone. We selected his affirmative uses, then we turned to use these so as to speak about his use of them. Just now we saw that his use involves both what happens in the examples, and what happens here as they give meaning to the phrase "what happens." Now we will go further, but without leaving him behind. By using his phrases as he did not do, we are saying what he did.

4. Elevating the redundancy; it is a new mode of saying:

Since we cannot help but speak about both the happening in the examples, and the happening of the examples, let us consider it as a positive mode of saying.

Since "what happens" acquires its meaning (like every phrase) from what happens when we say it, we find that "what happens" says how it acquires its meaning. The phrase can be about how it is used. Then it says how it is used.

Similarly, in this context, his word "show" says the showing by which it shows what it means. It can be used to tell about this showing, without substituting a representation for it.

Although we seem only to have gone around the May pole, this turn is more than Wittgenstein allowed himself to do. Although this use of words adds no information, the words can say how they are used and meant. They enable us to speak about how they work. Before we go further, let us repeat this turn with another of his phrases:

Wittgenstein could not have failed to notice that he used "family resemblance" as a metaphor. Since he constantly warns against metaphors but did not refrain from this one, we can infer that his usual objection does not apply. Wittgenstein says that a metaphor is misleading when it presents a picture of something that is not actually there when we examine what happens. But a "family resemblance" is precisely not a picture. When you look at the pictures in the family album, there is a resemblance across them, but there is no picture of the resemblance, and there could be no picture of it, since that would only be one more. This has been well understood. The absence of a picture prevents this metaphor from being misleading, but a mere absence is not enough here. The use of words is not determined simply by not-being-pictured. Let us try to use "family resemblance" in order to speak about what "family resemblance" does mean.

The attempt to use the phrase about its own use can reveal something striking: "Family resemblance" is itself a case of family resemblance! There is only a family resemblance between Wittgenstein's use of this phrase and its usual use.

If we ask what "family resemblance" means, the answer can be told in terms of "family resemblance:" The answer would be: "You have the answer already, right here: A new word-use by family resemblance comes about just as it is doing here in the new use of 'family resemblance.'"

What we have found may seem obvious and redundant. It may seem obvious because family resemblance applies to all word-uses, and redundant because we seem not to find out more about how it works when we apply it to itself. We seem unable to say how it works.

Or, we can let "family resemblance" say how it works. We can answer: "Look, it says that. And, in addition, you also have what you are asking about. It is happening right here, so you are in the best position to look and see what happens."

Is this only redundant? And, is it something that we cannot say? Or is it a way of saying how it works, moreover a way which also brings its working?

The double duty of "what happens" and of "family resemblance" comes because we are using these phrases to speak about what happens when we use them. Can we permit ourselves to do that?

On the one hand one might object that I am doing what Wittgenstein warned against, namely speaking about language; (about his use of his words). On the other hand, isn't this just the task we set ourselves at the start? It was our purpose to find a way which Wittgenstein admittedly did not give us, namely a way to discuss his standpoint in a discourse that would not assume objectivism or intellectualism. If this is possible at all, would it not have to be a discourse Wittgenstein did not undertake, and yet one in which one could say what he did?

But I can offer an even better defense in favor of allowing his words to speak of their use: We can ask why Wittgenstein was so opposed to speaking about language. Then we can see if we are making the mistake he warned against. It was that the attempt would ignore the fact that words about language are themselves dependent on the very practice they pretend to explain. Representations of language ignore the obviously ongoing language in which the rendition is attempted. And, not only that. What the words would seem to say about language would be false and misleading. The words would be about referents which do not exist, as Wittgenstein strove to show.

Now it is quite clear that this is not what we are doing. When we let a word or phrase speak about how it is being used just then, this is not the aboutness which Wittgenstein opposed. The words are not about nonexistent things that the grammar creates. Such a misimpression is precisely what we avoid by letting the word be about how it is used just then, about what is happening.

But even so, did not Wittgenstein warn against any way of letting words be about? No, he was not opposed to the grammar that enables them to be about something, to refer to something. He only pointed out that this grammatical power of words sometimes seems to create referents that do not exist. For example, one might speak of the family resemblance as if it were a thing. We might then think of it as determining when a word can be used. But this is precisely what we do not do when we let "family resemblance" say what happens in the very use of it. Then we can look and see that the family resemblance is not there in advance; the word comes and is used. Only afterwards can we notice a family resemblance. The use of words about their use protects us from the kind of "about" that Wittgenstein opposed.

