article

RAE LANGTON (B. 1961)

Langton is Professor of Philosophy at the University of Cambridge and has made influential contributions tothe history of philosophy, ethics, metaphysics, and feminist philosophy. She is the author of Kantian Humility(2001) and Sexual Solipsism (2009).

IGNORANCE OF THINGS IN THEMSELVES

1. Skepticism and Humility

Many philosophers have wanted to tell us that things may not be quite as they seem: manyhave wanted to divide appearance from reality. Democritus wrote in the fifth centuryBCE:

by convention sweet and by convention bitter, by convention hot, by convention cold, byconvention color; but in reality atoms and void.

1

Plato argued for a different division: the imperfect, changeable things we see around us aremere appearance, and reality is an independent realm of perfect, invisible, eternal forms. Muchlater, Descartes wondered whether the familiar world of stoves and dressing gowns, streetsand people, might turn out to be mere appearance—not because it is less real than the realm ofatoms, or Forms, but because, for all we know, the stoves and dressing gowns, streets andpeople, don’t exist at all.

2 Perhaps I am dreaming or deceived by an evil demon who interfereswith my mind (like that evil neuroscientist of science fiction!) so that what appears to me isnothing like what’s really there.

That demon still haunts the halls of philosophy, despite Descartes’s own efforts to banishhim. The mere possibility of his deceptive machinations persuades some philosophers thateven if there is actually no demon, and appearance captures reality very nicely, we neverthelesscan’t be quite sure that it does. This means we don’t know what we thought we knew. Weconfront skepticism. We don’t have knowledge of “the external world.” We are ignorant of“things in themselves,” in some sense of that phrase: we lack knowledge of things independentof our minds.

Kant described skepticism as a scandal, and in 1781 he published his Critique of Pure Reasonto set the scandal to rest.

3 The Critique is a brilliant but formidably difficult work. In it, Kantaims to show that skepticism is wrong because, roughly, we could not have thoughts at allunless we had thoughts about things. Perhaps he was trying to say that appearance just isreality: for provided we are thinking at all, we can’t be wholly ignorant of things.

Whether Kant set skepticism to rest is one question. Whether he was really trying to, isanother. For Kant said something else as well, famously and often. Although we have knowledgeof things, these things are only “phenomena,” and “we have no knowledge of things inthemselves.” Are those the words of someone offering a cure for skepticism? The skeptic sayswe have knowledge only of appearances: Kant says we have knowledge only of phenomena. Theskeptic says we have no knowledge of things independent of our minds: Kant says we have noknowledge of things “in themselves.” Appearance is not reality after all. It looks like Kant issaying just what the skeptic says—doesn’t it?

Evidently, it depends what Kant means by “things in themselves.” If “things in themselves”means things independent of our minds, then being ignorant of them is a way of being askeptic. Instead of having a cure for skepticism, Kant has the disease. To be sure, Kant’sproposal relieves the symptoms: he offers a wealth of arguments about the very specialknowledge we have of objects—but they are phenomenal objects, mere appearances. What adisappointment, if we were hoping for knowledge of reality, of things independent of ourminds. What consolation is it to learn that we have special knowledge if it’s knowledge of mereappearance? Kant’s subtle arguments about the conditions of our thought look irrelevant ifthey deny knowledge of reality, and again land us in skepticism.

What, though, if Kant means something quite different by “things in themselves”? Thenignorance of “things in themselves” needn’t be skepticism. Knowledge of phenomena needn’trestrict us to knowledge of mind-dependent appearance. That is exactly the idea we’re going topursue here. We won’t go into Kant’s famous arguments against skepticism—about how wecan’t think unless we think about objects. We’re going to look instead at what “ignorance ofthings in themselves” amounts to. We’ll take seriously the possibility that you and I, right now,are ignorant of things in themselves, just as Kant said—and that we can welcome thisconclusion without thereby welcoming skepticism.

The key idea is this. The phrase “things in themselves” does not mean “things independent ofour minds.” It means “the way things are independently”; that is, independently of theirrelations not just to our minds but to anything else at all.

