The evolution of antievolution policies after Kitzmiller v. Dover

Creationism, whatever variety, became unconstitutional after the rise of evolution in science, higher education, and law. Only in the 20th century did judges "suddenly see" the Constitution as reflecting their bias.

Study tracks the evolution of pro-creationism laws in the U.S. LA Times:
http://www.latimes.com/science/sciencenow/la-sci-sn-evolution-of-creationism-in-schools-20151217-story.html

Congratulations! Brilliant approach to legal analysis. Were you inspired by the "cdesign proponentsists" phylogeny?

I can't stop chuckling at the awesome methodology combined with the inherent humor in demonstrating the intentionally obscured creationist origins of legal texts via evolutionary analytics.

Doesn't get much better than this.

Oh, and not to mention the title: the evolution of anti evolution. Gotta love it.

So these yahoos are completely busted. We don't need no stinkin cdesignproponists, we can demonstrate the relationships between the different bills and thus their theological underpinnings using basic evolutionary theory. Cool.

Of course they could defeat this strategy by just developing legislation independently using original thinking and creativity instead of just blindly copying and pasting previously used crap. Oh who am I kidding? If they were smart enough to do that, they would already realize that were wrong.

I think that this method is significant beyond its application to this example case.

Paywalled, unfortunately. Was there much reticulate evolution?

Ray Martinez said: Creationism, whatever variety, became unconstitutional after the rise of evolution in science, higher education, and law. Only in the 20th century did judges "suddenly see" the Constitution as reflecting their bias.

Yardbird said:

Ray Martinez said: Creationism, whatever variety, became unconstitutional after the rise of evolution in science, higher education, and law. Only in the 20th century did judges "suddenly see" the Constitution as reflecting their bias.

Nice try, dipwad. Creationism is not now and never has been unconstitutional.

Congratulations! You are getting a lot of press mentions already.

Ray Martinez said:

Yardbird said:

Ray Martinez said: Creationism, whatever variety, became unconstitutional after the rise of evolution in science, higher education, and law. Only in the 20th century did judges "suddenly see" the Constitution as reflecting their bias.

Nice try, dipwad. Creationism is not now and never has been unconstitutional.

From the link in Nick Matzke's post: "Matzke is no stranger to the battles over teaching evolution in public schools. He spent three years at the National Center for Science Education, where he aided the parents of public school students from Dover, Pa., who filed a federal lawsuit to remove intelligent design from their school district’s curriculum. The case, Kitzmiller v. Dover, was decided in favor of the parents in 2005, with the court ruling that attempts to insert Biblically inspired creationist theories into public school classrooms were unconstitutional."

Dr GS Hurd said: Congratulations! You are getting a lot of press mentions already.

Yardbird said:

Ray Martinez said: Creationism, whatever variety, became unconstitutional after the rise of evolution in science, higher education, and law. Only in the 20th century did judges "suddenly see" the Constitution as reflecting their bias.

Nice try, dipwad. Creationism is not now and never has been unconstitutional.

Ray Martinez said:

Yardbird said:

Ray Martinez said: Creationism, whatever variety, became unconstitutional after the rise of evolution in science, higher education, and law. Only in the 20th century did judges "suddenly see" the Constitution as reflecting their bias.

Nice try, dipwad. Creationism is not now and never has been unconstitutional.

From the link in Nick Matzke's post: "Matzke is no stranger to the battles over teaching evolution in public schools. He spent three years at the National Center for Science Education, where he aided the parents of public school students from Dover, Pa., who filed a federal lawsuit to remove intelligent design from their school district’s curriculum. The case, Kitzmiller v. Dover, was decided in favor of the parents in 2005, with the court ruling that attempts to insert Biblically inspired creationist theories into public school classrooms were unconstitutional."

Ray, does any one in the US government stop you from going to church?, reading the Bible?, writing and publishing religious literature? Please let us know if anyone does, so we can contact the ACLU and they can stop the government from violating your 1st amendment rights. I would be happy to help you out in any of those cases.

Ray Martinez said:

Yardbird said:

Ray Martinez said: Creationism, whatever variety, became unconstitutional after the rise of evolution in science, higher education, and law. Only in the 20th century did judges "suddenly see" the Constitution as reflecting their bias.

