Michael responded that the two were not likely to be so far apart in practice: “You can hardly develop a credible problematique without knowing some details.” Thill responded that this depth of interpretation necessarily “involves also an explanation of Kant’s argument for his views or claims!!!… What else could ‘why he said it’ mean or refer to?”

Thill’s question appears to be intended as rhetorical (especially given the laughs that precede and follow it in his comment). But it shouldn’t be. There is always much more to the reasons a philosopher says anything than the arguments that she makes for it. Certainly the arguments matter. They always do. But they are not the only thing that matters. Michael is right that depth of interpretation requires a serious attention to detail – but arguments are not the only details.

So what else could we be speaking of here, other than arguments? I don’t think it’s too hard to imagine what that could be. An argument consists of premises leading to a conclusion. But where do those premises come from? Sometimes from other arguments – but not always. We can follow a chain of reasoning back from one argument to another argument to another, but eventually it’s going to stop somewhere. There will be a premise that is simply asserted – or at least as often, and this is particularly important, a premise that is not even stated but merely assumed. And if one merely understands the structure of a thinker’s arguments but not the assumptions that underlie them, one will not have understood that thinker.

I should note that there’s nothing inherently wrong with an assumed premise, or one asserted without argument. Indeed, one has to do it at some point; one cannot say everything, or one would run out of space. It’s just that if one is going to assume or assert a premise successfully, it must be an assumption that is shared by one’s intended audience. That’s the point that is typically missed by overeager campus missionaries: you are not going to get anywhere by telling me that Jesus is God’s only son because the Bible says so, since I don’t accept your assumption that the Bible as an authority on that matter. If I did, your argument would be sound; but I don’t, so it isn’t.

Within analytic philosophy, when these shared assumptions are highlighted it is usually with the term intuition. I find that term highly inappropriate, because it suggests that these “intuitions” are something more than mere shared assumptions. But it’s not wrong to ground one’s arguments in those shared assumptions that get called “intuitions” – simply because, again, one has to start somewhere. On the “continental” side this point was one of Gadamer‘s key insights: new knowledge is always measured against the “prejudices” (Vorurteilen) we already have. (I find Gadamer’s “prejudices”, or Martha Nussbaum’s “prevalent ordinary beliefs” – a term derived with reference to Aristotle’s phainomena – all much more appropriate terms than “intuitions”. For the purposes of this discussion, I think it’s fine to call them “assumptions”.)

Now where all of this gets us into trouble is when we start dealing with thinkers who don’t share our assumptions (and we don’t share theirs). Such thinkers exist even within our own time and place (as with the overeager campus missionaries). But the greater the distance in time and space, the greater the disconnect of assumptions is likely to be – and the more crucial it is to consider not merely the explicit arguments but also the assumptions of the thinkers we hope to learn from.

Figuring all this out was crucial to my own dissertation work. Śāntideva, I noted there, believes that material goods are harmful and still urges one to give them to others for their benefit. If I’d merely considered his explicit arguments and nothing more, I would have had to have stopped there: Śāntideva is a fool who contradicts himself, and there’s an end on’t – and in that case, why bother studying him any further?

But I didn’t do that. Instead, I followed the method of looking for coherent authorship, as stated by Thomas Kuhn: I tried to ask myself how an intelligent person could have written such an apparent absurdity. And that required looking deeper into Śāntideva’s assumptions: the things he believes but doesn’t say. Key among these was the idea that gifts benefit the recipient through the gift encounter and not the gift object. I argue in the dissertation that if you look at the things Śāntideva does say, you can infer that Śāntideva believes this, and it makes sense out of his explicit arguments in a way that you can’t get from looking at the arguments alone. Such an approach, I think, is crucial to making sense of any philosopher outside of one’s own immediate cultural milieu. If all you’re going to consider is the arguments, you might as well not bother. And indeed, most analytic philosophers don’t bother much with thinkers from distant times and places, which, considering their method, is just as well.

But that is not to say analytic philosophy is worthless. Not at all! It just doesn’t prepare you very well for studying the history of philosophy (which is why that history tends to be relegated to the sidelines of analytic departments). What it does very well is attempt to get to truth within a given context, namely ours – to take the incoherent mess of “intuitions” or prejudices, with which we must always begin our philosophical reflection, and start to hammer them into something that actually makes sense. For that reason I often refer to analytic philosophy as the scholasticism of the liberal tradition. Like medieval Christian scholasticism, analytic thought provides an extraordinary level of detailed reflection within one given context, which is necessary if those within that context are going to seriously strive to reach a truth about their lives. But it also makes that thought look parochial from a foreign context; I strongly suspect that the majority of analytical reflection will look as bizarre to people 500 years from now as Christian scholasticism looks to us today. Those people of the future may well be able to benefit from the argumentative details of 20th-century analytic philosophy; but it will require someone with the interpretive approach of a continental philosopher to figure out just what it was the analytic philosophers were going on about.

Rosen argues that we need to see a much closer association between two fields of study often thought separate: logic and psychology. At first glance, the two might seem to have little in particular to do with one another. Logic concerns itself with the proper formal relationships between statements in arguments; psychology, with the empirical investigation of mind and behaviour.

But more basically, what are logic and psychology? Both, really, are the study of thought. One might narrow logic down a bit and say it is the study of reasoning; but reasoning is very much a part of psychology’s subject matter as well. Their differences are not in their subject matter. There is certainly a difference in method; but that difference flows from a more fundamental difference between the two.

Namely: psychology (or at least certain branches of it) tells us how we do in fact reason. Logic tells us how we should reason. Psychology tells us about the circumstances under which we do or do not follow the rules of reasoning that logic sets down as proper. Which is to say: logic is a normative discipline. That is, logic, like ethics and aesthetics, is concerned with goodness and value – with what should and should not be the case, not merely with what is and is not the case. The idea of a truth that is better than falsehood, and of the methods to discover that truth, is part of what makes logic possible.

The relationship between logic and psychology, taken in this way, is more or less the relationship between ethics and ethics studies, between the philosophy of science and science studies, or between theology and religious studies as religious studies is very often conceived. The latter member of each pair tells us how we do reason (or act); the former tells us how we should. But where the ideals of science, ethics and theology tell us how to reason within their particular fields of inquiry, logic tells us how to reason in the general case, including all the others.

