Three Ways Of Reading The Republic

If you just read the Republic it seems simple enough, but when you try to pick it apart to see how exactly it is supposed to work as philosophy things aren’t so clear. Often there appear to be large jumps in reasoning, and sometimes complicated claims are simply asserted, as if no one could doubt them once they were put forward. And when you spend enough time thinking about it the direction of the argument itself is no longer clear; sometimes it seems as if the conclusions precede the premises. Obviously these are defects in the Republic, but I think that they are partly responsible for its popularity. What makes a philosophical work popular is rarely its careful development of theories from grounds that ensure their truth, or a rigorous testing of theories tentatively put forward. That sort of work bores people, what they want are grand ideas that resonate with what they already believe; whether those assertions are likely to be true doesn’t even factor into their considerations. By simply skipping many of the steps that would have reigned in his wild speculation Plato is able to develop grand theories, which make the Republic popular, but they also make it extremely opaque.

Still, we can try to connect the dots as best as we are able, to tie the Republic together into some kind of coherent investigation of one sort or another. Because the Republic is so full of gaps, as I have mentioned, there are three ways of doing this (at least). Of course it is unlikely that any of them actually reflect what Plato thought, because they all treat the Republic as if it must fit into a rigid metaphilosophical framework, such that deviations from that structure would constitute flaws. And that is far too systematic an approach to philosophical problems to attribute to Plato, or any philosopher of his time. Philosophy at that time is characterized best as a kind of patchwork enterprise, where whatever reasoning strategies and principles seemed convenient are used to arrive at conclusions. But our goal here is not to get inside Plato’s mind, and uncover what he really thought. First of all Plato is dead, so doing that is impossible in principle, there will always be ambiguity about what Plato really thought. Secondly Plato was just a man, and a relatively uneducated one by today’s standards. Thus Plato almost assuredly made mistakes, and so our goal is not to imitate Plato but to surpass him. And to that end considering these three different ways of interpreting Plato’s argument is worthwhile in its own right because, first of all, it makes us sensitive in a critical way to philosophical method. And, secondly, it gives us a concrete example with which to consider the advantages and disadvantages of each way of pursuing the argument, to see what they could possibly accomplish.

Let’s consider first the most natural interpretation of the course of the Republic, as a deductive enterprise where the claims established previously, as well as some additional premises, entail the subsequent conclusions. I call this the most natural interpretation simply because this is how we are used to reading philosophical arguments: the author puts forward some premises, which are supposed to seem beyond doubt, and then conclusions are drawn from them, further conclusions from those, and so on until the author has established what they set out to. If we were going to read the Republic in this way we would probably have to begin the argument with Plato’s functional theory of virtue, the claim that a virtue helps something perform its function, and that the function of a city is to serve our needs. Then Plato deduces from those premises that the city that he describes in great detail, which I will refer to as the ideal city, is best, because it best satisfies our needs (indeed it is developed by always considering what would make the city better at doing that), and thus best performs its function. And, given that city, and the definition of virtue, we could deduce what the virtues of this city are by seeing what aids it in serving our needs. But here things get complicated for the deductive interpretation, because, given what we have established so far, all that we could conclude would be something like: the division of labor is a virtue of the city. We couldn’t say what particular virtue was, as we usually name them. Plato proceeds at this point by simply asserting that if the city is good that it has four specific virtues, but this assertion apparently comes out of nowhere. If this investigation is proceeding deductively that claim must, therefore, be an additional premise. But that doesn’t solve all of our difficulties. It is also hard to see how these virtues are to be identified with particular aspects of the city that help it function well. When we look at what Plato actually does it looks like he just fumbles around until he finds an identification that might be intuitively acceptable. But clearly that is not sufficient when it comes to a deduction from premises, deduction proceeds only by logical steps, not by intuitive leaps. Thus the identification of virtues with the things that help the city function well must be additional premises. Together those additional premises establish what justice is in the context of the city, but not what the just person is like. To do that Plato returns to a way of arguing that is easier to interpret as deductive; he begins with a principle about how to identify when something has distinct parts, another premise, and together with facts about human nature deduces that the soul has three parts. But to draw conclusions about just people he must lean on yet another premise, that just people are like just cities in their organization, which allows him to transfer his results from his investigations concerning the city onto individuals. Of course that is not the end of the Republic, Plato goes on to make more assertions, but let us set those aside for the moment. So that is the deductive interpretation. Obviously there are significant problems with it, not because we can’t cram the Republic into it, but because it has such large gaps, specifically in the premises that a good city possesses four particular virtues (and no more or less), that the virtues are to be identified with the aspects of the city that Plato claims, and that the just city is structurally similar to the just person. These premises are simply too “large” to be part of a proper philosophical deduction. The whole point of deducing our conclusions from premises is that our certainty concerning the premises carries over to the conclusions, such that we become sure of them in virtue of the argument (or at least consider them more likely). But the three premises listed are highly questionable, and so, because the argument relies on them in an essential way, such that if any one of them were to be disputed the conclusion could not be established, the deduction doesn’t really support the conclusion. It may be valid, but it falls far short of being sound, and thus is a failure as an argument.

