## Order = Signal. Random = Noise.

Joshua Swamidass from Washington University invited me to respond to conversations on his blog about Evolution 2.0. I wrote two blog posts in response.

We are asking: Are the mutations that produce useful changes in evolution random? Or not?

Joshua says they’re random. I (along with scientists like Nobel Prize winner Barbara McClintock) insist

cells guard against random mutations. When stressed, cells engineer deliberate, non-random changes. They do this through orderly, natural genetic engineering systems like transposable elements.

Appendix 1 of Evolution 2.0 is called “All About Randomness.” Longest chapter in the entire book. I took great pains to define randomness. I quoted some of it here.

I have likewise asked Joshua to define his use of the word “random” four times. He won’t answer my question. Not only that, he told another guy on his forum to not answer my question either.

Is a crisp definition an unreasonable request? Is it not normal and customary in all intellectual pursuits for people to define their terms first… and then debate?

Here, I gave a formal definition from information theory. Here’s a practical every day example of randomness: White noise from a TV:

For all practical purposes, this TV noise is random. Both picture and sound. Wanna know what random looks and sounds like? It looks just like this.

Randomness = noise. Noise always destroys your signal.

What is a signal? A signal is the order that you are trying to transmit. Whether encoded in a TV transmission or strand of DNA.

All signals have noise. Noise is the enemy of your signal. Always. It’s never your friend.

This video is a noisy TV signal. As you watch this, you can clearly distinguish the signal from the noise:

Joshua Swamidass said to me, “I think you have a false conflict in your mind between random and order. That needs to be stamped out if you want to make sense of this.”

Joshua, you have this exactly backwards. In transmitting and receiving information, there is always a conflict between random and order. The conflict is not false. It is real. Joshua, you are mixing randomness with order, then labeling your conflation of the two ‘random.’ I am making a clear distinction between the two.

Order = signal. Random = noise. Signal is not noise. Noise is not signal. I do not know how to be any clearer with my terminology.

Cells detect noise. They have at least three layers of error correction. They devote enormous resources to this. They fight noise (copying errors, mutagens) at every turn. Evelyn Witkin discovered bacteria even switch on more error protection when they anticipate noise might be coming!

Organisms also harness noise at times, holding most parts of DNA fixed while changing small sections, as they search for a solution. Your immune system does this.

Joshua Swamidass, for the fifth time… will you please define what YOU mean by the word random?

### 13 Responses

1. I find that post to be unnecessarily aggressive. It appears you are trying to create controversy where none need exist. When you are ready to communicate with me directly, let’s talk. I’m not willing to engage in a blog post war with you.

• Define “random.”

• S. Joshua Swamidass says:

Drop the theatrics and come discuss with me. You can even give me a phone call if you like. Stop pretending I haven’t answered too. Ive answered several time. I’m happy to explain it to you.

• Joshua I apologize if you have defined randomness elsewhere and I have not seen it. You are welcome to re-post your answer here so everyone will know what it is.

2. Mark Chenoweth says:

This seems to be semantics. If there is a process that is successful 75 percent of the time, Swamidass says it is 25 percent random and 75 percent ordered. Maybe you two could agree on this. But my guess is, if we had to use only one word to describe the process, Swamidass would use random (since we are assuming all randomness has some order), and Marshall would probably choose Ergodic?

I feel like you’re both acknowledging the facts, but using different terminology to describe them. Marshall’s explanation seems to me much more common-sensical, but as long as random is defined the way Swamidass defines it (as allowing for order within the randomness), I guess if someone wants to continue to use the the word random, I see no problem with it other than it would extremely confusing to the non-scientist.

• This is not merely semantics. The terminology we use in science (or any other discipline for that matter) is extremely important.

You’re in theology. Is theology just “semantics”? Or do the words you use actually matter?

Do words in theology have any meaning?

Joshua still hasn’t told us what he means by the word randomness, not that I’ve been able to tell. (Please correct me if I’m wrong.)

Random has a very, very specific meaning in engineering and mathematics. My background is communication systems and control systems; almost everything in biology relies on both of those things. One branch of control systems theory is stochastic control systems. An airplane is a stochastic control system for example, because the pilot has no control of the weather.

In our field these terms have very precise meaning because if you say “random” when you really mean “ergodic” or “semantic” or “9.8m/s^2”, the airplane crashes and 250 people die. And you lose your job and possibly spend the next ten years of your life in lawsuits.

Biologists use such terms in a very sloppy way. Especially in evolutionary biology. Much of the time when they say “random” they’re not even sure exactly what they mean. Joshua is not even saying what the pattern to the randomness is. (Hint: It’s not pink noise.)

When something is ergodic, you don’t call it random. When a computer or human or semiotic biological system exhibits a particular pattern, if you just call it “random” you’re not doing your job.

