Re: [asa] Information and knowledge

On 4/14/07, Rich Blinne <rich.blinne@gmail.com> wrote:
<Quoting Yockey>

The transmission of genetic messages for more than 3.85 billion years
> since the origin of life (Mojzsis et al. 1999; Woese, 2000), with
> modification and diversification by evolution could be done *only*
> because the message in the genome is *segregate, linear, and digital*
> (Chapter 12). It is impossible to remove the effect of noise in analog
> signals. Early analog records of the glorious voice of Enrico Caruso
> (1873-1921) do not compare with the modern digital recordings of the
> Three Tenors; Placido Domingo, Jose Carreras, and Luciano Pavaroti.
> Shannon's Channel Capacity Theorem (Shannon, 1948) showed how to
> eliminate the effect of noise as much as we wish by digitizing the
> signal. The digital revolution has now provided digital television
> eliminating noise almost to the theoretical limit. Even cameras are
> now digital. [emphasis in the original]
> <end quote>

<snip>

First, he bought the hype of my industry. So-called digital recordings
> use lossy compression such as MP3, AAC, and H.264. Analog channels
> have bandwidth too and digital encodings need to sample at the Nyquist
> frequency which is related to that analog bandwidth. In addition you
> have to worry about aliasing if you sample with too few bits. This all
> before the digital signal is compressed. It also works in the other
> direction where you need sufficient radio frequency bandwidth to
> transmit the digital Wi-Fi signals most of you are using right now.
> Yockey also confused communication noise with the surface noise of the
> Caruso recording.
>
> Second, anything can be made into segregate, linear and digital by
> sampling and encoding. In fact, the "code" we are all discussing is
> created precisely by that kind of operation. As I also showed above
> analog signals can be dealt with in classical information theory. You
> can also have error correction in quantum information theory, e.g.
> CSS codes. There is really nothing magical about digital information.
> In fact, our really high speed serdes (read high bandwidth) are all
> analog.
>
> In summary, while the genetic structure of life is sufficient to drive
> the common descent with modification we are observing it is by no
> means necessary. This wrong conclusion is driven by the popular
> misconception that digital is perfect or near perfect but analog is
> noisy.

I'm not sure I agree with you there. I think Yockey's statement that it is
impossible to remove the effect of noise in analog signals is a little
misleading, however. It is true that both analog and digital signals are
noisy - digital signals always posess "quantization noises", because the
true (analog) value of the signal is truncated to the nearest integer.
However, the key point (surely for the transmission of genetic information)
is that the digital (noisy) signal may be copied perfectly (ie having
exactly the same noise signal) from generation to generation in an unlimited
fashion without further degradation. However, when you make an
analog-to-analog copy, new noise is introduced at each copy, due to the
inevitable non-perfect S/N ratio in the electronics, so the signal
degradation grows. If one were to make a chain of 50 copies from an initial
recording using cassette tape-to-tape copying, I'm guessing that the final
50th generation copy would be so noisy that you wouldn't be able to
recognise the original. So I would argue that digital IS necessary for the
conservation of genetic messages.

Iain

To unsubscribe, send a message to majordomo@calvin.edu with
"unsubscribe asa" (no quotes) as the body of the message.
Received on Sat Apr 14 11:58:23 2007