Nice try, dipwad. Creationism is not now and never has been unconstitutional.

From the link in Nick Matzke's post: "Matzke is no stranger to the battles over teaching evolution in public schools. He spent three years at the National Center for Science Education, where he aided the parents of public school students from Dover, Pa., who filed a federal lawsuit to remove intelligent design from their school district’s curriculum. The case, Kitzmiller v. Dover, was decided in favor of the parents in 2005, with the court ruling that attempts to insert Biblically inspired creationist theories into public school classrooms were unconstitutional."

attempts to insert Biblically inspired creationist theories into public school classrooms were unconstitutional

John Harshman said: Paywalled, unfortunately. Was there much reticulate evolution?

Nick Matzke said: Replying to serious people...

John Harshman said: Paywalled, unfortunately. Was there much reticulate evolution?

The main "reticulation", which was known beforehand, is the introduction of Ouachita text into the AFA tradition. I did a fair amount of work trying to characterize this, since I thought this would be too important to model in the traditional phylogenetic way, but it turns out if you just have 1 tip in the tree representing the "incoming" lineage, then it's really just another tip, just one where a bunch of characters change from absent to present. So you get a spike in rates at that point in the tree. Other than that, there could be some other lateral transfer, but I didn't do an extensive analysis since (a) lateral transfer would look like homoplasy in a traditional vertical-inheritance phylogeny and (b) all the stats I ran suggested that homoplasy was pretty darn low for most characters. But, I wouldn't exclude the possibility of reticulation/lateral transfer completely -- especially for the new DI Model Bill (dated "2015-2016") which seems to have been put up recently (after I did the original analysis in August) -- this new model bill definitely has SEA features but also may re-instantiate older features (like the Academic Freedom Act). The Supplemental Material has all the details of the analysis and should be free online, but I noticed one of the figures seems to be funky on my machine (but looks fine on the original I submitted). Email me if you have trouble getting it - nick.matzke@anu.edu.au

Also the fact that the creationists have given up actually challenging science and just trying to indoctrinate kids. They are probably paying more for lawyers than they are investing in their own science
If Creationism is real science why not get Liberty University to start offering degrees in the various science streams. They start pumping out creation biologists and creation geologists by the bucket load and produce hundreds of papers that challenge accepted science. Instead we have Behe at a University that has disowned him and Dembski turning his back on ID to do other things.

The only reason they don't is that deep down they know that it is all snake-oil and the only ones who might accept it is the kids.

Independent event: Eric Rothschild podcast on Kitzmiller v. Dover: https://soundcloud.com/penn-arts-sciences/kitzmiller-v-dover-podcast

I suggest that Nick's analysis of the "horizontal meme transfer" of blocks of language in the bills is very conservative. Reading the Supplementary Materials to the paper, which I can do as my university subscribes to the online version of Science, I see that each such feature was coded as two states (absent and present, so in effect 0 and 1), or else as just a few states. The probabilistic model used for the likelihoods was the Mk1 model, which envisages symmetrical change among these few states. This means that parallel acquisition of the same block of text in two different lineages is quite possible.

In real life, it is very unlikely that two identical blocks of text will arise independently. If we see "to be or not to be, that is the question", we pretty well know that this block of text originated with Shakespeare. It is extremely unlikely that someone who never heard of Shakespeare dreamed it up on their own.

So in concluding for copying between bills, Nick is making a very, I would say overly, cautious inference.

Have I got this right, Nick?

Ray Martinez said: Creationism, whatever variety, became unconstitutional after the rise of evolution in science, higher education, and law. Only in the 20th century did judges "suddenly see" the Constitution as reflecting their bias.

Forgive me, the non-biologist, non-scientist here. I think I understand the series of trees presented. I suspect that the primary "tree" is the "best fit" for all of the various "Character" traits, with the presence and "inheritance" of each "Character" represented in a separate instance of the tree.

But in your various trees, what do the alpha-numeric codes at the "leaves" represent? Is there a translation table somewhere? And, are the "Characters" described somewhere else?

Thanks.

