|
If you have 5 choices, and can rule out 4 of them, that's an advance.
Peirce pointed out that there's no end to abduction--the creation of hypotheses that available data are consistent with. However, abduction isn't a valid logical process--it proves nothing, since none of the abductions is more valid than any other, and, in fact, they may all be wrong. But abduction is a good place to start (and while it's a bad place to finish, it's where many stop thinking).
Then science turns to testing the hypotheses--making sure that they actually do handle all the data, making sure the data is what it appears to be, and finding new data.
When it comes to testing hypotheses, however, there's a practical, pragmatic constraint: Finite resources, whether that be material, time, or researchers. So researchers don't test all the hypotheses, just the ones that they (and funders, as it turns out) think are the most promising. However they define that.
It's also good form to not only show that your hypothesis accounts for all the data, including predicting new data (which is subsequently found to exist, as predicted), but also that the competing hypotheses make predictions that are false according to the new data, or according to new data found specifically to falsify the predictions you think are wrong. Doing the latter and not the former is progress, but of a negative kind (it can end up with 'all the theories are wrong', which is an important, but not always award-winning, advance).
This shows one competing theory survives the latest round of data, and is likely to be correct--the crests were easily involved in sound production, and the critters easily had the cognitive "chops" to process the sound. The theory made a prediction, which proved valid. Moreover, it shows that there's not a tie with a competing theory, smell, because the critters almost certainly couldn't process the information such an olfactory organ would have yielded; that theory made a failed prediction. They've incorporated the visual effects into their conclusion as an additional role (never underestimate reproduction as a factor in design--after all, it's at the heart of evolution). Has for temperature regulation, my reading says they've just left that to one side, perhaps it was trashed before, perhaps they just didn't unearth relevant data.
|