Received: by alpheratz.cpm.aca.mmu.ac.uk id QAA11685 (8.6.9/5.3[ref pg@gmsl.co.uk] for cpm.aca.mmu.ac.uk from fmb-majordomo@mmu.ac.uk); Sat, 13 Jan 2001 16:05:46 GMT Message-Id: <5.0.2.1.0.20010113085606.00af84a0@pop3.htcomp.net> X-Sender: mmills@pop3.htcomp.net X-Mailer: QUALCOMM Windows Eudora Version 5.0.2 Date: Sat, 13 Jan 2001 09:54:46 -0600 To: memetics@mmu.ac.uk From: Mark Mills <mmills@htcomp.net> Subject: Re: DNA Culture .... Trivia? In-Reply-To: <000c01c07d19$562825e0$b463b8d0@wwa> References: <5.0.2.1.0.20010112165041.01c8b310@pop3.htcomp.net> Content-Type: text/plain; charset="us-ascii"; format=flowed Sender: fmb-majordomo@mmu.ac.uk Precedence: bulk Reply-To: memetics@mmu.ac.uk
Lawrence said:
>Can you say a bit more about what you are thinking about here, Mark? How
>would ethics be connected to the 80/20 ratio? Are you suggesting that 80% of
>the people adhere to 20% of the ethical choices?
>
>Might this suggest that only 20% of memes are going to be successful?
>Interesting.
Mark:
Try this out:
If words are neural-meme phenotypes, then a distribution of word frequency
(a few words used very often, many only once) represents a phenotype
population study. The processes of doing the study is fairly simple, just
pick a book or article and count how many times each word occurs. What the
article on word-use discovered was a changing power law distribution
depending on length of the book.
Though power-laws have observed in word usage for many years, it isn't
initially obvious that word-use distributions should be Zipf-Pareto power
laws. Why not a random frequency (all words used equally)? A Gaussian
(normal distribution - a few used once, a few a lot, the rest in a bell
shaped curve)? Log normal? These alternatives to Zipf-Pareto are more
familiar.
Since I'm considering words to be neural-meme phenotypes, the appearance of
a Zipf-Pareto distribution tells us something about phenotype
production. Since the phenotypes are produced many times, it is actually a
study of phenotype reproduction. The underlying process controlling
reproduction is not random (anything goes), nor Gaussian (some errors
around a central prototype). Something else is happening.
Solomon and Richmond have a paper showing that Lotka-Volterra models
produce population distributions which follow Zipf-Pareto (power-law)
distributions http://xxx.lanl.gov/html/cond-mat/0012479. You have probably
seen examples of Lotka-Volterra models. A popular version is called
'shark.' The program lets you set the number of fish, the number of sharks
and reproduction rates, fish consumption rates, etc. then you start the
model and watch fish and shark populations fluctuate. Too many fish and the
shark population soars, pushing the fish population down. Then the sharks
die allowing fish populations to soar, etc. Another version uses foxes and
hares. If you have a broad number of species (many different fish) the
phenotype frequency turns out to be a Zipf-Pareto distribution (power-law),
just like word frequency.
Getting back to neural-memes and words, the Solomon-Richmond paper suggests
the brain is using neural memes in something like a Lotka-Volterra model,
competition seems to exist. This might support Edelman's neuronal group
selection theory.
Now, consider the 'Driver's in Rio' paper. It shows a power-law
distribution of tickets and says something about driver ethics. If we use
Solomon and Richmond's idea that power-law distributions infer the
existence of multi-scaled competition.
I'll stop here and see if this is making any sense.
Mark
>From: Mark Mills
>
>SNIP
>
> > On field work, here is an example I came across recently:
> >
> > http://xxx.lanl.gov/abs/cond-mat/0101172
> > Title: Distribution of Traffic Penalties in Rio de Janeiro
> >
> > This study found that roughly 20% of Rio drivers got 80% of the
> > tickets. The distribution followed Pareto-Zipf's law (power-law
> > distribution). Everyone has equal access to the traffic laws, especially
> > after getting a ticket, why not something more like a normal distribution
> > for getting tickets?
> >
> > I also came across an article about word usage and power-laws:
> > http://www.santafe.edu/sfi/publications/Abstracts/00-12-068abs.html
> > Title:Two Regimes in the Frequency of Words and the Origins
> > of Complex Lexicons:Zipf's Law Revisited
> >
> > Why do traffic tickets in Rio have a similar frequency distribution to
>word
> > usage? I think something neural memetic is going on. Lawrence mentioned
> > 'ethics' a few messages ago, maybe word usage and ethical driving are
> > linked by Zipf's laws?
http://www.htcomp.net/markmills
===============================================================
This was distributed via the memetics list associated with the
Journal of Memetics - Evolutionary Models of Information Transmission
For information about the journal and the list (e.g. unsubscribing)
see: http://www.cpm.mmu.ac.uk/jom-emit
This archive was generated by hypermail 2b29 : Sat Jan 13 2001 - 16:07:18 GMT