Logic is a normative discipline – a discipline concerned with value – because it is fundamentally concerned with truth. And it is part of the nature of truth to be a value, to be better than falsehood. To deny the intrinsic value or goodness of truth makes no sense, as I argued before in claiming that truth has a value independent of pleasure.

The normative nature of truth holds true whatever one might understand truth to be. Analytic philosophers most commonly identify three theories of the truth of statements (or more generally propositions): correspondence, coherence and pragmatic. Speaking broadly, on the correspondence theory, propositions are true if they correspond to reality; on the coherence theory, propositions are true if they cohere with other propositions we hold; on the pragmatic theory, propositions are true if they are effective. But what is presumed by all three theories is that truth is a good thing: other things being equal, propositions that correspond to reality or that cohere with other propositions or that are effective are better than propositions that do not do these things. And when this is true of the more limited analytical theories of truth, how much more so of more expansive theories of truth that do not limit themselves to propositions, like those of Augustine or Gandhi. Augustine probably goes the furthest on this point: truth is goodness – which is God.

The fact that truth and logic are normative and value-laden has important consequences. For one thing, it gives the lie to simplistic claims of “value-free” science or social science (including psychology itself). All intellectual inquiry is predicated on at least one value, namely truth itself. More sophisticated defenders of “value neutrality,” like Max Weber, will argue that the scholar of a scientific field needs to put truth above other values – but we must recognize that this argument is itself a value argument, an ethical argument for the importance of truth relative to other values, at least within certain areas of inquiry. The argument that one should place truth above other values is a normative and value-based argument.

Now the discussion above should not imply that all true statements are good statements, or all true things are good things. It is true that outlying areas of Bangkok were recently hit by disastrous floods; it is not good that this happened. The relation between truth and goodness is and must be more complex than that.

The fact that not all true things are good probably seems obvious to anyone who isn’t a philosopher; but it nevertheless poses some problems. For someone like Augustine who identifies truth with goodness, it would seem to be at the very heart of the problem of bad. If truth is goodness, how can there be things that truly exist but are nevertheless bad? Augustine ingeniously deals with this problem (or at least this aspect of the problem) by identifying badness as a lack of the existence of good – in the same way a modern physicist identifies cold as simply a lack of heat. Absolute evil, on his account, looks very much like absolute zero.

And so with the perennial questions, to which I regularly return on this blog. Central to the idea of a perennial question, as I have expressed it, is that the answers have never come easily. People across cultures, in different places and times, have asked the question – but in each place, people have come up with opposing answers.

To observe this diversity of opinion is humbling. Here are some of the greatest minds in human history, people smarter than I will ever be, reading each other’s work and still coming to opposite conclusions. Can an answer then ever be found?

The quickest, easiest and most tempting response is to throw up one’s hands and say no, or effectively say no: there’s no way to decide between these different answers. This is the postmodern or relativist response, and it’s one to which undergraduates gravitate very quickly – and understandably – when faced with the big questions. But this answer very quickly reveals itself to be both incorrect and unsatisfying – for reasons beyond the performatives I have previously discussed.

For to say “there is no answer” is itself an answer, and an answer that is itself in disagreement with those very great minds. Plato and Aristotle might disagree significantly on the answer to the question of Ascent and Descent, but they will certainly agree that there is an answer to be found. Take the Descent and you will reject Plato; take the Ascent and you will reject Aristotle; say there can be no answer and you will reject both. There’s no way around fundamental disagreement with at least one of the great thinkers on any perennial question.

Or is there? There is another way to address such questions, but it is more complicated than any of the options discussed so far: taking one side over the other; adopting one thinker’s solution as truth; rejecting attempts to find an answer. Skholiast nailed it in his response to my first post on perennial questions. On perennial questions like that of Ascent and Descent, there is in the great thinkers always a dialectic: an attempt not merely to refute the opponent’s position but in some way to incorporate it. Skholiast describes the dialectical process using Hegel‘s complex but key German term Aufhebung (which is the noun form; the verb is aufheben in the present tense, aufgehoben in the past). Aufheben is often translated ineffectively with the word “sublate,” a word which has no real English meaning other than as a translation of aufheben. Ken Wilber renders it as “transcend and include,” which provides a much more helpful understanding of what the German term gets at, but is wordy enough to be awkward. I prefer “supersede,” which covers a lot of the sense of the German word. The new edition of a book (ideally) supersedes, aufhebs, the old. It cancels the old in a sense, moves beyond it and makes it unnecessary, but does so by preserving what is most important in the old while adding things that are new and better.

In the case of Plato and Aristotle, it’s easy to fall into the temptation of portraying them roughly as Martha Nussbaum does in The Fragility of Goodness, or as Raphael does in The School of Athens: as polar and mutually exclusive opposites, Plato seeking only to escape the fortunes of the world and Aristotle to embrace them. But as Skholiast notes and as I have tried to emphasize in my own posts, there is always a Platonic element to Aristotle, an attempt to embrace and incorporate Plato’s transcendence within a philosophy whose overall tendency is more worldly. This Platonic Aristotle comes out above all in sections X.6-8 of the Nicomachean Ethics, where Aristotle says that the contemplative life is the highest and best because it is the most godlike. This is a passage that Nussbaum has a hard time dealing with; she says effectively that Aristotle is contradicting the rest of his work (Fragility 375-7). But she agrees that he feels the power of Plato’s Ascent ideal, and is trying to consider it. It strikes me that his goal was very likely to supersede Plato, to transcend and include him, to be not merely a Descender but a Descender who includes Ascent within his thought. If Nussbaum’s interpretation is right, it may mean primarily that he failed at that task.

The point I’m trying to make is that the perennial questions are best addressed through a dialectical synthesis. What the greatest thinkers do when they address a perennial question is not merely to take a side, Ascent or Descent, ātmanism or encounter. If they do take a side, they will attempt to incorporate the best of the opposing side in their view.

There are two critical elements to the process of dialectical synthesis. First, it is an attempt to find synthesis, not compromise; it is not about finding a middle ground. The middle ground can turn out to be a vicious mean and not a virtuous one. (Compromise, I have argued, has its role in political practice but not in philosophy.) More important is to take seriously the underlying concerns that animate each side and bring them to where they are, and answer those concerns in a way that could be genuinely satisfying to those who have them.

