The first step is to make clear that there really are issues in language that need to be addressed. We can notice the following things. Technology was fairly primitive, compared to what we regard as commonplace today, before physics existed as a formal tool. So at the very least, as a base for technology, physics has provided new possibilities that people seem to have found valuable. And unlike any material commodity, physics is a resource composed entirely of ideas, which would seem to make it available for nothing more than the effort of communication. For these two reasons alone, that it can be communicated, and that it is useful in enabling technology, one would expect that it should be a priority to disseminate physics as widely as possible, because there is substantial advantage to be gained and relatively little trade-off in cost. And indeed, nominally, this is a priority of our culture, and of most cultures in the modern world. Physics (in name) is considered a necessary inclusion in any complete `education', and quite a large volume of text has been written, from classroom to layman level, in an attempt to communicate both its basic points and the advances made at its current frontiers.
Next point: it is clear that many of the people who write about physics know their subject well, because they not only use it, but often have been the ones responsible for building parts of it from their discoveries. And some of them are even very good writers. So there seems to be no lack of competence on the part of the people presenting the subject. Similarly, there is often no lack of diligence or intelligence in students or laymen trying to `understand' what they are being presented.
Yet despite the effort, basic competence and good intention put forward by both sides, it is clear that the essence of physics, as well as most of the knowledge contained within it, has not been communicated from the body of practicing physicists to the fabric of the social culture. By ``the social culture'' is meant, the standard set of tools and abilities that the culture is expected to offer to its members and to each new generation. Enough of the technological applications of physics seep out that people can do useful work producing, maintaining and using the technology at some level, but there is much more power and content to physics than is represented in a few technological prescriptions. Physics is special because it gives people ways to think and reason that can make them both more effective and more independent. These alone would have ramifications for the efficiency and interdependency of people's decision-making processes. In addition, even at basic levels, many things within the body of physical ``knowledge'' about the natural world could make day-to-day life easier for people in general, if they were understood and used. The ways that we, as members of a culture, place our dependency on each other for our understanding of what is right and wrong and for our decision-making processes, and the extent to which we continue to make mistakes in daily life that have long been unnecessary because of things about nature that have come to be understood, indicate that only shreds of whatever is central about physics have penetrated the culture. Moreover, the reports from the frontiers in the laymen's press have come to be regarded with such an air of unreality that they are almost presented more as entertainment than as something real that is intended to be taken seriously (in particular witness the lurid almost tabloid-style of even the reputable New York Times science reports.)
From one point of view, it might seem strange that with this much effort being made at communication, in a field where ``nothing more than communication'' is required, and this much value to be communicated, yet relatively little has been accomplished. But this viewpoint inherits a dis-service, of thinking that mere ``communication'' is more elementary than perhaps it is. The fact that the value of something is contained entirely within ideas, and the fact that a person understands those ideas, as evidenced by the fact that he can make productive use of them, does not guarantee that he will automatically be able to deliver them to someone else with the vehicle of language, even if the language is ``very good''. The problem is that language is not a complete, or even necessarily a very good, vehicle for ideas. An example of this was the way people at one time tried to convey the important point that there is an order to the physical way of asking questions, by presenting it in words as a formalized ``scientific method''. The method was hammered into students, composed of the steps of `observation, hypothesis, something, something, synthesis and something'. At the end the students could recite six words in order, but they have since grown into the adults who still have been reached by relatively little of the content of `science'.
Presently, the oversimplification of a ``scientific method'' as a highly concrete prescription, more popular a few decades ago than now, is slowly being replaced by a more sober and balanced presentation of something more like a ``scientific guideline'' or set of ``scientific habits''. This more appropriately represents the processes of science as the typically human activities they are, though unfortunately in the attempt to present a more complete perspective that encodes their uncertainties, something seems also to have been lost in the emphasis on just how important those guidelines or habits are in improving the efficacy of human thought. It is a useful retrenchment that people no longer try to present a sequence of six words as if it were the essence of scientific inquiry. But it leaves unaddressed the problem of communicating what is different about that method of inquiry that makes it special, and also the problem of why the words failed to convey that point, since that clearly was their intent.
