Van Fraassen on microscopy

Philosopher of science Bas van Fraassen is well known for defending the “constructive empiricist” position that the success of a scientific theory does not compel us to believe in the existence of the theoretical entities it posits.  In his view, a scientific theory aims only to be empirically adequate, in the sense of correctly describing the world of observable things.  When a theory is successful insofar as it makes accurate predictions or opens the door to technological advances, this gives us reason to believe only that it is empirically adequate, not that it tells us anything about a realm beyond what is observable.

Scientific realists take the contrary view that the success of a theory does give us reason to believe in the theoretical entities it posits.  Among other arguments, they sometimes appeal to the idea that entities that at one time were unobservable later became observable with the rise of new technologies, such as telescopes, electron microscopes, and ordinary microscopes.  This shows, they argue, that the boundary between observable and unobservable entities is not sharp enough to justify skepticism about the latter.

In chapter 4 of his book Scientific Representation, van Fraassen presents a very clever and interesting challenge to this sort of argument.  Taking on scientific realists on what might seem to be their strongest ground, he denies that even ordinary microscopes must be seen as confirming the reality of previously unobserved entities.

First a terminological point.  Consider the world of everyday experience – of tables and chairs, rocks and trees, reflections in water and shadows on the ground and on walls, rainbows and clouds and stars in the sky, and so on.  These are all examples of “phenomena” as van Fraassen uses the term.  In holding that the aim of science is empirical adequacy, views like his are often characterized as holding that science is concerned only to “save the phenomena.”  Scientific realism, by contrast, holds that science gives us reason to believe in an invisible world behind the phenomena.

The claim on the table, then, is that a microscope amounts to “a window on the invisible world,” as van Fraassen puts it (p. 93).  It allows us to peer behind the phenomena to what previously had been unobservable.  Van Fraassen allows that this is one possible interpretation.  But he argues that it is not the only one, so that instruments like microscopes cannot be appealed to as somehow settling the dispute between constructive empiricism and scientific realism.  Considered by themselves they are neutral, insofar as what they yield can be interpreted in ways consistent with either position.

Scientific instruments, notes van Fraassen, play several roles.  Sometimes that role is representative, as in the case of clocks, balances, and measuring rods.  Sometimes it is imitative in the sense of mimicking effects that occur in nature, as when an electrical generator is used to bring about an imitation of lightning.  Sometimes the role is instead productive, in the sense of bringing about phenomena which had not previously been observed, as when Faraday discovered electromagnetic induction.

Now, a microscope is a scientific instrument.  What role does it play?  The scientific realist, in conceiving of it as a kind of “window” on otherwise unobservable reality, thinks of it as playing an essentially representative role.  But whether or not it plays that role, van Fraassen says, it certainly plays the productive role of creating new phenomena, phenomena that had not been seen before.  And he wants to argue that there is nothing in what they show us that compels us to see them as doing anything more than that.

Here he appeals to the idea of a “public hallucination” (p. 101).  Hallucinations like the one Macbeth experienced when he thought he saw a dagger are private, directly knowable only to the individual experiencing them.  Contrast that with a rainbow, which, like Macbeth’s dagger, is not something really out there in the world, but is nevertheless experienced by many individuals at once. 

But not all public hallucinations are quite as distant from really existing things as that.  Van Fraassen draws a distinction between four types of “images” we distinguish from reality, but which are not all equally distant from it.  At one extreme we have private images such as after-images, dreams, and hallucinations, which are not only not actual things in the objective world, but are accessible only to the individual person having the experience.  At the other extreme we have graven images such as paintings, photographs, and sculptures.  These are not really things of the same sort as what they resemble, but they nevertheless are things in their own right really existing in the world outside the mind. 

In between are public hallucinations, but these can be divided into two sub-classes.  There are public hallucinations that are, as van Fraassen puts it, “not ‘copy’-qualified” (p. 104).  What he means is that they are not images of some thing that is really out there in the world.  Rainbows and mirages would be examples.  They are a step up from private images because they are shared rather than subjective.  But there are also public hallucinations that are “copy-qualified,” in the sense that they are images “of” something really there in the world.  Examples would be shadows and reflections, such as the reflection of a tree in the water.  These are a step down from graven images, because a reflection is, like a rainbow, not an actual thing out there in the world.  But it is of a thing (for example, the tree) in a way a rainbow is not.  Hence it is a step up from a public hallucination of the kind that is not copy-qualified.

It is in this last category that van Fraassen wants to put the images seen in a microscope.  What a microscope gives is, he argues, is a public hallucination of the copy-qualified kind, like a tree’s shadow or reflection in the water.  By no means does he want to deny that we do, of course, get much useful information from microscopes and other scientific instruments.  But this, he argues, is due to “our latching onto significant regularities in the phenomena” rather having to do with unobserved realities (p. 109).

As always with van Fraassen, there are further intriguing nuances, examples, and insights, and I can’t convey them all here in this summary.  But I think I have captured the gist of his argument.  It seems to me that the scientific realist can make the following points in response.

First, while there are similarities between the case of a tree’s reflection in water and the case of an image seen in a microscope, there are also important differences.  As I was reading van Fraassen’s book, I sipped a martini.  Through the glass I could see the books on the shelf behind it – only very obscurely through the part of the glass that still had liquid in it, a little bit more clearly through the part that did not.  But though the books were obscured, it seems correct to say that I was looking at the books themselves – even though I was doing so through the medium of the glass – rather than that I was looking at an image in the glass.  I was also wearing eyeglasses, and when I peered past the martini glass to the books it seems even more obviously correct to say that what I was looking at were the books, even if I looked at them through the eyeglasses (rather than saying that I was looking at an image in the eyeglasses).

These cases are obviously different from the case of looking at a reflection of a tree in the water.  In that case, what I am looking at is not the tree itself, but something caused by the tree.  As van Fraassen emphasizes, the reflection of the tree is not a thing there in the water in the sense in which a floating log is a thing there in the water.  The reflection will, for example, seem to move as you keep looking at it while you move along the shore, whereas a floating log might not do so.  That’s why it is plausible to classify the reflection as a kind of “public hallucination.”  By contrast, the books I see through the martini glass and through the eyeglasses are things there in the world.  And even given that the martini glass distorts the appearance of the books, the result is not a hallucination of any kind, precisely because I really am nevertheless looking at the books.

Now, the case of the microscope seems obviously closer to the cases of the martini glass and the eyeglasses than it does to the case of the reflection of the tree.  What you see when using a microscope to examine a cell, for example, is the cell itself – even if you’re looking at it through the microscope – rather than merely an image.  If that is the case, though, then it also seems plausible after all to conceive of a microscopes as a “window on the invisible world,” rather than merely as something that creates new phenomena.

A second point in reply to van Fraassen would be this.  Suppose the argument I just gave is wrong, and that van Fraassen is right to characterize what a microscope produces as a “copy-qualified public hallucination.”  It is not clear that that would actually support constructive empiricism over scientific realism.  After all, what makes an image “copy-qualified” is that there is a real thing that the image is an image of.  For example, a tree’s reflection in the water is copy-qualified because there really is a tree that is causing it.  The fact that the reflection is a “public hallucination” therefore would hardly support anti-realism about trees!

But then, if an image in a microscope is also a “copy-qualified” image, then that would mean that there must a real thing that the image is an image of.  And that would arguably be sufficient grist for the scientific realist mill, even if we don’t think of microscopes as windows on the invisible world.

Related posts:

Hanson on observation

Cartwright on theory and experiment in science

Cartwright on reductionism in science

Dupré on the ideologizing of science

The particle collection that fancied itself a physicist

Fallacies physicists fall for