And second, this process of “taking seriously” is a dialectical one: one starts from the positions one tries to supersede, and shows their inadequacies from within, making the opposing positions part of the process of reaching one’s own. It is in this sense that Nussbaum’s and Wilber’s major works are not themselves dialectical, though I think they may aspire to be; the endpoint of the inquiry has already been reached at its beginning. In their works, opposing positions are discussed only to be refuted. Nussbaum tries to make a movement from Plato through various other thinkers and ending in James Joyce; but by the time she gets to Joyce, there isn’t any Plato left.

Not much of what I’ve said here today is new; I’ve made most of these points in the various posts I have linked to above. But I’m trying to bring them together just because I do see my project as one of trying to work out some answers, however tentative they must be, to perennial questions – and I do not believe I’ve found those answers yet. In some respects this post is an attempt to remind myself, and hopefully others with me, of the best ways to think about the great questions – just because dialectical synthesis is such a difficult path to follow, and I think I’ve typically fallen short of it so far myself.

I’ve briefly discussed this question before, with a focus on ethics. Afterwards, following James Doull, I examined how it gets works out in the history of Western philosophy after the Sophists – in ethics. But as Doull knew, there is an epistemological story that parallels the ethical. God in the Hebrew Bible is the arbiter of truth as well as ethics; the Sophists reduce not merely justice and goodness, but truth, to the subject that knows them. (In a similar way, for Doull, Plato’s Sophist dialogue makes essentially the same point as the Statesman, even though it would appear that one is entirely about metaphysics and the other about politics; both are asking how ideals can relate to physical reality.)

In ethics, as I noted in the earlier post, analytic philosophers tend to use the terms internalism and externalism to describe the opposing positions on this question. In ethics, internalists say that genuine reasons for action must come from our own motivations or desires; externalists say there can be reasons that come from outside us. To use Doull’s examples, the Sophists are the ultimate ethical internalists, the Hebrew Bible the ultimate ethical externalist text. (Consider Ecclesiastes, which repeatedly advises us to fear God and follow his commandments even though it repeatedly denies the possibility of our attaining any benefit for doing so, in this life or the next.)

Now, as well as in ethics, analytic philosophers also use the terms “internalism” and “externalism” to describe positions in epistemology (the theory of knowledge). Ethan Mills discussed this other form of the distinction in comments on a recent post, and I’ve been trying to learn some more about it. What follows is some preliminary attempts of mine to think through the distinction between internalism and externalism in epistemology, and relate it to the distinction in ethics. Apologies if the results are somewhat unclear, as I’m still thinking it through.

Before reading this analytical literature, I had already saw an important internal/external distinction of sorts in epistemology, parallel to the one in ethics. (Doull influenced my thinking here too.) In ethics, we may ask whether we acting subjects are the measure of goodness; just so, in epistemology, we may ask whether we knowing subjects are the measure of knowledge. Ethical internalists say our reasons for action must all come from within us, from our motivations; ethical externalists say we can have reasons to act independent of our motivations. There seems to be a parallel set of questions in epistemology: can the reasons for our beliefs be independent of the ways we come to know them? For example, can we logically speak of a truth that no subject is capable of knowing? Or as the Sophists would put it: is man the measure of truth?

It is not quite clear to me, though, that these questions are what analytic philosophers mean when they speak of “internalist” and “externalist” epistemologies. In Ethan’s comment, the key question dividing internalism and externalism is: Can we really be said to know something if we don’t or can’t know why we know it? Externalists say we can, internalists say we can’t. In online introductory works on internalism and externalism, the problem is phrased in terms of justification rather than knowledge: can our beliefs be justified even in cases where we don’t or can’t know why we hold them?

Now, when the distinction was put in this way, it was not immediately clear to me why these positions were even called “internalism” and “externalism.” In ethics the terminology seems clear enough to me: internalists say that any reasons for action must come from our motivations, which are inside us. But what is “inside us” about knowing how we know or knowing why we hold our beliefs, the hallmark of epistemological “internalism” in the analytic sense? Another article, by Joe Cruz and John Pollock, helped clarify. Internalism on Cruz and Pollock’s view “is the view that all the factors relevant to the justification of a belief are importantly internal to the believer”; our knowing how we know or why we hold our beliefs is itself internal to us. An example of the contrasting, externalist view would be “reliabilism”: the view that a belief is justified if it comes from a source that is in its nature likely to be correct, even if we don’t know that it comes from that source. (So in recent debates “common sense” was defended on the grounds that it is
“reliable.”) The distinction might be illustrated with an example from Laurence BonJour:

Norman, under certain conditions that usually obtain, is a completely reliable clairvoyant with respect to certain kinds of subject matter. He possesses no evidence or reasons of any kind for or against the general possibility of such a cognitive power, or for or against the thesis that he possesses it. One day Norman comes to believe that the President is in New York City, though he has no evidence either for or against his belief. In fact the belief is true and results from his clairvoyant power, under circumstances in which it is completely reliable.

A reliabilist would need to say that Norman’s belief that the President is in New York City is justified, because his clairvoyant power is reliable – even though Norman doesn’t know it’s reliable. (Perhaps it’s one of the first few times the power has manifested.) By contrast, an internalist, like BonJour, says that in such a case Norman’s belief is not justified – even though it’s true. Which is to say, I think, that Norman has no reason to believe the President is in New York.

It’s on this matter of “reasons to believe” that I suspect the two analytical internalism/externalism distinctions dovetail the most. The internalism/externalism question in epistemological justification asks: what counts as a good reason to believe something? In ethics, it asks: what counts as a good reason to do something? In both cases, the internalist says that the good reason must be within us.

Ethan notes in his comments that the South Asian Nyāya school is similar to the reliabilists: what matters is that knowledge is formed by the right kind of process, not anything within us. We can know without knowing how we know. So they are effectively externalists in the analytical sense at issue here. Yet at the same time, the Nyāya have a slogan that “whatever exists is nameable and knowable” – for something to exist objectively, it must be available to subjective knowledge. To speak of something which exists but we couldn’t know – that is meaningless. Here, man (or other subjective knowers) is in some sense the measure of truth, as the Sophists would have wanted to have it. On this score, the Nyāya view seems more comparable to ethical internalism, the view that good reasons must come from within us – truth is in some sense within us knowers as well. Perhaps one could describe the Nyāya as externalists epistemologically, but internalists metaphysically or ontologically?

