While rereading Berkeley's A Treatise Concerning the Principles of Human Knowledge for some work on nominalism I am doing I stumbled upon an interesting moment. In section 58 Berkeley considers an objection to his proposal that things exist as ideas without existing as material objects potentially independent of any mind whatsoever. Specifically, the objector states that science has shown that the earth moves, but as no one can directly observe the movement of the earth science has asserted, and seemingly proven, a claim about something which no mind experiences. Berkeley responds by stating that what the well supported scientific theory should be understood to be about is what would be observed were one to place oneself in a given position in the future in relation to the earth. In other words, the content of the scientific theory is a prediction about future observations and applications based on past observations. Here is the text:
"58. Tenthly, it will be objected that the notions we advance are inconsistent with several sound truths in philosophy and mathematics. For example, the motion of the earth is now universally admitted by astronomers as a truth grounded on the clearest and most convincing reasons. But, on the foregoing principles, there can be no such thing. For, motion being only an idea, it follows that if it be not perceived it exists not; but the motion of the earth is not perceived by sense. I answer, that tenet, if rightly understood, will be found to agree with the principles we have premised; for, the question whether the earth moves or no amounts in reality to no more than this, to wit, whether we have reason to conclude, from what has been observed by astronomers, that if we were placed in such and such circumstances, and such or such a position and distance both from the earth and sun, we should perceive the former to move among the choir of the planets, and appearing in all respects like one of them; and this, by the established rules of nature which we have no reason to mistrust, is reasonably collected from the phenomena."
On reading this I was struck by a surface resemblance to the problem of the arch-fossil in Meillassoux's After Finitude. There Meillassoux claims that some science presents us with "ancestral statements" and the discovery of "arch-fossils". The first of these is understood as a statement about how things were before the development of thought in the universe. Imagine, for example, a hypothesis about the first few minutes following the Big Bang. The second of these are evidences which support ancestral statements. Consider, for example, light reaching the earth from a portion of space over thirteen billion light years away. Such light would be just now reaching us from a time comparably close to the time of the Big Bang and thus almost certainly from a time before the development of any consciousness or thought. Ancestral statements, then, speak of a universe without, and before, consciousness or thought and arch-fossils are evidences of such times. This point is central for Meillassoux for several reasons, not the least of which is that it allows him to claim that the completely honest correlationist must admit to rejecting science. (By a "correlationist", as I have discussed in other posts, Meillassoux means someone who claims that being and thought or being and consciousness are necessarily connect such that it ultimately makes no sense to speak of the one without the other. To be is to be given to consciousness.)
Of course Meillassoux stresses in his book that the problem of ancestral statements is not the same as the problem of the unobservered, i.e. the old problem of whether an event occurs if there is no one there to witness it. He wants to stress that ancestral statements and arch-fossils are not about things which just happen not to have been witness but are rather about things which based upon their very temporal nature could not have been witnessed. Their temporal being necessitates unobservability by consciousness. This being the case, Berkeley's example is not precisely the same. However, his answer is certainly the answer that many a correlationist is likely to jump to and one which Meillassoux, who so often takes Berkeley as a model for the most extreme of his enemies, is familiar with. What is odd, however, is that Berkeley's answer presents a plausible and coherent understanding of what the scientist is actually up to. Even today any number of scientists seem unlikely to object to his description. Meillassoux insists, however, that scientists mean their ancestral statements literally and it is only the sophistic correlationist philosopher who feels the need to hedge the meaning of their statements. Meillassoux states:
"Thus, a philosopher will generally begin with an assurance to the effect that his theories in no way interfere with the work of the scientist, and that the manner in which the latter understands her own research is perfectly legitimate. But he will immediately add (or say to himself): legitimate, as far as it goes. What he means is that although it is normal, and even natural, for the scientist to adopt a spontaneously realist attitude, which she shares with the ‘ordinary man’, the philosopher possesses a specific type of knowledge which imposes a correction upon science’s ancestral statements– a correction which seems to be minimal, but which suffices to introduce us to another dimension of thought in its relation to being."
The question, then, is what a scientific hypothesis actually means or, to put it differently, whether science is even able to formulate ancestral statements when we avoid short cuts or common language translations in our formulations of what a hypothesis or theory states?