So perhaps it would be better to find some other way to capture the method of the Republic. Thus we come to our second interpretation, what I am tempted to describe as a kind of ordinary language philosophy. Under this interpretation what we are to understand Plato as doing is not as engaging in a rationalist-esque project, where certainty trickles down from a certain starting point. Rather we are to see Plato as thinking that we already “know”, in some confused sense, what justice is. Thus he is trying to develop that knowledge, captured in our intuitions and the structure of our language, into a precise theory. And if this seems too modern let me assure you that it isn’t; the idea that we only need to remember truths is in the air, which supports taking Plato in this way. If that is what Plato is doing we should start in the middle, with the claim that the city, if good, possess four particular virtues. The virtues Plato lists are the accepted virtues (minus piety), and so that list can easily be construed as a statement of something that we should intuitively just know. And Plato’s subsequent identification of the virtues with particular aspects of the city also can be understood as justified by a kind of ordinary language thinking, because in each of those identifications he explicitly invokes the idea that what he will identify a particular virtue with is why we describe the city as possessing that virtue; he is relying on the way we ordinarily use language to confirm his identifications. That being established it lends support to the claim that the ideal city described earlier actually is ideal, because now we can see that it has all the virtues. And to establish the nature of the just man Plato again leans heavily on the assumption that in some way the ordinary use of language can be depended on, as he justifies the claim that the just man is similar to the just city by explicitly stating that since they are both called just that they must have something in common. Furthermore when he is done describing what the just man is like he states that his theory can be shown to be correct because someone he describes as just will act in the ways that we commonly attribute to just people, again evidence of Plato leaning on the assumption that we already have a pretty good idea of what the just individual is like. So the merits of this interpretation is that it doesn’t suffer from large gaps like the deductive interpretation did and that it has significant textual support. But it also suffers from a serious defect: Plato wants to make some extremely unintuitive claims about how we should organize our cities. If we can lean on our intuitions then we are perfectly warranted in rejecting Plato’s theory the moment it starts being unintuitive. Just because it has some intuitive support doesn’t meant that it can legitimately overthrow our other intuitions, because who is to say that those aren’t the better ones and that the ones Plato uses to support his claims are defective? This is a problem that any philosophy that attempts to ground itself in intuitions and ordinary language suffers from.

Finally we have the dialectical interpretation. Since I have written at length about this interpretation in other places, although about a more limited part of the Republic, I will say just a few words about it. Under the dialectical interpretation we can take Plato to be engaged in a dialectical ascent from his description of the ideal city to the claim that the good city possesses those four particular virtues. And we can take his conclusions about the just man and his soul to be a hybrid of deduction and dialectical descent (and possibly the method of hypothesis as well), since Plato himself admits that he doesn’t perfectly follow the dialectic. Instances in this process where it seems that Plato is appealing to intuition or ordinary language can be construed instead as cases where Plato is testing his conclusions against background beliefs (as part of the method of hypothesis). The dialectical interpretation has the virtue of taking Plato at his word about how the Republic should work, but it too is problematic. Most pressingly it is simply not clear how the dialectical ascent is supposed to work or why we should have any faith in it. Even as Plato describes it, it is not the sort of thing that can separate correct from incorrect claims, rather, as it would be actually used, it would just validate whatever beliefs we began with. And if that’s not a problem for a philosophical method than I don’t know what would constitute a problem, since surely the point of philosophy is not just to reinforce what we already believe to be the case.

So none of these interpretations work as an interpretation of the Republic (in the sense of making the argument work, not necessarily in the sense of being faithful to the text), but we didn’t really expect them to. A better question to ask is whether any of them could work, whether we could productively do philosophy under them, or some repaired version of them. That is obviously an open question, but certain aspects of the dialectic remind me of some metaphilosophical ideas of my own, so I am tempted to assert that the dialectic, as wacky as it is, might be turned into an adequate method for philosophy, although obviously our repairs would require revising a majority of it.

Cheap Philosophy

You may be wondering what cheap philosophy is, and given that it is a term I have just invented perhaps I should begin by explaining myself. We might define normal philosophy in terms of certain topics or certain kinds of investigations that are distinctly philosophical that meet certain standards of thoroughness and correctness. This is similar to how we might define normal science as investigations concerning nature using the scientific method that meet standards concerning evidence and objectivity. Cheap philosophy is something that pretends to be normal philosophy, by sharing a similar method or topics, but fails to meet the standards that we hold normal philosophy to. I think many of what most people are exposed to as philosophy is really only cheap philosophy, not because of any malicious design, but because as philosophical ideas pass from person to person they often become mangled, turning into cheap philosophy even if the original was worthwhile. And cheap philosophy tends to stick around simply because most people aren’t as aware of the relevant standards in philosophy as they are of scientific standards. Which is not to say that cheap science doesn’t also exist, certain popular ideas such as intelligent design are examples of cheap science, but they tend to be recognized as lacking, while cheap philosophy usually passes unnoticed.