What if Isaac Newton had watched the apple fall out of the tree and said “That’s random?” How would he have ever discovered the theory of gravity?

You can model a tree as a stochastic control system and the wind as a random variable. Does it help anyone to say “The apple fell out of the tree randomly?”

What new information does that statement confer that you didn’t already know?

Wouldn’t it be better to say that we have a model for the force of gravity – and also a model for the strength of the apple stem – and a random variable to represent the behavior of the wind – and then we figure out when the force tugging the apple exceeded 2 Newtons, the apple falls to the ground at 9.8 meters per second squared?

The RANDOM part of the model is the least useful part. The important part of the model is the patterns you can define. The pattern of gravity, the strength of the stem.

It is not only sloppy to say “apples falling out of trees is random”; It is outright misleading, because it did happen for a reason. If you are modeling systems to better understand their behavior, you must separate the random variables from the orderly processes as much as humanly possible.

If you mix them together you learn nothing.

Science is the discovery and classification of ORDER in nature. True or false?

3. Mark Chenoweth says:

I agree with just about everything you said and that’s why I think biology should adopt the more precise terminology of other fields of study.

Perhaps semantics is the wrong word. You are right, I have not seen Swamidass offer a concrete definition of random, but from everything he has said, it is clear that by random, he does NOT mean a process devoid of order and even admits that mutations can be “biased” in certain ways while I would assume being “random.” The closest I’ve seen Swamidass come to giving a definition would be when he said that to be devoid of randomness would be to be 100 percent accurate. He said, “It has to be 100% exactly to not be random.” So here is my guess for Swamidass’s apophatic definition of random: any process, function, or action that isn’t 100 percent determined. From the perspective of Christian theology, my first goal is to let people know that when biologists like Swamidass use the word, they DON’T mean purposeless, aimless, indiscriminate, etc., Given this definition, being random is even consistent with being ordered and purposeful. Yes, allowing the word to hold within it words that seem to be its exact opposite seems ridiculous to me. But my first goal is to simply let people know that when biologists say random, they don’t mean what we usually think of, and this is great news for Christians!

Yes, this seems like a crazy definition, but whether or not I like the definition, this is great news for Christian theology because it means that the evolutionary process could be highly teleological. That’s my main concern. But I support you in your goal of pushing that sloppy definition out of biology.

Have you found any biologists willing to adopt to more helpful terms like ergodic, etc.? If so, who, and what have they written with this terminology?

• Mark,

Extended Synthesis people are simpatico with terms like ergodic. There is an entire field of biology called semiotics (=viewing biology through a linguistic lens) which is also simpatico. As is Systems Biology. “Ergodic” is the most precisely descriptive term I’ve seen and it incorporates the fact that mutation patterns are not deterministic or entirely predictable.

But one should never confuse ergodic with random. They are immensely different.

Evolutionary biology uses the word “random” an unconventional, imprecise and misleading way – which undermines the field’s very credibility. It alienates people who would otherwise be supportive and creates all kinds of ideological problems.

And don’t forget, there remains a hard core group of atheism / reductionism / scientism advocates who absolutely believe evolution is random, purposeless, and non-teleological. They are a dying breed but they’re hanging on. They have a vested interest in retaining this language. This is the same crowd that touted “Junk DNA” for 30 years.

Allowing any word to hold within it words that seem to be its exact opposite IS ridiculous. I have zero tolerance for this. Sloppy terminology impedes everything it touches. From basic research to disease treatment to biotech to philosophy and religion.

• Mark,

As of December 15, I have been unsuccessful in my requests to get a definition of “randomness” from Joshua. https://discourse.peacefulscience.org/t/perry-marshall-what-is-random/1419/98

4. Eric Holloway says:

Dr. Swamidass believes anything that can be represented as a random variable is “random” including a deterministic function.

https://discourse.peacefulscience.org/t/garte-the-meaning-of-random-mutation/3583/10

‘A boundary case in random variables is a variable that can take on only one value. I.e. they are fully determined. We would still call this a “random” variable, one with a particular sort of distribution, one without any entropy. Yes, the term “random” is that flexible.’

• Eric,

If the term “random” is THAT flexible, then the word is meaningless.

• Eric Holloway says:

Yes, Dr. Swamidass is defining the problem away. sygarte in that thread says as much. I’ve yet to see a satisfactory response to the issue you raise. Anyways, I look forward to what you turn up. You are on the right track.

https://discourse.peacefulscience.org/t/garte-the-meaning-of-random-mutation/3583/13

“OK, I think we have gotten to the end of this long discussion between you and Perry. Its all about definitions which for “random” are about as distant from each other as possible. I would just add that if everything can be described by any particular quality (like being random) then the quality loses most of its value as a description for any particular thing.”

• If everything is “random” including Mark getting out of bed in the morning, and what time his alarm clock goes off.. then random has no meaning. I can’t do much better than this last sentence you quoted right here.