In previous posts, I discussed means of assessing whether a transmission from outer space implied intelligence. The intelligent design theorists use as an example a transmission from space that follows a pattern of the prime numbers in consecutive order. Such a transmission would be immediately recognizable as a communication, or sign of intelligence. Similarly, the ID backers say, a proto-cell is astronomically unlikely to have fallen together spontaneously.
I have argued that we can't necessarily accept that a prime number transmission implies intelligence. We first have to know about primes. But anyway, it occurs to me that cryptographers and communications engineers would have little problem with such a challenge -- if by intelligence we mean something to which humans can relate.
As is well known, every English letter occurs with a specific frequency. Hence each pair in the alphabet A (each element of AXA) also has a specific ratio. This also holds for words and for symbols that are part of some language L in general. That is, if a symbol is used in some form of communication of ideas, it bears a stable relationship with every other symbol in L.
So if a set of discrete pulses arrived at some bandwidth, we might use scatter diagrams to check each possible pair of frequencies at time interval t and see whether, for pair (x,y) a football shape emerges, and to check the level of correlation. Of course, if the correlation is too weak, either the sample is insufficient or the pair (x,y) cannot be construed as a grammatical structure in L. If the correlation is too linear, then we would have to consider mechanical rather than symbolistic cause, as with pulsars.
Noise would yield a group of scatter diagrams most of which lack a football shape. Those with the shape would be checked more closely to see whether a language is implied.
This technique should work in most cases, though I am not sure whether it would work in the specific case of a prime number sequence. More thought necessary. But, the argument may still proceed.
The question arises: can we discover such a symbolistic code at the level of the origin of life? We are not necessarily talking about the DNA or RNA codes, but about whether the makeup of the first cells might indicate an intelligence.
How might we determine whether some system has been designed by an intelligence rather than being a consequence of the "hidden hand" or regulator of chaotic processes? We might model the system's workings with a blueprint, perhaps a Boolean circuit whereby the symbols represent logic gates. If such a circuit C's symbols are all paired (CXC) and scatter plots are done, we would check for the telltale football shapes.
If they showed, we would be fairly sure that some intelligence was "speaking" through the blueprint. But a lack of such evidence would not rule out intelligent design, of course.
A study might be done comparing Boolean circuits of artificial and inorganic natural systems to see whether there are significant differences. This information might be used for comparison with organic or near-organic systems. If the symbol language of an organic system was fairly close to that of an artificial system yet different from an inorganic system, that would be big news. Still, lack of such evidence would not disprove intelligent design.
So the point here is that the odds against a fluke outlier may be extraordinarily high but that isn't necessarily analagous to an improbable signal from space. In the latter case, we can apply statistical methods to determine the likelihood of non-random activity associated with communication versus the random pulses associated with noise.
What statistical methods can be used to check for non-randomness in the "signal" of early life forms? I gave one method, though I do not think a positive result is likely.
Yet, we cannot foreclose the possibility that a strong statistical approach might yield surprising results.
No comments:
Post a Comment