But just because I call it cheap, and complain that it doesn’t meet the usual philosophical standards, that doesn’t necessarily mean that cheap philosophy is a bad thing. I can imagine someone arguing that philosophy rarely has immediate practical consequences, and thus whether someone subscribes to a cheap philosophy might not matter in most cases, and it is possible even that this philosophical pretender might make them happier than the real thing would. I will grant my imaginary opponent what they claim, but cheap philosophy does have undesirable effects, even if they aren’t of an immediate and practical nature. First of all cheap philosophy degrades the real thing. If there is a lot of cheap philosophy out there then the real philosophy tends to go unnoticed simply because it is drowned out. Moreover if cheap philosophy is popular than intelligent people may start ignoring philosophy as a matter of principle, as a result of a reasonable generalization from over-exposure to cheap philosophy. And that is undesirable because philosophy does have its benefits, even if they aren’t immediate and practical. A second problem is that the standards that we hold philosophy to are rather general standards that could easily apply to any exercise of reason. Thus cheap philosophy doesn’t just undermine real philosophy, but reason in general, by making bad reasoning commonplace. And so I think we have some obligation to do what we can to rid the world of cheap philosophy, even if that amounts to just calling it out as such when we find it.

I divide cheap philosophy into three major forms, although this list may not exhaust the ways philosophy can be cheap it covers the most prevalent forms, at least in my experience. The first kind is philosophy that pretends to be justified by describing itself as the only way to avoid a ridiculous alternative. Often this means that the obvious facts which compose part of this philosophical position are asserted as if to take an alternative position would amount to denying these facts. But usually what is really the case is that the position put forward and the alternatives are differentiated by disagreement over something else, and that the focus on what is not in dispute simply serves to distract from that fact, and to mislead people into thinking that disagreeing with that position would amount to disagreeing with that obvious fact. For example, we can imagine an argument against sentimentality that argues that the objects of sentimentality are just things like any other things and that they have no special properties. This implies that those who are sentimental must incorrectly think that somehow the objects of their sentimentality are physically distinguished from others like them, which we know not to be the case. But that is not really where the distinction lies at all, the distinction is that those who accept sentimentality believe that the emotional attachment that the sentimental people have developed to various objects is worth something in its own right, even if the objects aren’t special by themselves, while the opponents of sentimentality deny that this emotional attachment has any value. And rather than actually discussing this difference the assertion has simply distracted us from it. Cheap philosophy is also common when it comes to how the economy should be run. For example, proponents of pure capitalism will often assert that everyone has the right to the fruits of their labor. Which embodies this kind of cheap philosophy perfectly because it suggests that those who would disagree with them hold that people have no right to the things they produce. But the real debate is usually not over that, what is really a matter of contention is whether there are other obligations that may conflict with that right. By ignoring them the argument attempts to deflect us from the real issues. Perhaps the central defect with this kind of cheap philosophy is that it avoids actually supporting the claim under discussion in favor of merely giving the appearance of doing so. And that problem can arise more generally whenever we argue for a particular position by arguing against alternatives; it’s simply a bad way to go about things. Indeed even real philosophy can occasionally be somewhat cheap in this way, for example it is easy to argue for materialism by arguing against dualism. But that is a distraction from the real issues, which is how exactly the hypothetical material mind is supposed to work. It is best to avoid even the possibility of being cheap by simply avoiding arguing against other positions and simply arguing for what we think is correct.

The second common way for philosophy to be cheap is simply by leaving out justification altogether and relying on people accepting it for non-rational reasons. Obviously we should only accept a claim only if we have reason to believe it to be true, but, unfortunately, we are not perfect reasoners. Often our judgments about how well a position is supported are clouded by whether we would like it to be true. This is why intuition has the strength that it does, people dislike changing their beliefs, and so we would prefer it if things that agree with what we already believe, namely our intuitions, were to be true. Thus we will falsely judge positions that are “intuitive” to be better supported than they actually are, simply because they don’t force us to revise our judgments. And of course we would also like theories that promise us happiness, immortality, or that wrongs will be righted to be true. Obviously cheap philosophy of this kind tends to shade into religion, which itself might be called cheap philosophy in certain cases. Thus it is hard to give examples of such cheap philosophy that are clearly philosophical and not religious, but the opinion that the world contains some kind of karmic balance, or that human consciousness plays a special role in nature, can both be described as philosophical. And in their usual forms they are almost always cheap because they are rarely argued for, but instead our emotional bias in favor of them is relied on to lead to their acceptance.

