Evolution Explained with Weasels, Flashcards, and Telescopes.

One of the barriers that seem to keep Intelligent Design proponents and Creationists from understanding the theory of evolution has to do with a complaint about how the process of random mutation and natural selection could produce anything new.

You can often hear the objection voiced in a few different ways, but it is used as justification that a process that does not involve help from an intelligent agent could not possibly create new information in the DNA of an organism.  I propose in this article a demonstration for how that is possible.

The best articulation of the complaint that I have seen goes as follows.  The theory of evolution claims that life became so diverse through the process of random mutation, natural selection and inheritance.  This suggests two questions:

  1. How can a random process produce anything new at all in terms of information?
  2. Natural selection is a filtering process.  All it can do is remove information from a system.   It has no ability to create new information.   So how can natural selection contribute to increasing information in the genome?

This is actually an excellent set of questions which deserve an answer.    My aim is to answer the questions with the simplest scenario I can find that demonstrates how this is possible.  The scenario needs to model a random process that produces a random set of possibilities working through a selection process which removes possibilities.

Suppose we have two mountain tops a few miles apart.   Our goal is to send a message from someone on one mountain top to someone on the other one.   On one mountain top is a guy with a telescope.  Let’s call this Telescope Guy.   On the other mountain top is a guy with a deck of large flashcards each having a different single letter from the alphabet on them.    This is Flashcard Guy.  And his deck of flashcards has many copies of any given letter so he has plenty of cards.

So now contrary to what seems to make sense, we are going to try to send a message from Telescope Guy to Flashcard Guy.   It seems backwards, doesn’t it?   But let’s begin anyway.

Someone comes up to Telescope Guy and gives him a message to send, which contains the single word, “weasel.”    So at a given time, Flashcard Guy starts holding up cards one at a time from his randomly shuffled deck of cards.   And Telescope Guy watches the flashcards through the telescope as they are held up one at a time.

Now given that Flashcard Guy has no way to know if any letters are part of the message or not, all he can do is hold up a letter for a minute and then put them down side by side.    And all  Telescope Guy can do is watch through the telescope in frustration as random letters are produced one by one and laid down.   After a while Flashcard Guy just stops because he thinks he is done.    The sequence he has gotten is,


Not a very impressive messaging system, is it?    The problem here is that all we are doing is watching a random process.   What is the probability that this process will produce the message “weasel” for Flashcard Guy.  It might happen someday, but we might as well say “never” for our purposes.   Cards off the top of a randomly shuffled deck really don’t produce messages or anything else of interest.   So the complaint is correct, according to this scenario, a random process cannot produce anything new of any value.   It certainly cannot be used by itself to send a message.

But then the two guys get an idea.   Telescope Guy equips himself with a very bright red lanterns.  The lantern is bright enough so Flashcard Guy can see it if it is lit..   And the two guys have prearranged that when Flashcard Guy holds up a letter, Telescope Guy will light the red lantern if the letter is to be rejected.

Now we begin again, and we get the letter “y”, Telescope Guy sees it through the telescope, lights the red lantern and Flashcard Guy puts the letter back in the stack.   Next letter is “b”, so it gets the red light again.   Next letter is “w”, and Telescope Guy does not light the red lantern and Flashcard Guy puts the letter down on the table as part of the message.   This goes on until the entire word, “weasel” is spelled out at which point Telescope Guy lights both lanterns to signal that we are done.

So there we go, message sent.  It’s not very efficient, but the message does get sent.  It is important to note that what turned a random process into a creative one is the addition of the selection process.   The combination of the two are perfectly capable of sending a message of arbitrary complexity, yet each one alone is not.

So how does the mountain top scenario relate to the process of evolution?.   In evolution the DNA is represented by the table where the weasel message will end up.   The random mutations in DNA is the drawing of random letters from the shuffled flashcard deck.  And the Telescope Guy with his lantern represents natural selection.

So random mutation creates a new possibility in the DNA, which codes for some change in the traits of the organism.  Nature rejects that new possibility if it decreases the probability that the organism can survive to reproductive age..

Notice that I started with a message analogy and ended with evolution.  This is not inconsistent with what information theory tells us about information in general.   In this backwards sort of way, nature is sending a message to DNA.   Its content is information about the state of the organism and the environment throughout the history of ancestry of the organism.

And just like the message in my story oddly goes from the Telescope Guy to the Flashcard Guy, so does a message go from the environment into the DNA.

One more thing.   One might complain that the weasel message is something that only intelligent people can read and understand.   It is a message that conveys meaning.   But consider that I happened to choose an arbitrary message using an English word.  Suppose a French-speaking person passed the same message in French to the Telescope Guy who does not read French.    And suppose the Flashcard guy does not read French either.    Notice it doesn’t really matter, though.  The French word belette is just as easily sent as any English word, regardless of whether any of the guys speak French or not.

