Ecological emergence and control and Dirty Data

CSGnet Archive CSG1992
1

mmt@BEN.DCIEM.DND.CA [Martin Taylor 920707 14:45] calls me to task:

I was basing my answer earlier today on your statement that the length
distribution had been found to be constant ove all fish even though the
distribution within a species was not. Do I understand you now retract
that statement.

(Now you say that the data are not that clean, and you can't tell
whether the Test has been applied. Right? Then the questions in your
original posting are based on a hypothetical situation, and my answer
applies to that hypothetical situation, not the real one as determined
by inadequate measurements.)

Did anyone else understand that I meant to imply that data on the total
biomass of fish are exact numbers? To all who did so, my apologies.

Martin and perhaps others will perhaps be surprised to discover that we
have yet to come up with exact methods for sampling fish in the ocean.
I clearly erred in assuming that the readers of this group knew that.

Just for future reference, real data are rarely exact. Saying that this
is due to "inadequate measurements" seems a bit unfair (anyone here ever
hear of the uncertainty principle, which makes some physical
measurements not exact?). But honestly, I thought that control theory
had at least some relevance to the real world!

Bill

ยทยทยท

--
Bill Silvert at the Bedford Institute of Oceanography
P. O. Box 1006, Dartmouth, Nova Scotia, CANADA B2Y 4A2
InterNet Address: bill@biome.bio.dfo.ca

2

[Martin Taylor 920708 14:15]
(Bill Silvert, still not dating his postings)

I was basing my answer earlier today on your statement that the length
distribution had been found to be constant ove all fish even though the
distribution within a species was not. Do I understand you now retract
that statement.

Did anyone else understand that I meant to imply that data on the total
biomass of fish are exact numbers? To all who did so, my apologies.

Martin and perhaps others will perhaps be surprised to discover that we
have yet to come up with exact methods for sampling fish in the ocean.
I clearly erred in assuming that the readers of this group knew that.

Sorry, but I am confused. What does length distribution have to do with
total biomass? You said, around 920707 (why don't you date your postings?):

Past studies have shown that
the length-frequency distributions (curves showing the number of fish of
each size) for all the fish in an area show much more regularity than
the distributions for individual species.

That's all I have ever been responding to. Why do you keep bringing in
other factors. If your statement is true, then my points about The Test
hold. The Test has been performed, and if interpreted in the usual way,
it shows control. If yours was a hypothetical statement, my points hold
in the hypothetical case.

And in respect of exact numbers, may I ask what percept in the whole world
is immune to sampling error? Only the magnitude of the error variance in
relation to the perceptual magnitude is different from one percept to another.
Even if we were talking about the perception of total biomass, which we never
have been, I would have thought readers of this group knew that. I clearly
erred.

It is nice, sometimes, to respond to arguments or points that are made, rather
than to bring new ones out of thin air and use them to castigate.

Do you want to return to the thread about the informational limits on control?
We let it drop last year, partly because it got played out without resolution.
Ask Gary for the archives on it if you missed it the last time through.

I must have missed some postings on this issue, because I have received none
(and certainly sent none) to which your last paragraph can refer:

Just for future reference, real data are rarely exact. Saying that this
is due to "inadequate measurements" seems a bit unfair (anyone here ever
hear of the uncertainty principle, which makes some physical
measurements not exact?). But honestly, I thought that control theory

had at least some relevance to the real world!

Could you copy back to us something that this connects with?

Martin