And, finally, philosophy can also be cheap by being devoid of real content and simply misleading the reader into thinking something deep is being said. This is most easily accomplished by intentionally confusing the reader, by placing before them something they cannot easily find a sensible interpretation for. Since it is presented to them as meaningful philosophy and since they cannot immediately understand what is being said the reader thus assumes that the message is something really deep, something that they can’t immediately grasp because of its subtly. Thus they make a great effort to find some interpretation for that puzzle that seems deep and meaningful to them, at which point they become convinced that the original puzzling statements really were philosophically worthy because they led them to that deep insight, and they pass them on to others, completing the circle. The central problem is that what the reader takes to be the deep message of what is claimed as philosophy is really their own invention, and because they invented it, partly guided by the standard that it must seem deep and insightful, that message will seem correct to them, even though no support for that claim whatsoever has been provided. Again we have a trick that leads people to accept claims without any rational support for them, which seems to be a common theme. Generally any intentionally obscure writing can serve as an example of such cheap philosophy, but “mystic” slogans such as “all is one” or “he who is first is last” also fit the bill. If we take them seriously puzzling over such apparent contradiction will lead us to some strange interpretation in which they might be true, despite their apparent falsity.

Since cheap philosophy seems to rely on psychological tricks there is probably no way to completely wipe it out, at least not as long as people are still able to be deceived. However the three forms of cheap philosophy described here all share the common flaw of not actually arguing for their philosophical conclusions. So it might be possible to diminish the effectiveness of cheap philosophy by always demanding reasons to believe it to be true, and never to be satisfied with an insistence that it obviously correct once it is really grasped.

Sources Of Philosophical Error

In philosophy disagreement is common and agreement is rare. And since these disagreements are substantive ones we must conclude that at least some of the parties disagreeing with each other must be in error, if they aren’t all in error. That is an awful lot of error, and it may make us wonder whether we have any hope of arriving at philosophical truths or if we are doomed to essentially make random guesses at what they might be without any hope of confirmation. Since such a high level of error is undesirable it seems reasonable seek out the origins of these errors in order to repair the defects that give rise to them, and so by eliminating them reduce the number of substantive disagreements. Of course it is easy to tell a psychological story about the sources of these errors. Philosophers, we might observe, don’t begin with certain facts and proceed from them to other certain facts. Rather they begin with certain ideas about the way they think things should be and then reason backwards to find simpler premises that justify those claims. Such a process allows every position to be put forward as serious philosophy, since no matter what claim we are considering there are always some simple principles to be found which will support it, although whether these simpler principles are any good is a different question. But these psychological sources of error are not our targets; psychological errors can never be eliminated, investigators in every subject often proceed in ways we wouldn’t rationally endorse. However, we expect there to exist certain sanity checks, that the way the claims are presented and pursued formally forces them to be satisfactorily supported, making the matter of how we actually got to them irrelevant.

Thus what we are really looking for is some defect in the formal practice of philosophy that allows mistakes rooted in psychology to look the same as philosophically justified positions to an unbiased observer. Because in disciplines that don’t suffer from such widespread error it is usually the case that given any two claims it is pretty easy to establish in an objective way which one is better warranted than the other, and thus, while someone could put forward a claim justified only by a vague sense of what was right, it probably wouldn’t be taken seriously. Given the way philosophy is usually pursued we can say that this problem must either result from an ambiguity in the way conclusions are drawn from premises or in which premises are acceptable. Obviously debates do arise about which conclusions really follow from a set of premises, however I get the feeling that this is not really the root of the problem. Such arguments arise in the first place because someone is attempting to refute some position by constructing a reductio for it, and defenders of that claim come forward and try to show that the reductio doesn’t really follow. But the reason that such endeavors are even attempted is because of the very problem that we are trying to address here, that we lack any kind of objective way of separate the correct from the incorrect philosophically. This means that the only remaining option if we wish to show that a certain position is incorrect is to try to defeat it from within, to show that it simply can’t be true. Naturally if we fix the underlying problems it will still be possible to make deductive mistakes, but I simply don’t see such mistakes as really being a significant problem, and outside of the context of refutation by reductio there is usually a consensus, or close to it, concerning which consequences follow from which assumptions. Therefore the real problem seems to be in determining which assumptions are acceptable and which aren’t.

But philosophers consider themselves rational and so clearly they don’t just begin with an arbitrary set of assumptions, they think that their assumptions are reasonable, such that anyone who disagreed with their assumptions would be wrong. (And obviously every philosopher thinks that, despite the fact that they all begin with different assumptions.) Now sometimes these assumptions are simply put forward as “obvious”, but I think it is obvious that “obviousness” is insufficient. People disagree about what is obvious, and there is a history of things that appear obvious at first later turning out to be wrong. Perhaps because of this “obviousness” is often cloaked, sometime with words such as “conceptual analysis”. But not every philosopher feels comfortable leaving their assumptions unexamined. Clearly the assumptions can’t be supported by deduction, since that would leave us with more assumptions, so they must be justified by contradiction, by showing that it is impossible in some way to reject them. Sometimes this is simple logical impossibility, showing that the assumptions are in some way tautologies, but more often it is the case that their contradiction is shown to make some obvious or well known fact impossible. For example, sometimes assumptions are supported because it is claimed that without them it would be impossible to have knowledge, that if things weren’t as claimed that we would be unable to know any facts. Or possibly that it would be impossible for us to communicate, to be conscious, or for the world to exist. Again the claim that certain facts are “obvious” has slipped in, but this time I think it is slightly more acceptable, because these seem more the of the kinds of facts that we can be certain of if we can be certain of anything at all.

