This is not a frivolous example. For thousands of years before Australia was discovered, the only swans people in the West had seen had been white. Consequently, the entire Western world took it for granted that all swans were white, and expressions like “white as a swan” abound in Western literature. It was only when Europeans landed in Australia that they saw, for the first time, a black swan. What was previously considered a scientifically inviolable truth had to be retired.
At this point one might expect today’s champions of science to start patting themselves on the back and saying, “Yes, and this is the wonderful thing about science. It is always open to correction and revision. It learns from its mistakes.” Sure enough, Carl Sagan praises scientists like himself for their “tradition of mutually checking out each other’s contentions.” Sagan’s view is echoed by Daniel Dennett, who writes, “The methods of science aren’t foolproof, but they are indefinitely perfectible…. There is a tradition of criticism that enforces improvement whenever and wherever flaws are discovered.”8
To say this is to miss Hume’s point, which is that science was not justified in positing these rules in the first place. All scientific laws are empirically unverifiable. How do we know that light travels at the speed of 186,000 miles per second? We measure it. But just because we measure it at that speed one time, or ten times, or a billion times, doesn’t mean that light always and everywhere travels at that speed. We are simply assumingthis, but we don’t know it to be so. Tomorrow we might find a situation in which light travels at a different speed, and then we will be reminded of black swans.
But can’t scientific laws be derived from the logical connection between cause and effect? No, Hume argued, because there is no logical connection between cause and effect. We may see event A and then event B, and we may assume that event A caused event B, but we cannot know this for sure. All we have observed is a correlation, and no number of observed correlations can add up to a necessary connection.
Consider a simple illustration. A child drops a ball on the ground for the first time. To his surprise, it bounces. Then the child’s uncle, a graduate of the Massachusetts Institute of Technology, explains to the child that dropping a round object like a ball causes it to bounce. He might explain this by employing general terms like property and causation. If these are not meaningless terms, they must refer to something in experience. But now let us consider a deep question that Hume raises: what experience has the uncle had that the child has not had? The difference, Hume notes, is that the uncle has seen a lot of balls bounce. Every time he has dropped a ball it has bounced. And every time he has seen someone else do it, the result was the same. This is the basis—and the sole basis–of the uncle’s superior knowledge.
Hume now draws his arresting conclusion: the uncle has no experience fundamentally different from the child’s. He has merely repeated the experiment more times. So it is custom or habit that makes him think, “Because I have seen this happen many times before, therefore it must happen again:’ But the uncle has not estab- lished a necessary connection, merely an expectation derived from past experience. How does he know that past experience will repeat itself every time in the future? In truth, he does not know. In this way Hume concluded that the laws of cause and effect cannot be validated. Hume is not denying that nature has laws, but he is denying that we know what those laws are. When we posit laws, Hume suggests this is simply a grandiose way of saying, “Here is our best guess based on previous tries.”