The claim here is that from another perspective, that of the way words inter-relate with the actions, experiences and history of the people who use them, even the expectation that language should somehow be complete in itself seems somewhat naive. Not only is it not surprising that many fine books and lectures should fail to have much impact in some audiences, but if they do not incorporate a certain crucial awareness of the potential fundamental inadequacies of language, it is actually somewhat surprising when they can accomplish any impact at all.
The subject could, for instance, be considered this way: A particular language does not seem to be something intrinsic to the structure of a person or the experience of living in the world, as evidenced by the observation that people can grow up in different countries and learn different languages, while their structures as people and the substances of their lives consist mostly of the same traits and activities. Therefore, language is far from self-evident; in fact it is rather the opposite, something created in somewhat arbitrary forms or inherited from a person's peers and predacessors. So it is not natural to expect the form of language itself to have a great deal of intrinsic content. Of course we find language useful, so it must obtain a real content from something. The natural and reasonable origin to credit for that content seems to be our shared experiences and actions. Some of the experiences are shared simply by virtue of our similarities as living people, and others by virtue of shared or similar circumstances. We note that this is consistent with the observation that, even among languages that have developed widely separately, there are similarities of structure that usually make one intelligible in terms of another, and make it possible for a person to translate many ideas back and forth across them.
Even there, though, the actual meaning of the structure is relatively weak, because starting from nothing but the form of the language itself as a code, interpreting its meaning is a highly uncertain process. For example, among very old languages of which no living representatives exist with whom to match actions, anything that doesn't directly represent (e.g., in pictures) some action or object tends to be highly resistent to decoding (with one interesting exception usually being representations of counting and numbers). Therefore we conclude that, while a language can contain a great amount of content, that content is almost entirely not something intrinsic to the form or structure of the language, but something inherited by it from a base of shared experience.
This view carries negative implications for using language to communicate anything new. The whole assertion that something is new amounts to the claim that not all the experiences associated with it are already familiar. So while it may turn out that language is surprisingly effective at representing it, it would be unnatural to expect that it could do so completely, because precisely here some aspect of the base of shared experience from which it draws its meaning is missing.
At the same time, the whole point of teaching in any form, and much of the advantage of living together in groups and societies, is that what has been learned in experience by one person can be communicated to other people, so that complete repetition of all of the experiences is not necessary. Therein lies a synergy. Obviously, our assertion that what has been learned within physics should be communicated, so that it can be used, implies that we believe it can be. (We claim that this is not in fact a dilemma, though at first it may look like one.)
So first, is it even true, and more importantly, is it relevant, that language is not a complete tool even when one understands one's subject, and if so how does this find expression in the failure of good language (already constructed about physics) to present an important subject? We will present evidence about this shortly. Secondly, even if it is true that language fails in important ways when one talks about physics, how it is possible to espouse both sides of this issue, to recognize that failure and still to claim that it is possible and should even be regarded as necessary to communicate physics so that much of what is known finds use in many people's everyday resources? And thirdly, if one insists on communicating in the only highly efficient way we have, which is by means of exchanged language, especially in a very pure form for it like a book, what has to be added to the language, what has to be changed from what has been done before, to overcome its intrinsic limitations and reconcile them with the fundamental need to communicate?
Much of the answer to this last question constitutes what has been learned within physics itself as a necessity to overcoming its own confusions. The same difficulties that afflict attempts to teach the results of physics to anyone have always plagued physicists trying to explain what they have learned to each other. Our central claim is that much of their success, not only in working cooperatively but in thinking at all, has been possible precisely because they have adopted new ways of using language that are the most effective yet at overcoming its limitations. What those ways are and the structure that they have given to the physicist's language are the main subjects of this chapter. In a way, these are the first and most fundamental part of what could be called the ``methods of physics''.