Unfortunately proceeding in this way brings with it its own problems. Foremost among them is that often the supposed contradiction between the negation of an assumption and some obvious fact rests on a lack of imagination. It is claimed that the obvious fact simply couldn’t be the case were the assumption not to hold, but this presumes that we have surveyed all the other reasons the obvious fact might be the case and determined none of them to have been sufficient. For example, some would have once claimed that knowledge is impossible without completely certain foundations from which the deduction of facts might proceed from, and thus that they are warranted in assuming such foundations exist. But we now know that there are plenty of other possibilities, and that knowledge may not come about deductively at all, completely eliminating the need for such foundations. This goes hand in hand with the other mistake inherent in such arguments from contradictions, namely that too much gets read into what is obvious. I will grant that it is obvious that we have knowledge, but it is not obvious what knowledge is exactly. And often to show that the assumption must be true given some obvious fact a particular understanding of terms such as “knowledge” or “experience” or “existence” is leaned upon. All this does, effectively, is sneak in a new assumption, about what knowledge or its equivalent is, but which is not acknowledged as an assumption.

But if we aren’t to defend our assumptions using reasoning from contradiction, or something similar, it isn’t clear how we are to justify them at all. And I do think it is a bit hasty to throw contradiction our completely. It might be possible to use contradiction profitably if we pay attention to exactly what is obvious, resisting over-interpreting those facts, and if we considered the strict logical negation of our assumption, which obviously encompasses or is part of any alternative proposal, we might be able to avoid those pitfalls. On the other hand, it may not be unambiguous when exactly we have proceeded within those limitations, and when we aren’t being blinded by our lack of imagination. Moreover, such failures of imagination can be dangerous in other ways, they can lead us to erroneously lean on certain hidden assumptions, because we simply can’t imagine them being false, or to make mistakes about what follows from a fully acknowledged premise, for similar reasons.

The fundamental source of philosophical error thus seems to be the deductive method itself, the idea that we proceed to philosophical conclusions on the basis of an argument from premises. This observation complements another I made on a previous occasion, that deduction can’t arrive at any genuinely new knowledge, that anything new must come from additional premises. Thus we might be motivated to simply reject deduction as a way of establishing anything philosophically substantive. But, unfortunately, it is not clear what we should replace deduction with, all we can really say is that, whatever the replacement is, it must not proceed on the basis of some foundation, otherwise the problems with the deductive method will simply crop up again. Of course there may be no suitable replacement, but in that case significant philosophical progress might simply be impossible, because there would be no objectively sound way of determining which previous philosophies are worth building something more off of and which can be safely ignored, and thus every philosopher would find themselves starting from scratch again.

How Philosophy Proceeds (2)

Previously I set out both to describe some of the characteristic features of philosophy and to investigate how we might construct a coherent picture of how philosophy is supposed to work on the basis of them. One of those distinguishing characteristics is the focus on definitions, especially of terms that seem extremely vague as we normally understand them, so much so that giving a “correct” definition may seem impossible in principle. Without passing any judgment on philosophy on the basis of that fact alone we can consider the limitations inherent in any attempt to provide precise definitions for vague concepts, which is what I discussed yesterday. Today then the task before us is to look at where philosophy goes with those definitions and at what the significance of philosophy might be if it proceeds as described, given the limitations uncovered.

Suppose then that we had developed a precise definition of justice that satisfies the conditions mentioned yesterday. What are we to do with it? Obviously definitions by themselves are basically useless to anyone but writers of dictionaries, and so if philosophy has a point it must be because we don’t take definitions as the final product. In the context of philosophy such definitions serve as the foundation for what is sometimes described as the deductive enterprise of pulling out further conclusions from them. For example, from our definition of justice, as well as other precise definitions, we might develop further conclusions about justice, one of which might be about whether justice is desirable in absolute terms or just in certain situations. I would resist, however, characterizing this as a purely deductive enterprise. To really draw significant conclusions from our definitions requires us to load into them more and more content if we are really proceeding deductively. Not only does this somewhat undermine the benefit of precise definitions (namely the ease with which we can work with them) but it makes it extremely difficult to determine whether there is actually anything in the world that is described by the definitions, because each additional property we add is one more property that we have to find and to verify that it can be found alongside the others, and which adds additional complications when determining to what extent the new precise definition refers to the same things as the vague concept we started with did (adding additional definitions also gives rise to similar problems). Since, as part of the development of our definitions, I think it is better to understand the process of drawing out further conclusions from them as partly empirical. Since we have found examples of things that satisfy our precise definitions there should be no inherent difficulty in examining these things and by generalization developing further facts about the objects we have precisely defined. Thus we get to further conclusions by a mixture of deduction and investigation, where one can feed off the other. Whenever we perform a deduction from definitions and established facts we must check the results of that deduction to see whether it is actually true, and verifying it in this way lends support to the definitions and facts it followed from (or refutes them as the case may be). And of course any investigation that turns up new facts can open up new avenues for deduction.