These court decisions are anti-creationism. They are purposed to censor creationism. Creationism is not nti evolution TEACHING in schools. just equal time until they are discredited and then rid of.
I think the ID revolution with YEC and the general public awareness of state control on information or truth of origin subjects makes bugger, funner, juicier court cases on science censorship a good thing to bring about.
Asking lawyers in black robes what science is just a waste of time.
It should be about the moral and legal rights to tell the truth on any subjects dealing with the universe WHERE the subject is being discussed.
Censorship of ideas is opposite to a free people and nation. Ideas about God and Christian doctrines being censored is absurd.
REALLY!!!
Yes evolution would struggle to compete before students in some curve on the graph but thats too bad.
YES I think a bigger court case is on the horizone.
YES celebrate this court decision as it brings up the issue of court decisions.
Stoke the fire indeed. ID brings it up with expection it suits them.
Mpre cases please, We are creationists.

Joe Felsenstein said: I suggest that Nick's analysis of the "horizontal meme transfer" of blocks of language in the bills is very conservative. Reading the Supplementary Materials to the paper, which I can do as my university subscribes to the online version of Science, I see that each such feature was coded as two states (absent and present, so in effect 0 and 1), or else as just a few states. The probabilistic model used for the likelihoods was the Mk1 model, which envisages symmetrical change among these few states. This means that parallel acquisition of the same block of text in two different lineages is quite possible. In real life, it is very unlikely that two identical blocks of text will arise independently. If we see "to be or not to be, that is the question", we pretty well know that this block of text originated with Shakespeare. It is extremely unlikely that someone who never heard of Shakespeare dreamed it up on their own. So in concluding for copying between bills, Nick is making a very, I would say overly, cautious inference. Have I got this right, Nick?

Scott F said: Forgive me, the non-biologist, non-scientist here. I think I understand the series of trees presented. I suspect that the primary "tree" is the "best fit" for all of the various "Character" traits, with the presence and "inheritance" of each "Character" represented in a separate instance of the tree. But in your various trees, what do the alpha-numeric codes at the "leaves" represent? Is there a translation table somewhere? And, are the "Characters" described somewhere else? Thanks.

"But in your various trees, what do the alpha-numeric codes at the “leaves” represent? Is there a translation table somewhere? And, are the “Characters” described somewhere else?"

Tip labels indicate AFA or SEA, year, state, bill number (SB, senate bill; HB, house bill), and versions (a, b, or c, for legislative revisions; t or f, teachers or faculty targeted).

Nick, this is important work. Are you allowed to self-publish an online copy? If not, is it worth considering publishing in an open access journal next time? I saw this referenced on Vox, but their summary leaves a lot to be desired.

Erik M�ller said: Nick, this is important work. Are you allowed to self-publish an online copy? If not, is it worth considering publishing in an open access journal next time? I saw this referenced on Vox, but their summary leaves a lot to be desired.

(I will also note the irony of creationists banning each other from creationist sites, and the banned creationists then showing up on pro-science sites, which do not ban them, to complain about censorship.)

The DI has responded with in their usual manner - disregarding the actual claims made in your article, Nick, and charging you with grants misappropriation. Laughable.

http://www.evolutionnews.org/2015/12/did_nick_matzke101761.html

Hey Discovery Institute, instead of trying to sue and legislate your way into the science classroom, why not do some research and publish your way in? You know, the way *we* do it.

I. Hadi said: The DI has responded with in their usual manner - disregarding the actual claims made in your article, Nick, and charging you with grants misappropriation. Laughable. http://www.evolutionnews.org/2015/12/did_nick_matzke101761.html

There is a report in Scientific American blog which refers to this as "tongue in cheek, or rather panda's thumb in creationist's eye".

Forgive me, but I know practically nothing about phylogenic analysis, and absolutely nothing about the details, and seem to have no access beyond the basic "abstract" of Nick's paper. Perhaps if you'll permit me to put it into my own words, to make the abstract and (to the layman) obscure statistical techniques more comprehensible.

I then have a couple of questions.

My rudimentary understanding is that you have in hand a bunch of objects that you believe may or may not be related to each other through a hierarchical ancestral tree structure. The intent of the analysis is to see if this assumption is correct, and to see if it is possible to identify the structure of that branching tree and measure (to some extent) the "relatedness" of the objects in hand. This kind of analysis is often performed on biological "objects", but can be applied to any "objects" that may be related to each other.