We are often interested in the relations one thing bears to another. Sometimes the relevantrelations are spatial: the tennis ball flew over the net and over the white line. Sometimes therelevant relations are biological: Jane is Jim’s cousin and Joan’s granddaughter. But when wetalk about the relations a thing has to other things, we tend to assume there is something moreto the thing than those relations. There is more to Jane than being Jim’s cousin and Joan’sgranddaughter. There is more to the tennis ball than its passage over the net and the whiteline. Now, Kant sometimes uses the word “phenomenon” to mean, quite generally, an object “ina relation” to some other object. And he sometimes talks about this assumption that there mustbe something more to an object than its relations to other things:

The understanding, when it entitles an object in a relation mere phenomenon, at the sametime forms, apart from that relation, a representation of an object in itself. (Critique, B306,emphasis added)

He also says:

Concepts of relation presuppose things which are absolutely [i.e., independently] given, andwithout these are impossible. (A284/B340)

This absolute or independent thing, which isn’t exhausted by its relations to other things, issomething to which we can give the name “substance,” which just means an independent thingthat has an independent, or intrinsic, nature:

Substances in general must have some intrinsic nature, which is therefore free from allexternal relations. (A274/B330)

Putting this all together, the idea that there is a thing “in itself” turns out to be the idea thatthere is something to an object over and above its relations to other things: something more toyou than being the son or daughter of A, the cousin of B, the grandchild of C; something moreto the tennis ball than its spatial relations to nets and lines on a tennis court. A thing that hasrelations to something else must have something more to it than that: it must have someintrinsic nature, independent of those relations. It is this something else, this something more,that is the thing “in itself.”

• we take this idea at face value, it promises to solve the difficulty we face. Ignorance ofthings “in themselves” is not skepticism. It doesn’t rule out knowledge of things independent ofour minds. It rules out knowledge only of a thing’s non-relational, intrinsic properties.
• can know a lot! Appearanceis reality: things as they appear to us are things as they reallyare. But something is still ruled out; namely, knowledge of how things are independent of theirrelations to other things. Appearance is reality, but it’s not allreality. We can know a lotabout the world, but we can’t know everything about it: we can’t know its independent,intrinsic nature.
• say we can know a lot about something, but not everything, is not skepticism but a kindof epistemic modesty—so let’s call it “humility.” And since it is at the center of Kant’s philosophy(or so I’m arguing), let’s call it “Kantian humility.” 4 what follows, I’m going to say why Kantbelieved it. And then I’ll say why you, too, should believe it.
• Humility in Kant
• ve suggested that Kant’s distinction between “phenomena” and “things in themselves” is acontrast not between “appearance” and “reality,” but between extrinsic and intrinsic aspects ofsomething. On this usage, if we say a tennis ball fell over the white line, we ascribe to it arelational, hence “phenomenal,” property, whereas if we say it is spherical, we ascribe to it anintrinsic property, which concerns the tennis ball as it is “in itself.” Let’s summarize thedistinction this way:

Distinction: “Things in themselves” are things that have intrinsic properties; “phenomena”are their extrinsic, or relational, properties.

Against this backdrop, ignorance of things in themselves is not skepticism but ignorance ofcertain properties—intrinsic properties:

Humility: We have no knowledge of the intrinsic properties of things.

This could be construed as the idea that we have no knowledge of any of the intrinsicproperties of things, and that (I think) is the idea we should ascribe to Kant. Admittedly, itsounds odd. If an intrinsic property is a property something has, independent of its relations toother things, then many of those seem perfectly accessible to us; for example, the sphericalityof the tennis ball. But Kant himself seemed to think we lack knowledge of any intrinsicproperties: we do have knowledge of certain physical properties of things, such as their shapeand their powers of attraction and impenetrability, but he thinks these are not intrinsic, as thefollowing passages illustrate.

The Intrinsic and Extrinsic. In an object of pure understanding the intrinsic is only that whichhas no relation whatsoever (so far as its existence is concerned) to anything different fromitself. It is quite otherwise with a substantia phaenomenon [phenomenal substance] in space;its intrinsic properties are nothing but relations, and it itself is entirely made up of mererelations. We are acquainted with substance in space only through forces which are active inthis and that space, either drawing other objects (attraction) or preventing their penetration(repulsion and impenetrability). We are not acquainted with any other properties constitutingthe concept of the substance which appears in space and which we call matter. As object ofpure understanding, on the other hand, every substance must have intrinsic properties andpowers which concern its inner reality. (A265/B321)

Substances in general must have some intrinsic nature, which is therefore free from allexternal relations. . . . But what is intrinsic to the state of a substance cannot consist in place,shape, contact, or motion (these determinations being all external relations). (A274/B330)

All that we know in matter is merely relations (what we call its intrinsic properties areintrinsic only in a comparative sense), but among these relations some are . . . enduring, andthrough these we are given a determinate object. . . . It is certainly startling to hear that athing is to be taken as consisting wholly of relations. Such a thing is, however, mereappearance. (A285/B341)

Kant thinks the physical world is made up of matter, “phenomenal substance,” but that mattersomehow “consists wholly of relations.” He is drawing on a dynamical account of matter(further developed in his works on physical theory) according to which matter is constituted byforces. He has a proto–field theory, which had an important historical role to play, influencingscientists who went on to develop field theory proper in the nineteenth century.