Obviously the view that philosophy is to proceed to further conclusions in this way, essentially using the precise definitions as tools to investigate and explain the world with, contradicts certain other views about philosophy. For example, some see philosophical definitions as primarily resolving problems raised by the way we use language. Because our use of language reflects the vagueness of our concepts the possibility for contradiction arises, especially when a number of associated concepts come into play. And some have even held that all philosophical “problems” are really the result of such linguistic difficulties. Obviously if you see the philosophical problems as linguistic problems then the task of philosophy will be to provide linguistic solutions. And thus the philosophical point of providing precise definitions is to dissolve those apparent problems by showing that no such problems arise when we use those terms according to the new definitions. There are, however, a few problems with this vision of philosophy. The first is simply that many philosophical problems are not apparent contradictions, but questions, and thus would be resolved by answers not by redefinitions. Now this might still be hammered down to fit into the model by arguing that once we have these precise definitions the questions will not even arise. A more serious difficulty is that this position seems to undermine itself in certain substantial ways. In order to assert that philosophy is a linguistic enterprise, one centered around improving language, we have had to presuppose that all philosophical problems are really linguistic problems. But, if philosophical problems are linguistic problems then they aren’t really problems at all, and don’t really need answers; once you point out that a contradiction is caused by a linguistic imprecision there is nothing more you need to do. And, additionally, there is a tension between the way such philosophy actually proceeds and its stated task. If the goal is just to resolve linguistic problems then any precise definition will do just as well as any other, we might as well just stipulate that all philosophically vague terms are now to refer to various numbers. And, if we take them in this way, clearly none of the contradictions or questions previously troubling us will arise. Obviously that is an absurd solution, but why it is absurd under such a view of philosophy is not clear. If we can’t radically redefine the words in this way it must because they were designating something of importance, something that it is useful to talk about. But if that is the case then it might very well be that there are interesting questions and interesting problems about the things referred to by those terms. And obviously those questions and problems cannot be resolved linguistically any more than scientific or mathematical questions might be. Thus it seems obvious that, if we aren’t rejecting the way philosophy is conducted out of hand, that there must be more to it than a systematic revision of language.

But, given the way I have portrayed it as involving empirical investigations at a number of steps, some may fear that philosophy will collapse into science leaving nothing for philosophers as such to do. In some ways this is a bit of a strange worry, because the revelation that all biology could be explained in terms of chemical processes or that all chemistry could be explained in terms of physics did not turn biology into chemistry or chemistry into physics. Just because we could explain everything, in some sense, without appeal to biology doesn’t mean that we should do away with biology as something distinct. Biology works at a higher level of abstraction than physics and so while we could, in principle, establish every result in biology from the physical theories alone it wouldn’t be practical to do so. In fact showing that biology was really very complicated chemistry actually gave biology a better handle on many distinctly biological problems, by allowing biologists to give explanations of certain processes in chemical terms and to develop further biological results by appeal to facts about chemistry. For example, finding DNA, which only makes sense from a chemical point of view, was useful in explaining how traits were inherited. But, more importantly, by understanding the chemistry behind DNA more precise predictions about how traits are inherited, mutations, and so on could be produced. These biological claims could never have been developed if biology had been kept perfectly distinct from chemistry. And I think the same holds for philosophy. Just because philosophy can be seen as a kind of empirical investigation doesn’t mean that we can automatically brush it aside in favor of the established sciences, nor does it mean that philosophers will now need laboratories. But it does mean that philosophy might be able to build off certain scientific theories to reach philosophical results. If we want to see what philosophy might look like simply consider economics. Economics too is an empirical discipline, but at the same time it is a very abstract one (and thus a very philosophical seeming one). And, like philosophy, economics proceeds from certain definitions and assumptions, developed initially from vague concepts such as wealth and profit, from which it attempts to produce theories that describe the world. And, like any form of empirical investigation, it attempts to evaluate its theories by seeing whether they reflect reality, although in the case of economics this is not a simple matter since the things it deals with cannot often be easily measured. This is a model of how philosophy might work and how it may remain distinct from science as a whole despite embracing empirical ideals.