What Wittgenstein warned against does not happen because the about (the referring) is taken up with the ongoing working of the words, rather than being about a presented referent. This protection holds even when the grammar would seem to create a nonexistent referent, for example when we speak of the family resemblance as "it."

5. So used, a word says something also about other words:

Another objection can lead us to a further step. Someone may see something tricky in using a phrase about itself. It may seem reminiscent of medieval second intentions and mystifying Cretan, Russelian, and Richardian Paradoxes. So it will be better always first to emphasize that all words mean what happens in the situation in which we use them, and only then say that "what happens" is no exception. Similarly, we can first emphasize that according to Wittgenstein the different uses of any word share only a family resemblance; then we can point out without mystery that "family resemblance" is simply no exception to this.

If we are asked to define what "intricate" means, we need not immediately answer that "intricate" means precisely how it's own meaning is more intricate than a definition or a single referent. Rather, let us first say that what may happen with all words is more intricate than definitions or seeming denotative referents. Then we can add without mystery that this is so also with "intricate." It has no denotative twisted object. Rather, it means and says the intricacy shown by Wittgenstein's examples.

Why did Wittgenstein not apply his approach to itself? It would not have been the sort of thing he opposed. But, of course it would not have supplied what people asked of him, an explanation in other words.

I think Wittgenstein went very far, indeed further than we can easily follow even after years of study. So it is not surprising that he left it at "only showing," without asking how words could possibly show without also saying. We have now pursued this question and found that in his use "not saying" meant not giving a substitute version. We can now rescue the word "saying" from saying only the kind of saying he opposed. Why insist that "saying" can only say the misleading substitution of nonexistent referents? That is not what "saying" usually says in ordinary language.

Let us permit "saying" to say what happens when we say something. What Wittgenstein rightly avoided is only one kind of saying. Now we can solve the problem about only showing. When words show, of course they also say. If words acquire their meaning by how they work, of course this will be so also with words about how words work. If used about themselves, they show and say how they work.

In a recent book, Schneider shows in great detail how Wittgenstein constantly denies that categories underlie anything. It is misleading to map a pattern from one situation to another. "Wittgenstein warns us against concluding that if the grammatical forms (for example the subject-predicate form) are the same, the relations of the content must be the same." [5]

What do Wittgenstein's examples have in common? What do they exemplify? We fail to find a generalization to encompass them. But since they are examples, it seems that they must have the category "intricacy" in common. It would be a mistake to assume that intricacy must be a class or category. Indeed it would be just the mistake which he is trying to correct by using this very word. So the phrase "more intricate" ("verwickelter") says at least that we are not wrong when we find no commonality across his examples.

"More intricate" says something more intricate than a common class. We can allow "intricacy" to say a kind of kind that does not reduce to categories.

6. All words about the use of words can say how they work:

Now let us take a large step further, still within the zone of discussing Wittgenstein in his own way. Rather than being locked into Wittgenstein's few affirmative words and their showing and happening, we will now see that a great many words — the whole language — can become free for this kind of saying.

Isn't it contradictory when I speak of a kind of intricacy that is not a kind? It is very like a grouping by a resemblance that is not a picture. Indeed it is the same grouping. Wittgenstein's examples do of course have in common that they are all what might happen when we use the given expression (for example "expect.")

We can say that these are examples of what "more intricate" means, if we allow the word to say how they do function, although they do not form a common concept or class. We can also say more about how they function. They are the kind of variety we find, when we look and see what happens. But we can give this explanation only by letting the word "find" say this kind of finding in which we find this kind of happening. We can say this because "kind" (and every word) comes with this kind of grouping which is not conceptually identifiable by some pattern that is always the same.

If we were to insist that saying is only possible in the usual conceptual categories, we would seem to have only denials. We do not find the referents of the grammar; we find a group of happenings that do not constitute a kind. Then it seems we cannot say what happens in the examples, nor what happens here with them.