5 And familiarphysical properties—shape, impenetrability, attractive power—count, for him, as extrinsic orrelational properties.

Whether that is the right way to classify them depends on how we understand theintrinsic/relational distinction. We have said, loosely, that an intrinsic property is one thatdoesn’t depend on relations to anything else. Some philosophers have tried to make this moreprecise by saying a property is intrinsic just in case it is compatible with isolation; that is, itdoes not imply the existence of another wholly distinct object.

6 On this way of thinking, atennis ball’s sphericality will be intrinsic: a tennis ball can be spherical and be the only thing inthe universe. And the tennis ball’s bounciness will also be intrinsic. After all, a tennis ball can bebouncy and be the only thing in the universe, although, to be sure, it will not bounce unlessthere is something else for it to bounce off! On the face of things, shape properties likesphericality and dispositional properties like bounciness are intrinsic: something could havethem and exist all on its own.

If Kant nonetheless describes them as relational, perhaps he has a different conception ofintrinsicness in mind. Perhaps he thinks a thing’s shape properties are relational because theydepend on a relation to the parts of the thing. For example, the sphericality of the tennis balldepends on how the parts making up its surface are equidistant from its center. Perhaps hethinks dispositional properties are relational because they depend on how something wouldrelate to other things if they were there. For example, whether something is bouncy dependson what it would do in relation to something else—if, say, it were dropped on the ground orthwacked against a tennis racket.

Some metaphysicians like to ponder the distinction between intrinsic and relationalproperties, but we needn’t settle it here. All we need is that Kant denies us knowledge of anyintrinsic properties, in some defensible sense of “intrinsic,” and that this is what he means bydenying us knowledge of “things in themselves.”

Why does Kant deny us knowledge of intrinsic properties? The answer, I suggest, has twoparts. First, he thinks, as many philosophers do, that our knowledge is “receptive”: our mindsneed to be causally affected by something if we are to have knowledge of it.

The receptivity of our mind, its power of receiving representations in so far as it is affected inany way, is called “sensibility.” . . . Our nature is such that our intuition can never be other thansensible, that is, it contains only the way in which we are affected by objects. (A51/B75)

Our knowledge of things is receptive, “sensible”: we gain knowledge only through beingaffected by objects.

Receptivity: Human knowledge depends on sensibility, and sensibility is receptive: we canhave knowledge of an object only insofar as it affects us.

This simple fact about our knowledge, he seems to think, dooms us to ignorance of things inthemselves.

Properties that belong to things as they are in themselves can never be given to us throughthe senses. (A36/B52)

It is not that through sensibility we are acquainted in a merely confused way with thenature of things as they are in themselves; we are not acquainted with that nature in any wayat all. (A44/B62)

Why should the “receptivity” of knowledge imply ignorance of things in themselves? Manyphilosophers have wondered about this on Kant’s behalf, and some have criticized him roundlyon the topic. P. F. Strawson wrote:

Knowledge through perception of things . . . as they are in themselves is impossible. For theonly perceptions which could yield us any knowledge at all of such things must be theoutcome of our being affected by those things; and for this reason such knowledge can beknowledge only of those things as they appear . . . and not of those things as they . . . are inthemselves. The above is a fundamental and unargued complex premise of the Critique.

7

Is Kant really taking this for granted, as a “fundamental and unargued complex premise”?Perhaps not. Perhaps he has good reason to connect receptivity to ignorance—but a reasonthat has gone unnoticed.

Here is a simple suggestion. According to receptivity, we have knowledge only of whataffects us. But things “as they are in themselves” do not affect us: the intrinsic properties ofthings do not affect us. If Kant believed this, then that, together with his commitment toreceptivity about knowledge, would certainly explain why we are ignorant of things inthemselves.