But this view of philosophy, as a very abstract science, raises an additional question, namely what philosophy deals with better than science, or what there is that it is uniquely suited to handle. It seems to me that philosophy essentially steps in where there are gaps in science, where there are important subjects that science refuses to address (often because of normative associations). But since philosophy isn’t a mature discipline, as it is still working out exactly what its subject matter is, some might assert that there is nothing for philosophy to be about as I claim, and because of that philosophy must be something radically different from science. Simply consider, however, how useful even concepts such as ethics, justice, society, consciousness, and so on are even in their vague and unrefined forms. Their usefulness, the fact that we continue to use such concepts despite the progress of science, suggests that they reflect, in an imprecise way, interesting and useful divisions of reality that we might want to theorize about in order to explain the world and to exploit for practical gain. There are no guarantees of course, but the same skepticism could have been raised to a systematic treatment of wealth and trade when economics was first under development, and so the mere possibility of skepticism tells us nothing. And if this is indeed the case then it provides us with a way to evaluate philosophical theories, given that testing them against hard data is always going to be extremely difficult. A philosophical theory is successful when it captures a feature of the world that we deem important, such that we can see uses for the theory. Of course this is not meant to be an attitude of pure pragmatism, but simply a general guideline for determining when we have found the interesting subject matter that was behind the original vague concept. For example, a theory of justice may be useful because it tells us when justice is desirable and to what extent, as well as what is just. And so we may wish to modify our society in light of that theory, to more or less approximate the ideal justice in certain situations. And, judged by such standards, we thus see that any possible worries raised by the fact that there is no one right way to create precise definitions of vague terms are without substance, because ultimately what matters is not how well the precise definition mirrors the vague concepts but what work it and the associated theory does. So even a “bad” definition may the basis for good philosophy, so long as that bad definition captures something genuinely interesting.

How Philosophy Proceeds

I have heard philosophy characterized as strong opinions in the absence of any information. And I can’t say that is an entirely unfair description, because even when considering philosophical claims that seem correct it is always a bit of an open question as to how the author arrived at them or what justifies them. I’m not exactly sure what to do with philosophy as such, we might embrace it and accept that philosophy is wildly different from most worthwhile intellectual enterprises (in the sense that correctness mostly falls by the wayside). Or we might say that there is something wrong with them, that they only approximate the best way of doing philosophy in some loose sense, and that philosophy is a discipline in which we have yet to come to an unproblematic understanding of what we are supposed to be doing. I opt for the second. As such I often begin with theories about the way knowledge works and the way investigations in general must proceed and then try to apply those standards to philosophy in order to reach some conclusions about what we should and shouldn’t be doing. The problem with that, naturally, is it runs the risk of collapsing philosophy into science. Which is not by itself a bad move to make, because there is probably some broader category under which philosophy and science genuinely are united. But being blinded by that fact is undesirable, just as it would be undesirable to be blinded by the differences between the sciences themselves and collapse them into a single discipline. So today I am going to focus on what is characteristic of philosophy, rather than of investigations in general. And what is characteristic of philosophy is an unusual, some might even say unhealthy, emphasis on definitions, and on conclusions that follow essentially from those definitions.

Of course inquiries in every discipline involve definitions, generally new words are needed to hide some of the complexity developed as part of the investigation so that we have a manageable way to deal with the subject on levels beside the most fine-grained. Imagine how tedious biology would be, for example, if we were unable to use the labels used to divide life into groups and instead had to fully describe in each case what the distinguishing characteristics were exactly which two species were united into a single category by. Such definitions are always stipulative; there is never any question about their correctness because the definitions are just abbreviations. But in philosophy definitions are often challenged, which means that, whatever they are, they are not simply convenient abbreviations. In fact almost every philosophically interesting definition seems to be an attempt to turn a vague concept into something precisely defined. Let us bracket for a moment the point of doing that, and instead direct ourselves to issues that may arise as a result of proceeding in this way. To make the discussion more concrete let’s also consider an actual example, justice, an exquisitely vague word, defined as follows: justice consists in adjustments made so that in interactions between parties any special advantages a person has accrued from acting unethically and any special disadvantages a person has accrued from acting ethically or from interacting with an unethical person are compensated for after the fact. (With what is ethical being precisely defined somewhere else.)

The first observation we might make is that, since these concepts are vague to begin with, not everyone will have exactly the same vague concept. Of course since we can successfully communicate there must be some overlap, but that by itself doesn’t provide us with a neutral starting point. Although the vague concepts of any two people may have a great deal in common the amount of overlap given larger and larger numbers of people may shrink until there is no substantial universally common ground remaining. Secondly, since these are vague concepts determining where the overlap is may be a Herculean task in its own right. And, finally, trying to proceed on the basis of some shared overlap means that an empirical investigation into what people commonly take those vague terms to mean must be conduced before any philosophical attempt at preciseness, and clearly that is not anything like how philosophy proceeds. This is certainly the case when it comes to justice, our example. Everyone’s conception of justice overlaps to some extent, at the very least that certain laws demonstrate it, but beyond that the nature of justice is largely up for grabs, so long as justice is superior to injustice by some standard (because justice is a normative concept). And so obviously the precise definition being entertained will be more in line with some vague conceptions of justice than others.