The basic method is that one first identifies a set (or multiple sets) of "characters", features of the objects that may be related, or may have similar morphology or function, but that may differ to some extent from object to another: color, shape, texture, function, etc.

The results of this analysis is a tree structure, showing the relationship (or the sense of "relatedness") among the objects being analyzed. The result is typically a tree structure, not because of any preconceived or biased assumptions in the analysis, but because that is typically what the data actually shows. Presumably, if there was more than a single "root" to the tree, if one subset of the objects was not related to the others in any way, then this type of analysis would show that. That is, the results might end up being several rooted trees, or a whole orchard of trees. Presumably there is also a full set of numbers behind these trees, showing the "confidence" or "uncertainty" in the various branches of the tree. (One seldom sees "confidence" levels attached to these trees, but I assume that they exist.) Also, presumably, if the data is ambiguous enough, the analysis might generate more than one possible tree structure, but the one with the highest "confidence" level is the one presented.

My understanding is that one would like to test these data analysis methodologies, in order to see if they actually produce what they purport to produce. To that end, one would like to take a set of objects whose historical roots and relationships are known prior to the analysis. Then, apply the methodology: choose and measure the set of "characters", construct the model (or whatever it is you do), run the analysis, compare to known results, then tweak the process until the two agree. Hopefully, if the "modeling" is good and the analysis sounds, then the results ought to compare well to the known history of the objects. In the best case, in addition to validating the tools, the analysis should also be able to provide new insights into the known sample "objects", allowing one to infer relations that were not previously "obvious".

I presume that is what was done in this case. As Nick describes, there is a relatively new variation on the existing methodology (within the past year or so?), and it would be well worth while to test this new analysis technique on a known set of objects, in order to confirm the validity of the methodology. In this case, the chosen set of objects was the various anti-science bills in the various state legislatures, over the recent past.

Is that a fair, though exceedingly rudimentary summary?

If so, I have a couple of questions.

Can the analysis methodology identify more than a single "tree", if the two subsets of objects are indeed unrelated? Or will the methodology always produce a tree, perhaps with branches of low "confidence"? Or is there any difference between "separate trees" and branches with low "confidence"?

Is there an assumption that all of the objects in question are indeed "leaves" of the tree? All pictures of these trees that I've seen, seem to suggest so, as all of the identified objects appear at the far right of each branch.

Expressed another way, does the analysis methodology allow one to place one of the objects *on* one of the branches, rather than at a leaf? That is, can the analysis identify a "direct ancestor" to a "leaf" object? I presume that perhaps the "direct ancestor" relationship might be represented by the pattern of branches for "b2013 MO HB179", "… HB1587", and "… HB486". This pattern appears to show a "least common ancestor", but in the case of HB486 and HB1587, the latter shows no variation from the "least common ancestor", suggesting that HB1587 is (for all practical purposes) as close to a "direct ancestor" of HB486 as the analysis method can suggest.

Is that a fair reading of the data?

Thank you for your patience.

I'd think that to determine direct ancestry, one would need sufficient DNA samples of the objects in question to do paternity tests on those samples. When talking about a fossil, that doesn't seem likely.

Scott F said: Forgive me, but I know practically nothing about phylogenic analysis, and absolutely nothing about the details, and seem to have no access beyond the basic "abstract" of Nick's paper. Perhaps if you'll permit me to put it into my own words, to make the abstract and (to the layman) obscure statistical techniques more comprehensible. I then have a couple of questions.

John Harshman said:

Scott F said: Forgive me, but I know practically nothing about phylogenic analysis, and absolutely nothing about the details, and seem to have no access beyond the basic "abstract" of Nick's paper. Perhaps if you'll permit me to put it into my own words, to make the abstract and (to the layman) obscure statistical techniques more comprehensible. I then have a couple of questions.

As far as I know, the methods Nick is using will always return complete trees (meaning that two or more separate trees are not on the table. If there really are multiple trees, this will as you suggest show low confidence for some branches, but there are many possible reasons for low confidence. I believe his method does allow for direct ancestry, though, and this would be represented on his tree by a terminal branch of zero length. (Note that only the vertical branches actually count as branches, by which I refer to the branches generally parallel to taxon names; horizontal branches are just there to make the tree legible.) The method also doesn't allow for reticulation or horizontal transfer, though that may be a good hypothesis to account for conflict between data and tree.

