221
edits
Changes
Jump to navigation
Jump to search
Line 1,831:
Line 1,831: + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
Talk:Essay:Best New Conservative Words (view source)
Revision as of 17:23, August 3, 2010
, 17:23, August 3, 2010→A summary with graphs...: new section
: Finding the best new conservative words is like drilling for oil. Of course it is not a random process. That would be silly. But the lack of randomness does not mean that no intelligent conclusions can be drawn. We can properly conclude that there is more oil in the Arabian Peninsula than in Pennsylvania, where oil was first discovered, despite the lack of randomness in drilling oil wells. Our conclusion would be based on the difference in output, and the implausibility that ''any'' reasonable selection technique would produce such a differential if there were not also an underlying difference.--[[User:Aschlafly|Andy Schlafly]] 13:06, 1 August 2010 (EDT)
: Finding the best new conservative words is like drilling for oil. Of course it is not a random process. That would be silly. But the lack of randomness does not mean that no intelligent conclusions can be drawn. We can properly conclude that there is more oil in the Arabian Peninsula than in Pennsylvania, where oil was first discovered, despite the lack of randomness in drilling oil wells. Our conclusion would be based on the difference in output, and the implausibility that ''any'' reasonable selection technique would produce such a differential if there were not also an underlying difference.--[[User:Aschlafly|Andy Schlafly]] 13:06, 1 August 2010 (EDT)
== A summary with graphs... ==
On June 9, 2009, Andrew Schlafly proclaimed his [[Essay:Conservapedia's Law]]: <i><blockquote><b>Conservapedia's Law</b> is the observation that conservative insights increase over time at a geometric rate, as in 1-2-4-8-16-etc.<br>
For example, there is a doubling in effective new conservative terms per century. </blockquote></i> This remarkable precise observation was bolstered over the time by a list of 300 <i>conservative words</i><i><blockquote>Powerful, insightful new conservative terms have grown at a geometric rate, roughly doubling every century. For every insightful new conservative term originating in the 1600s, there are two new terms originating in the 1700s, four new terms in the 1800s, and eight new terms in the 1900s, for a pattern of "1-2-4-8". This implies a more conservative future and a correlation between conservatism and truth. The year 1612 is our starting point: the King James Version of the Bible was published in 1611, and William Shakespeare had written nearly all his plays. </blockquote>
{|
|The pic on the right shows the percentage of terms per century in which they were first mentioned.
|[[File:Bcw-001.png|thumb|Best Conservative New Words]]
|-
|The distribution of these ''conservative words'' is even more remarkable as it doesn't reflect the general trends of creating new words. To exemplify this, I took a sample of ≈ 42,000 words from the word-book of my ubuntu-distribution and checked their creation date with the Merriam-Webster (the gray areas of the two diagrams overlap)
|[[File:Bcw-002.png|thumb|distribution of 42,000 words]]
|-
|To convince Andy Schlafly's that the distribution of the words in his list is just an artifact generated by .[[sample bias]], I at first showed that it doesn't hold for shorter periods of time than centuries. But Andy Schlafly told me that he didn't see ''any merit'' in my challenge: <i>The entry observes that new terms are generated at higher rates during productive periods within decades, for example just after or during religious awakenings.
|[[File:Bcw-003.png|thumb|Cons. Words per Decade]]
|-
|Interestingly, I couldn't observe any effect of the <i>Great Religious Awakenings</i>, neither in the number of <i>conservative words</i> nor in the number of all words.
|[[Image:Bcw-004.png|thumb|All Words per Decade]]
|}
No methodology is given how the <i>conservative words</i> are detected and gathered. To put the whole thing on a sounder ground, I asked Andrew Schlafly to take an unbiased text: I gave him a list of 500 words, which were - according to the [[Merriam Webster]] - first mentioned between 1600 and 1999. Andrew Schlafly objected to this sample as newer words are not well represented in word-books. When I propelled a second sample, consisting of 500 words all originated between 1600 and 1899 (300 years should be enough to prove his law), he was willing to mark the words he things to be conservative.
I'm very grateful that Andrew Schlafly took part in my little experiment: Thank you!
{|
|The five hundred words were taken at random from the suitable subset of the 42,000 words which I had dated, and so the sample distribution matches the overall distribution
|[[File:Bcw-005.png|thumb|500 Words per Centuries]]
|-
|Andrew Schlafly identified 41 words (8.2%) as conservative. The distribution of these 41 words over the centuries does <b>not</b> show a geometric progression.
|[[Image:Bcw-006.png"|thumb|Identified as Conservative]]
|-
|In fact, their distribution mirrors the distribution of the whole sample, as each century ≈ 8% of the words are identified as <i>conservative</i>
|[[File:Bcw-007.png|thumb|Percentages]]
|-
|This linearly dependence can be found for shorter periods of time (here for steps of 20 years) - and is found to be statistically significant.
|[[File:Bcw-008.png|thumb|Periods of 20 years]]
|-
|A striking contrast to this is the absence of any connection between Andrew Schlafly's list of over 300 <i>conservative words</i>, and a general distribution of the creation of words.
|[[File:Bcw-008.png|thumb|Periods of 20 years]]
|-
|However, if one maps the <i>percentage</i> of words in - e.g. - a 20 years' period in the corresponding century (24% of all 20th-century word are from the 1900s, 18% from the 1920s, 26% from the 1940s, and so forth...), one finds again a positive correlation.
|[[File:Bcw-010.png|thumb|percentages over 20 years]]
|}
Andrew Schlafly explains the discrepancies as follows:<blockquote><i>That's a
fascinating analysis, but its meaning is simply this: roughly 8% of all new
words are conservative in nature. That is greater than the number of words I
would have identified as liberal in nature. </br>But very few of these words
qualified for our list, which expressly consists of the "best" new conservative
words. Those words are being generated at a geometric rate.--Andy Schlafly
11:55, 28 July 2010 (EDT) </i></blockquote>
I've to take exception to this: Andrew Schlafly has shown that he is able to
generate candidates for his list for any time period as it is needed to fit his
prediction. He could as easily make a list following a 1-3-9-27 (or 2-3-5-7-11)
pattern.
Even a claim like: ''the number of Best New Conservative Words'' is not corroborated in any way by the list of the [[Essay:Best New Conservative Words]], and the geometric progressions seems to be just the approximation of a phantasy.
[[User:RonLar|RonLar]] 13:23, 3 August 2010 (EDT)