John Bell on 'Subject and Object': an Exchange (original) (raw)

2021, arXiv (Cornell University)

This three-part paper comprises: (i) a critique by Halvorson of Bell's (1973) paper 'Subject and Object'; (ii) a comment by Butterfield; (iii) a reply by Halvorson. An Appendix gives the passage from Bell that is the focus of Halvorson's critique. Bell then says that (3) is a serious defect that makes quantum mechanics "vague" and "intrinsically ambiguous" and "only approximately self-consistent." (We have included the complete text of the relevant passage in an appendix.) Let me begin by saying that I simply deny (1), i.e. that quantum mechanics is fundamentally about the results of measurements. I'm afraid that Bell has himself made a logical leap from "the quantum mechanical formalism needs a user" to "quantum mechanics is fundamentally about the results of measurements." There is a wide range of possibilities between these two extremes-e.g. that the quantum-mechanical formalism provides a means for translating facts about subatomic reality into a language that human beings can understand. I will grant that Bell is correct about (2), that the subject-object distinction is needed for quantum mechanics, but unfortunately, Bell has misunderstood the sense in which it is needed. He seems to think that quantum mechanics must describe the world as bifurcated into two parts-subject and object. If that were correct, then I would completely understand Bell's unease with the distinction. If the theory describes a world with two parts, then the theory should offer some guidance about what belongs to each part. But if you think about the meaning the word "subject", it quickly becomes obvious that it's not supposed to play the role of a predicate in the theory (unlike, say, "electron"). Rather, the idea is that a subject uses the theory to describe objects-and in the case at hand, these objects fall under the laws of quantum mechanics. The theory sees no subjects, it sees only objects, and so it has no need for specifying where and when the subject-object split occurs. Such a split is a necessary prerequisite to physical theorizing, when a subject decides to use a theory to try to say something true about the world. Now what about the complaint that quantum mechanics does not specify who the subject is, or when and where and how she decides to use the theory? But wait a minute. Is there any theory that does that? What an amazing theory it would be! Indeed, such a theory would fulfill Hegel's aspiration of finally unifying the subject and object. In other words, such a theory would "theorize itself." Is Bell suggesting that quantum mechanics is defective because it doesn't yet achieve the Hegelian Aufhebung of the subject-object distinction? So, in short, Bell is correct that quantum mechanics, as it stands, needs a subject. But that is true of every theory that has ever appeared in physics-i.e. these theories need subjects to decide when and where and how to describe things. Bell's subsequent rhetoric in the article is effective only against the backdrop of his false assumption that the subject must appear in the quantum-mechanical description. For example, Bell raises a question for which quantum mechanics doesn't appear to have an answer. Now must this subject include a person? Or was there already some such subjectobject distinction before the appearance of life in the universe? (p 40) But quantum mechanics is simply not interested in the question of what counts as a subject. If you ask me what counts as a subject, then my answer is that anyone who can use a theory to describe things is a subject-no other qualifications are necessary! If your dog can theorize, then he is a subject, and if an artificial intelligence could theorize, then it would also be a subject. And to Bell's second question, I suspect that before the appearance of "life" in the universe, there were no things that could describe other things, and hence no