© NASAEurope's Planck satellite will measure the big bang's afterglow with unprecedented precision
Surely a victory for science? Only, it seems, if cosmologists can resist the temptation to gorge themselves on all those goodies.
A trio of astronomers have warned that, unless we use the information sparingly, we risk squandering a once-in-eternity opportunity (see arxiv.org/abs/0909.2649). If the whole data set is released at once, as is planned, any new ideas that cosmologists come up with may have to remain untested because they will have no further data to test them with.
This is a problem unique to cosmology. In other sciences, additional information is always available: you can always reset and rerun an experiment, or go out into the field to collect more data. Because of our fixed location in the universe, however, cosmology doesn't have that luxury. There is only a finite amount of information we can gather about the universe, and once we gather all there is to know about one aspect of it - in this case the temperature fluctuations in the cosmic microwave background - the well runs dry.
There is only a finite amount of information we can gather about the universe
In 2005, astronomers discovered a mysterious alignment of hot and cold spots in the CMB, which they dubbed the "axis of evil". If the phenomenon is real, it has important implications for our understanding of the universe. The Planck data will be used to test it. But imagine if cosmologists find another, similar, mystery buried in the data. What will they use to test that one?
The answer, according to Roberto Trotta of Imperial College London, is to be frugal with what you let the cosmologists see. Instead of giving out all the data at once, the supply should be rationed. Drip-feeding will allow the development of new hypotheses which can be tested as more of the Planck information is released. If we don't adopt this approach, we risk wasting the finest cosmology data set we have ever had, and remaining forever in the dark.
This is ridiculous.
I understand that it can help to start with imperfect information, see patterns and then make predictions about further information, but be realistic here. The more information that is out there to be studied in different ways, the more possibilities there are of finding connections.
If any theorists think that they would be helped in their theorizing by blinding themselves to any part of the available data, that should be up to them as individuals. Maybe they would never be able to "prove" that their theories weren't spun together from the full set, but if their theories have any merit at all, it won't matter how they came about them, they'll stand on their own. If theorists aren't imaginative enough to consider a partial set and put aside the rest to use later as verification, the aren't very good theorists.
This smacks of another ridiculous reason to hoard information instead of sharing it.