Scott F said: Hi John. Thanks for the update. That raises a couple more questions. I presume that all of the trees would share a similar time axis?

"the methods Nick is using will always return complete trees" Is that too restrictive an assumption? Wouldn't it miss (or fail to highlight) potentially independent acts of "creation", especially in human created objects like this?

You say [underlining added]: "only the vertical branches actually count as branches" And: > [direct ancestry] "would be represented on his tree by a terminal branch of zero length" I got the impression that the distance of the vertical separation between objects was purely arbitrary.

Perhaps I'm misunderstanding??

Am I correct in assuming that there is some "weighting", some quantitative measure of the relatedness between objects? I mean, this is just a visual representation of a "best fit" graph, but the computer had to be chunking away on actual numbers. Do such graphs ever label the "confidence level" of these branches (or whatever measure of uncertainties are involved), or would that just be too much information on a single graph? Thanks again.

John Harshman said: The way the algorithms work, the method wanders around in tree and parameter space, sampling all sorts of trees and parameters. They spend proportionally more time in regions that the data like and less in regions the data don't like.

John Harshman said:

Scott F said: Hi John. Thanks for the update. That raises a couple more questions. I presume that all of the trees would share a similar time axis?

I don't think that's the case. The trees for individual character reconstructions appear to be parsimony trees, and their branch lengths represent the number of character changes. Bills earlier in time would tend to be more "primitive" and thus there should be a rough correspondence to a time axis, but accent on the "rough".

"the methods Nick is using will always return complete trees" Is that too restrictive an assumption? Wouldn't it miss (or fail to highlight) potentially independent acts of "creation", especially in human created objects like this?

It's too restrictive if and only if any of these bills are independent acts of creation. I doubt that's true.

You say [underlining added]: "only the vertical branches actually count as branches" And: > [direct ancestry] "would be represented on his tree by a terminal branch of zero length" I got the impression that the distance of the vertical separation between objects was purely arbitrary.

Sorry for misleading you. "Vertical" and "horizontal" are arbitrary directions and different trees can be rotated through 90, 180, 0r 270 degrees from each other. By "vertical" I meant "along the axis of change", whether that axis represents time or character change. So your vertical here is my horizontal.

Perhaps I'm misunderstanding??

Just switch the words "horizontal" and "vertical".

Am I correct in assuming that there is some "weighting", some quantitative measure of the relatedness between objects? I mean, this is just a visual representation of a "best fit" graph, but the computer had to be chunking away on actual numbers. Do such graphs ever label the "confidence level" of these branches (or whatever measure of uncertainties are involved), or would that just be too much information on a single graph? Thanks again.

I'm not sure what you mean by "weighting" or "relatedness" here, but if you are referring to the degree of confidence that a particular branch actually exists or, more properly, that the data and evolutionary model being used provide strong support for that branch, given their assumptions, then yes. Bayesian methods do produce such a measure of support and it's often printed on the tree figure. The way the algorithms work, the method wanders around in tree and parameter space, sampling all sorts of trees and parameters. They spend proportionally more time in regions that the data like and less in regions the data don't like. If you randomly sample trees every so often as the program runs, you will get a sample of this space, weighted towards trees the data like. And a commonly used measure of support is a consensus of these trees, with each branch labeled by the proportion of sampled trees in which that branch appears, from zero to one. I presume Nick has such a consensus, and since the paper itself is paywalled I don't know if his published figures report it.

Ironically, creationists apply descent with modification to their tactics: “In what is almost a too-clever illustration of how evolution works, a scientist at Australian National University has created a chart to show us the evolution of anti-evolution bills. The study was published last week in Science, on the 10th anniversary of the historic Kitzmiller v. Dover Area School District trial, which struck down the teaching of intelligent design, an attempt to mask creationism with pseudo-scientific language. Evolutionary biologist Nick Matzke revealed how these bills have evolved over time to avoid potential predators such as the pesky Constitution and public outcry.”