Furthermore, the message can consist of arbitrary symbols as long as those symbols are in the flashcard deck.   And finally, the two guys can be replaced by a machine or a living organism and DNA mutations for the flashcard deck.

Another complaint is that it took intelligent people to set up the experiment.   Therefore, it doesn’t demonstrate that the process can work without intelligence.    This argument is offered from time to time.  However, setting up the scenario and operating it are two different things.  This scenario happens to have been set up by intelligent people, but it strictly models the processes as specified, which is a purely random process followed by a process that only removes possibilties.

Another complaint might be that this doesnt’ prove that evolution can do this.  Or it doesn’t prove that this could do enough to explain all of evolution.  Those are valid question that have answers to them.  But keep in mind that the original question or challenge was that there is no way any random process and a selection process could increase the information content of anything.   I have demonstrated that it can.


4 thoughts on “Evolution Explained with Weasels, Flashcards, and Telescopes.

  1. Problem
    This is far to simplistic, the problem with evolution is complete working gearboxes have to be running perfectly in one big jump not in single steps. The assumption is also that the red/green lights must be non random to build any message of sense. The end message must itself make sense to the receiver ie what does the word ‘weasel’ mean to the receiver unless he has intelligence. Random transmition is random unless at least one end has ‘intelligence’.

    • Owen, thanks very much for the reply, and thanks for reading over the article.

      I have a few comments on what you said. I will take them one by one.

      “the problem with evolution is complete working gearboxes have to be running perfectly in one big jump not in single steps.”

      The modern theory of evolution does not predict that things need to come together in one big step. In fact it predicts completely the opposite. Some people have a misconception about this, but they are not professional biologists.

      “The assumption is also that the red/green lights must be non random to build any message of sense.”

      Yes, precisely. The red/green light must be non-random to build a message. And that is what happens when the Telescope Guy rejects or allows a symbol and lights the appropriate lamp. That is the whole point of the example. It models natural selection, which is definitely non-random. Natural selecton selects variations in traits because the organism with that variation may or may not survive depending on how that variation affects the survivability of the organism.

      So yes, selection has to be non-random. But that is the whole point of the example. When a random process is followed by a non-random selection process, information can be transferred.

      “The end message must itself make sense to the receiver ie what does the word ‘weasel’ mean to the receiver unless he has intelligence”

      Actually, no. Notice that the word, “weasel” could be transmitted whether either guy speaks English or not. Consider that the two mountain top guys are just like telegraph operators whose job it is is to pass a message. Maybe even a coded message where they don’t understand the content.

      The point of the example is that information can be transferred by a rote process that has no awareness of what the message will accomplish.

      This was the big breakthrough that created the field of information theory. Claude Shannon created the entire field around 1948 with his landmark paper. It established the fundamental properties of information and linked them with thermodynamics. Here is a link to his paper.


      It is only humans who acquire, store, and send information in which they find meaning and purpose. Nature acquires, stores, and sends information all around us from natural process to natural process.

      So the two mountain top guys’ job is to merely send information from one place to another using a random process followed by a selection process.

      In evolution, this is how information about the environment ends up encoded in the genome. Random mutation has no sense of meaning. It is just DNA making bad copies of itself. Some of the DNA goes on to code for traits of the organism, and nature selects or rejects the variations from that mutation.

      No meaning is interpreted anywhere.

      “Random transmition is random unless at least one end has ‘intelligence’.”

      And that is the point. No intelligence is required to send and receive a message containing information.

  2. One key point you mention in passing but haven’t explored in detail is that the ‘information’ encoded in DNA is ‘information about’ the environment. More specifically, it is ‘information about the environment of all ancestors of the current organism’. A naive treatment of information theory deals with the ‘information to unambiguously record’ some message (message complexity information), but really what we’re dealing with here is some sort of mutual information between some aspects of the history of environments and the behaviour of the organism as captured in the DNA sequence.

    ‘Mutual information’ between A and B behaves quite differently than the ‘message complexity information’, and behaves much more like the intuitive notions of ‘information’ that the creationist arguments appeal to. However, the attempts of creationists to place their arguments from ‘information’ on a mathematically formal footing seem to doggedly use ‘message complexity information’ metrics, and hence, get their maths in a muddle and reach befuddled conclusions.

    • Mathew,
      Thanks for the excellent comment. I agree that information in the DNA of an organism is

      some sort of mutual information between some aspects of the history of environments and the behaviour of the organism as captured in the DNA sequence.

      If you don’t mind, I will correct the article with that in mind. Also, thanks for the comments on information as message content. That is the way I see it. A message is sent from the DNA mutation to the environment, and the environment either rejects or accepts the proposal. Accepting the proposal causes the mutation to persist.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s