Let me first state that it seems that we can find non-temporal arch-fossils and ancestral statements that allow us to clarify the problem by glancing at some of the major developments in twentieth century physics. Subatomic particles are, in their very being, un-able to be directly observed by anything like consciousness. They are small enough objects or events that it seems fairly safe to assert that their very being forecloses all unmediated observation. Statements about subatomic particles are, then, ancestral statements about a region of being, admittedly spacial and not temporal, which by its very nature is independent of any correlation with consciousness or thought. How, then, do scientists understand their hypothesis about subatomic particles? Do they always and simply intend them "literally" and in a naive realist sense? This seems pretty clearly not to be the case. Models of, for example, the structure of the atom are taken to be precisely that, models, partaking more of the nature of metaphor than of factual claims. The models attempt to capture in intuitive forms, as far as possible, the predictive and observational consequences of various theories while suffering from the various drawbacks associated with metaphors, i.e. they work in certain regards but not in others. Look at the debates between Bohr, Heisenberg and Schrodinger concerning wave/particle duality or the uncertainty principle and you will see clashes over how much of the metaphors are useful and how much not. For example, does the uncertainty principle say something literal about subatomic particles, i.e. that they REALLY EXIST as something like waves of probability until they are observed and collapse into a given determined form, or does it say something about the limits of our observational capabilities and the inadequacy of our concepts (metaphors?) "wave" and "particle"? These are old problems and the general way out of them is to state that the meaning of the theories in question is their observational and predictive consequences and not the common-language stories about what the subatomic world is like which we attempt to spin out of them. One focuses on the differences which make a difference in scientific practice.
When I was an undergraduate student at Boston University I studied some physics with Dr. Lawrence Sulak, the discoverer of the finite mass of the neutrino. I was reading a book on super-string theory and asked professor Sulak what he thought about the theory. His response was that it was a beautiful fairy tale. His reasoning was that, at the time, super string theory had no observational or predictive consequences and so didn't exist as an actual scientific theory at all. Clearly here there are also echoes of Popper's proposal that a theory which can never be falsified, because it has no predictive consequences, is not science. One could envision using Popper's guidance to distinguish between the scientific and non-scientific content of a hypothesis which would be a practice of distinguishing between the observational-predictive content of a hypothesis and whatever else they may contain, such as the metaphorical content used to make a hypothesis more imaginable for both the theoretician and common public.
In theoretical physics it is not uncommon to find a very sharp and powerful distinction drawn between the real hypotheses, which tend to consist of mathematical formulations and observational predictions, and the every day common language translations of these hypotheses. To put this in more pragmatic terms, and to connect it with Berkeley's original claim, the hypothesis is about what one can observe now or in the future and what one can do with the hypothesis. This is, arguably, the scientific content of the hypothesis from the standpoint of at least some, and I suspect many, physicists. This point has become more and more important in physics because the hypotheses with which the physicist works have been getting harder and harder to capture in everyday descriptive language.
Astrophysics, because it works at a macro rather than micro level, tends to be more friendly to common language formulations and descriptions. Nonetheless it often enough finds itself forced to admit that the actual content of its theories are mathematical formulations conjoined with observational predictive content and not common language descriptions. Cases of talking about "the warpage of space-time by mass" are a fair example here. At best what we are dealing with in this case are mostly failed metaphors, failed because impossible to imagine even though formulated in everyday descriptive language, for a mathematical formulation of observational predictions. This, then, will be equally true of hypotheses which are taken to be about the universe before the development of consciousness. The scientific content of such hypothesis is conceivably held by some scientists to strictly be their observational predictions and applications, not their everyday descriptive formulation into a story about the early moments of the universe. For an ancestral statement, then, it is precisely not the ancestral part of the statement which seems to be its scientific content. For this reason we might even say that science conceived in this manner is unable to formulate ancestral statements.
I tend to side with Feyerabend on these issues to the effect that there is little to nothing that can be said about Science in general. However it seems to me that we have good reason to think that Berkeley's understanding of scientific practice may be less violent to it than Meillassoux's and, in fact, that Meillassoux may himself be caught in the position of having to admit that he is playing the role of the sneaky philosopher making scientific statements say something other than they actually tend to say.