So far I’ve explored two major perennial questions: ascent versus descent and intimacy versus integrity. I have taken these as two different axes along which philosophies can be classified – in their ethics and soteriology as well as their metaphysics and epistemology.

But why should we treat these as exhausting the perennial questions? I think there’s value in limiting the number of questions we treat as perennial – in being prepared to say “those are different aspects of the same question” or “those are different ways of asking the same question” rather than allowing the questions to proliferate randomly. But that’s not to say the number of questions should be limited to merely two – though it’s certainly interesting to consider the two as axes on a single graph.

For there are other questions which are similarly widespread and have similar ramifications. A little while ago I pointed to Mou Zongsan’s distinction between “perfect” and “separation” theories; these map onto the distinction I discussed earlier between ātmanism and encounter, but Mou effectively tries to show that ātmanism-encounter is its own perennial question, distinct from the integrity-ascent and intimacy-descent positions they might seem to map onto.

Other perennial questions are significantly better known than the debates I have discussed above. One of these is human nature: the question that finds its most classic expression in the ancient Confucian debates between Mencius and Xunzi, but is also well expressed in the West in Rousseau and Augustine, among others. So too, I suspect it is at the heart of the changes in Buddhism as it moved from India to Mencian China. At its heart, this is a metaphysical question about what human beings are and what makes them so – a question which is also open to at least some empirical verification or falsification. But it is also an ethical question. If human beings are naturally good, they need far less ethical correction, need to watch themselves or be watched far less, than if they are systematically prone to error and wrongness. It extends into soteriology: a good human nature makes sudden liberation more plausible. And at several points the recent debates over “common sense” extended this question into epistemology. To what extent are human reasoning processes naturally good enough to lead us to the truth, and to what extent are they so prone to error that they need regular and systematic correction?

Then there is the similarly metaphysical question of free will – much less subject to empirical verification. The empirical methods of natural science assume that the world is made of causal processes whose workings can be ascertained; this very assumption begs the metaphysical question at issue. But it too has significant ramifications in ethics and politics. Free will is a fundamental assumption behind the characteristic organizing concepts of modern liberalism: rights, respect, autonomy. The idea that individual choices are to be respected qua choices – as opposed to their being instrumental to other goods like happiness – implies that something about these choices gives them a different status from other phenomena in the universe. So you can’t get even close to a Kantian ethics without free will – but consequentialist ethics can do fine without it. I’m told that Fyodor Dostoevsky even saw this point as the fundamental difference between the worldviews of Protestantism and Catholicism: Protestants sacralize individual autonomous choice even if it leads to overall misery; Catholics want an order that produces general happiness even if it leads to tyranny over individual choice. (Whether his characterization was accurate, let alone whether Eastern Orthodox churches provide the appropriate synthesis he thinks they do, is a separate topic.)

The idea of free will has been particularly important in the West, but it has not been limited to that context. It is important enough to Śāntideva that he spends several difficult verses refuting it. Very much like Nietzsche, Śāntideva believes that the idea of free will is harmful and dangerous because it leads us to blame others: their actions have causes just like a stomach upset does, so we should not get angry at them any more than we get angry at our stomach bile. And I think points of view like Śāntideva’s tend to frame the left-right axis in Canadian politics, and in other countries where God is not a serious political issue. The right believes criminals make free choices, and so deserve their punishment, while the left seeks to reduce the causes of crime; and if people’s fates in society largely come down to their free choices, then the government has less of a duty to help those whose fates turned out poorly.

The questions I’ve listed – ascent/descent, intimacy/integrity, ātmanism/encounter, free will, human nature – hardly exhaust the list of perennial questions either. In future weeks I’m hoping to examine others. But I’m returning to the idea of perennial questions now because I suspect that it may form part of a highly fruitful method in cross-cultural philosophy. Too much cross-cultural philosophy so far has been dominated by the idea of a philosophia perennis, a single universal philosophy shared across cultures. That idea is usually taken to refer to some sort of Advaitic mystical monism, a single cosmic truth that can be known through mystical experience. And while ideas of that sort are indeed present in many cultures, they’re rarely all that widespread. Most people do not believe this so-called perennial philosophy. Moreover, there’s an odd parallel between that sort of perennialism and the view of “common sense” recently advocated on this blog by Thill Raghunath and others. Though Thill describes “common sense” as excluding “religious” ideas (which I suspect includes the “perennial” mystical monism), he shares with the perennialists a common view of human access to truth: all humans, across cultures, share an innate faculty which allows them access to truth, but most humans access this faculty so little that they are enmeshed in delusion. (As I noted above, epistemologically this seems to put both Thill and the perennialists on the side of the human nature debate that stresses our natural goodness.)

What is truly universal to me in philosophy, it seems, are not the answers but the questions; and that is why I think the cross-cultural study of philosophy should devote more time to these questions. To the extent that the answers are universal as well, it seems to me that multiple and contradictory answers are universal: both mystical Ascent and a “common sense” Descent are found across cultures. The student of cross-cultural philosophy should pay attention to both sides.

In August I will be taking some vacation time with my wife and my friends. So there will be no blog post next week; posts may be sporadic for the rest of the month as well.

The question, or some form of it, came up a number of times in recent discussions of “common sense”: if common sense isn’t reliable, I was asked, what is? I’m going to try to avoid the word “reliable” as I think its different uses became confusing in the previous debate; I have little stake in its use as a term. But the basic question of determining truth from falsehood is a crucial one and worth asking.

That’s not to say, however, that it admits easy answers, for I don’t think we should expect easy answers on the most basic philosophical questions. If the answers were easy, it would be a stunning and bizarre fact that so many intelligent people have spent so long trying to answer them and explain them without coming to a resolution (as indeed has, so far, been the case in the recent debates, though these have lasted only weeks and not centuries). This is one reason why I don’t identify knowledge of the truth as deriving from a single source like “common sense” – though my posts and comments should make clear I have many more specific problems with that concept, especially as defined by Thill and other commenters on this blog.

