Why not broaden the vision? This does occur, to some extent, in the sections ahead. Even so, there are limits to how broad a philosopher's vision can be, at least when philosophy's task is defined in certain ways. Thus suppose philosophy's "purpose...is...to discover what is common to all modes of understanding." If the philosopher's vision is so broadened as to include everything properly called an instance of understsanding, then very likely the modes of understanding have nothing distinctive in common. The term `understanding' is a family-resemblance term, like `game', its various uses related by a series of similarities rather than by a single set of necessary and sufficient conditions. As with games, the only thing common to all instances of understanding might well be that they are a human activity. Nothing distinctive would have been learned about understanding in general (even assuming there is such a thing). Far better to confess to an irreducible multiplicity of ways of understanding.

Of all the ways of understanding and changing ourselves and our worlds, why concentrate on discourse and its domains? Suppose I were to say I am interested in domains of discourse because I am interested in domains of truth, and truth is defined in the first instance for such linguistic items as sentences, statements or propositions. This is a respectable response, hallowed by long tradition. The trouble is that `truth' itself is a family-resemblance term, used for many things other than propositional truth. Truth is not really defined "in the first instance" for discourse (or for anything else), and to say so can prevent us from seeing that the strategic choices have already been made when we declare our interest in truth, meaning propositional truth. On the other hand, suppose we are interested not only in the glories of propositional truth, but also in its potential evils. Then we need not pretend that `truth' is not a family-resemblance term, or that we have not already accorded priority to discursive ways of understanding and being. On the contrary, part of what we are interested in is whether nondiscursive ways also enjoy some sort of priority over the discursive.

One of the potential evils of the pursuit of propositional truth is that we may thereby "deny ourselves a place in the world." A reason is that propositional truth usually means objective truth, and objectivity is widely believed to be somehow inimical to authentic human being, by omitting the subjective point of view, without which we cannot express or experience death as our own death, time as our time, and our choices as our own. But we see in §7.1 that this is a false problem. There are evils attending the unbridled pursuit of propositional truth, but they are not substantially different from those which attend any unbridled pursuit.

A special case of the pursuit of propositional truth is the attempt to unify all things in relation to a particular domain of truth. And physicalism is a special case of this way of unifying all things. But of course there are many otherways of seeing things whole, and some seem far more compelling -- different ways for different people, and occasionally for the same person at different times. Does one represent the way things are in themselves? The way the world is? The meaning of Being? The ultimate nature of existence? Or are such questions based on false presuppositions? Indeed they are, if we may believe §§7.2-7.3. Hence a traditional aim of philosophers, namely to "reconstitute the shattered picture of the world," is misguided, to the extent it presupposes we are to strive for the picture, some one all-inclusive portrait, emblematic of the ultimate nature of things.

Still, can't we at least say that "the aim of philosophy...is to understand how things in the broadest possible sense of the term hang together in the broadest possible sense of the term"? Not necessarily. Often the aim should be not only to understand things but to change them. Moreover, Sellars thinks the key sort of understanding is a matter of fusing into one vision -- into a single stereoscopic view -- diverse images or perspectives or conceptions of human-being-in-the-world. But the metaphor of stereoscopic vision will not bear the load. Fusion of two perspectives into a single stereoscopic view occurs only when they are of the same thing, are not too widely separated (about as far as our eyes), and use compatible lenses. In any case we cannot accept Sellars' implicit suggestion that this single stereoscopic, synoptic vision would constitute the true, complete account of the world, corresponding presumably with the way the world is; for there is no such thing.

Sellars is right that philosophical understanding is a matter of "knowing one's way around" with respect to diverse perspectives, conceptual domains, and so on; and that knowing one's way around is mainly a form not of knowing that but of knowing how. Yet knowing one's way around with respect to the domains is not happily expressed or achieved in terms of stereoscopic vision, and probably not in terms of any visual metaphor. Rather it is a matter of knowing what discourse to use on what occasion for what purpose, or whether instead to use some nondiscursive way of representing and changing ourselves and our worlds. It is therefore part of the general problem of knowing how best to live and be.

Construed this way, metaphysics and ontology become more a matter of values than of what there is. Traditional metaphysicians aspire to trace what there is, meaning what the ultimate reality is on which everything is dependent. They are convinced their questions are more fundamental than any others, since they ask about the most basic or real nature of things behind the subject matter of particular inquiries.

The priority supposedly enjoyed by their candidates for the ultimate reality -- Substance, Form, Spirit, Matter, whatever -- is absolute. But there is no such priority. Our characterizations of what there is reflect prior choices and inventions of (or acquiescence in) particular vocabularies together with entrenched metaphors (§5.3). In this sense "vocabularies acquire their privileges from [those] who use them rather than from their transparency to the real." For there is, to repeat, no such thing as the real to which they could be transparent. Characterization of something as the real, or as "reality," reflects a prior choice of vocabulary-cum-metaphor. These choices in turn reflect value judgments, often subconscious. In a sense, therefore, valuing is prior to characterizing something as "real," and thus prior to metaphysics and ontology. Since the values are ours, there is a sense after all in which we are is the measure of all things.

These cryptic remarks demand explanation, of course, which the chapter tries to provide. Meanwhile, bear in mind that calling us the measure of all things does not mean truth is relative or subjective. Nor does calling us the measure conflict with physicalism. In particular, there is no need to revise or retract the physicalist's tough-minded insistence on objectivity as a prerequisite for claiming that certain unifying entities exist. Instead, calling us the measure is a dramatic way of expressing a priority of value over fact. To some it will seem melodramatic. But this sort of priority might just be what Protagoras had in mind in the first place. The Sophists contributed to the rise of Periclean democracy, and Plato and other aristocrats might well have misrepresented Protagoras' "man the measure" by omitting context and qualifications. In any event, to say value is prior to fact is not to say that the priority is absolute, as we shall see, but conditional. Why bother to dethrone (or defrock) the notion of the way existence is, only to have it reinstated in the guise of talk about value?