A second observation is that by giving a precise definition we must necessarily contradict the vague concept that it was developed from. For starters with any vague concept there will always be borderline cases, cases where whether something does or doesn’t fall under that concept isn’t completely clear. By giving a precise definition we automatically do away with all such cases since there is now a clear-cut matter of fact whether something counts as such under the new definition, and so it differs from the vague concept to at least that extent. Furthermore, the vague concept tends to be shaped to a large extent by the examples that we originally abstracted it from in our psychological development, and by what we would like to be the case when it comes to normative concepts. Because of this it rarely follows any simple set of properties, and thus there will be cases in which the vague concept will plainly contradict the more precise definition. But such conflicts are simply unavoidable unless we wish to develop extremely long definitions, perhaps pages and pages long, which probably make them useless for any purpose. With respect to our example of justice I suspect that we could construct some situations in which people act ethically or unethically without the intention to do so, or possible the opposite intention, and which we don’t think they should be punished. Or perhaps our intuitions lead us to think that unethical behavior should be punished even when the person acting unethically doesn’t gain anything by it. But such contradictions can’t invalidate the definition of justice, because, if they were allowed to, then it would simply be impossible to define it, and thus to do philosophy at all.

A third observation is that when a vague concept is given a precise definition there may not be anything that the new definition actually refers to. For example, there is a vague concept of water that conceives of it as a continuous clear fluid with certain properties. But, if we gave it a precise definition following those guidelines, as a continuous clear fluid, and so on, there would be nothing that was actually water (because of the atomic nature of matter). In fact it is reasonable to say that almost every vague concept is defined almost exclusively by appearances, that what is X before we set out to say what X is precisely is simply those things that present themselves to us in one of a number of ways. Thus while the concept is guaranteed to latch onto something, simply because it could really be latching onto nearly anything, the precise definition is not. Of course how to determine when something exists is a matter of some debate. And what we may be considering could very well be a kind of ideal, such that what we want to show is not that it exists but that things naturally tend to approximate it, which adds additional complications. The easiest way to show that something exists is simply to show that it has already been discovered in some way by science. In fact that is probably the only way to show that something exists, because if we can’t find it in an objective way through science then that is a good reason to believe that it doesn’t exist. Fortunately there are no such difficulties when dealing with our example here. That parties interact and that they can receive advantages and disadvantages depending on how ethically they act in their dealings is not a matter of much debate. Of course justice, described as a perfect balancing, can only be an ideal, but it is also fairly trivial to show that societies have at least some motivation to approximate that ideal: because ethical behavior is good for society it is beneficial to encourage it by protecting the ethical from being taking advantage of by the unethical, which also has the benefit of discouraging unethical behavior.

Fourthly we observe that even if a precise definition may manage to refer to actually existing things the things it picks out may not be what the vague concept referred to. To return to the example of water given earlier it may be possible that some weird substance is found in an odd corner of the universe really does fit that definition (and which is thus infinitely divisible, and so on). But just because that definition now manages to refer to something doesn’t mean that it does a good job of being a precise definition of the vague concept water. So not only do we have to show that something exists that is described by the precise definition we must also show that what is picked out by the vague concept is often referred to by the precise definition as well. About the only practical way of doing that is simply taking a number of samples of things that fall under the vague concept and demonstrating that they are picked out by the precise definition as well. In our case this means simply bringing a number of cases that seem to embody justice before our minds, such as just laws, and seeing whether they usually are cases of correcting imbalances resulting from unethical behavior or not (or at least approximate such corrections). Naturally listing a number of such examples would be tedious, but I think you can easily confirm for yourself that many of them have that property.

Finally, on the basis of the previous four observations, we can see that however constructing precise definitions is supposed to work it must be a matter of finding a best fit or a good fit to the vague concept, and this that there is no “correct” precise definition, and that our precise definitions may always be subject to revision as long as we are motivated to bring them more in line with the vague concept. This is because, as the first and second observations pointed out, there is no single vague concept to make precise, nor is there a single best way of making it more precise. And thus we will always have a number of options open to us for developing precise definitions. And, as a result of the third and fourth observations, we can see that there is always the possibility that what seemed like a perfectly acceptable precise definition based on its own merits may be discovered to need serious revision if it turns out that nothing actually existing is described by it, or if a large number of cases come to light in which it vastly misses the vague concept (or what actually satisfies that definition turns out to be other than was initially expected).

Given these facts about how precise definitions can be developed on the basis of vague concepts we must examine what we expect of them, why we are making our vague concepts precise, to see whether these limitations raise any problems for philosophy. Or perhaps, to be more honest about how I am proceeding, we must see whether we can come to an understanding of the philosophical enterprise where they aren’t problems. Originally I intended to include that discussion in this post, but I have gone on long enough for one day, and so I will leave for tomorrow the task of describing how we might proceed on the basis of these definitions and what the point of such investigations is.