How should we identify truth instead? The question of how we should discern truth is closely linked to the question of how, in practice, we do discern it. I like to say that we start where we are: we assess new information learned by reasoning out its coherence with the information we have already accepted. The new information comes in through sense perception one way or another, though the perception might be of someone else’s testimony: I observe you tell me something.

So I think the Vedānta schools are probably right when they describe the means of knowledge (pramāṇas) as perception, inference and authority – that is, the testimony of sources we trust. But that’s not to say any of these sources are always right. Rather, they’re right often enough to be worthy of our belief unless there is some reason to mistrust them in a particular case: for example, I would normally believe my eyes telling me that there is a large yellow stick floating in front of me, but I can’t touch the stick and I have heard that this perception is a symptom of eye diseases, so I don’t.

When a particular belief is in question, though, it’s not enough to refute it merely by saying we learned its contrary through any of these means of knowledge; for they can be, and often are, wrong. Moreover, this is not a matter of one means taking precedence over another. Yes, my senses tell me that the sun revolves around the earth; but because I trust the authority of trained astronomers, I know that this is not the case. Or alternately: a scientist friend (in this case our esteemed commenter Ben) tells me there’s a new article in a refereed psychology journal telling us that caffeine doesn’t actually increase alertness; but I don’t accept this claim because it is so completely contrary to my felt and observed experience of caffeine’s effects on myself. The conclusions must have been misreported, or something wrong with the methodology, or the sample unrepresentative, or the definitions of “alertness” something very different from what I understand by it.

But how do I, or should I, make the decision in those cases where means of knowledge conflict with each other or with themselves? I don’t think a hard-and-fast rule can be provided. Providing an easy and definitive answer to the question “How can I tell true from false?” is like providing an easy and definitive answer to the question “How can I become a better fiddle player?” Discernment of true and false is a virtue, a skill learned with time and practice; there is a wealth of tips and advice one can offer about how to do it better, but one can’t provide a formula for it that will settle disputes in advance. (Or rather, one can; it’s just that one will be wrong.) In saying this, I’m expressing agreement with a contemporary school of analytic philosophy known as virtue epistemology.

Thill disputes such claims:

If you have easy answers to determine what is unreliable, indeed, if you can go to the absurd length of deeming common sense (on which you rely for your very survival) unreliable, you can surely specify what you consider reliable and what you depend on to function in the world…. your claim that it is not easy to ascertain what is reliable implies that it is not easy to ascertain what is unreliable. This is at odds with your easy dismissal of the appeal to common sense on the grounds that it is unreliable.

But I’ve made no such easy dismissal. The easy answer Thill asked for, as far as I can tell, is a statement of “that which is X is reliable and that which is not-X is not,” an exaltation of one single source of knowledge in the way that Thill exalts common sense, which is what I’ve refused to provide here and elsewhere. My refutation of “common sense” as a reliable source of knowledge didn’t rely on a single-sentence knockdown; more importantly, it didn’t say simply “all X is true and all Y is not,” but tried to show us the complexity of the world and of knowledge. I have never said that the items of knowledge included in “common sense” are always wrong; indeed, I suspect most of them are right. The point was that we do not have any special reason to believe a claim based on the fact that it is said to belong to “common sense” (in the sense of knowledge learned without training).

If my alternative view can be described in a sentence, it is probably this: we need to engage in the complex process of knowing as best we can. And if that sounds vague, that’s because it is, intentionally. You should be suspicious of anyone who claims to give you a single easy tip that sums up the whole of how to play the fiddle, do successful biology experiments, or pick up romantic partners. You should be similarly suspicious of anyone who claims to easily sum up how to tell truth from falsehood in the general case.

There is, of course, plenty to be learned in each of these practices; that’s one of the reasons they’re not easy. There are various tips and tricks that can aid in each: play emphasized notes with a down stroke of the bow; control as many variables as you can; groom your hair carefully; trust the conclusions of scientists with expertise in their fields. All of these tips are generally wise, but still admit exceptions: there are two emphasized notes of the same pitch in a row; controlling an additional variable would cost so much that you’d need to hire fewer staff and make careless mistakes; you’re courting someone who likes the dishevelled look; the scientist misspoke because she’s having a bad day. And in each field there is also advice offered that is well meaning but inappropriate, advice we should not take: play as fast as you can; fudge your data a bit and nobody will notice; pretend to be wealthier than you are; treat a claim as true because one can learn it without specialized training. The acceptance or refutation of one of these tips may be a relatively simple matter by itself; but that doesn’t make the whole practice an easy one.

Is this a definitive account of how we can discern truth? No, it’s just a start. But that’s the point.

As I did last week, I will assume that my readers have read the two posts that preceded this one on the subject; I will not assume that you have read the comments to those posts. In his first comment, Thill very helpfully gives us his definitions of three key terms whose meanings have so far been elusive in this debate:

The word “plausible” also has the meaning “worthy of being accepted as true or reasonable” and this is exactly sense in which I am using that word. Interpreting “plausible” in terms of “apparent truth”, as Amod does, is at odds with this sense.

The word “reliable” means “credible; trustworthy; dependable.” That which is plausible (= worthy of being accepted as true or reasonable) is, therefore, also reliable in this sense.

The word “infallible” means “indubitable; exempt from and incapable of error”. That which is true is also infallible. Truth excludes error and doubt. Hence, knowledge of truth also excludes error and doubt. Therefore, truth and knowledge of truth are infallible.

The distinction made here was surprising to me. As it is described here, the distinction between infallibility (on one hand) and plausibility or reliability (on the other) appears to be a distinction between truth and justification. If something is infallible, that means that it is actually true. If it is merely plausible or reliable, that in turn means that it is worthy of being accepted as true, worthy of our trust, credible, believable – that is, we are justified in believing it. Plausibility and reliability are about justification, not truth as such. And there must be a distinction between the two, for Thill’s entire argument depends on there being a significant difference between infallibility and reliability (or plausibility), and with these terms defined thus, that requires a distinction between justification and truth. If we are only justified in believing those things that are actually true, then only the infallible (that which must be true) is reliable (that which we are justified in believing); but that is exactly what Thill’s argument requires him to deny. For Thill there must exist some claims which are reliable but not infallible; and according to the definitions above, these are claims which are at least potentially false but which we are nevertheless justified in believing. (Unless, of course, the ground of these definitions shifts beneath our feet.) If we are never justified in believing false things, then the distinction between reliability and infallibility – as expressed here – collapses.