It may seem strange to begin a book about physics with a chapter on something as seemingly tangential to the main topic, and as seemingly familiar, as language. This preamble on the essential inadequacy of language has been made to try to show why we have found it necessary to do so. If the entire content of what physicists have learned about language were represented simply in an improved use of that language, it would suffice to write incorporating those improvements. Yet in large measure that is what has been done already, because physicists, once in the habit of using language in a certain way, usually find it intolerable or impossible to go back to using it in more conventional ways. Yet this has not solved the problem of communication, as people still do not (evidently) make use of what is known. Hence we are forced to the conclusion that the language in itself is not enough, now matter how it is used. The other thing that is needed is a particular relation of the reader to the language read, which is not encoded intrinsically in the choice of words itself. That relation cannot be provided by a book, any more than a book can provide the effort required of the person who reads it. But it is intended that this chapter can provide a sequence of signposts, to indicate what about the language is special, and what can be provided by the reader in his relation to it, so that the two together will make communication possible.
So, before trying to understand why the solution to this problem of language takes the form it does, or even claiming that such exists, we try to gain a sharper appreciation of the problem itself, by seeing what can go wrong when language is used in a traditional sense when important parts of the base of shared experience are missing. For a demonstration of this effect, it is hard to find a more fertile ground than modern physics, because at its current frontiers, physics has expanded the set of experiences that a person can consider far beyond what is generally realized, though not quite so far beyond what is actually shared, if one knows where to look.
Aphorism: We give meaning to the language we use through the base of our shared experience.
A difficulty with words is that one can hear them with very little consequence. The difference between understanding and not understanding is not as simply recognized a division as one of physical comfort or pain is in other circumstances. So to even demonstrate that something is lost, one first has to make claims about what can be, and in fact should be, expected from the speaking and hearing of words. In other words, one has to provide a way even to identify what it means to understand something. Like everything else in this book, the reader is left to decide from his own experience whether the criterion given here is a useful and appropriate one.
Our approach will be very mechanistic. There are those who may object to it, claiming that language and thought should not be treated mechanistically, but for us it serves the purpose of identifying a distinction unambiguously and allows us to make a point. Speach impinging on the human body, or reflected light from the pages of a book, has relatively minor physiological effects in itself. And speach and argument have vast grey areas in which no definite landmarks identify when one has either understood something new or closed an issue. So in some sense, the only reliable, definite and reasonably easy-to-identify measure of the effect of words is the actions they lead to in the person who has received them. If I ask directions when I am lost, I suffer no uncertainty whether I have understood the response I get, because I then either know how to go to the place I asked about, or I don't. And, assuming that the reason I asked directions in the first place is that I had some intention to use the information reasonably soon, the person who gives me directions also suffers no uncertainty whether he has been understood, because either I take the correct action, or I take no action, or perhaps I try to bluff my way through the exchange without understanding, in which case it is likely I will take some other action than he described. This level, and even precisely this kind of unambiguity is what we want to be our measure of the efficacy of language, so we will use this example of directions as our paradigm.
It is not surprising that the language of engineering takes this form, because engineering is generally understood as a set of methods, plans and directions for building machines, of one kind or another. Since people build machines with the intent that they will serve some function, there is no ambiguity whether the design of the machine was correct or not, because either it works as it was expected to or it does not. Similarly, as long as one is actively engaged in using what one knows to do something, he knows whether he has understood a rule or an instruction, because either he then knows how to build the machine, and can see that it works, or he does not or cannot. A fact that seems not to be appreciated about physics is, that no matter how great the difficulty of translating a physicist's statements into the concrete plans for building a machine, and even if they are never translated that way, the essential concreteness of the tests for whether one has understood them or not is exactly the same. There is always a measure of the meaning of words in the actions that they can be expected to produce. Thus the paradigm of directions is quite literal. And in physics, as with any other kind of directions, understanding gives power, by means of which the understanding may be recognized.
So we will look at three pairs of statements, to illustrate the kinds of power in action that we can expect from language, and the difference between our relation to the language when if provides us with these from the cases in which it does not.