The Evolution of Modern Metaphysics: Making Sense of Things (76 page)

and

• various versions of the distinction between necessary truths and contingent truths, including those to be found in the early and the later Wittgenstein.
^36
,
37

Quine is not averse to acknowledging associated distinctions of degree. He readily concedes that, among the statements we currently accept, some would be more resistant to rejection than others. In the eponymous metaphor of his co-authored book
The Web of Belief
, these statements are closer to the centre of the web of what we currently accept than the others are, and hence more directly connected to more of the rest of the web. So their rejection would necessitate more rejections elsewhere. And the more of the web we reject, the harder it is for us to maintain our grip on what we come to accept. Hence Quine’s ‘maxim of minimum mutilation’ (
Pursuit of Truth
, p. 14). The fact remains that the sharp distinctions of kind that he finds in his predecessors are abhorrent to him.

There is one feature of these distinctions, which is accepted even by his empiricist predecessors, that may in any case always have seemed an offence to empiricism – it certainly seems that way to Quine – namely the idea that not all knowledge is grounded in sense experience.
³⁸
On Quine’s view,
all
knowledge is indeed grounded in sense experience (albeit no individual item of knowledge is grounded in any individual episode of sense experience). Quine’s empiricism, he would insist, is empiricism of the purest strain, empiricism without any unempirical accessories.
³⁹

5. … and a New One Introduced

In repudiating the various distinctions that we have just seen him repudiating, Quine at the same time repudiates all that subserves them, notably the Fregean notion of sense (see e.g. ‘Meaning’).
⁴⁰
This in turn leads him to draw
a distinction of his own. This distinction is best illustrated by appeal to what he calls ‘radical translation’, that is ‘translation of the language of a hitherto untouched people’ (
Word and Object
, p. 28).

Imagine that you are engaged in a project of such translation: you are trying to compile a bilingual dictionary for English and the language of some people with whom neither you nor any other English speaker has had any prior contact. And suppose you have got as far as speculating, perhaps on excellent grounds, that one of the statements they accept can be translated as ‘All swans are white.’ Now suppose that you see, for the first time, a group of
them encountering a black swan, though this does not stop them from accepting the statement in question. There are all sorts of hypotheses you might form. Perhaps they allowed for this possibility all along and your translation was imperfect. Perhaps they take themselves to be subject to an illusion. Perhaps they have started using one of their terms in a new way, say the term that they previously used to denote swans. Perhaps they operate with some bizarre logic. To be sure, some of these hypotheses will come to mind much more naturally than others, depending on what exactly these people go on to say and do. But if Quine is right about the araneous nature of what they accept, and of what you accept, then
in principle
all of these hypotheses, and more besides, can, with suitable compensatory adjustments elsewhere, be kept alive. And this in turn will allow for incompatible ways of translating from their language into English, none of which is precluded by what you observe, or could observe, as you go about your interpretative project. If there were Fregean senses attaching to the expressions of their language, and to the expressions of English, then the correct way to translate their statement would be by means of an English statement that had the same sense, and your inability to choose between these various options would just betoken (irremediable) ignorance on your part. As it is, Quine does not believe that there
is
anything to make one option correct rather than any other. There is, as he famously puts it, ‘no fact of the matter’ concerning which is correct (‘Things’, p. 23).

He means this quite literally. As we saw in §2 he has a physicalist understanding of what a fact is. And he is denying that any
physical
features of how these people are disposed to use their language or to interact with their environment rule out any of the aforementioned options for translation. The whole point, in a way, is that we have the conceptual resources to discriminate more finely than the facts themselves can.

This is very reminiscent of the underdetermination of truth by evidence to which we saw Quine accede in §3. There is, however, a crucial difference. In that case Quine was prepared to say the very thing that he is not prepared to say in this case, namely that there
are
facts beyond what is settled by the evidence. Indeed, that is precisely what underdetermination consists in.

The picture is as follows then. The evidence leaves open different options for what to say about certain matters. In some cases this is because the facts discriminate more finely than the evidence can. Such is underdetermination. In some cases, however, including the translation case, it is because not even the facts can discriminate finely enough. Such is what Quine calls
indeterminacy
. (See
Pursuit of Truth
, §43.) And it is this distinction, the distinction between underdetermination and indeterminacy, to which I was referring.

But what motivates Quine to draw it? Why does he insist that the facts slice just so thinly and no thinner? (This question is especially pertinent in view of the regulative version of his physicalism, which warns against forming an overly narrow preconception of what the facts are.) It is not enough to advert to his repudiation of Fregean senses. Either his repudiation of Fregean senses deprives translation of just one source of determinacy, in which case the question remains as to why he is so sure that it has no other, or Fregean senses are construed as
whatever
would make translation determinate,
⁴¹
in which case the question remains as to why he is so sure that they are to be repudiated.

The key, once again, is his naturalism. To determine the general character of reality, or to determine what the general facts are, is the very business of the natural scientist. It is not the business of the radical translator. The radical translator is trying to find a way of interpreting certain people, and perhaps also of communicating with them. This is a practical exercise. It does not, in the present context, count as an exercise in making sense of things.
⁴²

6. Quinean Metaphysics I: An Overview

It is comparatively easy to see where Quine’s views must leave metaphysics. Insofar as he regards metaphysics as a legitimate exercise at all,
⁴³
he must
regard it as the sort of thing that I have defined it to be, that is to say the most general attempt to make sense of things. In his own terms, he must regard it as the most general of the sciences – and therefore as continuous with all the other sciences. As Quine himself says towards the beginning of ‘Two Dogmas’, referring to the eponymous doctrines that he is about to abandon, ‘one effect of abandoning them is … a blurring of the supposed boundary between speculative metaphysics and natural science’ (p. 20).