So assuming the distinction between truth and justification in this way (thus allowing for the distinction between infallibility and reliability), let us continue to “common sense” – in Thill’s definition of the term, as beliefs which can be learned by human beings without special training (which has also not yet been defined). Thill, as I understand it, wishes to claim that common sense is “worthy of being accepted as true or reasonable” and “credible; trustworthy; dependable” – qua common sense. That is, insofar as something can be learned without specialized training, it is worthy of being accepted as true or reasonable.

Now, let me return to my favourite counterexample. Since we learn without specialized training, from the evidence of our senses, that the sun goes up and down as a thrown baseball does, this fact clearly belongs to common sense as Thill defines it. (And I will reiterate that if common sense merely tells us that the sun appears to go up and down, then it must be superseded by specialized training when it comes to the actual truth, for it tells us only about appearances and not truth. If common sense is to have any of the philosophical weight claimed for it, certainly if it is to be considered reliable, then it must tell us about reality and not merely appearance.) It is for that reason – it has been in response to this claim – that Thill has already accepted or at least implied, repeatedly, that common sense is not infallible. As must be the case, for in this case the conclusions of common sense are simply false.

Now what of reliability and plausibility? If common sense qua common sense is “worthy of being accepted as true or reasonable” and “credible; trustworthy; dependable,” this too must include the false claim that the sun literally rises and falls. Thill introduces the distinction between infallibility on one hand, and reliability or plausibility on the other, in order to claim that every single common-sense claim is, if not infallible, still reliable and plausible. But this set of claims includes the claim that the sun rises and falls. The claim of the sun’s rising and falling, because it is a member of the set of commonsense claims, must therefore be considered “worthy of being accepted as true or reasonable” and “credible; trustworthy; dependable” – even though we have already agreed it to be false. We cannot avoid such absurdities so long as we consider a commonsense claim “worthy of being accepted as true or reasonable” merely on the grounds that it is common sense. (And if you don’t like this example, I reiterate that if common sense is indeed not infallible, there must be cases where it is wrong, and those cases may be substituted here mutatis mutandis.)

Now several of the critiques that the commenters have made to my posts have suggested that they assume common sense is all or nothing: if I say (as I have) that common sense as a category is not reliable, that must imply that every member of the category is unreliable. But, as Ben has rightly and repeatedly noted, this assumption is a pretty basic logical mistake. I have never said that everything which falls in Thill’s category of “common sense” is false, or even that most of it is. I am merely saying this: the bare fact that a claim falls within the category of common sense is insufficient reason to consider the claim worthy of being accepted as true or reasonable. Each claim must be accepted on its own merits, based on the variety of sources of knowledge we have available to us (logic, perception, trustworthy authority). The fact that something is learned without specialized training does not make it worthy of belief, any more than the fact that it is learned with specialized training.

This point (in addition to brevity) is why I entitled the earlier post “lack of training is not reliable” rather than “beliefs achieved without training are not reliable.” Some beliefs obtained without specialized training are indeed reliable, in the sense discussed here; but their reliability does not stem from the absence of specialized training. I reiterate: the fact of a belief’s being learned without specialized training does not make that belief worthy of being accepted as true or reasonable – let alone actually make the belief true.

One further note: So far I have been pushing ahead with objections to the common-sense advocates’ views and their logical flaws. I have not yet addressed a central objection that they have made to my view: that ways of knowing other than common sense (such as science) themselves depend for their reliability on common sense itself. This point should be addressed, especially given some of the claims I have just made in this post, and I intend to do so. (Ben has already made some important points on the topic.) I intend to take it up in a post soon, but this one is already long enough. Let us discuss the matters here in the meantime.

In those comments I challenged Thill to define the term “reliable,” which he had previously introduced to the discussion. I structured the post around the term “reliable” because in Thill’s previous comment, it had been at the centre of his only serious response to the point that “common sense” can be wrong (as in the case of sunrise and sunset). He said: “The fact that it is not infallible does not support the conclusion that it is not reliable!” No doubt I should have probed the definition of “reliable” further in the post – examining what Thill could have meant by it; I did not. I tried to make up for that lack in a later comment, where I asked Thill to define “reliable.” Thill responded that the onus was on me to define “reliable” since I had advanced a thesis relating to it; but my supposed thesis was intended as a response to his own thesis about the reliability of common sense, a word which, again, he introduced to the discussion. So I noted that I am happy to drop the term from the discussion as long as he, too, is willing to refrain from using the term “reliable” to refer to the epistemological status of so-called common sense. (That also applies to the others, Jabali108 and Neocarvaka, who have been exalting “common sense” in recent discussions.)

If we drop “reliable,” where are we left? We have established that “common sense” is not infallible. And within this discussion we may no longer describe common sense as “reliable,” unless someone wishes to reopen that can of worms – and anyone who does so had better define “reliable” and be prepared to defend the definition. (I note that Thill did briefly identify “reliable” as meaning “not likely to be justified or true” – but as I noted here, he had earlier claimed that concepts of likeliness or probability do not apply to the kind of philosophical claims most at issue in these discussions, such as the Madhyamaka claim that the visible world is illusory.)

So is there any way that common sense differs from any other kind of belief? When we assert “common sense tells us that X,” do our listeners have any additional reason to believe this claim beyond the bare assertion of X?

It is in the following comment, I think, that Thill updates his position on such questions in a way that does not rely on “reliable”:

As I have made clear a couple of times, it is a mistake to think that all of common sense is infallible or none of it is. Some of it is infallible, e.g., fire burns unprotected human skin. Some of it is plausible belief, e.g., there will be sunrise tomorrow.
Plausibility, is of course, consistent with fallibility. It is plausible to believe that there will be sunrise tomorrow, but it is a fallible belief.

