|
Editorial Review Product Description
Ten papers by prominent philosophers of science which challenge van Fraassen's thesis from a variety of realist perspectives. Together with van Fraassen's extensive reply, the articles provide a comprehensive picture of the current debate in philosophy of science between realists and anti-realists. ... Read more Customer Reviews (2) 
Convincing
When I read van Fraassen's "Image of science" (1980), it made perfect sense.The image van Fraassen was presenting was making perfect sense, and it made it much easier to understand how I could improve the research design for my PhD in Information Systems.However, in order to feel fully convinced, I also wanted to hear what the critics of van Fraassen were saying.As I read this book, I saw no arguments that made van Fraassen's position less convincing.In fact, some of the realists sound incredibly stupid, as if they don't understand science at all.Very useful book.Very convincing.
Constructive empiricism
----Constructive empiricism is a third option between positivism and realism
Van Fraassen's antirealism, constructive empiricism, is based on the claim that the goal of science is not truth but empirical adequacy. A theory is empirically adequate "exactly if what it says about the observable things and events in this world, is true---exactly if it 'saves the phenomena.'" Constructive empiricism is realism minus truth, so to speak, in the sense that van Fraassen accepts a number of claims with which the realist used to bash antirealists of a positivist type, e.g., that a theory cannot be identified with its empirical consequences; that existence and truth is independent of observation; and that it is impossible (because of the inevitability of theory-ladenness) to build up science from neutral observation statements. All these things can be admitted, van Fraassen says, without taking the leap to truth. (Van Fraassen rightly condemns this characterisation of his views as pseudohistory.)
I summarise some arguments in the form of a dialogue between a constructive empiricist (vF) and a realist (R).
----Constructive empiricism is realism freed of metaphysics
vF: in rejecting positivism, we both agree that a claim C can have a truth value without having any empirical consequences (e.g., "The Absolute is lazy"). Now suppose we conjoin any empirically adequate theory with C. I can say "C is pointless because it does not improve empirical adequacy." But you cannot say the same with respect to truth unless you profess to be able to assess truth values in some mysterious metaphysical fashion, e.g. by claiming that there is a link between truth and simplicity. R: the link between truth and simplicity need not be purely metaphysical. A principle of simplicity can be corroborated at the metalevel through its successful applications.
vF: if I believe in a theory's empirical adequacy and you in its truth then I can be right and you wrong but not vice versa. You take a greater risk. But empirically we take the exact same risk. This shows that you allow the intrusion of metaphysics into science: to you theory choice is not determined by evidence alone. R: for you to be right and I wrong, there must have been multiple empirically equivalent and empirically adequate theories (among which I picked the wrong one). This is a mere logical possibility; it never happens in practice. vF: an example of a spectrum of equivalent theories is the Newtonian dynamics of the solar system conjoined with the assumption that the centre of mass of the system is moving with a particular velocity. R: this equivalence was removed with the introduction of electromagnetism (which would have enabled measurement of absolute velocity, had there been such a thing). Thus the alleged equivalence was removed by extension of the theory. It is surely not unreasonable to expect this to be the case in general.
----Theory/observation dichotomy problematic
R: your definition of empirical adequacy is based on a dubious distinction of the observable from the unobservable. How does one draw this line? vF: Observable means observable with unaided senses under some conditions (e.g. observing the moon of Jupiter from up close). Of course this is vague but since it has clear instances and noninstances it is still meaningful. R: this is a relative predicate; if we had, e.g.,electron microscope eyes then more things would be observable. vF: yes, our epistemological situation depends on our specific human sensory capacities. R: why should so much philosophical import be given to accidental properties of human physiology? vF: is it so strange that human knowledge is limited by human capacities?To point out that our knowledge may be different if we had different senses is merely to observe the truism that our knowledge may be different if different evidence was available to us. R: You make your view sound reasonable only by confounding observation and observability. For example, you allow me to infer the existence of a mouse from scratching noises, disappearance of cheese, etc., but not the existence of electrons from just as strong evidence. Thus the distinction you draw is not one between observed and unobserved by me or by humans in general; rather it is between observable and unobservable in principle. It seems that the former rather than the latter in the natural distinction to draw with regard to epistemological import.
R: In fact, the above example reveals an even more fundamental weakness of your theory. If knowledge depends on observation, how do you know that electrons are unobservable? Since no one can observe them to be unobservable you must, according to your own theory, remain agnostic on the matter. Thus your distinction between observed and observable in principle collapses.
R: Why should observation be the foundation of epistemology while theory is treated with the utmost scepticism? After all, people have argued for the reverse at least since Plato.
----Explaining the success of science problematic
R: realism explains the success of science. vF: I can explain it too: "any scientific theory is born into a life of fierce competition, a jungle red in tooth and claw. Only the successful theories survive." R: this is not sufficient for it explains only why we have ended up with good theories, not what it is about these particular theories that make them successful. vF: How does one explain the white fur of the polar bear? By pointing to its adequacy with respect to the (empirical) circumstances facing the bear; not by appeal to some extraempirical goodness of white fur.
----Conjunction of theories problematic
R: Scientists often conjoin theories and derive things from the conjunction. This makes perfect sense from a realist point of view: two truths conjoined make another truth. But the conjunction of two empirically adequate theories need not itself be empirically adequate or even consistent (e.g., T&C and T&¬C with the C above). vF: conjunction is a legitimate heuristic tool for generating new and potentially empirically adequate theories. R: fine, but that explanation is available equally to us both; so my position is superior because I have an extra reason to do conjunction. vF: Anyway, conjunction rarely works straightforwardly; rather it involves simultaneous corrections, so your argument does not apply. R: This does not undermine my point but rather reinforces it: it is precisely because a conjunction of truths cannot be false that the falsity of the conjunction prompts correction of the original theories.
----Whether explanation provides the link to truth
R: The goal of science is not only to describe and predict phenomena; it is to explain them. Your emphasis on empiricism makes you miss this further dimension. And here of course truth is crucial since an adequate explanans must be true. vF: You beg the question by assuming that truth in necessary for explanation. In fact, this is contrary to the common-sense meaning. For example, I can say without contradiction both that Newton explained the tides and that his theory was false. R: Sure, we can speak of false explanations, just as we can speak of a law being false, even though this of course implies that it was not a law after all. But an explanation is admitted as adequate only if it is believed to be true. vF: Or believed to be empirically adequate. What extra work is being done by adding truth? R: Look at it the other way around: explanatory power vouches for truth. Other things equal, a theory with greater explanatory power is more likely to be true than a theory that merely accommodates the data. vF: The concept of explanation cannot carry the epistemological burden with which you shoulder it. Explanation is relative: it is an answer to a why question (posed for subjective reasons) that is relative to background knowledge, an implied contrast-class ("why P rather than Q, R, etc.?"), and so on. As such, the explanatory power of a theory is a purely pragmatic virtue, analogous its enabling control over nature for certain practical purposes. It has no special epistemological import.
... Read more
|