A case can be made that this is just what Kant believes. I confess that, as a matter ofinterpretation, it is controversial. Here is not the place to do it justice. It involves detailedinvestigative work of a kind we historians of philosophy find strangely thrilling, although we’reaware not everyone shares our enthusiasm. But here at least are two small gestures in thisdirection. In an early philosophical treatise, Kant argues:

A substance never has the power through its own intrinsic properties to determine othersdifferent from itself, as has been proven.

8

According to Kant, the causal powers of a substance are something over and above its intrinsicproperties. At this stage of his thinking, he believes that they require an additional act ofcreation on God’s part—an act that is “obviously arbitrary on God’s part.” He took this ideaabout the insufficiency of intrinsic properties for causal power to imply not only thecontingency of causal power, but also the inertia of intrinsic properties. The idea returns in theCritique of Pure Reason:

When everything is merely intrinsic . . . the state of one substance cannot stand in any activeconnection whatsoever with the state of another. (A274/B330)

9

Receptivity requires that if we are to have knowledge of something, we have to be in activecausal connection with it: but we’re not in active causal connection with a thing’s intrinsicproperties. Receptivity means we can be acquainted with the causal powers of things, the waysthey relate to each other and to ourselves: but however deeply we explore this causal nexus, wecannot reach the things in themselves.

3. Why We Are Ignorant of Things in Themselves

Kant said we are ignorant of things in themselves: ignorant of the intrinsic properties of things.The picture I have painted on his behalf is, I hope, appealing, in certain respects: Kantianhumility does not, at least, condemn us to skepticism. But the picture will not appeal toeveryone. Kant’s conclusion seems too strong: we have no knowledge of any of the intrinsicproperties of things. His reasons invoke a seemingly idiosyncratic conception of intrinsicness: atennis ball’s sphericality and bounce are not among its intrinsic properties. And they invoke aseemingly implausible causal thesis: intrinsic properties are causally inert. So however thisinterpretation succeeds as a way to understand Kant, it is unlikely to succeed in reaching awider audience.

But wait. There is a conclusion very similar to Kant’s that is significantly closer to home.Kant says we are ignorant of the intrinsic properties of things, and he is right, though not forquite the reasons we have been looking at. And if he is right, of course, then you too shouldbelieve we are ignorant of things in themselves.

Imagine that a detective is investigating a murder case. She puts together the clues. Themurderer had a key, since no windows were broken. He was known to the dog, since there wasno barking. He had size 10 shoes—there are his footprints. More and more of the picture beginsto be filled out. He wore gloves, since there were no fingerprints. He had a tame parakeet—there are green feathers on the rug. The detective learns a lot about the murderer. Does sheknow who he is? She knows who the murderer is in relation to other things—houses, shoes,parakeets, and so on. She knows that the murderer is whoever fits this role. But does she, so tospeak, know the murderer in himself ? Is there something more to the murderer than being apossessor of keys, parakeets, and shoes? Of course. There must, in the end, be more tosomething or someone than their merely relational properties, as Kant pointed out. Andultimately, let us hope, the detective finds herself in a position to identify the person who existsindependently of these relations to other things: “Aha! There is one person who fits this role.There is one person who is known to the dog, wears size 10 shoes, has a tame parakeet, couldhave a key, and that person is . . . Pirate Pete!” Then she knows who the murderer is. Then sheknows who fits the role: she knows, so to speak, the murderer “in himself.”

Some philosophers have suggested we are in a situation rather like that of the detective. Animportant recent attempt to show this is that of David Lewis, whose “Ramseyan Humility”explicitly claims inspiration from “Kantian” humility.

10 We are trying to find out, not about amurderer, but about the fundamental features of the world. We put our best theorists on thejob, and they tell us a lot. Suppose we want to find out about our tennis ball. They tell us that atennis ball is whatever fits this sort of profile: it’s something that can be hit across a net with atennis racket, something that has a specific degree of resistance and elasticity, something thatis readily visible to human eyes in normal conditions, something that will roll smoothlydownward when placed on a slope. These relational descriptions, suitably filled out, give us astory about the role something must fit if it’s to be a tennis ball. They capture the relational or(if you like) “phenomenal” aspect of the tennis ball. Is there something more to the tennis ballthan this relational role? Of course there must be, just as there is something more to PiratePete than his relations to keys and parakeets.