Here, some common sense is infallible, and some of it isn’t; the latter is merely plausible. “Plausible,” as I understand it, means that something appears on the surface to be true. And I have no beef with the latter; I don’t believe I have ever said that common sense (or even any portion of it) is implausible. But even the Mādhyamikas agree that common sense is plausible: the majority of common sense turns out on reflection to be false, but we believe it in the first place just because it is so plausible. So for them, the plausibility of common sense is exactly what’s wrong with it.

Now what of those parts of common sense that are not merely plausible but infallible? Thill does not tell us how we are to distinguish infallible common sense from merely plausible common sense. But clearly, this distinction cannot be made merely on the grounds that it is common sense; the fact that something is common sense does not itself make it infallible, it only makes it plausible. Rather, there must be some criterion according to which some kinds of common sense are determined to be infallible and others are not – and this criterion cannot be the fact that they are common sense, since it has been agreed that there are kinds of common sense which are not infallible. By itself, the fact that something is common sense (by Thill’s definition) tells only that it is plausible; and something which is merely plausible may well be false.

In sum, it does not seem that, even on Thill’s view as developed here, common sense qua common sense carries any epistemological weight beyond mere plausibility. There is some “extra-commonsensical” criterion or criteria according to which common sense may be judged infallible or not. Once we hear what that is, we can debate whether it is correct that this criterion allows us to declare certain beliefs infallible. Regardless: according to this quote here, certain kinds of common sense are proposed to be infallible; but it is not and cannot be the fact of their being common sense that makes them so.

So when Thill or others use a phrase of the form “common sense tells us that X” (as for example here), are we to understand this as meaning only “X is plausible”? Which is to say, “X appears true to the untrained eye, but could easily on further reflection prove to be false (unless established to be infallible by some separate criterion of infallibility)”? I focused before on the concept of reliability because Thill’s use of it seemed to imply something much more significant than this sort of plausibility. But if it is indeed the case that (for Thill and others who refer to common sense) the fact of something being common sense indicates merely that it is plausible in this sense, then I will cease criticizing the concept, for it turns out we have no significant disagreement on that score. We may move on to other matters.

I am very busy with work this week, and don’t expect to be able to respond to comments at the length and frequency I did with the previous post. But I will read them and think about them.

Let me start out by noting that I see some philosophical value in appeals to common sense defined in a certain way. This is the sense that I outlined in my first post on the topic: the prejudgements one brings to a given inquiry, especially as they come out of shared assumptions of one’s own society. My commenters seem to have something quite different in mind, however. Thill said so explicitly in his first reply: he understands “common sense” not as socially shared assumptions or presuppositions, but as something else, something one might describe as more objective. His most recent (and probably clearest) statement on the subject contrasts common sense with “special training, e.g., items of scientific, technological, or aesthetic knowledge, scientific, technological, or aesthetic ways of knowing, and scientific, technological, or aesthetic standards of reasonableness,” and also with the paranormal or supernatural; but this ordinary or untrained sense, he claims, still provides us with reliable access to knowledge. (I discuss Thill’s exposition of “common sense” here because he has so far spelled it out more explicitly than the others.)

One must first note that the information that one learns without special training can be wrong. The point is clearest with respect to natural science, for the information one learns with scientific training so often contradicts the information one receives without it. I have returned repeatedly to a key example: untrained inference tell human beings that the sun goes up and comes down in the sky, in a manner similar to a baseball being launched and landing. This is a clear case in which “common sense” as Thill understands it is demonstrably wrong. Thill twice gave a wholly inadequate reply, based around the inarguable point that “Science does not deny that we perceive sunrise and sunset.” But that evades the important issue. The question I asked multiple times in response is: what is it that our supposed common sense tells us about the sun? Does it tell us merely that the sun is perceived to rise and fall (even though it actually doesn’t)? Then we can depend on common sense only to tell us about appearance and not about the truth of the matter; specialized training supersedes common sense when it comes to truth. Or does common sense tell us that the sun really does rise and fall? Then common sense is wrong in this case and can be wrong in other cases too, and we need specialized training to be able to tell when it is wrong and when it is right.

We made some progress in a discussion tangential to a more recent post. Here, as far as I can tell, Thill effectively admitted for the first time that common sense is fallible. It can be wrong; the fact that something is perceptible to the untrained, or widely understood among them, does not automatically make it correct.

This position appears to be a (welcome) change from a previous definition of common sense, as “the stock of Truths pertaining to the world naturally accessible to normal human beings anywhere on this planet and the faculties of the ordinary human mind employed in gaining access to those truths…” Fortunately, Thill is no longer treating common sense as true by definition, as he did in that previous definition. For then it would have been a useless tautology for establishing the truth: if common sense is necessarily true, then it does us no good to say that something is true because it is common sense, for that is merely saying it is true because it is true. To establish that something was common sense, we would need to establish its truth first, or we would not really have established that it was in fact common sense.

Rather, Thill’s admission that common sense is fallible comes in the context of distinguishing the infallible from the reliable. He compares common sense to visual perception: our eyes are fallible, as through optical illusions or hallucinations. But despite the existence of illusions and hallucinations, we can generally rely on our the evidence of our eyes, and we will be right the vast majority of the time.

But what reasons do we have to believe that “common sense” is indeed reliable at all, let alone reliable to the degree that visual perception is reliable? The latter claim is a most extraordinary one – and one for which I have seen precious little evidence or argument proffered. If the world can be known so easily without training, one wonders why one would ever bother with any training at all – including scientific training, which Thill has explicitly included among the kinds of training not necessary for common sense.

Science proves “common sense” wrong in a great many ways; sunrise and sunset are only the most obvious. Common sense tells us that a piece of rock is a continuous, solid whole; science tells us that in fact it is made of separate atoms that do not touch each other but are kept apart by force. Common sense even tells us that something as perfectly suited for its work as the human hand or eye could not have happened merely by random chance: the only non-biological phenomena we see with that degree of adaptation are the products of deliberate intentional design, by humans or other animals (such as a beaver dam or beehive). One can see this point without any specialized training; and when one does so, one is wrong.

But such situations, of course, are exactly what one should expect. People who think and train hard to learn about a given matter for a long period of time learn things about that matter that untrained people do not. It should come as no surprise that scientists know the workings of the natural world better than those with no scientific training, for they have worked long and hard at establishing conclusions that are better than everyday ones. To say that they do not is the lowest form of know-nothing populism. And similarly, though their methods are often different from scientists and though they are often wrong themselves, the views of the great philosophers are nevertheless usually more adequate and more profound than those of untrained “common sense.” Thinking about something longer usually makes you understand it better; if anything should be obvious to the untrained, it’s that point.