The idea that philosophy, or more particularly metaphysics when it is reckoned a proper branch of philosophy, should be both a contribution to sense-making and yet something of a fundamentally different kind from empirical science – the idea which we have seen so many of the rest of our protagonists embrace – is a direct breach of Quine’s naturalism. And some of the specific tasks that have been assigned to philosophy in the name of this idea, such as providing a once-for-all
a priori
vindication of empirical science (Descartes), or promoting clarity of thought as opposed to discovering truths about reality (early and later Wittgenstein), or clarifying certain linguistic frameworks by articulating analytic truths that depend on the rules of those frameworks (Carnap), or addressing external questions about the costs and benefits of adopting such frameworks (Carnap again), are but further indications, for Quine, of just how wrong-headed the idea is. If metaphysics is anything, then it is maximally general science, different only in degree from the rest of science.
⁴⁴

Very well; what examples are there of tasks that can be properly assigned to metaphysics on this conception? Significantly, much of what we have already seen Quine undertaking serves as a paradigm. For Quine, as for so many of the rest of our protagonists, an integral part of the most general
attempt to make sense of things is the self-conscious attempt to make sense of making sense of things.

But there are other tasks. Many of these have to do with the upkeep of our most general conceptual apparatus, that which is common to all the sciences. They involve the pursuit of various desiderata in this endeavour, the very desiderata that Carnap would have said we needed to pursue when choosing a linguistic framework: power, clarity, elegance, familiarity, user-friendliness, and the rest. Each science involves its own pursuit of these desiderata in the upkeep of the conceptual apparatus that is peculiar to it. The metaphysical pursuit differs only in being more general. And,
contra
both Carnap and Wittgenstein, it ‘is not to be distinguished from a quest of ultimate categories, a limning of the most general traits of reality’ (
Word and Object
, p. 161). Considerations that Carnap would have said were relevant only to one aspect of the project of making sense of things (choosing a linguistic framework) are on Quine’s conception part of a package of considerations that are jointly relevant to the entire project: the single unified project of determining how things are.
⁴⁵

One very typical example of such a task, which is concerned with a concept that is as integral as any to our most general conceptual apparatus, namely the concept of a physical object, is to determine whether physical objects are three-dimensional objects that endure through time or four-dimensional objects with temporal parts. The former view is pretty much the view of common sense. The latter view assimilates objects to what we ordinarily think of as their ‘histories’, where these extend into the future as well as into the past. Quine himself favours the latter view, which for various reasons he thinks makes for greater simplicity and clarity (
Word and Object
, §36).

A closely related and equally representative example is to determine whether statements concerning the future are (already) true or false. Quine’s view, consonant with his four-dimensionalism, is that they are (
Pursuit of Truth
, §38). This too, he thinks, makes for greater simplicity and clarity.
⁴⁶
It is striking, however, that Quine is also prepared to advert unashamedly to the view’s
ethical
payoff – as we saw in §7 of the Introduction. He evidently means what he says, in the very last sentence of ‘Two Dogmas’:

The considerations which guide [a man] in warping his scientific heritage to fit his continuing sensory promptings are, where rational,
pragmatic.

7. Quinean Metaphysics II: Ontology

Perhaps the most characteristic metaphysical questions, however – on Quine’s conception – are questions in ontology: questions about what exists. Quine’s view is that there are things of a given sort if some true theory is ‘ontologically committed’ to things of that sort, that is, roughly, if some true theory
says
there are things of that sort (e.g. ‘Universals’, p. 103). In a way this is trivial, as Quine himself would be the first to concede.
⁴⁷
It sounds less trivial, however, when the notion of ontological commitment is made less rough. To say that a given theory is ontologically committed to things of a given sort is to say that the theory cannot be true unless things of that sort are among the things about which the theory makes explicit generalizations once it has been suitably formalized, where an explicit generalization is a statement of the form ‘Everything is thus and so,’ or ‘Something is thus and so’ (
Word and Object
, §49, and ‘Existence’, p. 106). Anyone versed in contemporary formal logic will recognize this account as the purport of Quine’s famous slogan ‘To be is to be the value of a variable’ (‘On What There Is’, p. 15).

Very well, what sorts of things are there? What does exist? That, we now see, is an issue to be broached by ascertaining which theories are actually true. So it is an issue largely for various specialists in the natural sciences – but not exclusively for them. For there is also the question of how any given theory is best formalized.
⁴⁸
And that is in part (the part that has to do with the theory’s most general conceptual apparatus) a metaphysical question. Typically, it will involve what Quine calls ‘semantic ascent’ (
Word and Object
, §56): the shift from talking in certain terms to reflecting on those terms instead. Thus if the issue is whether positive integers exist, it is settled by deciding whether we do well to include any arithmetical apparatus as part of the formalization of any of our scientific theories.
⁴⁹
This is
precisely
the kind of shift that Carnap took to be constitutive of asking an external question about a linguistic framework. For Carnap, it was a characteristically philosophical move. There is a sense in which, for Quine too, it is a characteristically philosophical move. But Quine does not think that moves of this kind are only ever made by philosophers – any more than he thinks that moves made by philosophers are only ever of this kind. Nor, crucially,
does he think that moves of this kind only ever issue in decisions about how to speak. Provided all goes well, they issue in insights into how things are (see the previous section).