Thursday, January 17, 2008

"Once we have the evidence, we will pull the trigger"

The quote in the title is from 2004, and concerns a city's plan to shut down an adult business, once evidence of illegality is gathered. This sounds funny, but to be charitable, the quote is only funny if "evidence" means epistemic evidence, but in context it probably means "legally admissible evidence"--the speaker is a lawyer after all.

At times, people have a near-certainty that future scientific findings will vindicate a view that they hold. One meets this in contexts related to naturalistic assumptions. Thus, some have a near-certainty that future neurophysiology will find evidence that there is no physically inexplicable influence on the brain of a sort that could be reasonably attributed to a soul. And some have a near-certainty that future biological research will fill in present-day explanatory gaps without having to posit any radically different kinds of processes from the ones the current neo-Darwinian synthesis posits. An example from the past would be 18th century folks who might think that all the fundamental things we are ever going to discover in the future course of physics is more forces, and that this will fill in the currently unexplained phenomena (cf. this post of mine).

I want to suggest that to justify a view by reference to the expected future findings of science is epistemically unacceptable. For there are, I think, two sources of evidence for the expectation. The first is independent evidence for the view in question. Certainly, there can be times where one has evidence that a view is true, and evidence that the view is likely to fall within the competence of a future science, and then one can conclude that future science will likely vindicate the view. However, in that case the prediction about future science is a fifth wheel and does no justificatory work. Here the invocation of future science is no more than a rhetorical device, an attempt to apply the prestige of science to one's view.

But there may be times when one actually has reason to think that science will conclude to a particular view, and one's reason to think this is independent of evidence for that view. For instance, one might have an inductive argument from "the direction a science is heading" as to where it will end up. There are two ways of understanding this. One way is that we are simply extrapolating from past data. In that case we are not actually basing anything on future science. We are doing science now by taking past data and extrapolating from it--this is standard scientific practice. Again, future science is a fifth wheel epistemically. And if this is what we are doing, we have to be careful to adhere to all of the standards of the science in question--basically, our extrapolatory work has to be of a sort to count as good statistical work in the relevant field in order to invoke the authority of science for the findings.

There is, however, one remaining way to understand "the direction a science is heading." The sciences are human enterprises, and as such are appropriately studied by sociology and anthropology. One can thus make sociological and/or anthropological predictions about where a science is going. But these predictions, when made sociologically or anthropologically, unless they embody the sort of extrapolation that I discuss in the previous paragraph, do not justify one's believing that the predicted conclusions of the science are true. Thus, one might have anthropological reasons for thinking that future neurophysiologists are going to be by and large naturalists and are not going to accept any explanations involving non-physical processes. But this anthropological prediction yields no additional evidence for the claim that non-physical processes have no impact on neurophysiological functioning. One can equally well make the sociological prediction that future scientists will by and large focus on theories for defending which one can get good grant money, but this sociological prediction does not justify belief that theories for defending which one can get good grant money are more likely to be true. Basically, the sociological/anthropological prediction is based on our knowledge of the biases of scientists in the field.

In summary, invoking future science as evidence for one's view is epistemically irresponsible. One can make predictions about the future progress of science in three ways:

  1. Having independent evidence of a view and evidence that it falls within the purview of science
  2. Doing good science now in extrapolating past data
  3. Doing good sociology or anthropology of science
But on none of these approaches does the prediction about future science carry independent epistemic weight. Typically, I submit, it is just rhetoric. I also suspect that often the source of a near-certainty about where science is going is (3), which is the most epistemically objectionable as evidence for the truth of a view.

3 comments:

Drew Tatusko said...

Just picking one of the many arguments that follows in this way, one argument from Sam Harris' argument that goes something like...

Science has effectively disproven numerous religious ideas regarding events such as cosmogenesis.

Religion has yet to disprove anything in science.

Therefore we can predict that science will continue to disprove religious claims until there is no need for religion to "fill the gaps".

But based on the second premise and the conclusion drawn in the third, this seems to be a red herring.

Hans Georg Lundahl said...

"Science has effectively disproven numerous religious ideas regarding events such as cosmogenesis."

Effectively disproven ... qed?

Hans Georg Lundahl said...

effectively discouraged would be more on the mark

evolution of living species as well as of stars are effectively inculcated in an age when:

a) most of the boys really ready to question something on purely theoretic grounds are unattractive geeks

and:

b) most of the geeks desire approval at least of their teachers

that is effectively discouraging creationism and young universe, it is not effectively disproving them

and this is good socialogy of the sciences