I have argued before that this is the great problem with relying too heavily on “intuitions” in ethics as well: there should be no primacy given to untrained knowledge. The whole point of training in something is to do better at it than do the untrained. There is no reason why the burden of proof should be on the trained. Rather, it should be on those who go against common sense in the first sense: the existing assumptions and prejudgements with which a debate implicitly begins.

I fully expect to get debate on the points above; but let me offer one caveat before it begins. You will be wasting your time if you try to refute anything I have said here with a comment like the first one here: “It is a non-sequitur to conclude that something must be good or great on the basis that a few or many have been reading or subscribing to it for a significant number of years. Just take a look at the history of superstition, religion, theology, ethics, and philosophy!” That history is exactly what I have been looking at in considerable depth for many years myself, and it is exactly that study that has convinced me the great thinkers of the past are much smarter than you or me, however many points they might be wrong about. I have little patience for that sort of question begging.

My post recounted Carl Sagan’s claim that although “religions” claimed an ideal of humility, science was actually more humble; I argued that the two were in fact very similar. A comment from Ben acutely pointed out something I had been missing, a way in which Sagan was right that the tradition was different. Sagan, Ben points out, is defending “not the humility of individuals, but the humility of the whole tradition.” Science as a whole is able to admit when it is wrong, in a way that Christianity and Buddhism are not. In a following dialogue, Ben and I agree that science maintains an institutional humility that “religious” traditions do not, though those other traditions likely do a better job of promoting individual humility.

Other commenters took issue with this agreement, however. If you follow the comment threads on this site with any regularity, you will know that Thill and Jim Wilton do not usually agree on very much. But this time, they unanimously condemn the point shared by Ben and myself: “There is a category mistake here,” says Thill. “Traditions cannot be said to be humble or arrogant. Only individuals can be said to be humble or arrogant.”

And this is a question that well deserves further philosophical exploration. Can an institution or a tradition possess a virtue? Can a government be courageous? Can a corporation be honest? Can a tradition be humble?

The answer will necessarily be “no” if we define “virtue” (or any of its species) strictly, so that virtue is by definition individual. But I see no clear reason why we should do this. Going back to earliest accounts of the concept, Aristotle does not limit virtue to individuals; in explaining aretē, the word we translate as “virtue,” he speaks of the aretē of a knife: a virtuous (or excellent) knife is one that cuts well. Even thinking of common English usage, we can speak of an honest car dealership, one where all the sales staff are genuinely expected to be upfront with their customers and act accordingly. We can speak of a courageous action taken by a political party, when it adopts a platform that is politically unpopular but is nevertheless the principled thing to do.

Now common usage can and should be criticized; everyday speech is often inaccurate. Are these examples of category mistakes? Virtue is realized and expressed in action; if human collectivities can take action, that fact suggests that they can also be virtuous. But is it inaccurate to speak of an action taken by a collectivity? When we speak of an honest car dealership, a generous government or a humble tradition, is this merely an inexact way to say that these collectivities are generally made up of honest, generous or humble individuals?

I don’t think so, at least not necessarily. The idea that the virtues or actions of collectivities are merely those of their constituent individuals – this puts me in mind of Margaret Thatcher’s famous quip that “there is no such thing as society. There are individual men and women, and there are families.” But in this very quote Thatcher shows that she’s not ready to carry a reductionist individualism all the way: there are families, which she grants an existence distinct from the individuals who make them up. If families are not merely the individuals that make them up, then surely other institutions – including society itself – can also be more than their constituent individuals.

Collectivities can take on a life of their own. (I say “collectivities” rather than “groups” because the latter term tends to connote a mere aggregation of individuals, prejudicing the discussion in that direction.) We understand this point when we make the important distinction between the rule of law and the rule of men (or women). A government (or a corporation) works best when its members act not according to their arbitrary individual preferences, but according to the interest of the whole organization and the precedents that have been collectively established. When an organization successfully acts according to the rule of law, it is that organization as a unit and a whole, and not merely the individual members who make it up, that is acting justly. It is a just organization, not merely a bunch of individuals who happen to be just by themselves. To describe the organization as just is no category mistake; it is correct.

It is in terms similar to these that I think one may accurately speak of the humility of a tradition – and as something quite separate from the humility of individuals. As Jabali108 noted, defining the terms matters here. I set out a basic sketch of the idea of a tradition two weeks ago, as consisting of both a normative ideal and a set of institutions which often does not live up to that ideal. Thill, rightly I think, pointed out a third separable element of a tradition: its body of accumulated knowledge.

As for humility, I take it to mean the awareness of one’s limits and weaknesses, not only in an intellectual sense but also in a practical one – acting on the recognition that one is fallible and dependent on others. In a more specifically intellectual or epistemological sense, it means listening carefully, recognizing that one has never thought of everything, that others very often have something valuable to contribute – even when one maintains the courage to defend one’s own sincerely held convictions. Above all, perhaps, the readiness to admit when one has been wrong. A mean between the vices of arrogance on one hand and meekness or timidity on the other, as I said to Thill. (If this definition seems imprecise, that’s intentional: spelling out the nature of a virtue too precisely implies that one already knows exactly what to strive for, which in my books itself demonstrates a lack of humility.)

On these terms I defend my previous claim, developed with Ben: natural science maintains an institutional humility as a tradition, because it does not take its claims as infallible, is ready to see them overturned when better evidence comes to light. The ideals of scientific tradition encourage its institutions to act in a humble way. This institutional humility is a very different thing from encouraging the humility of individuals; and indeed the two are in distinct tension with one another. When a tradition emphasizes its own unchanging rightness, as Buddhism or Christianity does, it is much more likely to foster a sense of individual humility – a recognition that one as an individual doesn’t have all the answers, that one has been wrong before. I think this is typically a good thing for the individual within the tradition; but it’s not so good for the health of the tradition itself. Science is a whole made humble by its arrogant members; the “religions” are wholes made arrogant by their humble members.