And our theorists can tell us about this “something more.” They tell us not only that thetennis ball can land over the white line, not only that it is spherical and bouncy, but that it ismade of rubber and felt, which in turn are made of very tiny parts called molecules, which inturn are made up of tinier parts, called atoms, which in turn are made up of still tinier parts,called protons, neutrons, and electrons—and more, with names too peculiar to recount here.Our experts give detailed descriptions of the tiny parts something must have, and theirparticular arrangements, if that something is to fit the “tennis ball” role. They are giving us asplendid account of what the tennis ball is, “in itself.” What more could there be to know aboutwhat the tennis ball is, “in itself”?

We certainly know a lot about the tennis ball. In Kant’s terms, we know a lot about the tennisball as “phenomenon”—how it relates to tennis rackets, nets, and players. And yes: we alsoknow a lot about the tennis ball “as it is in itself,” what its parts are made of and how they arearranged. But now shift the question: What exactly do we know about those tiny parts of thetennis ball?

Take the electron, for example. Our story about the electron has something in common withthe detective’s story about the murderer. It captures a complicated relational profile. Anelectron is whatever it is that fits a distinctive role, whatever it is that fits the “electron” patternof relating to other things. An electron is the thing that repels other things we call “electrons”and attracts other things we call “protons.” It’s the thing that, in company with lots of otherelectrons, makes the lightbulb go on, makes your hair stand on end on a cold, dry day, and soon. The physicist will have a more detailed story, but it will nevertheless be a story that has thisrelational form. “Electron” refers to whatever fits the physicist’s relational “electron” role, justas “the murderer” refers to whoever fits the detective’s relational “murderer” role.

The detective discovers who the murderer is, “in himself” as we put it, when she discoverswho fits the relational “murderer” role; namely, Pirate Pete. The physicist discovers what theelectron is when she discovers what fits the relational “electron” role; namely, . . . what? Herethe analogy with the detective breaks down. The detective is able to find out who the murdereris, apart from the story about how the murderer relates to keys, shoes, parakeets. But thephysicist is unable to find out what the electron is, apart from the story about how the electronrelates to protons, hair, and lightbulbs. For the detective, there is something more to say. Forthe physicist, there is nothing more to say. The electron is, to borrow a phrase from Kant,merely a “something = x” about which we can say nothing, or rather nothing more than what’sgiven in our relational description. The upshot: we know the electron as “phenomenon,” so tospeak, but we don’t know the electron “as it is in itself.”

Here is another way to bring out the point. Suppose, inconveniently, more than one personfits the role given by the detective’s list of clues: suppose Pirate Pete, Pirate Percy, and PiratePeggy all have keys, parakeets, large shoes, and so on. Then although the detective knows a lot,she still doesn’t know who the murderer is. That, or something like it, is the situation we facewith the electron. Consider the thing, whatever it is, that fits the electron role. We aresupposing its intrinsic properties are not inert (we are leaving that part of Kant behind) but arethe causal grounds of its power to repel other electrons, attract protons, and so on. We are,indirectly, in causal contact with those intrinsic properties—but, receptivity notwithstanding,that is still not enough for us to know what those intrinsic properties are. Why not? Supposewe give a name to the intrinsic property responsible for this complex causal profile: let’s call it“negative charge,” or “NC” for short. Now let’s draw on Kant’s insight about the contingency ofcausal power, which is shared, in some form, by many philosophers today (including Lewis).This contingency means that NC could have been associated with a completely differentrelational, causal profile, and a different intrinsic property—call it NC*—could have beenassociated with the electron’s relational, causal profile. But now ask: Is the electron’s intrinsicproperty NC or NC*? We don’t know, any more than the detective knows whether themurderer is Pirate Pete or Pirate Percy. So we are faced with humility again.

Humility: We have no knowledge of the most fundamental intrinsic properties of things.

This is admittedly a modified version of humility: the conclusion is less drastic than Kant’s. Weare not denied knowledge of all intrinsic properties: the tennis ball is spherical and is made ofrubber and felt, which in turn are constituted by molecules, elements, and subatomic particles.We know a lot about the intrinsic nature of the tennis ball.

But we do lack knowledge of things in themselves. Kant was right to say it, and we need toaccept it. It’s not so bad. It’s not skepticism. It is what it is. We face the sad fact that we knowless than we thought: there are some intrinsic properties of which we shall forever be ignorant.And, sadly, they are the most fundamental properties of all.