...fields, when ordered by size, appear to yield an asymptote which is interpreted as evidence of an approaching limit... in fact, the asymptote appears to be nothing more than a statistical artifact--that is, use of a large population, ordered by size, will frequently yield an exponential curve with an apparent asymptote.
What I think this means: if Cambell had done the same analysis on another set numbers with a similar distribution, such as the number of bloggers that link to each blogger (the power curve), he'd have seen an approaching limit on that data set too. The key is the distribution - a few large discoveries (or popular bloggers) and many small discoveries (or ordinary bloggers).
It goes on - the Lynch study is quite readable and entertaining particularly if you visualize the blood on the floor. Figure 6 is great, showing how widely an actual production curve diverges from the doomsayers' predictions.
The other contrarian article simply debunks the common meme that TV and movies make kids more violent. Some miscellaneous evidence:
- "SUNY-Albany sociologist Steven Messner has found that metropolitan areas in which violent TV programs attract especially large audiences have lower rates of violent crime."
- "the homicide rate barely changed from 1945 to 1967; the big increase started in the late 1960s, suggesting that something other than TV was at work"
- "The murder rate remained constant in Italy and declined in France, Germany, and Japan following the introduction of TV.