Wittgenstein repeats the procedure with many expressions. Each time we do find this kind of array of examples of "what may happen." The examples do their exemplifying in this way; they form a group of happenings of this kind of kind. So his words and ours say how they are used. Can "say" be used to say this kind of "saying?" Indeed it can; it is the ordinary kind of saying.

Now the word "say" says this saying which may be different in each situation. We can say this, since "situation" now says what may happen, including all sorts of things some of which are called "subjective" because they are not observable from outside. Wittgenstein considers them part of the situation, just as words are part of a situation. "We make an effort to find the right expression for our thoughts" for example "in writing a letter." Now we can say that it is from situations that Wittgenstein speaks.

Each further word we use in this way is retrieved from other uses because it acquires its meaning from what happens when it is used here. The word "used" says what happened with it when it was used just now. Now the word "now" means a kind of "when" that is neither just momentary nor just timeless. It is not momentary since it will mean its use whenever someone reads these lines. On the other hand, this "when" is not eternity, rather only those moments. If we look to see, we find a more intricate variety of times, just as with anything else. What kind of time (and what kind of kind) is it, if we say that this "now" says the times when someone reads these words, which might happen for hundreds of years, but only then? So there are more intricate time characteristics than any scheme of time. Wittgenstein finds that most anything is more intricate than the few schemes we usually attribute to it. Can this intricacy be said? But the intricacy is precisely what is said.

If we look to see what happens ordinarily, we find that this more intricate saying is of course understandable. Although an expression may never have been used that way before, in that situation it says precisely what it says, and not something else. Ordinary saying has this more intricate type of order which is not in accord with a scheme, nor is it always the same. But now the word "order" says something about what happens when we use an expression in a situation. Although it is not a conceptually definable order, neither is it an unknowable Kantian thing in itself which we cannot find. It is rather what we do find. We recognize Wittgenstein's familiar examples, and can devise our own. It is the more intricate kind of order that we find in our situations with other people in the world in which we live. And "world" ... we could go on and on in this way, using and retrieving one word after another from the schematic assumptions.

Once a few words say their use, the further words we use to speak about them can also be taken as saying something from and about their own use. Instead of falsifying the use of words by saying something else about it, we let the words say (some aspect of) the use which is happening just then. In this way all words about the use of words can say something that their ongoing use instances, not only the few Wittgensteinian words with which we began.

Imagine a language in which what words say about the meaning of words could always be something their use also instances. What characteristics would such a language need? Words about words would need to involve two uses, so that one could instance the other. For this to be possible, we would need a language in which what words say depends on how they are used, so that what words say about the meaning of words would involve two uses. We recognize our own language of course. Our vocabulary about Wittgenstein and the use of words is now the whole language.

I do not argue that words about words must always say what they also do, just that they can. Although they certainly cannot say everything about their own use, they can always say something from it. To let them do that, we forgo for the moment their capacity to say something else, which they would merely be about.

It is not true that what Wittgenstein showed cannot be said. It seem so because it cannot be said as a substitution in a theoretical language. Of course it can be said, but only in the language he uses to show it, the same language in which we normally speak.

Wittgenstein stands beyond the reach of what is currently called "the post-modern dilemma," since he employed nothing that postmodernism has undermined. What he showed depends neither on clean distinctions, nor on the assumption of something present and given, which we can represent. Wittgenstein points beyond postmodernism if we can go further in his way. [6]

For example, Wittgenstein can speak intricately about what is commonly called the "self" or the "subject:"

"If someone has a pain in his hand ... one does not comfort the hand, but the sufferer: One looks into his face" (286).

In this mode Wittgenstein can speak about the intricate way in which what is usually called "the self" and "the body" are related, without setting up theoretical terms for how it is that the one we comfort is not in the hand, or how that one has (owns, observes, feels, lives in, lives with .....) the hand, and that we find (reach, communicate with .....) the person in the face. No existing theory approaches the intricacy of what Wittgenstein's simple statement says.

Of course there is nothing ineffable or unspeakable about what he showed. And, of course he said what he showed. One can say more in many further ways (for example with the words in my parentheses), but only by what I call "naked saying," [7] without covering it with a theoretical version which then claims to be what we really said. Such a substitutional explanation is the only saying that is made impossible by what Wittgenstein showed.